Datasets:

mimiklee

/

masterthesis_clean_text_20k

like 0

Introduction --- Background Sexual violence perpetrated by women against men is poorly investigated and often dismissed in the political and societal scene . SV is defined as any sexual activity committed against someone who is unwilling or unable to consent . This includes sexual acts, attempts to obtain sexual acts, unwanted sexual comments or advances against a person's sexuality using coercion . Coercive strategies may be hands-on or hands-off . --- Social Scripts and Women Perpetrated SV The literature on this topic often uses the term female and male when referring to the perpetration of violence and the myths that support it. In this work, we have chosen to use the terms "women" and "man" as we wish to highlight the social nature of these assertions and disengage from notions of biological essentialism, even though these were not those used in most previous research. SV perpetrated by women against men is often underrepresented in the official statistics and in the scientific literature, partially due to prescribed social scripts and the endorsement of stereotyped gender roles . Social scripts are prescriptive cognitive structures which ascribe traditional gender roles according to an individual's sex . Within these social scripts, rape myths are particularly relevant for this research, as they implicate the invisibility of offending by women and of victimization by men . One of these rape myths has been described as the myth of women's innocence, which concerns the invisibility of offending by women . This myth can in part be explained by the same gender stereotypes that associate women with passive, nonaggressive, caring, and nurturing roles . In turn, this narrative would be incompatible with aggressive, violent, and criminal behavior. Similarly, men rape myths assume that men cannot be forced to have sex against their will, that they are always willing to have sex, and are able to defend themselves if confronted with SV . Additionally, these myths assume that only gay men are victims and/ or perpetrators of men's rape, that men are not impacted by rape, that men's rape only happens in prisons, and that women cannot rape men . Thus, being a victim seems incompatible with stereotypes about men, which obliterates men's experiences of SV . In addition, when men use coercive strategies to obtain sex, they are seen as aggressive, while women using the same strategies are seen as promiscuous . Several studies have shown that women sex offenders are described more positively than men sex offenders, because their behavior is considered more acceptable, less dangerous, and interpreted as affective or romantic . Consequently, in a scenario where women are seen as nonaggressive and men believed to have total control over sexual interactions, men are unlikely seen as victims and more often blamed for the acts they have suffered . The fact that earlier conceptualizations of sexual aggression assumed men as perpetrators and women as victims, such as the ones held in the early version of the widely used Sexual Experiences Survey , or by the FBI until 1992 , created serious obstacles to the estimation of the prevalence of women-perpetrated SV against men . The only victimization of men that was considered was the one perpetrated by other men . The interviews of Sarrel and Masters of 11 men who were sexually molested by women pioneered the research in this field by recognizing that men could also be coerced into sex by means of non-consented sexual stimulation, eliciting fear and anger, which resulted in traumatic responses. --- Prevalence Rates Research focusing on women-perpetrated SV is scarce, but findings concerning both community and criminal samples indicate that women use sexually violent behaviors in their intimate relationships with men, particularly handsoff strategies, such as verbal coercion, psychological manipulation, and blackmailing . These relate, respectively, to pressuring someone with verbal arguments, questioning someone's sexuality, or threatening to use a position of power . Despite this, men and women seem to use the same hands-on and hands-off coercive strategies , and to act similarly in terms of variety and frequency of these strategies, even if their use is prompt by different motivators . A review from Williams et al. , which examined women-perpetrated intimate partner violence in university settings, found prevalence rates for women-perpetrated SV ranging from 1.2 to 46.2% in eight studies. A cross-sectional online survey of 260 women studying in college reported a prevalence of 35.8% of some form of sexual aggression against men . Similarly, in a sample of men seeking help for intimate partner violence, 48.6% reported being victims of women-perpetrated SV, among which 28% were victims of severe SV . Finally, in a meta-analysis including data from 12 countries, women-perpetrated sex crimes accounted for 12% of the ones captured by victimization surveys and 2% of those reported in official statistics . --- Relevance of the Research Topic The literature highlights the real consequences for adult men who suffer SV from women, including guilt, anxiety, depression, sexual dysfunction, somatization, sleep disturbances, trauma, or even suicide . Additionally, the impact of the stigma of being victimized by a women can exacerbate the trauma of the abuse, creating obstacles to help-seeking . Some male victims report fear of not being believed and being considered homosexual . In sum, social scripts, rape myths, and gender stereotypes seem to be contributing to the maintenance and invisibility of the sexual victimization of men by women . Describing how individuals, and particularly university students, perceive the phenomenon of SV perpetrated by women against men seems paramount to initiate the process of challenging these social mechanisms. Thus, previous studies have examined these perceptions in groups such as university students . --- Current Study The current study gathers the perspectives of university students on women-perpetrated SV against men. University students were interviewed after reading a description of a hypothetical situation of SV. The interviews conducted aimed to answer the following research question: How do university students perceive SV perpetrated by women against men? --- Method --- --- Procedures The study was approved by the Ethics Committe of the School of Psychology and Life Sciences at Lusófona University. All participants were a priori informed of the topic and sensitive nature of the research, and that they could withdraw at any moment. The results presented below are based on thematic analysis . This methodological approach is anchored in a reflexive constructionist background, which seeks to understand and problematize the way respondents make sense of the world around them, rather than seeking to objectively describe phenomena "as they are." Because of this, and considering the topic, special attention was paid to the interviewing process when it comes to making the participants feel criticized or afraid to express their positions and worldviews, regardless of how they might conflict, or not, with the extant literature on the topic. The interviews were not intended to ascertain whether participants had factually correct information about the topic, but how they made sense of it. The methodological design for this study involved the creation of a short vignette as in Fig. 1 which portrayed a fictitious situation where a woman is on a date with a man, at his place, and resorts to both physical and psychological strategies to pressure him into having sex. The storyline was based on real perpetration episodes of women against men described by Sarrel and Masters and on other vignettes used in previous research . The specific violent strategies included groping, pressuring, insulting, and threatening of posting on Facebook that the victim could not get and erection if he did not take off his pants, and were based on the items of the Sexually Aggressive Behaviors Scale . The vignette in this study was co-created by two of the authors , and informally discussed with a group of postgraduate students to ascertain whether the phrasing and language were understandable and context-appropriate for university undergraduate students. The vignette was used in the beginning of a face-to-face interview to spark interest, reflection and debate, and to act as an icebreaker, considering the sensitive nature of the topic at hand. It also allowed the respondents to be able to externalize or project any opinions or thoughts that they might consider to be socially undesirable to express as themselves, or any experiences they might have had, onto the characters in the vignette. The interviewees were of the same gender of the interviewers, to prevent, within reason, any potential discomfort for either the researchers or the interviewees, considering the sensitive nature of the topic being covered by the interviewing process, and attentive interviewing practices were used. Furthermore, as stated above, the participants were duly informed of the topic of the research before the participation. The institution had psychology services available to the participants, and this information was provided to the participants before and after their participation in the study. No participants disclosed any discomfort during or after the interviews, and no participants disclosed any previous experience of perpetrating or having been victim of sexual violence. Participants provided written informed consent for participation in the study and publication of the results before the interviews. All interviews were conducted in European Portuguese. --- Data Analysis Computer-assisted qualitative analysis was conducted through the deployment of NVivo 12 PRO. The procedure was followed as laid out by Braun and Clarke .This meant that interviews were first read to identify the main elements and characteristics that would stand out the most, and then a tentative coding scheme was formed from it. The research team collaborated on the creation and discussion of the coding scheme. As per the steps described by Braun and Clarke , the interviews were read repeatedly for the researchers to become familiar with the data, and initial codes were generated, which represent basic units of meaning within the dataset, privileging explicitor semantic-content and meaning, rather than implicit meaning. Each response was taken, in toto, as a unit of coding analysis, to better preserve the complexity and connectedness of those responses, especially considering that data collection was done through semi-structured interviewing. An almost-final version was then tested for coding consistency using NVivo, and any issues identified were then clarified and discussed among the team. After that, the coding scheme was re-discussed to optimize it, by joining together thematically similar codes until there was no overlap between similar codes. The final coding scheme was thus the product of the authors' iterative process. The discussion of the scheme was only complete when there was full agreement among the whole team about the themes, sub-themes and codes, their definitions, and what interview excerpts belonged to each code. As is typical with thematic analysis, the same excerpt could be coded in different codes simultaneously, as the same sentence can allude and refer to several ideas at the same time. --- Results --- This thematic analysis identified four main themes: Descriptions of Woman Aggressor and Man receiving Abuse, Identification of Sexual Violence Strategies used by Woman Aggressor, Gendered Consequences of Sexual Violence Perpetrated by a Woman, and Motivations and Contexts for Woman Perpetrating Sexual Violence. The first theme included two sub-themes focused either on the Woman Aggressor or on the Man receiving the abuse described in the vignette presented. The theme Strategies was divided into two subthemes while the theme Motivations and Contexts was divided into three sub-themes . The Consequences theme was divided into the sub-themes Man receiving abuse and Woman Aggressor and included perceptions about the characters from the vignette but also about abstracted "men" and "women" in --- Translated vignette Martha got ready that evening to meet John for the first time because she found him extremely attractive. John was waiting for Martha to watch a movie with her at his place. As they were watching the movie, Martha started getting closer to John, sat on his lap, and started biting his ear. She said she was "really eager to be with him", although he looked a bit uncomfortable. John asked Martha to move away. Martha did not seem to care, as she stayed where she was, and she started grabbing and groping John. He asked her again to stop and explained that he was tired and just felt like watching the movie in peace. John said "Afterwards, who knows, maybe we can do it" and smiled at her. However, Martha continued to provoke him, placing her hand inside his pants. John started getting an erection, although he kept pushing Martha away. After a while, Martha got annoyed, got on top of John, and said: "I want to do it here and now". As John seemed uncomfortable with the situation, Martha got mad and said that João must be a "fagot" to refuse. John replied that he only wanted to finish the movie in peace and that she knew very well we was not a "fagot". Martha replied "yes… but if you don't take off your pants, I will post on Facebook that you could not get an erection and you will regret it!". similar situations. Within each of the subthemes, we classified several codes. See Fig. 2 for a graphic representation of themes and subthemes. Further down in Table 1, a description and verbatim examples of all codes are presented. We have not included frequencies per Braun and Clarke's recommendations, since these might be misinterpreted as proxies for the relevance or salience of the data presentedour objective here is to understand the breadth of ways in which participants make sense of the vignette and, by extension, the topic of woman-on-man sexual violence; we make no claim to statistical generalizability, but we do note on the salience and relevance of particular aspects of our data. The examples included correspond to translated extracts of our participants' contributions in Portuguese. The translation was performed by L.O. --- Descriptions of Woman Aggressor and Man Receiving Abuse Participants tended to describe the Woman Aggressor in a mostly negative light, while describing the Man receiving abuse in a mostly positive light. Both characters were seen as defying gender stereotypes. In the transcript below, the participant voiced the expectation that girls in a situation where they felt their partner was uncomfortable with engaging in sexual acts would give up on it, while boys in similar situations would pressure girls into it, a scenario opposed to the one described in the vignette: I believe that there are many girls who, in situations like this, try to have something with boys, of course they do, but then they realize that they are not comfortable and will not force the situation. In the part of forcing the situation, I believe it is more frequently done by boys No, I just found it curious because it is usually the opposite situation, but I think that maybe the girls in this interview might be the opposite situation. --- Woman Aggressor The depictions of the Woman Aggressor were mostly negative , as expressed in the participants lingo describing her as "childish," "immature," or in BC's perspective, a man, "she is a bad person, so if she threatens John for not doing something he doesn't want to do, I think, okay, she is bad" . Participants also described Martha as homophobic, as seen in this example: "The fact that she started calling him a faggot… that language means she is being homophobic, isn't it? ". The woman was frequently described as being seductive, which entailed constructions from our participants that were somewhat neutral Maybe even sassy, not in a bad way…") , and others more negative and emotionally charged, as seen in the way some participants slut-shamed Martha . --- Man Receiving Abuse As for the depictions of the Man receiving abuse, these are mostly positive , present in the participants discourse in expressions such as "a normal person," "concerned," "a good person." Opinions regarding the man's self-expression were split, as some believed his posture was assertive despite the woman's sexual initiation strategies, and others believed he was introvert --- Identification of sexual violence strategies used by the women aggressor --- Physical --- Genital stimulation The woman stimulates the recipient's genitals without his consent or will Marta goes so far as to take João's pants in her hands, that is, there is already a sexual touch and an attempt to initiate a sexual act --- Non-genital unwanted touching The woman touches or gropes the recipient without his consent or will Is also criticized for having grabbed and groped even after he said he didn't want to and was uncomfortable --- Physical coercion The woman uses physical force so that the recipient acts involuntarily Acting violently and the other person wants to stop, and it doesn't stop --- Using an unconscious person The woman takes advantage of an unconscious or intoxicated person When they get extremely drunk people or, on any type of substance and, oops, the person goes, doesn't say no, doesn't oppose yes, but is not completely aware --- Psychological --- Blackmailing The woman threats to publicly reveal something compromising about the recipient --- A story of blackmail Harassment The woman pressures the recipient to go beyond his limits The psychological pressure she put on him Table 1 --- Main themes --- Subthemes --- Codes --- Description --- Verbatim Examples --- Humiliation The woman degrades or humiliates the man The fact that she insulted him and said that she put it on Facebook, I don't think anything justifies that --- Gendered consequences of sexual violence perpetrated by a woman --- Man --- Anger The man will become angry I --- Gender roles The situation is the result of as in internalized stereotyped gender roles or in "natural" gender tendencies We try to strive for equality, however we still continue to say that the stronger sex is the man and normally this type of behavior is adopted by the strongest link --- Homophobic assumptions The situation is a result of toxic masculinity He's the one who's worthless or not, that's it, who doesn't like women --- Recurrent situation The situation is frequent or realistic --- I believe they are frequent actions --- Sexual objectification The situation is a consequence of degrading ideas about people's bodies and sexuality I saw João as a person and not as a sexual object, although I may feel extremely attracted to the boy, he is not a sexual object to favor me and submissive. As an example, the young man's assertiveness is described in LV's voice : …there at some point he starts being more secure and saying, "you know perfectly well I'm not that, I just want to finish watching the movie" and that's it, I want to believe that he was going to keep the same stance, or even, okay, send her away or whatever. I don't believe he would submit to that. On the other hand, views of John's submission opposed, as in HB's perspective, a young man, John is "… having a little bit of a passive attitude, in my opinion. I don't think he should just say, «Oh, I want to see the movie». I think he should say «Stop! I don't want to! Get off me!» I think you should be a little less passive." Finally, one participant hypothesized that "John might not be heterosexual. That information is not here, that he desired Martha. But he did get an erection, so it does not make a lot of sense, but ok." . --- Identification of Sexual Violence Strategies Used by Woman Aggressor The students described varied coercive tactics used to initiate sexual contact to describe Martha's behavior, but also strategies used in other violent encounters. --- Physical Strategies Physical strategies were coded as genital stimulation, nongenital unwanted touching, physical coercion, and using an unconscious person or, in these students' voices, "put her hand down his pants," "gropes and grabs," "forced the act," or "when they catch extremely drunk people." Furthermore, for several participants, the use of physical force or direct physical stimulation was where they drew the line of the abuse. For them, when these strategies were in place, it seemed clearer they were in the presence of a situation of SV. As one of our participants described: Ok. I think this could be [sexual violence] … Because of this: because he says no, Martha continues to encourage the situation, John continues to say no, he just wants to see the movie and Martha goes so far as to take John's pants in her hands, that is, there is already a sexual touch and already the attempt to initiate a sexual act. [...] Against John's will. In other words, for me it is, it's already sexual violence. --- Psychological Strategies Psychological strategies mentioned by participants included blackmailing, harassment, and humiliation. Most participants used the word "blackmail" to describe the actions of the woman, others mentioned strategies that were less specific, often named as "psychological pressure," and others specifically focused on how the woman was degrading or pushed the man's boundaries. One participant described: It's like that, I think the most objectionable part there is after he says two or three times no, she continues. That's the worst, because there it is, everyone has their desires, and she had desires at first and there's no harm in trying... After a while it starts to be too much, there it is, passes that barrier and it is already too much. For me it's just that and the fact that she wants to put it on Facebook, that he didn't get an erection. I think it's the worst thing about her. However, it should be noted that not all participants would describe the situation portrayed as SV, some even employed commonly seen rhetoric blaming the victim for their situation: "I think sexual violence happens when one of the people ends up being forced. Here, John was not forced. Only when she said she was going to post on Facebook" . --- Gendered Consequences of Sexual Violence Perpetrated by a Woman When participants discussed consequences of the situation presented in the vignette, they tended to focus on the aftermath of the potential Facebook post. The use of this strategy was described as surprising to many, and often seemed to overrun other aspects of violence. Participants considered the threat regarding the Facebook post the form of violence portrayed in the vignette that would be less likely to occur in the real world. These consequences were coded under consequences for the Women Aggressor and for the Man receiving abuse. --- Women Aggressor In this context, individual consequences for the woman included personal distress or inconvenience and official charges. One participant described this distress like this: "John might look affected by the Facebook post, so she would still not get what she wanted, which would increase her frustration." . Some participants described the possibility of a formal complaint: "His friends would report it to an authority" . Within social consequences, some participants anticipated impunity and social support, while others anticipated a negative public image and retaliation. For instance, one participant justified Martha's anticipated impunity as follows: "I think if I saw that on my feed, I would have thought it was a joke, so I would keep scrolling. I think maybe nothing would happen to Martha, because everyone would have the same reaction as me" . Several participants mentioned that social support would be given to both Martha and John, as exemplified here: "I think his friends would support him and her friends would support her, but his public image would always be affected" . Regarding Martha's public image, participants believed this violent incident would have a significant negative impact, as described: "I want to believe that most people would point the finger at Martha and that they would have an attitude of rejection by saying «this is not correct»" . One participant mentioned the possibility of retaliation: "Another possibility would be that John would turn to violence against her. I would not rule that out" . --- Man Receiving Abuse Anticipated consequences for the man included impact for the self or negative emotions, including discomfort and anger. Some participants thought that the man might have an attitude of devaluation of the abuse and that this might be inconsequential for him, as exemplified here: "If he had a clear conscience, because this does not define his personality, he would not mind that much that she posts about it on Facebook. Although it is a defamation, him and his friends know that it does not define his personality" . Many participants believed that the man might give in to sex, eventually: "Maybe he really wants it, but now he is focused on the movie and then when the movie finishes… he accepts Martha's pressure and does it in that moment" . Participants anticipated both social support and social stigma. As an example of a description of social stigma, one participant mentioned: "I think it would destroy John's life a little. With social networks it is easy to expose someone and to be seen by millions of people. Even if it is a rumor, people believe it right away" . As an example of this stigma, some participants employed commonly seen rhetoric blaming the victim for their situation: He should have considered the possibility that this could happen… When I bring a girl home, it is likely that it happens. […] I think it is sexual violence, however, they already knew each other, and this could happen between them hum… I think that in this case it isn't, but if they were strangers or if they barely knew each other, it would be [sexual violence]. […] It depends on their relationship. Finally, one participant mentions several of the aspects referred under this theme: Probably John, if he used his head, would take Martha to court for having exposed such a situation because it is cyberbullying… Martha would be upset about this situation and the rest would depend on the people around her. If John had good friends, obviously they would defend him. If Martha had good friends, good people, they would say "Are you silly? This can't be done!" or else there would be very bad people who would make fun of John, and he would start to be a victim of bullying, he would be humiliated, and he could kill himself. --- Motivations and Contexts for Woman Perpetrating Sexual Violence Participants offered different levels of explanations for situations such as John and Martha's, including intrapersonal, interpersonal, and sociocultural explanations. --- Intrapersonal Within Intrapersonal motivations/contexts they focused on individual facets of Martha as personality, inner beliefs of homophobia, higher sex drive, or ways of coping with rejection. The following quote summarizes some of these aspects: "Martha might have been a very insecure person and feel like she was not desired by John." . Some students reflected that Martha could have been a victim or a witness of SV, and that she may be re-enacting past abuse, as exemplified: "Maybe someone has done the same to her" . --- Interpersonal Interpersonal explanations included relationship imbalances in power , in sexual attraction or concerning a sex drive discrepancy. For instance: "John was extremely attractive, and she wanted to be with him because that was her desire. Maybe John did not want to be with her so much. Maybe when he was with her, he did not feel that connection. But Martha desired him a lot, sexually." There seemed to be a split between those who felt Martha's actions were due to her feelings for John of romanticism , or, contrarily, due to her emotional detachment or their non-coupledom . Some students explained her action as acts of revenge against his seeming lack of interest in pursuing sex with Martha, as reported here: "Or maybe she did it out of revenge because she could not get what she wanted, so she will try to impose something bad on him, 1 3 in this case, sharing these lies with the intent to embarrass him" . --- Sociocultural Lastly, sociocultural explanations focused on the overarching context of abuse and the wider social norms, particularly on gender roles. As exemplified by this participant, some students approached SV using stereotypes regarding the sexual drive of each gender, among other gender stereotypes: "This situation seems uncommon to me. Men have more sexual and hormonal appetites than women and in this case, she is the one going after him, which is uncommon." . The students explained that disinformation, or the lack of education on consent and SV could be contributing to the phenomenon: "First of all, education at home and then at school. I think at school this is not talked about. Ok, a boy aged 7 or 8 knows it is wrong to touch a girl without her authorization, but that is about it. People are not informed enough." . Students also reasoned that SV could be a consequence of sexual objectification, as mentioned by this participant: If I was in Martha's position, I would see John as a person, not as a sexual object, although I might have felt extremely attracted to him. He is not a sexual object that needs to satisfy me. He is a person and has a person, I have to respect him. They also added that certain facilitating contexts may increase the risk of SV: "I think that these situations often occur in certain environments, at home, or sometimes in a bar or a night out, in which only one person wants it and starts provoking until the other one gives into it" . Many highlighted that the fact that SV is a recurrent situation, as exemplified here: "Unfortunately, I think this is a common situation, but in this case, we are seeing the other perspective. Normally, it is a man doing it toa woman, but it is true that it also happens the other way around" . In conclusion, although the situation depicted referred to the violence of a woman toward a man, the discourse often translated the standard narrative, that is, the victimization of women by men. One exception refers to changing times, when participants reflected on changes in societal values that might be responsible for a rise of women's initiative. As one participant described: In my time when I was younger 6 or 7 years ago it would be... I'm aware that it would be like I'm saying: men either put much more pressure on women and that women had… ah... they felt much more pressured and manipulated than exactly in this moment now we live in because now women are having relationships at an earlier age and as I said also because of social pressure. It's happening a lot more than it used to because, maybe, before, seeing women with 18 years old virgin was more normal than it is now. Without drawing any conclusion like that, because I can't draw without a statistical basis… But drawing a conclusion like that theoretically I think that now there are more situations as described in the text and there is more and more balance. --- Discussion The aim of the current study was to explore in depth the perceptions of a group of university students regarding SV perpetrated by women against men. This study provided a discussion focused on social scripts concerning SV, namely men's rape myths. This was the first study to evaluate perceptions regarding women-perpetrated SV against men in a community sample in Portugal. When describing the woman's features, participants attributed her characteristics such as childish, homophobic, hostile, impulsive, manipulative, and seductive, even calling her a "slut," and mentioning that she defies gender stereotypes. Here, participants admitted that they would tend to attribute these characteristics or these behaviors to men and not to women. In fact, several studies have shown that perpetrators of SV tend to be described differently according to their gender, with women usually being described more leniently . Participants were aware of this counter-stereotypical pattern. Similarly, the man was also described as defying stereotypes, with some participants describing his behavior as normal, assertive, and thoughtful, while others described him as introvert, submissive, passive. The same behavior was interpreted by some participants as assertive and by others as submissive. The participants that described the man as passive and submissive expected him to act more proactively and assertively, revealing an attitude of blaming the victim for the violent situation he was in, which has been previously reported . A few participants argued that the man could have acted differently while simultaneously declaring that he could not be blamed for what had happened, revealing an ambiguity in their evaluations of this situation. These statements mirror men's rape myths that assume that men cannot be coerced into sex, and that they should be able to act proactively in order to defend themselves from violence . Additionally, participants associated the victim's behavior with the possibility that he might not be heterosexual. In the vignette, when the victim refuses the woman's attempt at sexual contact, she uses a homophobic insult to coerce him. Thus, she expresses the men's rape myth that "real men" are always ready to have sexual contact with a woman and that those who are not, must be gay . Some participants agreed with this interpretation, although some did not endorse the violent reaction to the victim's lack of consent. Thus, by acting in a way that defies gender stereotypes, the man was assumed to be gay, and not necessarily a victim. This is a clear example of the definition of men's rape myth and of the impact that these myths can have on silencing the victimization experiences of men, as reported in previous studies . A qualitative thematic synthesis described the impact that sexual assault can have on men's perceptions regarding their gender identity and sexual orientation, which highlights the real impact of men's rape myths . The participants of the current study expressed other gender stereotypes such as the idea that men have higher libido than women and more difficulty inhibiting sexual impulses. These stereotypes were used in some instances to justify the higher prevalence of SV perpetrated by men against women. This highlights again the influence of social scripts on perceptions regarding what is expected in sexual encounters and how SV takes place . In this case, according to some participants, the man would not fit the stereotype because he would have low libido and/or high impulse control. Nonetheless, some participants expressed awareness of the invisibility of sexual victimization by men, referring the impact of stigma on social perceptions regarding its prevalence. The experiences of stigma in men survivors of SV have been explored in previous qualitative studies . Most participants identified the different SV strategies used in the vignette. Some participants recognized psychological abuse as SV strategies. Some were particularly surprised with the threat regarding the Facebook post, which seemed to be more impactful than other violent strategies. Participants assumed that this strategy would be less frequent than the other ones. These beliefs were anchored on the idea that this online post would be a form of public humiliation that could have adverse consequences for both the victim and the offender, as people might criticize and sanction the publisher's behavior while mocking and stigmatizing the victim. Violent physical strategies to initiate sexual interactions were also identified, namely genital stimulation, non-genital unwanted touching, physical coercion and using an unconscious person. Some participants only identified the situation as SV when abusive physical touch was involved. This mirrors the stereotype of SV as equivalent to hands-on abusive touch, particularly when the victim is a man, overshadowing other forms of violence . A higher likelihood of labeling abuse as SV when physically abusive touch is involved has been reported in previous studies, among various samples . However, not every participant would describe the situation presented as SV. A participant that explicitly verbalized that he would not consider this situation an SV incident justified that based on the relationship between the woman and the man. As they knew each other and there was an expectation regarding that interaction, the participant thought that the behavior of the woman could not be classified as violence. The influence of the relationship between the victim and the offender on the perceptions of seriousness of the abuse as also been described in previous studies . Participants could imagine different consequences of the abuse. When asked whether they thought the man would give in to sex with the woman, most participants thought he would not, as they believed he would be able to deal with the situation. A few participants expressed that the man would be "weak" if he did give in to sex after being abused. Others thought that he might, if he felt attracted to the woman, which could be a possibility, as he showed signs of arousal . Thus, some participants were not aware that physical arousal can happen in SV situations due to physiological mechanisms that do not reflect the individual's consent . This has been reported in previous studies, namely with samples of men SV victims . In this context, the current data shows that some participants placed the responsibility of the impact of the abuse on the victim, assuming that by having more control over the violent situation, the impact could have been diminished. Again, blaming the victim narratives were at play here, as described in previous studies . Some participants could anticipate adverse consequences of the abuse for the victim on an individual level, including anger and discomfort, and on a social level, namely social stigma. However, others thought that the abuse could have no consequences and that the victim could devaluate its impact. The underestimation of the impact of SV on men, especially in instances where physical SV did not take place, has been reported previously . For instance, in a study by Moore and Miller-Perrin , university students thought that the woman victim portrayed in the vignette experienced less feelings of enjoyment and greater feelings of shame, guilt, embarrassment, and trauma compared to the men victim. The anticipated consequences of the abuse for the woman reported by the participants of the current study were a negative public image, official charges, distress, and retaliation from the victim. Some participants were aware of the criminal framing of these behaviors and of the need for social sanctioning of such conduct. It was assumed by participants that the woman's behavior may be answered with impunity, because of her gender. This is a clear reflection of the myth of women's innocence. Indeed, several studies have been raising the issue of a gender gap in sentencing, proposing that a leniency effect toward women might benefit them . This is also in line with the results of the study by Moore and Miller-Perrin in which university students perceived the woman vignette offender as less guilty than the man perpetrator. In sum, the participants of the current study showed an understanding of the possible practical implications of gender stereotypes in this situation. Participants were also aware that some people might support the victim while others might support the offender. Thus, participants could picture several scenarios, some more in line with the ideal approach of supporting the victim and sanctioning the offender while others expected perpetrator's behavior to be devalued and the victim to be stigmatized due to their genders. The impact of stigma and social support in the recovery process of victims and survivors of SV has been widely documented . For instance, several men victims have reported being explicitly told by professionals or members of their support network that as men they could not be victims of SV, which exacerbated the negative consequences of the abuse . When discussing potential motivations, students mentioned both intrapersonal, interpersonal, and sociocultural factors that could help explain the occurrence of the abuse. They argued that the woman could have acted the way she did to cope with rejection, due to homophobia, due to her personality, to her sex drive or because she was re-enacting past abuse. The motivations mentioned by participants are mostly in line with the etiology of SV described in the literature, which involves a complex interplay of trauma, mental health and personality difficulties, and particular psychosexual characteristics . In particular, participants mentioned the re-enactment of past abuse, which shows a mature understanding of victimization and offending dynamics, as the literature reports a significant proportion of women sex offenders with histories of childhood abuse . Within interpersonal factors, participants mentioned a need for dominance, wanting revenge, emotional detachment, non-coupledom, romanticism, sex drive discrepancy, and sexual attraction. Thus, participants could explain this violent incident with both negatively charged dynamics such as dominance and revenge and with a disruption of expected positive dynamics such as romanticism and sexual attraction . This is in line with results from previous studies where participants attributed sexually violent behavior perpetrated by women to romanticism . Additionally, the feelings of the woman toward the victim were used to explain the abuse in two opposing ways: some participants thought the perpetrator was violent because she had feelings for the victim and felt rejected, while others thought she could do that because there was no emotional or romantic attachment. Most participants mentioned a feeling of rejection, with some resorting to stereotypes such as the idea that women tend to get all the men they want, so they do not know how to deal with rejection. Finally, on a sociocultural level, participants identified several facilitating factors including: gender roles, homophobic assumptions, disinformation regarding sexual education, facilitating contexts, sexual objectification, an evolution in societal values, and the fact that this would be a recurrent situation. These findings are in line with reports from previous research which highlight gender roles and homophobia as inescapable sociocultural factors involved in perceptions regarding women-perpetrated SV against men . Men's gendered experiences with SV were analyzed in a qualitative thematic synthesis which underlined the influence of gender stereotypes in violent strategies used against men, in their processing of victimization experiences, and in their behavioral and helpseeking responses to victimization . According to these authors, the social construction of gender and masculinity create a specific risk for the sexual victimization of men associated with the pressure to always be ready to engage in sexual activity. Alongside these, the fear of homophobic stigma seems to be an ever-present threat for men . Participants in the current study also pointed out disinformation regarding sexual education as a potential facilitating factor. Indeed, previous literature has advocated for the potential role of sexuality education in the prevention of SV . Within facilitating contexts, participants mentioned the influence of alcohol or drugs, in line with previous reports, as these are typically interpreted as facilitating factors . Finally, participants were aware of the association between sexual objectification and SV, which has been widely researched . The results of this study highlight the presence of gender stereotypes in the participants' accounts, although some did reflect on the need to challenge these and on an increasing tendency for these to be less endorsed and impactful. --- Limitations The results of this study should be interpreted in light of its limitations. A selection bias and a social desirability effect may have influenced the results of this study. In other words, students who participated in this study may have been more aware of SV and, in particular, of women-perpetrated SV, providing answers which were in line with their expectations of what a desirable answer would be. This study used a single vignette, which did not allow the exploration of participants' perceptions of SV in other scenarios, namely with different gender combinations. For instance, a vignette where the gender of the perpetrator was not specified but where the gender of the victim was explicitly of a men's, might have induced different responses. Thus, no interpretations can be drawn regarding participants' perceptions of SV in other contexts or offender-victim pairs. Nonetheless, we highlight that the vignette was designed according to reports of womenperpetrated SV against men shared in previous literature; hence, the scenario in the vignette should be realistic enough to elicit reactions comparable to those elicited by actual cases. Finally, no questions were asked regarding participants' perceptions of their willingness to intervene in similar situations, so practical implications from this qualitative data cannot be drawn. In other words, despite recognizing the situation as violence, it is not known whether participants would act as bystanders in similar situations in order to prevent them or intervene, for instance, calling out the offender or reporting to the police. Future studies are needed to further clarify the link between these perceptions and the willingness to intervene in SV situations. --- Recommendations Previous research has identified differences in the attribution of seriousness of women-perpetrated SV according to several factors, including gender and professional occupation . An interesting avenue for research would be the exploration of the associations between individual factors and perceptions regarding women-perpetrated SV, including men's rape myths . This line of research could also be enriched by exploring the intersections between the victim's gender and other situations which may increase their risk of stigmatization, such as belonging to the LGBTQ + community, having addiction or mental health issues, being involved in sex work, having a disability, being an immigrant, migrant or refugee, or being in a situation of homelessness, among others . Some participants in this study brought up the sexual orientation of the victim, questioning whether he might not be heterosexual. Thus, sexual orientation seemed to impact the participants' perceptions, which further justifies the exploration of these intersections. Thus, future studies could evaluate whether men victims are further stigmatized when they belong to other minority or vulnerable groups and whether blaming the victim narratives contribute to maintaining that stigmatization. --- Policy Implications This study highlights the impact of social scripts and of men's rape myths in the perceptions of university students regarding women-perpetrated SV against men. Rape myths have been identified in other research samples, including groups of professionals who contact with potential victims daily, such as police officers, health care professionals, and judicial professionals . Challenging rape myths among these groups is imperative to offering appropriate care and intervention. Thus, offering SV specialized training to these groups of professionals may be crucial to avoid the invisibility of victims and the impunity of offenders Finally, up-to-date evidence-based SV prevention programs targeting university students could be further refined in order to challenge men's rape myths and social scripts regarding women-perpetrated SV . Having a holistic approach that, while acknowledging the gender imbalance of SV, also raises awareness for victimization of men and offending by women seems paramount for ensuring the prevention of SV against all genders . The inclusion of these contents in prevention programs must be tested with rigorous methodologies, ideally in the context of randomized controlled trials. Public policies that foster the application of such prevention programs in university settings, such as the 2013 Campus Sexual Violence Elimination Act applied in the USA, may have an important role in ensuring the translation of this research into practice . --- Conclusions This study provides and in-depth exploration of the perceptions of university students regarding women-perpetrated SV, showing that gender stereotypes and rape myths play a significant role in the narratives of participants. Participants endorsed some gender stereotypes and used blaming the victim narratives, while also being able to challenge other gender stereotypes and rape myths. Participants could identify violent sexual initiation strategies, could anticipate potential consequences of the abuse and were able to imagine potential motivational and contextual factors that facilitated the abuse. These results highlight the importance of providing adequate information regarding women-perpetrated SV to challenge the invisibility of offending by women and victimization of men. Qualitative data on the general public's perceptions regarding women-perpetrated SV against men can inform preventative approaches targeting social norms that perpetuate the invisibility of this phenomenon. Author Contribution Joana Carvalho, Daniel Cardoso, and Nélio Brazão contributed to the study conception and design. Material preparation was performed by Joana Carvalho and Daniel Cardoso, data collection by Nélio Brazão, Mária Viegas, and Rita Vespasiano, analysis by Leonor de Oliveira and Daniel Cardoso, and interpretation by Leonor de Oliveira, Daniel Cardoso, Joana Carvalho, and Nélio Brazão. The first draft of the manuscript was written by Leonor de Oliveira, Eunice Carmo, and Daniel Cardoso and all authors commented/reviewed on previous versions of the manuscript. Project administration and supervision were performed by Joana Carvalho. All authors read and approved the final manuscript. Funding Open access funding provided by FCT|FCCN . This study is part of the project FEMOFFENCE-The myth of innocence: A mixed methods approach toward the understanding of female sexual offending behavior , supported by Fundação para a Ciência e Tecnologia through national funds granted to the Principal Investigator Joana Carvalho. --- --- Competing Interests The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. --- Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Introduction Sexual violence (SV) perpetrated by women against men is socially dismissed and underrepresented in research. The aim of the current study was to explore the perspectives of university students (women and men) on women-perpetrated SV against men. Methods A total of 19 undergraduate students were presented with a vignette describing a hypothetical situation of SV and interviewed. Results A thematic analysis was performed, identifying four main themes: characters' Features, sexual initiation Strategies, Consequences, and Motivations and Contexts. Participants endorsed gender stereotypes and victim-blaming narratives, but also challenged gender stereotypes and rape myths. Participants could identify violent sexual initiation strategies, could anticipate potential consequences of the abuse, and were able to imagine potential motivational and contextual factors that facilitated the abuse. Conclusions These results highlight the importance of providing adequate information regarding women-perpetrated SV and can inform preventative approaches targeting social norms that perpetuate the invisibility of this phenomenon. Recommendations for future research are discussed. Policy Implications Public policies that facilitate the training of professionals who contact with victims may help overcome the influence of rape myths that hinder appropriate intervention. Similarly, policies that support the prevention of SV in university contexts may contribute to translate the results from research into practice.

Background In 2021, the World Health Organization estimated that there were 247 million cases and 619,000 deaths of malaria in 84 malaria-endemic countries globally, and most of the malaria cases and deaths occurred in sub-Saharan Africa [1]. Children under 5 years bear a disproportionately high share of the global malaria burden, accounting for 67% of malaria deaths [2]. Children are at higher risk of contracting malaria and developing severe consequences, including death. Those children who survive often struggle with chronic anemia, seizures, or cognitive impairment, which severely hamper their growth, development, and attendance at school [3]. With evidence-based diagnostic and therapeutic procedures available, timely diagnosis and treatment among these children are vital to avoid the severe consequences above [4]. WHO recommends early diagnosis and prompt, effective treatment within 24-48 h of the onset of malaria symptoms [4]. Suspected malaria should be confirmed with microscopy or immunochromatographic rapid diagnostic tests . The results of such a parasitological test should be available within a short time of the patient presenting at the health facility [4]. Children with uncomplicated Plasmodium falciparum malaria should be treated with qualityassured artemisinin-based combination therapies [5,6]. The algorithm for uncomplicated Plasmodium falciparum malaria diagnosis and treatment is demonstrated in Additional file 1: Fig. S1. However, current evidence suggests considerable gaps in the access to the laboratory diagnosis of malaria among children in sub-Saharan Africa, though WHO recommends a confirmatory blood test for all suspected cases of malaria. First, febrile children often are not taken to health facilities for formal care [1, [7][8][9][10][11][12]. In 2015-2019, the mean treatment-seeking rate for febrile children under five years of age was 69% in sub-Saharan African countries, which only improved moderately from 65% in 2005-2011 [1]. The extent of this poor care-seeking behavior varies across countries. For instance, only 41% of febrile children from Ethiopia sought care in 2016, while 81% of febrile children from Liberia sought care in 2014 [13]. According to the current knowledge, care-seeking behavior can be affected by factors at the child level , factors at the caregiver level , and factors at the household level [e.g., socioeconomic status , type of residence , the travel time to health facilities [7,9,14]. Countries also vary in the distribution of related factors. For instance, it was estimated that the probabilities of care-seeking for fever at the nearest primary healthcare facility within 30-min travel time were 21% and 69% in Ethiopia and Uganda, respectively [7]. Second, blood tests for all suspected cases of malaria cannot be guaranteed even if formal care is sought, in which circumstances inappropriate treatment might occur thereafter [8,[15][16][17][18][19][20][21]. A cross-sectional study based on Malaria Indicators Surveys in 25 countries reported that 58% of febrile children under five years who had sought care received poor quality of case management for suspected malaria, with 62% receiving no blood tests, 82% receiving no antimalarial drug, and 72% receiving treatment more than 24 h after onset of fever [22]. This study suggested that regional disparities in malaria care quality could be driven by socioeconomic characteristics, but this study did not further examine these hypothesized associations. Another study using Service Provision Assessment surveys also found that the majority of children treated for malaria across the nine surveyed sub-Saharan African countries did not receive a blood test diagnosis and an appropriate antimalarial [20]. But this study primarily focused on facility-level factors due to the lack of data on patient characteristics. Other single-country studies also explored the potential predictors of quality from the perspective of the health system, such as the availability of diagnostic tools, antimalarial medicines, and health workers [18,19,21,23]. There is still much to know about the factors affecting the access to the laboratory diagnosis of malaria. Particularly, more remains to be known about how a patient's SES would affect the access to the laboratory diagnosis of malaria. Lower SES has been widely recognized to be associated with higher malaria incidence and lower accessibility to malaria control interventions [10,23,24]. However, research on the association between SES and the receipt of blood tests once a patient has arrived at a health facility remains limited and inconclusive. Furthermore, it is unknown how facility-level factors might interact with a patient's SES to influence the receipt of blood tests at a health facility. Lastly, most previous studies used national or sub-national data to examine the care-seeking and clinical management of malaria among children in sub-Saharan Africa [18][19][20][21]. Cross-country comparison is needed to understand how care-seeking behaviors and clinical practice vary across settings and possibly to find the common hurdles to the coverage of malaria diagnosis. This study attempted to assess access to health facilities and the laboratory diagnosis of malaria among children. To fulfill the aim of the study, we examined the variation in the access to the laboratory diagnosis of malaria among children at the national and sub-national levels. Furthermore, we investigated how SES affected the access to the laboratory diagnosis of malaria among children, and whether such associations would be modified by the type of health facilities patients visited. The following hypotheses were formulated: febrile children with higher SES would be more likely to receive formal care; febrile children with higher SES would tend to receive blood tests once they arrived at a health facility, and this association would differ across the type of health facilities. The rationale for these hypotheses was that parents with higher SES might have greater awareness about the disease and more resources to seek healthcare. --- Methods --- Study design We conducted a cross-sectional study in 19 malariaendemic countries in sub-Sahara Africa using the Demographic Health Surveys and the Malaria Indicators Surveys , both of which are repeated cross-sectional surveys [1]. We intended to include all DHS and MIS in a relatively tight time frame to facilitate cross-country comparison. We further restricted the study period to 2016-2018 to maximize the number of surveys included in this study. --- Data sources DHS and MIS are both nationally representative surveys that provide data on a wide range of demographic and health topics, including malaria treatment. The sample of each survey is selected using a stratified two-stage cluster design. In the first stage of selection, the primary sampling units are selected from the list of census enumeration areas with probability proportional to the size within each stratum. In the second stage, a fixed number of households are selected by equal probability systematic sampling in each PSU. Trained staff visit the selected household and collect information on the characteristics of the household as well as women and men of reproductive age in the household. Women also answer questions about their children under the age of five, providing important information on childhood illness, mortality, vaccination, and nutrition. As a subset of the DHS survey, MIS collects data on ownership of mosquito nets and the prevention and treatment of malaria during high transmission seasons. --- Study population and sample size To analyze the influence of maternal and family factors, we limited the sample to under-five children who lived with their mother and were usual residents of the family. We gathered data on 38,567 under-five children who reported being febrile in the last two weeks before the surveys. Among them, 22,180 children sought care. The unweighted samples for all treatment cascades are presented in Additional file 1: Fig. S2. --- Measurement of variables --- Outcome measures In malaria-endemic areas, malaria should be suspected in any patient presenting with a history of fever or temperature ≥ 37.5 ℃ with no other obvious cause, and suspected malaria should be confirmed with a parasitological test [4]. Accordingly, the first outcome indicated whether or not a febrile child was taken to a health facility which was defined as a public or private health facility except for a pharmacy. The second outcome indicated whether or not the child had blood tested for malaria parasites in any health facility. The first outcome was analyzed in all febrile children, while the second outcome was analyzed in children who sought care for the reported fever. --- Explanatory variables Based on the literature review and Anderson's model of health service utilization, we included explanatory variables which might affect care-seeking behavior and the receipt of blood tests [25]. The following variables were included as they were available in both DHS and MIS: child's sex, child's age, mother's age, mother and her partner's highest education level, number of children in household, household wealth index, type of residence , type of health facilities to which the child was taken. Among the explanatory variables mentioned above, the highest level of parental education and household wealth index were used to capture the child's SES. Education level was classified into four categories: having no education, primary education , secondary education , and higher education. Notably, as a composite measure of a household's cumulative living standard, the wealth index was generated with principal components analysis using easy-to-collect data on a household's ownership of selected assets, such as televisions and bicycles, materials used for housing construction, and types of water access and sanitation facilities. . We categorized the types of health facilities into public hospitals, public primary health care facilities, private hospitals, and private PHC facilities. Public hospitals indicated government hospitals. Public PHC facilities included government health centers, government health posts, mobile clinics, and community health workers that belonged to the public sector. Private hospitals indicated non-governmental hospitals. Private PHC facilities included private doctors, mobile clinics, and community health workers that belonged to the private sector. --- Statistical analysis All statistical analyses were conducted among complete cases, using Stata version 15.1 . Considering the small amount of missing data, bias due to missing data is unlikely to affect the results substantially . All the descriptive analyses and regression below were weighted by survey weights provided by the DHS and MIS to account for the complex survey design. First, we examined the percentage of febrile children who sought care and the percentage of care seekers who received blood tests. We assessed the variation in these proportions at the national and sub-national levels. Second, we investigated the association between SES and care-seeking behavior as well as the receipt of blood tests. For care-seeking behavior, multivariable logistic regression was used, controlling for child's sex, child's age, mother's age, mother and her partner's highest education level, number of children in household, household wealth index, type of residence, and transmission season . For the receipt of blood tests, the type of health facilities was further adjusted. Fewer than 3% of children who were taken to two or more facilities were excluded here, because DHS and MIS didn't indicate in which facility these children received/didn't receive blood tests. In all regression models, country fixed effects and year fixed effects were controlled for in all models to account for variations in malaria incidence across countries and years. We did not include the interaction term of education level and wealth index, which was not statistically significant . Notably, variables only available in DHS or MIS were not included, such as the mother's marriage, employment, and accessibility of health facilities. To assess the potential impact of omitting these variables, we controlled for these variables with only DHS data in the sensitivity analyses. Furthermore, stratified analysis was conducted according to the type of health facilities for the receipt of blood tests, to further examine whether the associations between SES and receipt of blood tests differ across facility types. 0.61% of children were taken to facilities that didn't belong to the public sector or private sector . These records were excluded from the stratified analysis. Post-estimation simulations were conducted to generate the predicted probabilities of receiving blood tests at each type of facilities with other covariates at means as well as their confidence intervals. --- Results --- Characteristics of febrile children under five years and their households We included 38,567 febrile children under five years from 19 malaria-endemic sub-Saharan African countries . Of these children, 76% lived in rural areas, and 26% belonged to households in the lowest wealth quintile. There were 33% of children who had no parents with formal education . --- National and sub-national variations in care-seeking behaviors and receipt of blood tests among febrile children under five years On average, 58% of the febrile children under five years were taken to health facilities. Among those who were taken to health facilities, 71% were taken to the public sector. As shown in Fig. 1, there were wide variations in case-seeking behaviors across countries. The highest percentages of children who were taken to a health facility were in Zambia , Burkina Faso , and Nigeria . The lowest percentages of children who were taken to a health facility were in Mali , Cameroon , and Benin . Variations at the national level were also observed in the receipt of blood tests. Among the febrile children under five years who were taken to health facilities, 55% were reported to have taken a blood test. Sixty-four percent of the children in public facilities have taken a blood test, compared with only 32% in private facilities. Burundi had the highest percentage for blood tests at 88%, while Nigeria had the lowest record at 18%. There were also extensive variations at the sub-national level. Tanzania showed the largest regional difference in care-seeking behaviors and the receipt of blood tests . --- SES and the access to blood tests for malaria As presented in Table 2, children from poorer households were less likely to be taken to health facilities. The odds of seeking care among febrile children from households in the second and lowest wealth quintile were only 0.850 and 0.775 times of the children from households in the highest quintile . In the meantime, febrile children whose parents had more education had a better chance of seeking care and receiving blood tests . Compared to parents with no formal education, the odds of seeking care for their febrile child were 1.182 times higher among parents with primary education, 1.413 times higher among parents with secondary education, and 1.830 times higher among parents with higher education. Parents' education level also showed a positive association with the likelihood of having blood tests. Compared to parents with no formal education, the odds of having blood tests were 1.142 times higher among parents with secondary education, and 1.695 times higher among parents with higher education . For the results of sensitivity analyses, please see Additional file 1: Table S5. --- The modifying effect of the type of health facilities As presented in Table 3, the type of health facilities modified the associations between socioeconomic status and receipt of blood tests. Among febrile children who visited public PHC facilities, compared to children from households in the highest wealth quintile, the odds of receiving blood tests were 1.294 times higher , 1.423 times higher , and 1.413 times higher among children from households in the middle, second, and lowest quintiles, respectively. However, among febrile children who visited private hospitals, the odds of receiving blood tests among febrile children from households in the fourth, middle, second, and lowest wealth quintile were only 0.626 , 0.388 , 0.490 , and 0.437 times of the children from households in the highest quintile . Meanwhile, higher levels of parents' education indicated a better chance of receiving blood tests in public PHC facilities . Children whose parents with primary, secondary, and higher education presented ORs of 1.183 , 1.238 , and 1.814 to receive blood tests, respectively . As presented in Fig. 2a, with other covariates at means, the predicted probabilities of receiving blood tests across facility types were highest at public PHC facilities for parents with at least some education, ranging from 0.671 As shown in Fig. 2b, with other covariates at means, the probabilities of receiving blood tests across facility types stayed high at both public PHC facilities and public hospitals across wealth quintiles, while these probabilities remained low at private PHC facilities, ranging from 0.104 to 0.125 by wealth quintile. Notably, the probabilities of receiving blood tests decreased from 0.629 for children from the highest wealth quintile to 0.397 for children from the middle wealth quintile, and these probabilities remained below 0.5 for the poorest two quintiles . --- Discussion Using recent DHS and MIS data, this study documented disparities in the access to health facilities and the laboratory diagnosis of malaria among children. Across the 19 sub-Sahara African countries included in this study, only 58% of the febrile children arrived at health facilities. Among those who ever visited a health facility, only 55% took a blood test. Significant disparities in the access to the laboratory diagnosis of malaria not only existed across countries and within a country but also manifested themselves on the SES ladder. Children of lower SES were disproportionately affected by the lack of access to quality malaria care. However, the influence of SES on the receipt of blood tests varied by where children sought care. Children from poorer households were more likely to receive blood tests in public PHC facilities, while less likely to receive blood tests in private hospitals. Consistent with previous studies [8,10,26,27], we found that febrile children with lower SES were less likely to be taken to a health facility. Such a relationship can be explained by Anderson's model as part of predisposing and enabling factors [25]. SES could influence healthcare utilization through various pathways. For instance, parents with better education might be more knowledgeable about malaria and seek care more promptly after symptom onset [11]. Wealthier families could have more financial resources for healthcare and other costs related to care-seeking [10]. Furthermore, the impact of SES even persisted even after febrile children had arrived at health facilities, while there were only limited previous studies that drew mixed conclusions over this association. This study found that febrile children whose parents had higher levels of education were more likely to receive blood tests in general. This is possibly because bettereducated parents had more knowledge about malaria and were more likely to report children's symptoms accurately. It is also possible that parents with higher levels of education had more financial resources to pay for blood tests. But in a previous study using SPA data to assess the quality of malaria case management for children under five years, having a caregiver with primary or some secondary education was not significantly associated with the receipt of blood tests and recommended medications for malaria, after controlling for other patient-level, provider-level, and facility-level correlates [20]. In recent years , children with a caregiver having some secondary education were even significantly less likely to receive blood tests and recommended medications for malaria [20]. In another study in Mali, caregiver education was not found to be associated with incorrect case management of uncomplicated malaria among children under five years at public and private sector facilities [28]. The discrepancies in the associations between SES and quality of malaria case management might be explained by differences in research methods . In other fields of healthcare, the influence of patients' SES on their clinical management has been documented around the world [29][30][31]. Low SES usually incur personal and financial strains for patients who do seek care [32]. However, the finding that the facility type modified how SES affected the quality of malaria case management suggested the potential protective role of the public sector in alleviating wealth disparities in malaria care. In public PHC facilities, febrile children from poorer families were more likely to receive blood tests. This might be a result of higher malaria prevalence among children from households with lower SES [33]. If this assumption held, among febrile children present at private hospitals, those who were from poorer households should also need blood tests more than their wealthier counterparts. In contrast, in private hospitals, children from wealthier families were more likely to receive blood tests. Thus, the wealth disparity Fig. 2 Predicted probability of receiving blood tests by SES across facility types. A Marginal mean probabilities and corresponding 95% CI of receiving blood tests by parental education level across facility types. B Marginal mean probabilities and corresponding 95% CI of receiving blood tests by wealth quintile across facility types. All estimates were calculated using survey weights. CI confidence interval could probably be explained by the differences in the ability to pay for blood tests. In public health facilities, blood tests might be more affordable to patients than those in the private sector, due to government subsidies. In other words, health facilities in the public sector might be more responsive to the health condition per se, while those in the private sector might be more sensitive to the ability to pay. This study has the following strengths. First, this study added to the understanding of how SES influenced the coverage of malaria diagnosis. Access to the laboratory diagnosis of malaria among children required prompt care-seeking behavior on the parent side as well as the appropriate response on the provider side. Previous literature primarily focused on how SES affected parental behaviors, but few examined the role of SES once patients had arrived at health facilities. Second, this study used recent nationally representative data covering 19 sub-Saharan African countries. Therefore, the findings of this study could have important implications for this region. This study has the following limitations. First, the associations reported in this study might not be interpreted as causal, as there might be uncontrolled confounders that were not available in both DHS and MIS data. For instance, variables on mother's marriage, employment, and accessibility of health facilities were not available in MIS. In the sensitivity analysis, we controlled for these variables with only DHS data and generated similar results . Second, reverse causality could not be ruled out with the cross-sectional study design. Previous literature has discussed the potential reverse causality between the risk of malaria infection and SES [33]. In this study, there might be potential reverse causality between the lack of access to malaria care and households' wealth index. Third, the information on malaria case management was self-reported by mothers, so there might be recall bias that can cause under or over-estimation of the response. Also, if a child was brought to two or more facilities, we cannot identify the facility which provided worse care. Fortunately, only 3% of children were taken to two or more facilities. --- Conclusions In sub-Saharan African countries, febrile children with lower SES were less likely to have access to care, and even when they were brought to health facilities, they were less likely to receive a laboratory diagnosis of malaria. Health facilities in the public sector might alleviate the socioeconomic disparities in the access to malaria diagnosis to some extent. To narrow the socioeconomic disparities in the access to malaria diagnosis, universal health coverage needs to be further strengthened to make formal healthcare in general and the laboratory diagnosis of malaria more accessible and affordable. Furthermore, since a significant minority of febrile children sought care from the private sector, the affordability and quality of malaria care in the private sector warrant further improvement. --- --- Abbreviations --- SES --- --- --- Competing interests The authors declare that they have no competing interests.

Background Prompt and appropriate clinical management of malaria is critical for reducing the continued high burden of malaria among children under five years in sub-Saharan countries. However, more remains to be known about how a patient's socioeconomic status (SES) would affect the access to diagnosis of malaria.In this cross-sectional study using the Demographic and Health Survey and Malaria Indicators Survey, we pooled the data of 38,567 febrile under-five children in 2016-2018 from 19 sub-Saharan countries. Multivariable logistic regression was used to assess the associations between SES and two binary outcomes: the visit to a health facility and a blood test for fever. Stratified analyses were further conducted by the type of health facilities (public hospitals/ public primary healthcare facilities/private hospitals/private primary healthcare facilities) for the latter outcome.Fifty-eight percent of the febrile children were taken to health facilities, among whom only 55% took blood tests. Compared to children from households in the highest wealth quintile, children in the lowest quintile were less likely to be taken to medical facilities [adjusted odds ratio (aOR) = 0.775, 95% confidence interval (CI): 0.675-0.889]. Parents with more than secondary education were more likely to seek care (aOR = 1.830, 95% CI: 1.561-2.145) and to have blood tests (aOR = 1.729, 95% CI: 1.436-2.082) for their febrile children than parents without formal education. The probabilities of receiving blood tests at public hospitals and public primary healthcare facilities stayed relatively high across parental education levels and wealth quintiles, while these probabilities remained the lowest at private primary healthcare facilities, ranging from 0.100 (95% CI: 0.074-0.127) to 0.139 (95% CI: 0.083-0.194) across parental education levels and from 0.104 (95% CI: 0.078-0.130) to 0.125 (95% CI: 0.090-0.160) across wealth quintiles. Conclusions Significant socioeconomic disparities existed both in the access to health facilities and laboratory diagnosis of malaria in children in sub-Saharan African countries. These disparities were particularly evident in the private sector. Universal health coverage needs to be further strengthened to make formal healthcare in general and the laboratory diagnosis of malaria more accessible and affordable.

Introduction In recent years, men who have sex with men in China have emerged as a group at high risk of contracting and subsequently spreading HIV [1][2][3][4][5][6]. Homosexual transmission contributed to 11.0% of the newly infected in 2007; this rate increased to 30% in 2008 [7]. The proportion of reported HIV cases among MSM increased sevenfold from 0.4% in 2005 to 3.3% in 2007 [8]. Our recent cohort study showed that HIV rapidly spread among MSM in Beijing, with an incidence rate of 2.6/100 person-years [5]. The confluence of high HIV prevalence and incidence, along with common high-risk behaviors, suggests the HIV epidemic among MSM in China has emerged as a significant public health challenge [2,[4][5][6]. In the past decade, the use of the Internet in China has grown rapidly [9]. Evidence has shown that the Internet is becoming a burgeoning sex-seeking environment [10][11][12][13]. Moreover, MSM are more likely to use the Internet to find sexual partners than heterosexuals [14]. A London study found that about 40% to 50% of MSM used the Internet to Qingchun Li, Yingjie Liu, Zhenhai Zhou contributed to this paper equally. seek sex [15]. Another report found that 81% of the participants had access to the Internet [16]. The use of the Internet among MSM to find sexual partners is increasing, including MSM who are HIV-positive [17]; young MSM who seek partners online also engage in other behaviors that place them at risk of acquisition and transmission of HIV and other sexually transmitted diseases [18,19]. Several studies have shown that MSM participants recruited online report higher rates of high-risk behaviors than those recruited offline [20,21], especially among HIV-positive men [22]. MSM who engage in high-risk sex were more likely to use the Internet for meeting sexual partners [23]. By the end of 2009, the number of Internet users had reached 384 million in China, the most of any country. Of Internet users, 54% were male [9]. This includes approximately 4-10 million MSM Internet users [24]. Currently, almost all cities have local gay-oriented websites that facilitate the communication and sex partner-seeking among MSM [25]. However, little is known about the use of the Internet among MSM in China. The objective of this study was to assess the correlates for sex-seeking via the Internet among MSM in Beijing. --- Methods --- Study Design and Study Participants Study subjects included all participants enrolled in the baseline of two cohort studies conducted among MSM in Beijing from November 2006 to February 2007 and from March 2008 to June 2008, respectively. Participants in both studies were recruited using three similar recruitment methods: study staff contacted potential participants and made appointments with willing participants who provided contact information through website advertisements by a non-governmental AIDS volunteer group ; peer recruiters were hired and trained to distribute flyers with study-related information at MSM-frequented venues ; and participants were encouraged to refer their peers to attend the study, as described elsewhere [2]. All potential participants came to a district HIV voluntary counseling and testing clinic in downtown Beijing. The location of the clinic was easy to access via various transportation routes. Confidentiality and privacy were ensured in the study. Eligibility criteria for these two cohorts included C18 years, male, had anal or oral sex with another man in the past 3 months, provided written informed consent and willingness to complete the study. Those who met the screening criteria were assigned a unique and confidential identification code and then completed a questionnaire-based interview. At the end of the interview, client-centered risk reduction counseling was provided on HIV, hepatitis B virus, hepatitis C virus, syphilis, and other STDs. Blood samples were tested for HIV and syphilis. Participants were instructed to return in one month for their results; a pre-assigned unique code was used to inform participants of their results. The reimbursement amount for participants was 50 Yuan RMB , 12 free condoms and one free lubricant. The study was approved by the Institutional Review Boards of the National Center for AIDS/STD Control and Prevention of the China Center for Disease Control and Prevention. --- Measures In both studies, identical questionnaire-based interviews were conducted on a one-on-one basis in a separate and private room of the district clinic. Trained health professionals conducted the interviews. Data collected included sociodemographic information , behavioral information and psychological information . --- Laboratory Tests HIV status was determined by an enzyme immunoassay screening and an HIV-1/2 Western Blot confirmation . Syphilis infection was determined using an EIA (Beijing Wantai Biological Production Company, Beijing, China --- Statistical Analysis Questionnaire-based data and biological testing results were recorded, double-checked and compared with EpiData software . After corrections, data were then converted and analyzed using Statistical Analysis System . Univariate analyses were conducted using Chi-square tests and Fisher's exact tests. Variables with P-values \ 0.20 were entered into a multivariable logistic regression model in order to identify factors independently associated with seeking male sex partners on the Internet. Only factors with P-values \ 0.05 were deemed statistically significant. --- Results --- Baseline Characteristics of the Participants A total of 901 eligible MSM were included in the data analyses. Of the 901 participants, 5.3% were infected with HIV, 22.3% with syphilis, 6.0% with HBsAg, and 0.7% with HCV. Of the participants, 38.6% were recruited via the Internet and 43.8% via peer referral; the median age was 26 years with 54.8% being B26 years; 92.1% belonged to the Han ethnic group; and 65.2% completed college or higher levels of education. Beijing residents accounted for 30.7%; the median monthly income was US$ 294; 18.9% were rural residents; 68.1% were single and 11.4% were married or cohabiting with female sex partners . --- Sexual Behaviors Participants who identified as exclusively homosexual and predominantly homosexual accounted for 53.3 and 41.2%, respectively, whereas 5.6% of the participants identified as predominantly heterosexual with only incidental homosexual activities. The median number of male partners in a lifetime was 10. In the past 3 months, 62.2% had C2 male sexual partners; 3.4% had sex with foreign male partners; 7.7% participated in male group sex; 7.6% had commercial sex with men; 4.9% were ''money boys'' ; 13.0% had sex with female partners. In the past month, 18.3 and 14.2% of the participants had UAI with regular male partners and casual sex partners, respectively; 6.2% had unprotected sex with female partners. Of the participants, 5.7% reported feeling discriminated against in the past month; 54.1% experienced loneliness in the past month; 44.5% experienced depression in the past month; and 42.3% have ever experienced sadness. Of the participants, 38.7% drank alcohol in the past 3 months and only 0.8% used illicit drugs , 10% attended an AIDS program in the past 3 months, 51.8% had ever received an HIV test and 26.9% had ever received an STD diagnosis by a clinic physician . --- Factors Associated with Seeking Sex Partners on the Internet Of the 901 participants, 73.0% had sought male sex partners through the Internet in the past 3 months. The proportion of online sex-seeking was 68.9, 81.3 and 65.2% among study participants recruited through peer referral, the Internet and MSM-frequented venues, respectively. In the univariate analysis, factors significantly associated with seeking male sex partners via the Internet in the past 3 months were younger age, higher levels of education, rural residency, married or cohabiting with female sex partners in the past 3 months, exclusively homosexual sexual orientation, more male sex partners in the past 3 months, drinking alcohol C1 times per week in the past 3 months and having an STD diagnosed by a clinic physician in the past . In multivariable logistic regression analysis, after controlling for five variables , seeking male sex partners on the Internet in the past 3 months was independently associated with younger age , higher levels of education and having more sex partners . --- Discussion Our study found an extremely high prevalence of online sex-seeking in the past 3 months among MSM in Beijing; this rate is substantially higher than in a previous meta-analysis [13]. The Internet has been playing an increasingly important role in people's daily lives, particularly as an efficient mode of communication in many developed metropolitan cities in China. Economic development and improved literacy rates have allowed a growing number of people to access the Internet. In China, MSM are highly stigmatized and they face social pressures from those who adhere to the traditional culture. Hence, the Internet is especially popular with MSM because of its accessibility, affordability and anonymity [26]. When seeking sex partners on the Internet, MSM can be open about their sexual orientation and yet remain anonymous. Besides the Internet, MSM seek sex partners mainly through bars, parks, bathhouses, and public washrooms [27,28]. However, these venues are less effective compared to online sex seeking, because of the relatively fewer opportunities to satisfy unique preferences in sex partners. Moreover, seeking sex partners offline has disadvantages due to transportation costs, time, potential disclosure of their sexual orientation, and privacy concerns. On the contrary, MSM can seek sex partners via the Internet more efficiently as they can search for other MSM who satisfy specific personal preferences. In addition, seeking partners online allows men the opportunity to get to know someone better [26] before engaging in sex. All of these advantages may lead to more MSM seeking male sex partners online. That the rate of online sex-seeking among MSM in Beijing found in this study is higher than any other report [12,13,15] might be a result of the high density of residents, rising individual incomes, technological development, and increasing availability, accessibility, and knowledge of computers in the past decade in Beijing. More tolerant views and broader sexual freedoms due to the dramatic socioeconomic transition in China, particularly in modern metropolitan areas such as Beijing, could be another contributor for extremely common online sex seeking. The growing popularity of the Internet suggests increasingly prevalent online sex-seeking among MSM in the near future. This indicates that the Internet could provide opportunities for recruiting study participants [29,30] into either qualitative or quantitative studies [26], as well This study found that younger age and higher levels of education were independently associated with online sex seeking among MSM in Beijing, which is consistent with other reports [12,31]. Young and educated MSM are more likely to adopt innovative technologies and learn the skills necessary to access the Internet [12,31]. Furthermore, most well-educated MSM with higher social status might be more afraid of disclosure of their sexual orientation, so the Internet becomes their preferred way to seek sex partners [26,32]. Results from studies in developed countries differed [20,33], in that community-based sampling recruited more educated participants than those recruited via the Internet. This might result from differences in Internet user populations and literacy patterns in China compared with developed countries. In China, the overall proportion of the population who use the Internet is relatively low; therefore, education could be a critical factor to accessing the Internet [34]. This study found that participants who had more male sex partners were more likely to seek sex partners on the Internet. Other studies have found that online sex-seeking is associated with high-risk sexual behaviors and increased rates of STDs [14,20,21]. Given the high prevalence rates of HIV and syphilis among MSM in Beijing, future studies need to achieve a better understanding of the different characteristics of venue versus non-venue communitybased HIV and syphilis disease distributions. This study provides useful preliminary data in guiding interventions and furthers research on Internet use among this group. Previous studies have found that social isolation, loneliness, desperation, and sexual orientation were the main reasons for men's use of the Internet to facilitate sexual encounters [35]. Loneliness in particular was the primary motivator for MSM to meet sex partners in online chat rooms [36]. MSM in China are traditionally marginalized, causing them to suffer long-term sexual depression and hide their sexual orientations, which in turn leads to even higher degrees of depression [37]. Although a statistically significant association between experiencing loneliness, depression or sadness and online sex seeking was not shown in this study, these psychological syndromes were common among participants. These findings underscore that intervention efforts should tackle prevalent psychological syndromes among this group; the Internet could be one of the venues to deliver such intervention measures. Some studies have found associations between meeting sex partners online and having unprotected sex, but others have noted that when studying specific online versus offline encounters there is no difference [23]. This study suggested that online sex seeking is associated with HIV/STD risk behaviors. Further research with longitudinal cohort designs is needed to better understand the lack of differences between men who seek partners online versus those who do not. We realize the limitations of this study. First, recall bias may exist due to the self-reported questionnaire. Second, we used dichotomous yes/no questions to measure discrimination, loneliness, depression, and sadness. The pilot survey showed Chinese MSM liked this dichotomous format, but further validation of the survey's psychological questions is needed. Third, only one-third of participants were Beijing residents; the interpretations of this study may not accurately pertain to MSM in Beijing. However, Beijing, like many other cities in China, has a significant proportion of people living and working in the city for months or years without a permanent resident identification ''hukou.'' These residents were classified as ''non-Beijing residents'' in this analysis. Participants were required to have physically lived in Beijing for 12 months or more, per the study inclusion criteria. Lastly, the nature of this crosssectional study design precludes causal association. In spite of these limitations, this study using multiple recruiting methods with sociodemographic mapping strategies provides a first examination of Internet sex seeking behaviors in Beijing. The Internet reduces physical, geographic, social, psychological, and financial barriers and enables people to communicate efficiently on a massive global scale. This study found that online sex seeking was extremely common among MSM in Beijing and that online sex seeking was independently associated with more sexual partners, higher education, and younger age. These findings indicate --- Conflicts of interest None.

To investigate factors associated with online sex-seeking behaviors among men who have sex with men (MSM) in Beijing, China. MSM participants were recruited from two cohort studies with multiple enrollment methods from November 2006 to February 2007 and from March to June 2008, respectively. Data collected included demographics and sexual behaviors. Of the 901 participants, 68.1% were single; 69.3% were non-Beijing residents; 94.4% considered themselves to be homosexual; 65.2% received college or higher levels of education; the median age was 26 years; 73.0% sought male sex partners via the Internet in the past 3 months; 66.2% had C2 sex partners. Younger age, higher levels of education and having had C2 male sex partners in the past 3 months were independently associated with seeking sex partners on the Internet. These findings indicate that Internet-based intervention programs could encourage younger high-risk MSM to use condoms and reduce their numbers of sexual partners.

Background According to the National AIDS Program , there were an estimated 2570 people living with HIV in Lebanon in 2019, and the HIV epidemic was --- Open Access *Correspondence: [email protected] 1 Dermatology Department, Clemenceau Medical Center Affiliated With Johns Hopkins, Beirut, Lebanon Full list of author information is available at the end of the article concentrated in men who have sex with men [1]. Sexual transmission risks among MSM are significantly shaped by inequitable social and structural contexts that influence individuals' sexual practices and access to HIV prevention. Individual-level acquisition risks include lack of condom use , multiple sex partners per month , sex work and lack of HIV testing [1]. Social contexts include habits and norms regarding sexual orientation and/or HIV-positive serostatus, including among families, friends, and cultural, religious and health institutions, which lead to the exclusion of MSM in HIV prevention, education and employment, and in healthier long-term relationships [2]. This has been supported by previous papers on Lebanese MSM, where men in relationships were much more likely to have been tested for HIV [3]. Structural contexts such as the criminalization of homosexuality and the healthcare system also shape the HIV response. Additionally, Lebanon has a community of migrant and refugee MSM who are considered even more vulnerable because they may lack or avoid access to the healthcare system due to stigma and discrimination by providers [4]. Individuallevel acquisition risks in migrant MSM include low rates of condom use , multiple sex partners per month , sex work and low HIV testing rates [1]. In local and foreign MSM communities, HIV prevalence was 12% and 3%, respectively, and knowledge of key HIV-related information was 8.7% and 6.3%, respectively [1]. For more than 10 years, the NAP has adopted free voluntary counselling and testing for HIV. Rapid diagnostic tools are procured by the NAP and provided to MSM, refugees and those who need them, and distributed by nongovernmental organizations and primary healthcare centers . The NGOs that provide free HIV testing services have a long-term partnership with the NAP. Healthcare workers who work at these NGOs are regularly trained by the NAP on risk assessment, counselling, diagnostic tools, HIV knowledge, stigma and discrimination, associated harm, and NAP reporting tools. Following these trainings, HCWs deliver VCT services in NGOs and PHCs, confidentially and anonymously, both on-site and through outreach activities. Moreover, NGOs distribute condoms and information, education and communication material for free, which is all procured by the NAP. The NGOs send a monthly report to the NAP with the number of tests conducted through the VCT services and the number of positive and negative cases. People who test positive are provided a list of physicians who confirm, assess and prescribe antiretroviral therapy which is provided for free by the NAP ART dispensing center. According to the NAP, VCT is the most frequently reported diagnostic screening tool, compared to more standard laboratory testing, among MSM because it offers free and anonymous counselling and HIV testing, which provides a sense of security, less stigma and more friendly environments towards their sexual orientation. In 2014, the Joint United Nations Programme on HIV/ AIDS released 90-90-90: An ambitious treatment target to help end the AIDS epidemic, setting three global targets to be achieved by 2020: 90% of all people living with HIV would know their HIV status; 90% of people with diagnosed HIV infection would receive sustained ART; and 90% of all people receiving ART would have viral suppression [5]. Additionally, to scale up HIV testing, WHO published the first global guidelines on HIV self-testing in 2016 [6]. HIVST is a process whereby a person who wants to know their HIV status collects a saliva/blood specimen, performs a test and interprets the result in private. HIVST is a screening test; it does not provide a diagnosis, and confirmatory testing is required if the initial result is positive [6]. HIV antibodies can be detected by HIVST following a "window period" of 1 to 3 months after exposure. HIVST was then endorsed by the WHO Eastern Mediterranean Region [7,8]. In line with these recommendations, the NAP adopted HIVST in the second half of 2018 to fill the gap between the first and the second 90s of the UNAIDS 90-90-90 target. This strategy aimed to reach the largest possible number of PLHIV who remained undiagnosed. HIVST is an example of a self-care intervention. In 2019, WHO published its first guideline on self-care interventions for health, with a focus on sexual and reproductive health and rights. WHO defines self-care as the ability of individuals, families and communities to promote health, prevent disease, maintain health, and to cope with illness and disability with or without the support of a health worker [9]. The concept of self-care recognizes individuals as active agents in managing their own health, including disease prevention and self-medication, and providing care to dependent persons. Self-care interventions for health, such as drugs, devices, diagnostics and/or digital products, can be used with or without the direct supervision of healthcare personnel. Additionally, quality self-care interventions should be accessible and available to everyone, and especially vulnerable populations [10]. Since the beginning of the coronavirus disease 2019 pandemic, HIV testing has become a significant challenge, and increased efforts were needed to augment access to and facilitate testing [11]. In Lebanon, due to COVID-19 restrictions , HIVST became the only diagnostic tool that could be easily used by people who wished to get tested. This is particularly true for Lebanon since the beginning of COVID-19 . During the lockdown, all NGOs were required to stop all on-site services, including VCT, and HIVST became the only available option for beneficiaries needing an HIV test. This paper has two objectives: to describe the effectiveness of implementing HIVST in Lebanon and to discuss how the success of HIVST implementation has been impacted specifically through a comparison of services pre-COVID-19 and during COVID-19 . --- Methods In mid-2018, the NAP conducted a series of workshops to introduce HIVST services, which included presenting the scientific evidence of the test, providing IEC material and HIVST hands-on support to different HCWs working in NGOs. A total of 280 HCWs and volunteers from different NGOs were reached by the workshops. The workshops' key messages about HIVST were that tests are distributed for free to the beneficiaries; tests provide more privacy since beneficiaries do not have to be physically in front of a counsellor during the test; tests are voluntary; test results are not shared directly with the NGOs, but beneficiaries should be encouraged to notify the NGOs about their test results, without any breach of confidentiality; and tests should target frequent testers and people who avoid classic VCT services because of fear or stigma. Following the workshops, several educational and informative resources were produced and shared with beneficiaries. First, an illustrated brochure was created in both Arabic and English and added to each test kit. The brochures explained the steps in performing the HIVST and how to interpret the result. In the leaflets, it was mentioned that any potential breach of confidentiality or forceful aspect or unwanted event should be reported to the NAP, similarly to the reporting path implemented since the beginning of VCT activities by NGOs. To mobilize MSM and their partners, HIVST was promoted by the NAP through all its educational sessions, awareness campaigns and events such as World AIDS Day. On each occasion, the NAP included a short presentation about the importance of HIVST, its confidentiality, the instructions and the steps to perform it, and where and how to access it for free through a list of NGOs with their contacts and addresses. Moreover, the NAP made HIVST available for free at the NAP ART dispensing centre for PLHIV who wished to test their partner. Similarly, the NGOs used the same material to promote the HIVST through their on-site HCWs, social networking platforms and online dating applications . Additionally, a hotline was set up at each NGO for beneficiaries to call about HIV counselling and guidance on testing. While distributing the HIVSTs, NGOs anonymously collected data about each beneficiary's age, consistency in condom use and HIV testing history. Additionally, each beneficiary was asked to call back and give the results of the HIVST to the NGO. The reporting tool was the same across all NGOs before and during COVID-19. HIVST data were reported to the NAP along with the standard reporting data on VCT services mentioned previously. Thus, HIV data reported to the NAP during the lockdown only included HIVST data, with no VCT data, because there were no VCT activities. At the NAP, the data were combined, aggregated and analysed. --- Results By the end of 2018, the NAP had procured and provided 1380 HIVST kits to the NGOs, and officially began collecting HIVST data in 2019. Out of the total 1380 HIVST kits that NGOs received from the NAP, 1103 were distributed to MSM in 2019. The data collected at HIVST distribution showed that users/beneficiaries were on average 26.3 years old , 73.1% used condoms inconsistently and 26% had never had an HIV test. Follow-up data were more challenging to collect. The data collected through follow-ups were impacted by a lack of beneficiary compliance to calling back and sharing their test results. Moreover, many beneficiaries distributed these tests to their peers or sex partners without any further contact with them, which impacted the data collected both at HIVST distribution and at the followup. Another challenge was that NGOs lacked the human and financial resources to actively follow up with their beneficiaries and thus relied on passive reporting of HIV results from the beneficiaries. This being said, 111 out of the 1103 beneficiaries who received an HIVST called back and notified the NGO about the HIV results. Among these, 2/111 users were reported HIV-positive. These cases were reported to the NAP and referred for confirmation and for treatment. They were enrolled into care at the NAP according to data from the NAP ART dispensing centre. From January through May 2020, a total of 780 HIVST were provided to the NGOs, and 625 were distributed to beneficiaries. This period was divided into the period from January to mid-March and from mid-March to May . Data collected during the distribution of HIVST from January to mid-March 2020 showed that beneficiaries were on average 29.1 years old, 71% of beneficiaries used condoms inconsistently and 21% of beneficiaries had never had an HIV test. However, compared to 2019, there was an improvement in the amount of follow-up data collected during COVID-19 because of the absence of VCT services in the NGOs during COVID-19, shifting more efforts towards HIVST. In the COVID-19 era, HIVST allowed communities to experience partner-to-partner testing according to what was reported to the NGOs by beneficiaries. In the pre-COVID-19 period, a total of 79 out of 242 beneficiaries who received HIVST provided their HIV results. Among the 79 beneficiaries that reported the results of their HIV test, only one case was reported positive. In the period during COVID-19, a total of 370 beneficiaries out of 383 who received HIVST kits reported their HIV result to the NGOs, none of which was positive. At the NAP and at NGOs, there have been no reports of unwanted incidents related to HIVST, which also includes the period during COVID-19. --- Discussion The adoption of HIVST as a self-care intervention in Lebanon has been shown to be acceptable and feasible. First, this self-care intervention was acceptable to local MSM who might face stigma in healthcare settings and by refugee MSM who lack access to healthcare. Self-care interventions such as HIVST are useful for all populations, and especially marginalized and vulnerable populations. Indeed, HIVST was found to decrease stigma about testing among MSM, as shown in a review of published data [12]. The fact that such high numbers of HIVST kits were distributed, and that beneficiaries reported their results , points towards a conclusion of general acceptance of the test by both native and refugee MSM. The distribution of HIVST for free had a positive impact, as otherwise the communities which were already socially or financially vulnerable would not have had any access to this diagnostic tool. The acceptability of HIVST in Lebanon and other countries in the region was explored in two qualitative studies conducted in 2016. MSM from Tunisia, Morocco and Lebanon were interviewed about their acceptance and willingness to use HIVST [7,8]. In both assessments, the participants acknowledged the potential for HIVST to contribute to a scale-up of testing but expressed concern about the accuracy of the test, selfharm and the absence of trained professional to interpret HIVST results. Access to HIVST via appropriate and safe delivery channels such as internet-based distributors, community centers, non-stigmatizing pharmacies, physician clinics, outreach workers or peers was acknowledged by participants as potential approaches for distribution. The direct assistance approach was preferred, which implies a major role of the HCWs in assisting the beneficiaries over the phone while they conduct their test. Our findings presented in the results highlight the major role of HCWs in HIVST implementation. Furthermore, with the lockdown having facilitated the follow-up on results, it is suggested that the HCW direct assistance approach is more productive: it allows the beneficiary to have a robust interpretation of his result and it limits lost or unreported cases. Thus, HIVST acceptability appears to be consistent in the previous WHO assessment [7,8] and in our findings in the pre-COVID-19 era and during COVID-19. Second, the implementation of HIVST was found to be feasible. Maatouk reviewed the published literature about the major steps to take before implementing HIVST including robust data collection, an update of testing guidelines, training of HCWs on guidelines, policies and regulations, and an update of HIV indicators to include HIVST-specific reporting tools [13]. These steps were adopted by the NAP to facilitate a more rapid scaleup of HIVST and updating of reporting and surveillance tools with HIVST. Moreover, the fear of safety regarding HIVST was obviated by the absence of any incident in 2 years. Our second objective was to discuss how HIVST implementation has been impacted by COVID-19. There was increased demand for HIVST during COVID-19 because VCT services were completely paralysed due to the general lockdown and the stay-at-home curfews. The lockdown, curfews and the unavailability of VCT services encouraged individuals who wished to get tested for HIV to use HIVST. Moreover, the highly contagious nature of COVID-19 led to a drop in demand for VCT because individuals were experiencing increased stress and fear from the pandemic and were afraid to go to health facilities and hospitals. Many NGOs and other PHCs diverted their resources to COVID-19 response. The increased use of HIVST also allowed communities to experience partner-to-partner testing, which can strengthen trust between partners. The responsibilities of HCWs at the NGOs shifted during COVID-19 towards promoting HIVST: HCWs were referring people more to HIVST and were encouraging people to use it. Consequently, people were using HIVST more, which enabled HCWs to be connected with beneficiaries and their HIV test results in different ways from before COVID-19. This shift in responsibilities allowed them to have more time for the follow-up with beneficiaries because they were unable to perform other duties such as VCT . This increase in follow-up with beneficiaries was possible because the majority of NGOs had to reorganize their work and responsibilities during COVID-19 to alter how essential services could still be delivered, including HIV testing. In fact, the NGOs suffer from a scarcity of human and financial resources with the necessary skills to implement, supervise and support client counselling . Having such resources would help having a focal point HCW in each NGO who would be responsible for HIVST distribution, promotion, follow-up and data surveillance. Thus, the reorganization of care and redistribution of HCW duties within each NGO that occurred during COVID-19 needs to be maintained as an adaptation to improve HIVST implementation. Despite the acceptability and feasibility of HIVST, however, there were several challenges faced during implementation. First, direct assistance with HIVST provided to beneficiaries by HCWs mitigates the fear of being alone during the stressful time of waiting for the test results. Second, some beneficiaries refused HIVST as a concept because of its long window compared to the period for the conventional VCT . These particular beneficiaries were advised to do the VCT instead. Third, some beneficiaries trusted the blood sample test over the saliva sample test used in HIVST. This challenge was solved by an explanation of the collection of antibodies through saliva to convince the beneficiaries of the solid scientific background of the test. --- Conclusion HIVST implementation in Lebanon serves as an example of a community-led effort to introduce a self-care intervention in the MSM community. Despite the impact of COVID-19 on individuals and health systems, the scaleup and dissemination of HIVST was successful. In order to continue providing the same, or increased, level of HIVST services, NGOs will need to shift HCWs' responsibilities and increase HIVST promotion. With continuous monitoring and evaluation of HIVST reported data, further implementation measures could lead to further expansion of HIVST services. Additionally, HIVST is not a solution to the stigma related to testing among MSM, which requires health equity and human rights laws in support of quality self-care interventions [10]. Continuing to target vulnerable communities, such as MSM, with HIVST will be important to reach those who are unable to access testing and other health services through traditional means. • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year --- • At BMC, research is always in progress. --- Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: --- --- Abbreviations --- --- --- --- Competing interests The authors declare that they have no competing interests. ---

Background: In Lebanon, HIV is concentrated in both native and refugee communities of men who have sex with men (MSM). For over 10 years, the National AIDS Program (NAP) has offered HIV voluntary counselling and testing through a partnership with nongovernmental organizations (NGOs). In 2018, implementation of HIV self-tests (HIVST) was introduced, and this self-care intervention has been further scaled up during the coronavirus disease 2019 (COVID-19) pandemic. This paper (1) describes the effectiveness of implementing HIVST in Lebanon, and (2) discusses how the success of HIVST implementation has been reflected during the COVID-19 pandemic.The NAP conducted a series of workshops (July-November 2018) to introduce HIVST services for healthcare workers working at different NGOs. The workshops highlighted that HIVST would be distributed for free, that it would be confidential and voluntary, and that participants were encouraged to notify the NGOs of their results, which would be kept strictly confidential. NGOs collected data anonymously and confidentially from beneficiaries (age, consistency of condom use and HIV testing history), who were asked to call back with the results of their HIVST. At the NAP, data were combined, aggregated and analysed.In 2019, the NGOs distributed 1103/1380 (79.9%) HIVST kits to their beneficiaries. The NGOs collected feedback on 111 kit results, of which two were HIV-positive. Feedback about HIVST results from beneficiaries was low (111/1103) due to noncompliance of beneficiaries and the lack of human and financial resources in the NGOs. From January through May 2020, a total of 625/780 HIVST kits (80.1%) were distributed. This period was divided into pre-COVID-19 and during COVID-19. The follow-up with the beneficiaries during COVID-19 was much improved because of the absence of on-site activities, shifting more efforts towards HIVST (449/625). There have been no reports of social harm related to HIVST.HIVST implementation in Lebanon serves as an example of introducing a self-care intervention as part of a community-led effort. In order to maintain HIVST services at the same improved level, reorganization of care is needed within each NGO following the adaptation process due to COVID-19, along with continuous monitoring and evaluation of HIVST reported data.

INTRODUCTION The large amount of data available today from Internet services and applications has allowed us to explore how entities relate to each other. As such, we can map entities and their links as social networks in order to provide new kinds of analysis, both structural and social [13], for instance, to characterize entities behaving like hubs [32] or acting as bridges by connecting different parts of a network [18]. Indeed, the so-called social network analysis has contributed to understand how highly connected complex networks work, ranging from graph theory to property rights [31,39,49]. In this regard, based on how entities play structural roles in networks, we contribute with a social-based perspective in order to better analyze the behavior and the strength of the ties involving such entities. We recall that in general social networks are constructed considering the existence of explicit relationships . In this static structural scenario, one approach is to explore the notion of social capital given by the position of the nodes in the social network structure [9,11,18]. For instance, Granovetter [18] defines the concept of weak ties as being those important relationships that make a network more cohesive by means of the creation of bridges between communities. As discussed by Aral [4], the most influential sociological theories explore bridging ties and cohesive ones , which provide more advantage when accessing information passing through a network. However, as explicit relationships evolve to other kinds of interactions , they become more complex, thus bringing more information about these social interactions. In this way, a more general approach is required to model these specificities by using only edge or node attributes [2,36]. By doing so, it is possible to promote more information about the social motivation involving each interaction, since individuals tend to change their attributes over time, whether in terms of location , relationships with other people or new skills acquired. This would enable us to understand the evolution of social structures, in which the persistence of attributes over time indicates the social value associated with each interaction. Additionally, several other works have investigated topological properties and patterns of social networks in order to define the behavior of their actors and measure the strength of their relationships [5,22,26,27,32]. Exploring the behavior and the dynamics of the actors in a social network is essential for a better understanding of its social structure, which is usually characterized by graphs that capture the social aspects involved [30,43,51]. Accordingly, Barabási [5] reinforces the importance of the network theory paradigm as fundamental for understanding the complexity that involves actors and their relationships. For example, Newman [32] measures the influence of the nodes in a network based on their proximity and the number of shortest paths among them. In a previous work [41,42], we analyzed how social aspects impact knowledge transfer in a network. Our proposed model allows one to represent the social dynamics of node-attribute relationships to capture the influence generated by knowledge transfer. Here, we go a step forward and focus on the strength of nodes and their dynamic relationships over time by means of social capital. More specifically, in this article, we propose to address together distinct aspects: actors, interactions, time intervals, attributes and social concepts. Our strategy to address theses issues consists in modeling nodes by associating them with their attributes in order to extract persistent features over time. Regarding its social perspective, our method is based on Burt's definition of social capital by considering two concepts: closure, the ability of aggregating actors with similar patterns, and brokerage, the ability of creating bridges with diversified patterns [10]. Blue edges emphasize the closure concept and red ones the brokerage concept . Black edges correspond to those regarded as innocuous . We have omitted multiple edges connecting the nodes for a better visualization. Best viewed in color. In a preliminary work [40], we presented two specific contributions to this kind of analysis: a node-attribute graph model that captures the social tie of individuals and their associated attributes, thus providing a dynamic attributed model that enables us to mine multiple interactions over time; and a new method to classify nodes and their relationships based on temporal node-attributes that considers the social role of the nodes and the social meaning of the edges. Here, we extend our previous work by first redefining the classes assigned to nodes and edges in order to reflect the social concepts behind the proposed model. For this reason, both nodes and edges are now thoroughly classified according to the concepts of closure and brokerage. We also contribute by applying our new social classification strategy to different types of social network, therefore providing a detailed discussion underlying social aspects based on additional experiments. In order to illustrate this, Figure 1 shows a Computer Science academic coauthorship network, including more than 79 thousand nodes and 263 thousand edges, built on data collected from DBLP . In this network, the edges classified according to the concept of closure are shown in blue, those classified according to the concept of brokerage are shown in red and those that express no social meaning are shown in black. Note that the edges based on the brokerage and closure concepts dominate the center of the graph, while the extremities tend to have a greater prominence of edges regarded as non-relevant. This means that edges strongly related to social concepts tend to be better positioned in a social structure , which provides early access to information passing through the network. Our approach allows to analyze networks in terms of their structural autonomy, which occurs when people are tightly connected to one another with extensive bridge ties beyond them, thus achieving high levels of innovation and productivity because there are both trust and cooperation between individuals who connect different parts of a network [10]. The rest of this article is organized as follows. Firstly, Section 2 reviews related work, whereas Section 3 introduces our proposed model and describes the methodology adopted for evaluating it. Then, Section 4 analyzes and discusses the results of applying our classification method to different social network scenarios, whereas Section 5 summarizes the results of our experimental validation. Finally, Section 6 presents our conclusions and some considerations for future work. --- RELATED WORK The study of the individuals' dynamics enables us to understand the evolution of social networks, thus providing models to characterize their behavior [6,23,30,39,43]. For instance, Barabási et al. [6] capture the social tie importance by observing the topology and the nodes' internal behavior in coauthorship networks, whereas Sun et al. [45] propose a model to analyze the social dynamics of science in terms of scientific disciplines. Yet, Iacobelli and Figueiredo [23] explore random walks on dynamic networks to better characterize and understand their structures. In another context, Silva et al. [43] characterize the moving properties and the behavioral profile of how researchers move around publication venues stratified in terms of their quality, whereas Brandão et al. [8] address how the social roles of researchers change over time. Also, traditional network metrics have been employed to identify the most important nodes within a graph, such as done by Newman [32] that uses centrality metrics based on shortest paths for determining the best positioned nodes in academic social networks. Several other studies analyze social networks based on particular social concepts such as tie strength [7,8,16,48], the friendship paradox [1,20] and social influence [25], as well as on structures such as triadic closure and social balance [17,22]. For instance, based on the information shared between nodes, Adamic and Adar [1] measure the strength of relationships by analyzing the similarity between messages exchanged between individuals. Yet, Gilbert and Karahalios [16] also consider temporal aspects, but modelling tie strength as a linear combination of different dimensions such as intensity by means of the number of words exchanged, emotional support based on positive words, social distance in terms of the political differences, among other aspects. Finally, Levchuk et al. [27] propose an approach to learn and detect network patterns such as repetitive groups of people involved in coordinated activities. Regarding social perspectives, several works have explored the notion of social capital given by the strategic positioning of a particular actor in a social structure [10,14,18]. Based on the premise that actors can make a network stronger by integrating different parts, Granovetter [18] defines the concept of weak ties as being those important relationships that make a network more cohesive by creating bridges between communities. Likewise, Burt [10] describes a structural hole as the gap formed by individuals who have complementary knowledge, and then defines as brokers those nodes that hold certain positional advantages due to their good location in the social structure. Considering such results, Feng et al. [14], for instance, used structural holes to identify the most central and bridging group of nodes in a network. In another context, Brandão et al. [8] investigate the strength of dynamic social relationships in academic social networks based on topological metrics, thus revealing that such relationships tend to have more weak and random ties than strong and bridges ones. Leão et al. [26] analyze the role of random interactions in the structure of communities, whereas Sanz-Cruzado and Castells [37] analyze the role played by strong ties and weak ties , thus showing that bridges work as enhancers of the structural diversity in social networks. In this article, we define social classes to better characterize social networks. In this regard, Yang et al. [50] propose a metric for expert identification in the StackOverflow Q&A site, which is based on the quality of its users' contribution. Specifically, they define two profiles: sparrows, as being highly active users on the network that contribute to the vast majority of its content, and owls, as being the most experienced users that provide useful answers. Furtado et al. [15] also characterize users' behavior in Q&A sites, but observing a dynamic perspective. They define ten profiles based on motivation and ability metrics, thus characterizing users as unskilled, expert, activist and hyperactivist. In another context, Vaz de Melo et al. [48] propose the RECAST algorithm that identifies random and social interactions based on network properties. Specifically, such algorithm explores topological and temporal aspects in order to measure the strength of the nodes' relations, which is derived from the neighborhood overlap and the persistence of the relationships. Doing so, it is able to classify the edges of a network by assigning them to one of the following social classes: friend, acquaintant, bridge and random. Our proposed method also relies on temporal aspects and on the regularity of the relationships over time, but differs from those discussed above by mainly exploring social concepts. In this regard, Alhazmi and Gokhale [3] measure the structural social capital of online social networks by defining closure and brokerage as closed and open triads, respectively. Alternatively, in a recent work [41], we classify nodes and edges by inspecting how knowledge is transferred across a network. For this, we depict a closure tie when two individuals are teaching to and learning from each other, whereas a brokerage tie characterizes a knowledge transfer between an expert and an inexpert user. Here, instead of describing an entire network or exploring knowledge-transfer behaviors, we aim at characterizing nodes and edges based on the strength of the social ties with their relevant attributes. Thus, we explore strong ties between nodes and their relevant attributes as representing the closure effect, and weak ties underlying the potential of the knowledge acquired by the nodes in the network as depicting the brokerage effect. --- SOCIAL-BASED PERSPECTIVE In this section, we first present an overview of our method for socialbased classification of multiple interactions in dynamic attributed networks [39,40]. Then, we present and discuss our classification scheme entirely based on social concepts. For this, we reinforced the notion of social capital to capture the importance of the nodes based on their positioning in the network structure and the social meaning of the relationships involved. --- Modeling Dynamic Interactions In our previous work [40,42], we modeled attributed networks by a dynamic multigraph 𝐺 = 𝑡 1 G 𝑖 that represents a temporal aggregated graph within a time interval [1, 𝑡]. In short, 𝐺 is the union of temporally disjoint graph instances G 𝑘 = constructed Algorithm 1 Extracting Relevant Attributes Require: 𝐻 , 𝑡 Ensure: Γ 𝑘 , ∀𝑢 ∈ 𝑡 𝑘=0 𝑉 𝑘 1: for all 𝑢 ∈ 𝑉 𝑡 do 2: A 𝑡𝑒𝑚𝑝 ← {} 3: for all 𝑘 ∈ [1, 𝑡] do 4: Γ 𝑘 ← {} 5: A 𝑡𝑒𝑚𝑝 ← A 𝑡𝑒𝑚𝑝 ∪ {𝑎| ∈ E ′ 𝑘 } 6: 𝑣𝑒𝑐𝑡𝑜𝑟 ← {} 7: for all 𝑎 ∈ A 𝑡𝑒𝑚𝑝 do 8: 𝑣𝑒𝑐𝑡𝑜𝑟 .𝑎𝑑𝑑 ) 9: 𝐼𝑄𝑅 ← 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 -𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 10: for all 𝑎 ∈ A 𝑡𝑒𝑚𝑝 do 11: if 𝑝𝑒𝑟𝑠 𝑘 > 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 + 𝐼𝑄𝑅 * 1.5 then 12: Γ 𝑘 ← Γ 𝑘 ∪ {𝑎} in time steps of size 𝑘 [48]. Likewise, in order to consider the dynamic temporal attributes, we defined a heterogeneous dynamic multigraph graph 𝐻 = 𝑡 1 H 𝑖 , where H 𝑘 = . This graph consists of two types of node: actors and attributes . In other words, this strategy is an abstraction that transforms the attributes of each edge into additional nodes, allowing an original actor node to be directly connected to these new attribute nodes. In this context, each actor is associated with a set of attributes that can change overtime. We also assume that all attributes are related to some knowledge, which can be seen as a skill acquired by the actors along the time. --- Extracting Relevant Attributes The next step in our approach is to determine the set of relevant attributes for each node at each time interval. We define as relevant attributes those that are closely connected to the nodes, i.e., persistent in their histories. The idea is to identify, for each actor, all attributes and evaluate them in terms of the set of attributes most strongly statistically associated with the actor nodes . For this, we analyze the nodes' interaction history in order to extract knowledge from the node-attribute relationships. We apply the concept of persistence of an edge along the time, which captures the importance of the relationship between two nodes in terms of their associated attributes. The persistence metric of an edge is defined as 𝑝𝑒𝑟𝑠 𝑡 = 1 𝑡 𝑡 𝑘=1 1 E ′ 𝑘 ) , where the indicator function is defined as 1 E ′ 𝑘 ) = 1, if ∈ E ′ 𝑘 , 0, otherwise. Note that this operation is performed on each attributed graph at discrete intervals and not on the aggregated graph. In other words, it captures the dynamics by observing the persistence in each temporal subgraph within the time interval [1, 𝑡]. More precisely, Algorithm 1 details the process of extracting relevant attributes. It receives as input the aggregated graph 𝐻 = {H 1 , ..., H 𝑡 } and the final time interval 𝑡. In summary, the algorithm inspects, for each actor, all attributes and evaluates them according to their persistence along the time by means of percentiles , thus identifying the set of attributes most strongly statistically associated with the actor's nodes. The idea is to filter such attributes that are exaggeratedly linked to a node in a specific period in comparison to the others, i.e., identifying the abnormal presence of certain attributes at each time point. In order to choose the appropriate statistical method to select the most significant attributes, we first check whether the values of the edge persistence metric follows a normal distribution. Then, we extract the relevant attributes based on the definition of an outlier given by the interquartile range . Another approach is to use the modified z-score for the same purpose [24]. Since the experimental results were similar for IQR and for the modified z-score, we chose IQR due to the possibility of applying different percentages by means of percentiles . As a result, this strategy builds a set comprising all attributes statistically relevant for each node 𝑢 ∈ 𝐺 at a time interval 𝑘 , which are referenced as Γ 𝑘 . Note that the sets , Γ 2 , ..., Γ 𝑡 ) are dynamically built according to the degree of persistence, i.e., different instants 𝑘 may contain completely distinct sets of attributes. --- The Classifier Our classification scheme reinforces the importance of social concepts as a relevant factor for better understanding the complexity that involves actors and their relationships in dynamic attributed networks. In summary, we classify nodes and edges as follows: • Node classification. This classification captures the expertise of the nodes by means of their relevant attributes. For example, in an academic social network, a node that has a long-lasting association with attributes like relational model, data definition and query language is likely to have an authority over them. Thus, this node can be classified as having a strong tie with the Databases domain. • Edge classification. This classification assigns a tie strength to the edges in order to represent the kind of their interactions. For example, in a social media network like Facebook, a strong edge may indicate a social tie between relatives, whereas a weak one may represent a social tie with a coworker. Based on the social structure that models the dynamic interactions along the time, the edge classification process assesses the degree of relevance of the attributes associated with each node by considering its past interactions . For this, it determines the dynamic state of a node at each time interval as representing a strong, weak or non-relevant association with a specific attribute . In this context, the strong state represents the importance of a node in terms of its expertise within a closed group, whereas the weak one captures its potential for connecting different parts of a network. The next step consists in mapping these dynamic states in order to determine the social classes of the edges. We propose three social classes for edges1 : closure, brokerage and innocuous. In this way, such social classes emphasize the strength else Δ) ← 𝑖𝑛𝑛𝑜𝑐𝑢𝑜𝑢𝑠 of the relationships as strong ties , weak ties and non-relevant , the latter when there is no relevant information passing through the edge. Formally, Algorithm 2 describes our process for classifying multiple edges. Note that we express the set of attributes of the edge formed by the nodes 𝑢 and 𝑣 at time 𝑘 as the function Φ 𝑘 ). In addition to function Φ, the algorithm receives as input the multigraph 𝐺, the final time 𝑡 and the function Γ that defines all relevant attributes for each node at each instant 𝑘 . First, the algorithm determines the nodes' dynamic states. These states are assigned independently at each iteration of the algorithm and considering each instant 𝑘 in which an edge is inspected. A node is assigned a state strong when there is a strong temporal link with its attributes at the exact moment of the interaction . However, if these attributes do not apply to the inspected edge, then the state weak is assigned to it . If there are no relevant attributes and the node is active in more than one time interval, then the state non-relevant is assigned to it . Once the dynamic states have been assigned to nodes 𝑢 and 𝑣, the class of the corresponding edge 𝑒 is assigned according to them . More specifically, the brokerage class can be seen as a social tie of nodes from distinct domains , whereas the closure one establishes a social role by demonstrating a high tightness between a node and its attributes . Finally, an innocuous class means that there is no knowledge being disseminated through the inspected relationship . For classifying the nodes, the same classes are assigned to them, in which case we mean by closure a node that has authority on certain attributes, by brokerage a node that has a weak association with its attributes and by innocuous a node that has an occasional presence in the network. The function Ω for this node classification they were not very discriminatory. Thus, here we map the classes according to Burt's social theory [10], which provided more representative results. is given by Ω =        𝑐𝑙𝑜𝑠𝑢𝑟𝑒, if |Γ 𝑡 | ≠ 0 𝑏𝑟𝑜𝑘𝑒𝑟𝑎𝑔𝑒, else if 𝑡 𝑘=1 1 V 𝑘 > 1 𝑖𝑛𝑛𝑜𝑐𝑢𝑜𝑢𝑠, otherwise, where the indicator function is defined as 1 V 𝑘 = 1, if 𝑢 ∈ V 𝑘 , 0, otherwise. In summary, the aforementioned social classes reinforce a sociological perspective based on their positioning in a social structure [10,18,19], i.e., by applying social concepts to better understand the strength of the node-attribute relationships. More precisely, we rely on Burt's definition of social capital [10] by considering the concept of closure as representing the importance of a node in terms of its expertise within a closed group, whereas the concept of brokerage captures its potential for knowledge transfer. In other words, by a closure edge we mean a high tightness between a node and its attributes, whereas a brokerage edge can be seen as a social tie of nodes with distinct relevant attributes. Likewise, when classifying a node, the closure class is assigned to it when there is a strong tie with some knowledge under its set of relevant attributes and the brokerage class when it represents a potential to acquire knowledge from attributes outside its own set of relevant attributes. Indeed, strong ties with certain attributes show an authority on them, whereas weak ties indicates a great potential to diffuse knowledge from its domain. Finally, the innocuous class assigned to a node or edge represents no skill acquired by an individual or a lack of relevant information passing through a relationship, respectively. --- CHARACTERIZATION OF NODES AND EDGES In this section, we characterize several social contexts based on our proposed classification method. We begin by introducing the datasets considered. Then, we analyze the overall results of our method when classifying the social behavior of the nodes and the social meaning of their interactions. --- Datasets We consider two specific scenarios, academic social networks and Q&A communities, whose datasets were built in our previous work [40,41]. Table 1 shows general statistics of the networks. Overall, they present distinct characteristics that allow us to contrast the effect of our classification method on each scenario. In summary, the academic scenario consists of 24 co-authorship networks derived from the ACM Special Interest Groups 2 , as well as the full network comprising all groups with more than 79 thousand nodes and 263 thousand multiple edges. In each network, we modeled authors as nodes, coauthorships in each paper as relationships, publications' year as time intervals and tokens taken from the publication titles as attributes [42]. Note that such networks are well-known in the Computer Science community, which enables us to carry out a more accurate discussion of their behavioral dynamics. Regarding the Q&A communities, we use the database built from the Stack Exchange network [41]. This dataset consists of 173 Q&A communities divided into six categories 3 . Specifically, nodes as representing community members and edges as representing answers to questions, comments to questions and comments to answers as described by Paranjape et al. [34]. In addition, each time interval has been configured to last one minute, and tokens taken from the questions and answers as attributes. The extraction process removes stop-words and reduces inflected words to their roots. 2 ACM SIGs: http://www.acm.org/sigs 3 Stack Exchange: https://stackexchange.com. --- Academic Coauthorship Networks The Full Network shows a substantial drop from 27.0% to 18.8% in the number of closure nodes when compared with the average of all 24 ACM SIG communities. Despite that, the number of innocuous nodes considerably increased from 57.8% to 69.0%. This was expected due to the fact that more active nodes tend to participate in more than one community. Thus, with more subjects covered, the likelihood of having many relevant attributes decreases. Even so, although there are fewer closure nodes, there is still relevant information flowing through them . In fact, such figures indicate that about 30% of all nodes have acted structurally to make the network stronger by means of cohesion and bridges . Note also that there are small variations between the means and medians for the 24 ACM SIG communities, but with a marked standard deviation. This means that there are different social behaviors according to the specificities of each community. Specifically, Figure 2a presents the distribution of the node classes for the 24 ACM SIG communities and the Full Network that includes all these communities. Overall, the classification shows a significant presence of nodes of the class innocuous . Indeed, an academic coauthorship network usually has a strong presence of new nodes . Despite that, there is also a strong presence of nodes of the class closure with percentages above 30% for more established communities such as CIKM, KDD, SIGIR, SIGMOD, STOC, SIGMETRICS, ISCA, PODC, POPL and MICRO. Particularly, most members from these communities tend to be coherent in the research topics addressed throughout their academic trajectories. In contrast, communities such as SAC, SIGUCCS, SIGGRAPH and SIGDOC show percentages below 18% for the class closure, which represents some lack of synergy among their members. Particularly, SIGUCCS and SIGDOC are two communities that address very specific topics. SIGGRAPH , although a well established scientific community, covers here only its editions up to 2003, since after that year their proceedings were discontinued and replaced by special issues of the ACM Transactions on Graphics. Generally, such percentages can be seen as evidence of the characteristics of each community. For instance, members of the STOC community have a tendency to show more competence in specific topics related to computation theory, thus the higher number of nodes of the class closure . On the other hand, SAC is a community mainly focused on applied issues, thus covering a wide range of topics, which justifies the high number of innocuous nodes . Regarding the edge classification, Figure 2b presents the distribution of the edge classes for the 24 ACM SIG networks and the Full Network, which comprises the 24 SIG networks altogether. As we can see, most of these edge classes carry some kind of information and have been characterized as closure or brokerage , thus demonstrating a strong social tie between the researchers and their relevant topics. On the other hand, edges without any social meaning tend to be less present in these networks. Again, specific communities show a singular behavior, such as ISSAC and SIGIR with the highest presence of closure edges. SAC, SIGUCCS and SIGDOC also stand out for having an expressive number of innocuous edges , thus reinforcing the fact their members show no regularity with their research topics. --- Questions & Answers Communities As we only consider frequent users in the Q&A communities , by definition there are no innocuous nodes in these networks [40]. With respect to the node classes, Figure 3a shows few variations in the percentages of closure and brokerage nodes across the communities . More specifically, the Vegetarianism and Buddhism communities show the highest proportions for the closure class , whereas Anime & Manga stands for 72.8%. In contrast, we notice that the full academic coauthorship network had 18.8% of its nodes classified as closure, 12.2% as brokerage and 69.0% as innocuous . Indeed, there are few closure nodes in an academic network compared with the other ones . However, in the Q&A communities, users are in general experts and enthusiasts about specific topics, which gives them some authority [28,35,38,46,47]. Considering the social classification of the edges in Figure 3b, the proportions by category and by community have significant oscillations, thus reinforcing a distinct behavior of our classification method on several topics. For example, the Buddhism community has 84.7% of closure edges, whereas the AI community has a much smaller proportion of edges belonging to that same class. There are also notorious divergences between communities in the same category such as Ask Patents and Quantitative Finance from the Business , where 𝑒 𝑖 is the number of edges between neighbors of 𝑖 and 𝑛 𝑖 is the number of neighbors of node 𝑖 category, Aviation and Freelancing from the Professional category, and Literature and Parenting from the Life/Arts category. As we only selected frequent users, it justifies the very low presence of innocuous edges. By comparing the Q&A distribution by community with the same figures from the academic ones , we observed that the entire DBLP academic coauthorship network had 38.5% of its edges classified as closure, 41.6% as brokerage and 20.2% as innocuous. That is, we note that both scenarios reveal very different proportions of assigned classes, particularly with a higher proportion of the closure class in the Q&A scenarios, whereas in the academic scenarios the most representative class tended to be the brokerage one. --- EVALUATION OF NODES AND EDGES Given the challenging task of analyzing social interactions in a social network in order to better characterize the social role of its nodes and the meaning of its edges, Newman [32] applies network properties to evaluate the robustness of his proposed method. More specifically, he applied centrality metrics to determine the bestplaced nodes in a network. As we are also dealing with non-labeled data, we evaluate our unsupervised classification by means of the nodes and edges acting as structurally important in a network. In others words, our methodology consists of quantifying how well nodes and edges are positioned in a social structure. For this, we explore network metrics as formally presented by Table 3, which their social aspects are discussed as follows: • Degree Centrality. As shown by Srinivas and Velusamy [44], this metric indicates influential nodes as, for example, a node with an immediate risk of catching a virus or getting some information. Thus, a node with high connectivity is more likely to have early access to knowledge. • Closeness Centrality. Nodes with higher closeness are, by definition, closer to the other nodes in the network. Then, we expect important classes to have high values for this metric, since they have better access to knowledge from other nodes . • Betweenness Centrality. Following Newman [32], nodes with a high degree of betweenness centrality are likely to be influential, since they act as an intermediary for other nodes . Thus, as nodes and edges with high betweenness centrality values play crucial roles in the spread of knowledge in social networks [29], then we expect high values for this metric for important nodes and edges assigned to the closure and brokerage classes. • Clustering Coefficient. As this metric reveals the fraction of a node's neighbors that are connected to each other [44], we expect low clustering coefficient values for the most important classes , confirming the behavior of connecting different parts of a network. In addition, we also use the PageRank algorithm [33] by considering that more important nodes tend to make stronger endorsements due to their connectivity and ties to other important nodes. That is, we also expect closure and brokerage nodes to have high values for this metric. --- Results By means of the aforementioned social properties, we now assess the importance of nodes and their dynamics relationships by contrasting the classes assigned to them with the network properties 4 . Considering the node classification, Figure 4 shows the distribution of the network properties by node class. In Figure 4c, we can see that the lowest clustering coefficient values are for nodes classified as closure, followed by those classified as brokerage. On the other hand, nodes classified as innocuous tend to have worse positions in the social structure. In this way, these results indicate that nodes classified as innocuous are highly dependent on their neighborhood, while those classified as closure and brokerage tend to diversify their relationships. In addition, Figure 4b and 4a confirm that nodes classified as closure and brokerage tend to have a better position in the social structure as, respectively, having more connections in the network and being on average closer to other nodes . Considering the importance of nodes in terms of the paths that pass through them and how much endorsement they receive, Figures 4d and 4e respectively show that nodes classified as closure and brokerage have more information passing through them and are seen as topologically more relevant by their peers. Note that all cases have a clear class distinction. Formally, all distributions are statistically different by means of the Mann-Whitney-Wilcoxon test and by the Kruskal-Wallis test [21]. As for the classification of the edges, Figure 5a shows the distribution of the betweenness centrality metric with respect to our classification. We clearly note that the brokerage and closure classes have more expressive values for this metric. Note that the distributions of closure and brokerage distinguish less than those reported for nodes, but now the brokerage class is slightly superior to the closure one in contrast to the classification of the nodes . Nonetheless, they are still statistically different according to the Kruskal-Wallis and Mann-Whitney-Wilcoxon tests. Moreover, even though the innocuous class accounts for 31.4% of all edges, their centrality values are very low. For the sake of comparison, Figure 5b depicts the same distribution for the RECAST classes. As discussed in Section 2, the RECAST algorithm assigns social classes to edges in temporal networks. For this, it explores the regularity of the relationships and the topological overlap existing among them over time. By comparing such regularities with random temporal graphs, it classifies social ties as friend, bridge, acquaintant and random. In this way, we expect the important social classes to have better network properties than those considered less important . Similar to our classification, the most expressive values of centrality are those assigned to the bridge class. On the other hand, RECAST classifies many structural edges as random, as well as several edges with low figures as friend. In conclusion, our method brings a new perspective and provides a more accurate analysis to characterize such relationships. --- Sensitivity Analysis Since we are dealing with temporal attributed networks, the relevant attributes and time aspects must be properly analyzed regarding the effectiveness of our classification method. Accordingly, we address these issues next. Discriminatory power of the attributes. In order to measure the strength of social interactions, Algorithm 1 ensures the function Γ containing the sets of all statistically relevant attributes for each node. In fact, if an attribute is associated with a node several times, then we can infer its importance. However, a specific statistical treatment can be added to this process in order to exclude attributes that, even if randomly distributed, were erroneously considered as relevant ones. This additional statistical step consists in making the function Φ, which associates each edge 𝑒 with a specific set of attributes, a random association Φ ′ . Then, we get Γ from different Φ ′ instances to measure the probability that each attribute has been erroneously classified as being relevant. Finally, we exclude such attributes that were considered as relevant with probability significantly higher than the level of significance 𝛼. In other words, we filter from our input the attributes that can interfere in the process of identifying the relevant ones. Even removing some of the data, we expect the proposed method to be robust enough to properly classify nodes and edges. As a result, both configurations are statistically equivalent by means of the distribution of network properties by classes. Existence time of the nodes. This sensitivity test consists in investigating the robustness of our approach to differentiate nodes with similar existence times. For this, we divided the nodes into the following annual time intervals: [1,5), [5,10), [10,15) e [15, ∞). Our method was able to distinguish the distributions of all network metrics by classes for all time intervals in terms of the Kruskal-Wallis test. However, for the time interval [1,5), the Mann-Whitney-Wilcoxon test did not differentiate the distributions between the classes closure and brokerage for the metrics betweenness centrality and clustering coefficient. --- CONCLUSIONS AND FUTURE WORK In this article, we reinforce the importance of the network theory paradigm for understanding the complexity that involves real world actors and their relationships [5]. Based on the structural autonomy that captures when people are tightly connected to one another with extensive bridge ties beyond them [10], we emphasize the concept of closure as representing the importance of a node in terms of its expertise according their associated attributes , whereas the brokerage one captures its potential for transferring its attributes . Then, we proposed a node-attribute graph model that captures the social tie of individuals and their associated attributes, thus exploring the importance of the persistence of node-attribute relationships over time. Overall, our classification method was able to reveal the social role of the nodes and the strength of the social meaning of their multiple interactions in different social contexts from academic coauthorship networks and Q&A communities. For instance, there is a contrasting social behavior when comparing the Theory of Computing and Applied Computing networks, as well as when we compare the Buddhism and Islam communities. In addition, based on Newman's experimental methodology [32], we statistically validated the assigned classes according to network properties, thus agreeing with their expected social meaning. As future work, we aim to apply our social-based characterization approach to the problem of community detection [12,26]. We also intend to propose a new strategy to explore its propagation behavior, mainly focusing on knowledge transfer aspects [42]. In addition, we plan to investigate the persistence of the nodes with respect to their neighborhood in order to identify the most influential ones.

How to characterize nodes and edges in dynamic attributed networks based on social aspects? We address this problem by exploring the strength of the ties between actors and their associated attributes over time, thus capturing the social roles of the actors and the meaning of their dynamic interactions in different social network scenarios. For this, we apply social concepts to promote a better understanding of the underlying complexity that involves actors and their social motivations. More specifically, we explore the notion of social capital given by the strategic positioning of a particular actor in a social structure by means of the concepts of brokerage, the ability of creating bridges with diversified patterns, and closure, the ability of aggregating nodes with similar patterns. As a result, we unveil the differences of social interactions in distinct academic coauthorship networks and questions & answers communities. We also statistically validate our social definitions considering the importance of the nodes and edges in a social structure by means of network properties.

Introduction Several studies have shown the ability of crowd workers to successfully contribute to the analysis and annotation of multimedia content, both based on simple perceptual skills ) and domain-specific knowledge . Musical content is no exception, and research has shown that the general crowd can be successfully involved in the annotation and evaluation processes of music-related data and methods. Plenty of music annotation tasks can be routinely found on microtask crowdsourcing platforms, mostly focused on descriptive and emotional tagging. Music, as a form of art, often requires a multifaceted set of skills to perform, and certain expertise to analyse its artefacts. There are cases that require advanced music perceptual skills and music-specific knowledge. However, both in literature and in practice, it is rare to encounter such crowdsourcing tasks. Consider, for example, annotation tasks targeting classical music, e.g. music transcription, performance evaluation, or performance annotation. Classical music is a genre featuring artworks with high musical complexity; it is no surprise that corresponding analysis and annotation tasks are often exclusively performed by musical experts and scholars. This unfortunately hampers current efforts to digitize and open up classical music archives, as scholars and experts are expensive and not easily available. Here, the ability to utilise microtask crowdsourcing as an annotation and analysis approach could bring obvious advantages. But how likely it is to find advanced music-related perceptual skills on crowdsourcing platforms? With the goal of answering this broad research question, in this paper we scope our investigation on the following two aspects: • [RQ1] To what extent are higher perceptual skills of melody, tuning, accent and tempo, present on microtask crowdsourcing platforms? • [RQ2] How are different perception skills and selfreported music-related knowledge distributed among crowd workers? Studies on human cognition and psychology, have shown that people can possess innate music perception skills, without previous formal training . However, the majority of those studies have been conducted in labs, under controlled conditions, and with limited amounts of participants. In our work we set out to measure the music sophistication and perception skills of crowd workers operating on the Prolific crowdsourcing platform. 1We designed a rigorous study that employs validated tools to measure the musical sophistication of the users and quantify their music perception skills: the Goldsmith's Music Sophistication Index questionnaire and the Profile of Music Perception Skills active skill test respectively. These tools allow for a general overview of musical ability characteristics, but also a more detailed understanding through their subcategories . By juxtaposing passive methods of assessment with the active evaluation of auditory skills, we aim to gather a better understanding of workers' actual skills on musical aspects, beyond their subjective selfassessment. With GMSI, we are able to evaluate a person's ability to engage with music through a series of questions focusing on different musical aspects. PROMS on the other hand, allows for a more objective way to measure a person's auditory music perception skills through a series of audio comparison tests. To the best of our knowledge, this is the first attempt to use PROMS in an online crowdsourcing environment, and the measured perception skills can offer valuable insights to the auditory capabilities of the crowd. Our findings indicate that pre-existing musical training is not common among crowd workers, and that music sophistication aspects are not necessarily predictive of actual music perception skills. Instead, we observe that the majority of workers show an affinity with specific sets of skills . As a whole, our study paves the way for further work in worker modelling and task assignment, to allow a wider and more refined set of microtask crowdsourcing tasks in the domain of music analysis and annotation. --- Related Work There is a long history of studies on perception and processing of music by humans; from the analysis of the sociocultural variables influencing a person's musicality amplitude , to the study of musicality from a genetics' base . In all cases, inherent music processing capabilities have been found in people and they seem to be connected with basic cognitive and neural processes of language since early stages of development . Even people with amusia, a rare phenomenon where a person can't distinguish tonal differences between sounds , they can still process and replicate rhythm correctly . In , we find a large scale study on musical sophistication through the use of the GMSI survey, on a unique sample of 147,663 people. GMSI is particularly calibrated to identify musicality in adults with varying levels of formal training. It is targeted towards the general public, and can prove less effective to distinguish fine differences between highly trained individuals. Musical sophistication in the context of that study, and ours, encompasses musical behaviours and practices that go beyond formal training on music theory and instrument performance. Their findings show that musical sophistication, melody memory and musical beat perception are related. The survey has been translated and replicated successfully in French , Portuguese , Mandarine , and German . Our study draws connections to those findings and aims to shed light into the musical capabilities of people on crowdsourcing platforms. The demographics and conditions of the studies presented so far, cannot be easily compared to those of online markets. Users on those platforms are participating in such studies through monetary incentives, and the conditions under which they perform the tasks cannot be controlled as in a lab environment . Currently, crowdsourced music annotation is primarily utilised for descriptive and emotional tagging. Large-scale music data creation and annotation projects such as Last.fm2 and Musicbrainz3 , are largely depended on human annotation, but from users of their respective online social platforms. A survey on the applicability of music perception experiments on Amazon Mechanical Turk , showed that online crowdsourcing platforms have been underused in the music domain and the status has not changed radically since then. Through our study, we want to examine the capabilities of the crowd on processing music audio and showcase their capabilities, in an attempt to encourage further research and utilisation of crowdsourcing in the music domain. --- Experimental Design The main focus of this study is to offer insights into the musical characteristics and perception skills of workers operating on crowdsourcing platforms. We therefore designed our experiment to capture these attributes through methods that can be used online, and that do not require preexisting musical knowledge. We used two methods: 1) the GMSI questionnaire to evaluate the musical sophistication ) of workers and 2) the Mini-PROMS test battery to evaluate their auditory music perception skills. We then compare the obtained results, paying specific attention to the overlapping aspects of musical sophistication and music perception skills. With this experiment, we are also interested in identifying "musical sleepers" and "sleeping musicians", a notion originally presented in . A musical sleeper is a person with little to no musical training but with high performance in the perception test, while a sleeping musician indicates the opposite. --- Procedure After a preliminary step where workers are asked basic demographic information , the study is composed of four consecutive steps , each devoted to collecting information about specific attributes corresponding to the crowd workers: 1) Musical Sophistication Assessment , 2) Active Music Perception Skill Assessment , 3) Self-Assessment of Music Perception Skills, and 4) Post-task Survey collecting information on workers audio-related conditions, and perceived cognitive load. --- Mini-PROMS Figure 1: The five steps in the music perception skills study. --- Questionnaires and Measures Capturing Musical Sophistication of Workers. Musical behaviours of people such as listening to music, practicing an instrument, singing or investing on vinyl collections, all show the affinity of a person towards music. The degree to which a person is engaged to music through these behaviours, constitutes the musical sophistication. Musical sophistication can be measured as a psychometric construct through the GMSI questionnaire, which collects selfreported musicality through emotional responses, engagement with music, formal training, singing capabilities and self-assessed perception skills. It is an instrument specifically designed to capture the sophistication of musical behaviours, in contrast to other questionnaires such as Musical Engagement Questionnaire , which measures the spectrum of psychologial facets of musical experiences. More specifically, the musical sophistication of people based on , is organised into the following five facets: Active Engagement: this aspect determines the degree to which a person engages with music, by listening to and allocating their time/budget to it; Perceptual Abilities: this aspect assesses the skill of perceiving elements of music. This is an important subscale in our study, since the self-assessed perceptual skills of the workers in GMSI can be directly compared to those we actively measure in Mini-PROMS; Musical Training: this aspect reports the years of training on aspects of music , which can indicate the formal expertise that a person has in the domain; Emotions: this aspect determines the emotional impact of music on that person; Singing Abilities: this aspect evaluates the ability to follow along melodies and tempo of songs. GMSI offers additional questions outside the subscales, which capture specific properties of the participant: 1) "Best Instrument", which represents which instrument the user knows to play the best, 2) "Start Age", which age the participant starting learning an instrument and 3) "Absolute Pitch", which indicates if the person can understand correctly the exact notes of a sound frequency. Absolute pitch is a very rare trait that develops during the early stages of auditory processing but can deteriorate through the years . As such, a person with perfect pitch perception, could have an advantage on a melody perception test, thus we included it with the rest of the subscales. The original GMSI questionnaire contains 38 main items and 3 special questions, and considering the rest of the study's parts, we chose to reduce its size, while keeping its psychometric reliability. For that purpose, we consulted the GSMI online "configurator"4 which allows to select the number of items per subscales and estimates the reliability of the resulting questionnaire based on the questions it selects. We reduced the size of the questionnaire to 34 questions, and preserved the special question about "Absolute Pitch", resulting in 35 questions in total. Table 1 presents the psychometric values5 of the final GMSI version we used; each sub-scale fares very-good to excellent reliability values . In the GMSI questionnaire, each question from the subscales, uses the seven-point Likert scale for the user's responses, with most questions having "Completely Agree", "Strongly Agree", "Agree", "Neither Agree Nor Disagree", "Disagree", "Strongly Disagree" and "Completely Disagree" as options. Few questions offer numerical options for topics . The workers is not aware of the subscale each question belongs to. The index of each subscale of GMSI is calculated with the aggregated results of the relevant questions. The overall index of "General Music Sophistication" is calculated based on 18 questions out of the total 34 items of the subscales; these 18 questions are predefined by the designers of the questionnaire; the question about "Absolute Pitch" does not contribute to the total index. Using the GMSI questionnaire is close to the typical methods used to assess the knowledge background of annotators in other domains. Especially the questions of "Musical Training" follow standard patterns to assess the formal training of a person in a domain, thus a certain objectivity can be expected . However, the rest of the categories are based purely on subjective indicators and self-reported competence, which can potentially misrepresent the true music behaviours and capabilities of a worker. For this reason, it is necessary to understand the best practices that could reliably predict a worker's performance to a music annotation task. To that end, we compare the workers' input in such questionnaires, and specifically on GMSI, to the music perceptual skills they might possess, which we measure through an audio-based, music perception skill-test. . Its original version is quite extensive, and its completion can take more than an hour, as it covers several music cognition aspects like Loudness, Standard rhythm, Rhythm-to-melody, Timbre, Pitch and more. Considering the possibly low familiarity of crowd workers with these tasks, and its inherent difficulty, we opted for a shorter version, the Mini-PROMS , which has also been adopted and validate in the context of online, uncontrolled studies. Mini-PROMS is a much shorter battery of tests , which still covers the "Sequential" and "Sensory" subtests. It can measure a person's music perception skills, by testing their capability to indicate differences on the following musical features: --- Melody: A sequence of notes, with varying density and atonality Accent: The emphasis of certain notes in a rhythmic pattern Tuning: The certain frequency of notes, when played in a chord Tempo: The speed of a rhythmic pattern The musical aspects selected in this test, are argued to well represent the overall music perception skills of a person, only in a more concise way. This version, retains test-retest reliability and internal consistency values close to the original PROMS test , validating it for our research purposes. Note that, will reduced in size, these four skills are required to enable a broad range of music-related research, such as beat tracking, tonal description, performance assessment and more. For each of the 4 musical aspects, workers receive a brief explanation and an example case to familiarise the user with the test. Each challenge after the introduction, presents a reference audio sample twice and a comparison sample once. The two audio samples can differ based on the musical aspect tested and the worker is asked if the samples are indeed same or differ. The authors of PROMS have put particular effort on distinguishing the musical aspects from each other, to make the skill evaluation as close as possible to the musical aspect tested. Finally, to minimise cognitive biases due to enculturation , the audio samples have been created using less popular instrument sounds, such as harpsichord and "rim shots". Meanwhile, the structure of audio samples and the aspect separation allow for a more precise measurement of a person's perception skill. The categories of "Melody" and "Accent" have 10 comparisons each, while "Tuning" and "Tempo" have 8. After the user has listened to the audio samples, they are asked to select between "Definitely Same", "Probably Same", "I don't know", "Probably Different" and "Definitely Different". The participant is then rewarded with 1 point for the high-confidence correct answer, while the low-confidence one rewards 0.5 point. The subscale scores are calculated through a sum of all items within the scale and divided by 2. The total score is an aggregated result of all subscale scores. During the test, the user is fully aware of the subscale they are tested for, but the name of "Tempo" is presented as "Speed" . Self-assessment on Music Perception Skills. Selfassessment can often misrepresent an individual's real abilities . For that reason, we employed a survey to study this effect its manifestation with music-related skills. After Mini-PROMS test, the worker has to input how many of the comparisons per subscale they believe they correctly completed -this information is not known to them after executing the Mini-PROMS test. Therefore, they are presented with 4 questions, where they have to indicate between 0 and the total number of tests per subscale . Finally, the results of this survey, are compared to the score of workers on the "Perceptual Abilities" subscale of GMSI, which also relies on self-assessment. We expect workers to re-evaluate their own skills, once exposed to the perception skill test. Post-task Survey. As a final step of the task, the worker is presented with three post-task surveys: 1) an survey on the audio equipment and the noise levels around them, 2) a survey on the cognitive load they perceived and 3) an openended feedback form. The audio equipment survey consisted of four main questions, to retrieve the type of equipment, its condition and the levels of noise around them during the audio tests. Insights on these can help us understand the to what extent the equipment/noise conditions affected Mini-Proms test, which is audio-based. More specifically, we asked the following questions: 1. What audio equipment were you using during the music skill test? 2. What was the condition of your audio equipment? 3. Does your audio equipment have any impairment? 4. How noisy was the environment around you? The options regarding the audio equipment were: "Headphones", "Earphones", "Laptop Speakers" and "Dedicated Speakers". For the condition questions and , we used the unipolar discrete five-grade scales introduced in , to subjectively assess the sound quality of the participants' equipment. Finally, for question on noise levels, we used the loudness subjective rating scale, introduced in . In the second part of post-task survey, the workers had to indicate their cognitive task load, through the NASA's Task Load IndeX survey6 . The survey contains six dimensions -Mental Demand, Physical Demand, Temporal Demand, Performance, Effort, and Frustration. Workers use a slider to report their feelings for each of the six dimensions. A low TLX score represents the music skill test is not mentally, physically, and temporally demanding, and it also indicates less effort, and less frustration perceived by the worker, while completing the entire study. Finally, we introduced an free-form textual feedback page, where users were encouraged to leave any comments, remarks, or suggestions for our study. --- Worker Interface The worker interfaces of our study is using VueJS7 , a JavaScript framework. The first page of our study, contained general instructions for the study alongside estimated completion times for each part of it. Each page thereafter, contained an interface for each of the steps in our study, as seen in Figure 1. To assist navigation through the GMSI questionnaire, we implemented the questionnaire interface to show one question at a time. We added a small drifting animation to show the next question, when they select their answer in the previous one. We also added a "back" button, in case they wanted to return to a previous question and alter their answer. They could track their progress through the questionnaire from an indication of the number of the question and the total number of questions . While we retrieved the questions for GMSI and implemented them in our study's codebase, for PROMS we wanted to use the exact conditions and audio-samples as in . To replicate their test faithfully, the creators of PROMS kindly gave us access to their Mini-PROMS interfaces . Mini-PROMS is implemented on LimeSurvey8 and users were redirected to it after the completion of GMSI. After the GMSI questionnaire, workers were introduced to the page seen in Figure 4. There, they had to copy their Participant ID and use it in the Mini-PROMS interface later, so we could link their test performance , with their entries in our database. At the end of Mini-PROMS, the users were redirected back to our study through a provided URL. In the final stage of our study, the participants were greeted and provided a "completion code", which they could submit on Prolific to complete the task. --- Participants, Quality Control, and Rewards On Prolific, we recruited 100 crowd workers to complete our study. We applied a participant selection rule for "Language Fluency": English, as all of our interfaces were implemented in English. Only crowd workers whose overall approval rates were higher than 90% could preview and perform our study. To assess the quality of the user input, we included attention check questions on the GMSI and NASA-TLX interfaces of the study. More specifically, we included three attention check questions in GMSI, asking the participants to select a specific item in the same seven-point Likert scale. In the NASA-TLX survey, we included a question asking the users to select a specific value out of the 21 available in the scale of the survey. Of the 100 workers recruited from Prolific, 8 of them failed at least one attention check question; 5 of them provided invalid/none inputs. After excluding these 13 invalid submissions, we have 87 valid submissions from 87 unique workers. We set the reward on Prolific for completing our study to 3.75 GBP . Upon the completion of our study on Prolific, workers immediately received the reward. The average execution time was 32.5 minutes, resulting in the hourly wage of 7.5 GBP , rated as a "good" pay by the platform. --- Results --- Worker Demographics Table 2 summarises workers' demographic information. Of the 87 crowd workers who provided valid submissions, 36 were female , while 51 were male . Age of participants ranged between 18 and 58 and the majority of them were younger than 35 . The majority of the workers were reported to be unemployed, while from those employed, 73.17% had a full-time job. Most workers had enrolled for or acquired a degree , with 51.47% of them pointing to Bachelor's degree. In total, we employed workers from 15 countries, with most workers currently residing in Portugal , United Kingdom , Poland and South Africa . --- Variables Statistics Gender --- Results on Worker Music Sophistication Table 3 and Figure 5 summarise the results of the GMSI questionnaire on our workers. We contrast our results to results of the original GMSI study , which covered a large population sample of participants n = 147, 663 that voluntary completed the questionnaire, on BBC's How Musical Are You? online test. Participants were mainly UK residents and, in general, from Englishspeaking countries , with 15.9% having non-white background. The sample contained a large spread on education and occupation demographics, where only 1.8% claimed working in the music domain. To some extent, this study is considered representative for the general population in the UK . As such, we can assume a certain disposition and affinity to music from GMSI's population sample, compared to ours where the incentives where monetary. In our study, the observed General Music Sophistication positions our workers pool at the bottom 28-29% of the general population distribution found in the GMSI study. We observe a similar effect also with the individual subscales with the exception of "Emotions", for which our workers fare a bit higher . The result indicates that the self-reported music sophistication of crowd workers is strongly below that of the general population. Most workers had received relatively little formal training in their lifetime . This finding is important for the rest of the analysis, as it indicates low formal expertise with music among the crowd workers. Most workers indicate relatively high perceptual abilities . Here, it is interesting that pre-vious studies estimate that less than 1% per 11,000 possess "Absolute Pitch". In our sample though, 9 workers indicated having this characteristic, little more than the 10% of our sample. This could indicate a possible confusion between quasi-absolute pitch which is related to the familiarity of a person with an instrument's tuning and timber , or with relative pitch. Relative pitch is trainable through practice and useful to professional musicians, as they can detect changes in pitch through the relations of tones . Table 4 presents the correlations between GMSI subscales. As the scores of each GMSI subscale follow a normal distribution , we applied Pearson's R test to calculate correlation coefficients. We observe that Perceptual Abilities shows positive correlations with most other subscales , especially with Music Training , Emotions , and Singing Abilities . This finding suggests that the listening skill plays the most important role in crowd workers' music sophistication. We also find significant correlations between Active Engagement and Emotions , and between Singing Abilities and Musical Training . The original GMSI study has shown that different subscales are strongly correlated . The difference we observe could be partly explained by the generally lower musical sophistication scores of the crowd workers in our pool. --- Results on Objective Music Perception Skills This section discusses the results of the Mini-PROMS active perception skill test, contributing to the answer of RQ1. Mini-PROMS categorizes perception skills as "Basic" if the total obtained score is lower than 18, "Good" if between 18 and 22.5, "Excellent" for values between 23 and 27.5, and "Outstanding" for values over 28 . The original Mini-PROMS study covered a total n = 150 sample of participants, all recruited from the university of Innsbruck, via email. Most of the participants were students with at least one degree , aged 27 on average. We observed an average of "Good" music perception skills for our workers . 48 out of 87 produced reasonably high accuracy in music skill tests (belonging to "Good" and better "Excellent" and "Outstanding" in Mini-PROMS's scale. In the following section we will analyse in greater detail the relationship between the measured music sophistication and the perception skills. A similar trend towards lower performance compared to the original Mini-PROMS study can be observed across the other musical aspects: workers correctly identified melody differences with 49.77% avg. accuracy , tuning differences with 52.73% avg. accuracy , accent difference with 51.95% avg. accuracy , and tempo differences with 64.3% avg. accuracy . The result of the music skill tests is in-line with the result of self-reported music sophistication from GMSI, suggesting that when compared to the populations covered by previous studies, crowd workers generally possess less music perception skills. To deepen the analysis, we calculated the intercorrelation of Mini-PROMS subscales, and made comparison with the original study . Since the Mini-PROMS scores across all the subscales follow normal distributions ), we carried out Pearson's R tests to get the correlation coefficients and corresponding p-values. We find statistical significance on all the intercorrelations. Especially, we find that workers' music skills related to melody are positively correlated with their accent-and tempo-related skills , while accent and tempo also shows a moderate correlation . In comparison with the original study, we do not observe large differences in the R values, while we did with the GMSI results. The results of the intercorrelation analysis suggests that worker melody, accent, and tempo skills are related with each other in our population too. This is a positive result, that suggests 1) the applicability of this testing tool also on this population, and 2) the possibility of developing more compact tests for music perception skills, for workers' screening or task assignment purposes. When focusing on the top 10% of workers, we observed an accuracy on "Melody" between 75% and 90% , while the top 5% scored higher than 85%. A person with "Absolute Pitch" would be expected to achieve high accuracy on this test. Only one person in the top 10% had indicated "Absolute Pitch", but their accuracy was one of the lowest in the group . This could indicate that the person is more likely to not possess such a characteristic. For the subcategory of "Tuning", the top 10% achieved accuracy between 81.25% and 93.75%, while the top 5% scored higher than 87.5%. On "Accent", the top 10% reached accuracy between 80% and 95%. Finally, on the subcategory of "Tempo" we measured accuracy of 87.5% and 100% in the top 10%, while the top 5% achieved perfect score of 100%. These results suggest the presence of a substantial fraction of workers possessing higher music perception skills than expected from their training, although differently distributed. For example, workers who perceived well changes in "Melody", didn't perform equally well on the other categories. This could indicate that music perception skills do not necessarily "carry over" from one music feature to the other; other workers will be good in perceiving changes in tempo, while others on tuning. This encourages the use of the appropriate set of tests, to identify potentially high performing annotators. Thus, if we take as example beat tracking annotation tasks, it would be more beneficial to focus on testing the rhythm-related perception skills, as the other categories have lower chance to capture the appropriate workers for the task. --- Comparison between GMSI and Mini-PROMS To answer RQ2, we compare the results between GMSI and Mini-PROMS by running Pearson's R correlation between the subcategories and total scores of GMSI and Mini-PROMS respectively . We observe moderate positive correlations between Accent-related music skills with worker self-reported music sophistication in terms of Perceptual Abilities , Musical Training , and the General Music Sophistication score . To a lesser degree, we see Melody-related skills to correlate with the same exact GMSI subscales, with equally high statistical significance. We can observe a weak positive correlation in the remaining categories, which indicates that in the studied population, GMSI and Mini-PROMS are only loosely related. The almost correlation close to zero between "Active Engagement" and "Emotions" to the Mini-PROMS categories replicate findings of a study targeting a larger sample , where GMSI was compared to melody memory and beat perception skills and they reported also low correlation values between them. These results indicates that the extent to which workers engage with music is not a predictor for a worker's accuracy on audio-based music tasks. The same would hold true for questions on how emotionally deep can the workers connect to music. In other words, it is not possible to trust the enthusiasm or commitment towards music that workers report, at it wouldn't be a good indication of their actual music perception skills. Finally, the low or in cases negative correlation of "Absolute Pitch" with the Mini-PROMS categories validates our scepticism towards workers who claimed to possess "Absolute Pitch" abilities. Figure 6 shows a scatter plot that describes the distribution of crowd workers distributed in terms of their music skill accuracy and their music training experience . With this analysis, we would like to understand if our sampled crowd workers pools "hides" 1) Musical Sleepers, who are not formally trained but can accurately perform music-related tasks, and 2) Sleeping Musicians, who have received years of musical training, but perform poorly in music-related tasks. Recall from Table 3 that most of our workers had not received formal training . Considering the presence of a substantial num- ber of highly performing workers , and the fact that very few workers received formal musical training , most of them can be qualified as proper music sleepers . The presence of these workers is very encouraging, as it shows that it is possible to deploy advanced music analysis tasks on microtask platforms and and finding high-value contributors. --- Self-assessed versus Measured Skill Levels With this step, we study whether self-assessment can truly reflect one's music-related skills. The self-assessment accuracy and the actual Mini-PROMS accuracy are displayed together as a scatter plot in Figure 7. In this figure, blue points represent individual workers. Workers who located on the brown line perfectly self-assessed their performances while doing the Mini-PROMS test . Clearly, the majority of the crowd workers who participated our study consistently over-estimated their performances on music skill tests, irrespective of their actual music perception skills. We therefore observe a Dunning-Kruger effect, similarly to what has been found in other types of crowdsourcing tasks . Post-task Survey: Equipment and Cognitive Workload The majority of the workers reported that, during the test, they used headphones , earphones and laptop speakers . All workers reported the quality of their equipment as "Fair" or better quality . 96.55% argued that their equipment either does not have any impairment or that the impairment is not annoying . Finally, the majority of workers reported near silence conditions, while 31.03% of them reported normal, non-distracting levels of noise. While these conditions are not comparable to lab setups, we consider them to be sufficiently good to accommodate the requirements of our study. We study workers' cognitive workload to understand if worker performance in music-related tasks is related to the perceived cognitive workload. We performed Pearson's R tests to calculate correlation coefficients between dimensions of cognitive worklaod and subscales of Mini-PROMS. Results are reported in Table 8 We observe negative correlations across all the dimensions and subscales, which suggests that workers who perform well in the music-related tasks also tend to perceive less workload, meaning they could feel less demanding , more successful, less difficult, and less frustrated. In traditional crowd tasks, workers need to pay more attention and put more effort to achieve high-quality performance. This result indicates that musicrelated tasks, high accuracy could be more "spontaneously" achieved, with less cognitive workload. --- Discussion In this study, we extensively measure the musical sophistication and music perception skills of crowd workers. We show that the self-reported music sophistication of crowd workers is below that of the general population and that formallytrained workers are rare. Nevertheless, we found surprisingly refined and diverse music perception skills amongst the top performers, which cannot accurately be predicted by questions regarding their engagement with music as a hobby. Studying the distribution of workers, we find evidence that supports the existence of workers with high accuracy and little to no formal training on crowdsourcing plat-forms, namely "musical sleepers", indicating the prospect of high-quality annotations by non-experts on these platforms. However, this opens the challenge of how we can identify these "musical sleepers" reliably during worker selection. --- Implications for Design Self-reported Musical Sophistication. The musical sophistication assessments is a useful tool to evaluate workers' capability in completing music-related tasks. It is however a lengthy questionnaire, which could result in extra cost and worse worker engagement. Reducing the number of question is possible, but with implication in terms of test reliability. For instance, the subscale of Musical Training is positively correlated to their actual music perception skills . As music perception skills are of primary relevance when executing music-related tasks, we suggest that in future task design, requesters could consider using the subscale of musical training which only contains 7 items. This could be complemented with novel methods to effectively and precisely predict worker performance to further facilitate task scheduling and assignment. Music Perception Skill Assessment. The Mini-PROMS tool appears to be an effective mean to evaluate worker quality in terms of music skills. Yet, it suffers from the same overhead issues of GMSI. In this case, we suggest to use PROMS or Mini-PROMS as a qualification test, possibly featured by crowdsourcing platforms. Workers could use this test to get the corresponding qualification, to obtain the opportunities to access more tasks, and earn more rewards. Music Annotation and Analysis Tasks. The results of this study indicate that knowledge-and skill-intensive musical tasks could be deployed on microtaks crowdsourcing platforms, with good expectations in terms of availability of skilled workers. However, performance on different skills appears to be unevenly distributed. We therefore recommend to analyse the capabilities of the selected crowd and tailor the design of advanced music annotation and analysis tasks to precise music perception skills. --- Limitations and Future Work A main limitation of our study is concerned with the size and diversity of the tested population. A larger and/or more diverse participation pool could potentially aid the generalisability of our findings and lead to more fine-grained insights. Even though our results are based on a population of crowd workers that have received less formal musical training than the average population used in similar studies , the use of standardised and validate tests, lend confidence to the reliability of our findings. Another potential confounding factor in our study, is the motivation for participation. We attracted crowd workers using monetary rewards, while in other studies people voluntarily performed their test . Such a difference could also explain the differences in observed distributions . However, monetary incentives are a feature of crowdsourcing markets, which makes them appealing in terms of work capacity and likelihood of speedy completion. In that respect, our findings are very encouraging, as they show the availability of both musically educated and/or naturally skilled workers that could take on musically complex tasks. As demonstrated in our results, workers who perform well in a certain perception category do not perform equally well in another . In future studies, we encourage the use of perception tests, adjusted and adapted for the specific music task at hand by using the appropriate categories, to accurately select potentially highly performing workers. Fatigue and distraction could have played a role in shaping the observed worker accuracy of the perception results. The relatively limited number of excluded workers though, and the limited cognitive workload experienced by workers, gives us an indication against this risk. In this study, we utilized standardized tools to capture domain-specific characteristics of the workers of a specific platform. Comparing results from their self-reported "connection" to the domain, with those from actively testing their skills, can paint a clear picture of the workers' demographics on a specific domain. While this work is specific to the music domain, we believe that similar workflows can be utilized to study the characteristics of workers on other domains. This holds especially true, as crowdsourcing platforms have diverse user-bases and direct comparisons cannot safely be drawn to studies with highly controlled population samples. --- Conclusion In this paper, we have presented a study exploring the prevalence and distribution of music perception skills of the general crowd in an open crowdsourcing marketplace. We measured and compared self-reported musical sophistication and active music perception skills of crowd workers by leveraging the established GMSI questionnaire and Mini-PROMS audio-based test, respectively. Our analysis shows that selfreported musical sophistication of crowd workers is generally below the general population and the majority of them have not received any form of formal training. Despite that, we observed a substantial number of workers who achieved a reasonably high accuracy in music perception skill tests, alongside a substantial presence of musical sleepers. Moreover, our analysis shows worker accessibility to adequate equipment. Together, these findings indicate the possibility of further increasing the adoption of crowdsourcing as a viable means to perform complex music-related tasks. Future work will focus on conducting experiments with a larger and more diverse pool of workers , to gain further insights and improve the generalisability of our findings.

Music content annotation campaigns are common on paid crowdsourcing platforms. Crowd workers are expected to annotate complicated music artefacts, which can demand certain skills and expertise. Traditional methods of participant selection are not designed to capture these kind of domainspecific skills and expertise, and often domain-specific questions fall under the general demographics category. Despite the popularity of such tasks, there is a general lack of deeper understanding of the distribution of musical properties -especially auditory perception skills -among workers. To address this knowledge gap, we conducted a user study (N = 100) on Prolific. We asked workers to indicate their musical sophistication through a questionnaire and assessed their music perception skills through an audio-based skill test. The goal of this work is to better understand the extent to which crowd workers possess higher perceptions skills, beyond their own musical education level and self reported abilities. Our study shows that untrained crowd workers can possess high perception skills on the music elements of melody, tuning, accent and tempo; skills that can be useful in a plethora of annotation tasks in the music domain.

Introduction The period of adolescence and the transition to adulthood can be difficult for child welfareaffiliated youth, as this population has a high prevalence of mental health problems accompanied by increasing underutilization of mental health services as youth age into adulthood . Contributing to these difficulties is the high likelihood of unstable housing experiences for this population . Unstable housing includes homelessness and other nonstable residences, such as living in cars . Most research on child welfare-affiliated youth, i.e. children who have had an open case in the child welfare department, has focused on unstable housing following foster care, and there is call for research on unstable housing to include youth in child welfare who remain with their parents . The present study examined the demographic, family, and maltreatment characteristics that were related to an unstable housing experience for child welfare-affiliated youth who remained with their parents as well as those who had a nonparent caregiver. Additionally, this study examined the relationship between unstable housing and subsequent delinquency and marijuana use in late adolescence. Family poverty is at the root of unstable housing in families involved with child welfare as rent is usually the largest expenditure in household budgets and families struggle to meet the need for stable housing . Families in general in the United States are getting poorer, and homelessness is associated with poverty even more now than in the 1980's . Thus, the effect of poverty on child welfareinvolved families and their children is imperative to consider in order to understand how these children end-up having unstable housing experiences. Additionally, for child-welfare involved youth, the type of maltreatment they have experienced has been found to be related to variability in the experience of unstable housing. In a small qualitative study, sexual abuse was identified by 39% and physical abuse was identified by 50% of the homeless youth . In a larger sample of 303 homeless youth ages 12-20, more than one-third identified physical violence as the reason for leaving home . The relationship of sexual and physical abuse to youth homelessness was also found in an earlier quantitative study conducted in Seattle, Washington, where one-half of the youth reported being physically abused and almost onethird experienced sexual abuse . For former foster youth, unstable housing experiences were associated with a history of physical and sexual abuse . Sexual abuse and physical abuse are the maltreatment types most often associated with unstable housing, although one study found neglect was associated with unstable housing in young adults . One of the negative outcomes associated with unstable housing among child-welfare affiliated youth is engagement in delinquent behavior . The stress of being in an unstable housing experience can lead youth to adopt "survival behavior" which includes meeting the basic needs of food and clothing; youth sometimes engage in delinquent behaviors to meet these needs . Additionally, unstable housing as well as delinquency can differ by the sex of the youth. In a longitudinal study of youth who aged out of foster care, being male and more engagement in delinquency behaviors were related to experiencing homelessness . Similarly, male former foster youth reported more homelessness and unstable housing than their female counterparts . Delinquent behaviors that begin in childhood are more frequent in males than females. This trend of sex differences in the onset of delinquency continues through adolescence, although the difference between sexes is smaller in adolescence than in childhood . Delinquency often cooccurs with substance use and both risk behaviors are outcomes related to child welfare involvement and experiences of unstable housing. In Canada, a population-based study found that youth who experienced unstable housing had higher rates of both problematic substances use and violence concerns than youth who were stably-housed . For adolescents who attended a HIV clinic, those who used drugs were more likely to use marijuana than other hard drugs, and those with a history of unstable housing were more likely to use marijuana compared to youth without a history of unstable housing . Therefore, we examined whether unstable housing predicted the most common untoward outcomes of delinquency and marijuana use in our sample of child welfare-affiliated youth. In thinking about how unstable housing contributes to risk for adolescents and young adults, the concept of housing careers for youth has been proposed . A housing career refers to the sequence of dwellings a person or family occupies and how this process of housing affects outcomes for the family members. Frequent moves resulting in a high number of residences has some risk for adolescents, but negative housing situations or having unstable housing, such a living in jails or being homeless, most often result in negative outcomes . Thus, housing instability and unstable housing are different concepts; housing instability is solely the total count of residences, whereas unstable housing is the lack of quality in the type of housing. Unstable housing, child welfare involvement, delinquency, and marijuana use in adolescence all appear to be interrelated and these relationships have been shown to differ based on individual and family factors. Most research on unstable housing in child welfare-affiliated youth is focused on how the foster care experience is related to an adolescent's subsequent homelessness, or how a family's unstable housing is a factor in causing child welfare involvement . There is very little research on unstable housing in child welfare-affiliated adolescents who have not experienced foster care. Ryan, Perron, and Huang noted that many of the risk factors for unstable housing are similar for child welfare-involved youth who remain in their home of origin compared to those who enter foster care and called for more research on unstable housing to be conducted on youth with a child welfare history who remain with their parents. --- Current Study Addressing this gap, the sample for this study included both youth who ever had a nonparent caregiver and those who remained with their parents. This study examined: whether youth demographics , family factors , and maltreatment type were related to an incidence of unstable housing; and whether unstable housing predicted subsequent delinquency and marijuana use after controlling for relevant confounding variables. A better understanding of risk factors for unstable housing can help social services, health, and child welfare professionals prevent unstable housing experiences during childhood and adolescence and possibly curtail subsequent risk behaviors. --- Methods --- --- Procedures Assessments were conducted at an urban research university. Institutional approvals were obtained from the university and juvenile court. After assent for adolescents ˂ 18 years old, and consent for the adolescents ≥ 18 years old and caregivers were obtained, the caregivers and adolescents completed questionnaires and tasks during a 4-hour protocol. The protocol for data collection was similar for each of the four time points. Both the child and caregiver were paid for their participation according to the National Institutes of Health standard compensation rate for healthy volunteers. --- Measures Demographics and caregiver type.-Caregivers provided information at T1 about the child's birth date, sex, ethnicity, and race. Birthdate was used to calculate the youth's age at the T4 assessment. Although the CWD identified the child's ethnicity/race as Latino, Black, or White , caregivers included multi-ethnic or bi-racial as a category. Caregiver type for the sample was dichotomized as youth who reported having a birth parent or adoptive parent as their primary caregiver at all four time points versus youth who reported living with a nonparent caregiver at any time point. We did not have foster care status, therefore we could only categorize caregiver type. At T4, 15.74% of the maltreated youth were not living with any caregiver. When no caregiver was listed at T4, we used caregiver information from Times 1, 2, and 3. Family income was reported by caregivers at all four time points. These categories include $14,999 and under, $15,000-$29,999, $30,000-$59,999, and $60,000 and above, and categories were assigned a number in ascending order. Income category was averaged across all time points. If no caregiver was listed at T4, this time point was excluded from the averaged income. Employment information was only collected from the caregiver who attended the assessment, thus we did not have a complete picture of employment of all caregivers in the household and employment status was not included. Maltreatment Classification.-Research assistants abstracted information from child welfare case records to classify the types of maltreatment experienced . Categories included neglect , emotional abuse , physical abuse , and sexual abuse . Definitions were derived from the Department of Children and Family Services . Four separate maltreatment variables indicated the presence or absence of that particular type of maltreatment and 76% experienced more than one type. The maltreatment classification was determined at T1 after abstraction of the case files at T1 and previous to T1. Maltreatment reports subsequent to T1 were not collected. Maltreatment type was dichotomized into those who had experienced physical and/or sexual abuse and those who had experienced neglect and/or emotional maltreatment but not physical or sexual abuse. We chose this dichotomization because physical abuse and sexual abuse were the maltreatment types most associated with experiencing unstable housing. Measures in the following sections came from data collected at T4. Residence number and type.-Youth completed a household stability questionnaire designed for this study to measure lifetime residence history only at T4. In this intervieweradministered survey, youth were asked to list their first residence and sequentially account for all residences until the present. Based on this self-report, the number of residences was totaled by counting all residences prior to the year before T4 . The household stability questionnaire also asked youth to list the type of housing for each residence . Type of housing was classified as either stable or unstable housing. For this analysis, we used the categories of unstable housing that policy-makers and researchers have used. . Unstable housing categories in this study included homeless, car, transition housing for foster youth, juvenile hall, jail, prison, camp, youth authority, group home, shelter, church, and motel. Only incidents of unstable housing identified prior to the year before T4 data collection were included . Stable housing categories included home, apartment, mobile home, military housing, job corps dorm, and college residence. Two variables were created for unstable housing: 1) number of incidents of unstable housing and 2) at least one incident of unstable housing. Delinquency.-Youth completed the Adolescent Delinquency Questionnaire via computer to protect participant confidentiality. The present study combined two scales from the original questionnaire: person offenses and property offenses . All questions had six possible answers: 0, 1, 2, 3, 4, and 5 or more times during the previous 12 months. Scores had a possible range of 0 to 110 on the combined delinquency scale. Marijuana use.-Adolescents also reported their own substance use on the Adolescent Delinquency Questionnaire. Marijuana use was taken from one question that asked about the frequency of marijuana or hashish use during the previous 12 months. Potential responses included 0, 1, 2, 3, 4, and 5 or more times. All analyses were conducted using SAS 9.4. Both unstable housing variables were tested for all analyses. The first variable was not significant in our models, and thus, the dichotomous variable was used for all reported analyses. Prior to modeling, we tested the correlations between of all study variables. Logistic regression was used to predict the odds of having an incident of unstable housing 1 year prior to T4. Predictors included demographics , type of caregiver, average family income category, physical and/or sexual abuse, and number of residences. Race/ethnicity was dichotomized into each racial/ethnic group reported versus remaining youth for all analyses. Ultimately, being Latino was the only significant predictor of unstable housing, and thus we used this dichotomization for all analyses. Linear regression was used to examine the housing predictors of delinquency and marijuana use, separately. We tested two housing predictors, an incident of unstable housing and number of residences. Since the measurement of delinquency behaviors included the year before Time 4, we limited the housing predictors to a year prior to data collection at T4. Covariates included age, sex, being Latino, type of caregiver, average family income category, and a history of physical and/or sexual abuse. --- Results --- Sample characteristics Table 1 reports demographic information for the participants and means of the outcome variables for the total sample. The final sample was predominantly Latino or Black , with a fairly equal representation of sex . The average age was slightly more than 18 years old . The average number of residences was 5.6 , and 17.1% of the sample had experienced unstable housing at least once. The breakdown for incidents of unstable housing was as follows: group homes , juvenile hall , shelter , motel , homeless , car , youth authority , and transition housing for foster youth . The correlations between all study variables are shown in Table 2. --- Model predicting unstable housing For the logistic regression predicting any incident of unstable housing, significant predictors were being Latino and number of residences . Latinos were less likely to have an incident of unstable housing compared to non-Latinos. When ethnicities other than Latino were included in the model, none were statistically significant. In terms of the number of residences, every additional residence increased the odds of having an incident of unstable housing by 1.5 times. --- Models predicting delinquency and marijuana use The results for the final model predicting delinquency and marijuana use separately are presented in Table 4. One or more incidents of unstable housing predicted higher levels of delinquency . Of the covariates, being male significantly predicted higher levels of delinquency and lower average family income predicted higher levels of delinquency . The results for the final model predicting marijuana use are also presented in Table 4. An incident of unstable housing did not predict marijuana use. Two of the covariates were related to marijuana use: physical and/or sexual abuse predicted more frequent marijuana use and male sex predicted more marijuana use . --- Discussion This study examined whether demographic, family, and maltreatment characteristics were associated with an unstable housing experience for a child welfare-affiliated population and whether an incident of unstable housing predicted delinquency or marijuana use. We found that youth who had more lifetime residences and were not Latino were more likely to have an incident of unstable housing. A previous analysis which included both child welfareaffiliated and comparison adolescents found that having more lifetime residences was associated with person offense delinquency , whereas in the current analysis we found that living in an unstable residence was also an important predictor of delinquency for child welfare-affiliated adolescents when controlling for number of residences. For this analysis, we only used the child welfareaffiliated sample because the comparison sample did not have enough variability in unstable housing incidents. The number of incidents of unstable housing did not increase the chances of an adolescent reporting delinquency, but ever having an unstable housing incident did. Unstable housing did not predict marijuana use. This study illustrates the importance of preventing one or more unstable housing experience for child welfare-affiliated youth during their childhood or adolescence since an incident of unstable housing predicted delinquent behaviors. Latino youth were less likely to have an incident of unstable housing. Latino youth might have had larger families, lived in more multigenerational houses, and had more extended family members who provided housing . The average size of a Latino family is larger than the national average . Swartz concluded that Black and Latino families are more likely to provide practical support and housing support than White families, although in this study Latino youth were the only group less likely to have an incident of unstable housing. It could be that many of the Latino youth in this study came from immigrant families. In 2014, California had 5.5 million foreign-born immigrants from Latin America . Immigrant families tend to pool resources to make sure that their families have social and economic support . Another possible explanation for the decreased likelihood of Latino youth to report unstable housing is the concept of familism, which is common in Latino families. Familism can be viewed as a protective factor and includes the idea that family cohesion, extended family networks, and high social support may mitigate against untoward outcomes for their youth, including unstable housing . Not surprisingly, in this study youth with more lifetime residences were more likely to have an unstable housing experience. The relationship between more residences and unstable housing was also found in a study of women with recent criminal justice history where multiple residences were more common among those in unstable housing than those in stable housing . When conceptualized broadly, unstable housing begins with high residential instability and financial distress which can lead to unstable home environments such as doubling-up, living in hotels, and ultimately homelessness . Even for youth in foster care, moving to a neighborhood where there is more residential instability increases the risk of delinquency . Most of the research about child welfare involvement and unstable housing is specifically about foster care . Unfortunately, the foster care status of the study population was not known, but we did have information about types of caregivers. As expected, our findings indicated that having ever lived with a nonparent caregiver did not predict an incident of unstable housing compared with always having lived with birth parents or adoptive parents. Therefore, caregiver type was not a correlate of unstable housing and youth who ever lived with relatives or foster unrelated caregivers were no more likely to have an incident of unstable housing than youth who always lived with their parents. Unstable housing predicted delinquency but not marijuana use. The positive association between unstable housing and delinquency is supported by other studies . Thus, a housing career that includes even one unstable residence during childhood or adolescence is a predictor of delinquency even when controlling for number of residences. Our unstable housing category included jail and youth authority facilities, and 12 of our participants experienced at least one incident of incarceration in addition to other forms of unstable housing. In order to test whether including incarcerated participants in the unstable housing category affected the delinquency reports, we performed a post-hoc analysis examining if incarceration impacted delinquency. No statistically significant difference occurred in delinquency scores between those who were ever incarcerated and those who were never incarcerated. All of the families in this study lived in similar low-income communities, thus community influence rather than housing influence might have been important in marijuana use. For example, researchers have found that youth living in deteriorating neighborhoods were more likely to use marijuana two years after high school than adolescents living in always-good neighborhoods . --- Limitations This study has restricted generalizability because the study population was primarily minority and urban. The adolescents self-reported both their number of lifetime residences and the type of housing, but their recollections may have not been completely accurate. The adolescents' recollections of time spent in each residence were fairly precise for recent residences, thus we were able to determine which residence changes and unstable housing experiences occurred within the last year. Recollections of time spent in each residence prior to the last year were imperfect, therefore we could not determine the length of time spent in each type of residence. Participants did not indicate whether their unstable housing experience occurred with or without their family. Adolescents in this study did not report their sexual orientation, and thus we could not explore whether a known factor for experiencing unstable housing, LGBTQ status , was a risk factor. Social expectations might have affected the adolescents' reports of delinquency behaviors and marijuana use. Although information about other substances was included in the survey, the frequencies of use of substances other than marijuana were not high enough to be tested. Adolescents reported the type of caregiver they were living with but did not report whether they were in formal foster care. The inability to determine the foster care status of participants made it more difficult to compare our findings to studies that examined unstable housing after or between foster care placements. Additionally, we did not have human subjects' approval to abstract child welfare records during the course of the study, thus we only had maltreatment data up until their enrollment in the study . --- Conclusions This study found that even one incident of unstable housing during adolescence was associated with subsequent delinquency in child welfare-affiliated youth. Unstable housing, including homelessness, during childhood and adolescence is receiving more attention nationally. Housing is one of the social determinants of health and supporting a secure home environment is one strategy to ensure better health and educational opportunities for youth . Obtaining a housing history for children who are part of the child welfare system could provide important information for the development of interventions for these populations . The federal government has developed a voucher system to support families with rental assistance in the private market, although only one fourth of eligible families are receiving vouchers . The Office of the Assistant Secretary for Planning and Evaluation in the U.S. Department of Health and Human Services commissioned an evaluation of 14 family homelessness prevention and treatment programs that combined human services with housing supports such as housing vouchers . In the evaluation, no programs reported that they targeted youth or families with high residential mobility or coordinated with CWDs. Some youth in this study who had an unstable housing experience might not benefit from family programs because they were already not living with their families. Some of the most promising models for decreasing youth homelessness are rapid rehousing programs for youth and young adults, which include moving youth quickly into supportive housing, providing financial assistance, and offering developmentally appropriate case management . These programs also need to target youth with high residential mobility and be connected to CWDs. It is especially important that housing programs target both child welfare-affiliated maltreated youth who remain at home with their parents as well as those who have non-parent caregivers. More research is needed to identify what types of events within the youths' home environment preceded an unstable housing experience. In addition, comparisons between unstable housing experiences in youth with a foster care experience and maltreated youth who remained in their home of origin are needed. b 0 = no incidents of unstable housing, 1 = at least one incident of unstable housing. c Among those with an incident of unstable housing. d Response options were 0, 1, 2, 3, 4, and 5 or more times during previous 12 months; 22 items; score range of 0-110. e Response options were 0, 1, 2, 3, 4, and 5 or more times during previous 12 months; one item; score range of 0-5. --- Table 2 Pearson correlation coefficients . Logistic regression predicting unstable housing status a . Categories include $14,999 and under, $15,000-$29,999, $30,000-$59,999, and $60,000 and above. f 0 = neglect or emotional maltreatment but not physical or sexual abuse, 1 = physical and/or sexual abuse with or without neglect or emotional maltreatment. --- Table 4 Linear regression predicting delinquency and marijuana use . Categories include $14,999 and under, $15,000-$29,999, $30,000-$59,999, and $60,000 and above. e 0 = neglect or emotional maltreatment but not physical or sexual abuse, 1 = physical and/or sexual abuse with or without neglect or emotional maltreatment. f 0 = no incidents of unstable housing, 1 = at least one incident of unstable housing. * p < .05. ** p < .01. --- Delinquency ---

The study examined whether youth demographics, family factors, and maltreatment type were related to unstable housing and whether unstable housing predicted delinquency and marijuana use. Participants included 216 child welfare-affiliated adolescents (mean age = 18.2 years). Youth with more lifetime residences were more likely to experience unstable housing although Latino youth (compared to White, Black, or multiethnic/biracial) were less likely to experience unstable housing. Unstable housing was associated with subsequent delinquency. Caregiver type (parent vs. relative/unrelated caregiver) was not related to unstable housing, thus homelessness prevention programs should include youth who remain with their parents and those with non-parent caregivers.

I. INTRODUCTION On November 16, 2016, some videos went viral on some social media of a young boy, beaten to a state of unconsciousness and set aflame at Alafia-Badagry, Lagos State, Nigeria. From the voices heard in the video, it was alleged that the boy stole a mobile phone from a passerby . It was also alleged that the boy wanted to stab with a knife the owner of the mobile phone before he was apprehended by the local people, beaten and burnt to death in public view. This Manuscript received July 12, 2018; revised March 7, 2019. doi:10.12720/jcm. 14.4.324-334 was purely a matter of jungle justice which is seriously frowned at by the Nigerian legal system. However, the most disturbing aspect of the event was the liberty and conscienceless manner with which one of the onlookers videoed and photographed this barbaric act, perhaps for the purpose of witnessing to the event through online media. The second scenario is the case of Odunfa versus Okepopo street fight which video also made waves on social media. It is the case of an annual Adakeja masquerade festival that turned bloody in the Adeniji Adele area of Lagos Island on October 19, 2016. The supporters of the masquerader clashed with some youths loyal to the local traditional ruler. A teenager, Charles Igbinovia, was run down, mulled over, and butchered by the supporters of the masquerader. One other man was also killed. Several others were reportedly wounded. By October 20, videos of some scenes of the battle, particularly the butchering of the young boy, flooded social media. Here, too, the most amazing aspect of the event was the presence, perhaps, of alone citizen spontaneously photographing and videoing the tragic events, using a mobile device, in order to bear witness to the conflict. In both cases, while the visual images are significant as each has a story to tell, the contradictions inherent in their very productions problematizes the duty of care the practice requires towards victims of tragic events. The interrogation of this paper is not on the realistic nature of these footages, but more specifically on the social responsibility inherent in and the moral dilemma sometimes faced by onlookers, who also double as citizen reporters, in the exercise of visual witnessing. --- II. A NARRATIVE OF WITNESSING AND TESTIMONY Though there is now a multiplicity of definitions of the term 'witnessing', in most recent times the concept "has emerged as a way to describe how digital technologies are transforming the capacity to bear witness, encouraging a number of productive lines of investigation" [1]. Frosh and Pinchevski [2:20-21] admit that media witnessing simply captures something that is central, not only to the practices of contemporary media, but also significant to the aesthetics, ethics, and politics of representation. The authors argue that media witnessing is principally concerned with "the systematic and ongoing reporting of the experiences and realities of distant others to mass audiences" [2:1]. It implies the simultaneous configuration and conflation of three distinctive things: "the appearance of witnesses in media reports; the possibility of media themselves bearing witness; and the positioning of media audiences as witnesses to depicted events" [2:1]. For example, the visual reportage of the Badagry boy's saga and the Odunfa-Okepopo conflict may be seen to, simultaneously, depict witnesses to the events, bear witness to those events, and turn bystanders into witnesses of the events However, at the core of most cases of media witnessing is what Frosh and Pinchevski term as the "crisis of witnessing" [2:3]whereby victims, because of forgetfulness or speechlessness resulting from the overwhelming nature of the experience, or the sudden death of the victims themselves, are unable to report directly about their traumatic experiences. In this case, the bystanders, convinced of the perpetrators' attempt to extinguish the internal witness of their victim, could build a discourse that bears out the traumatic processes in the form of documentary videos or photographs, so as to mediate on behalf of the victim. In this regard, the authors argue, the medium of video could be vital in documenting the personal memories of witnesses of traumatic events, so that what the perpetrators attempt to bury is given new visibility [2:4]. Frosh and Pinchevski's [2] study is vital to the understanding of the concern of this paper on a number of points: Firstly, the concept of 'witnessing', articulated by the authors, could be stretched to aid the understanding of a wide range of visual communication issues in relation to the representation of traumatic experiences through citizenship photojournalism in Lagos-Nigeria. Secondly, the word can enable us to see all forms of media practice as a kind of testimony to crisis situations, enabled especially by changes in technologies . Thirdly, 'media witnessing' offers, conceptually, a demonstration of the connection between 'embedded journalism' and 'citizen journalism', namely, the attempt by professional reporters and ordinary citizens to put an experience into a visual language for the benefit of those who were not there and for the purpose of mediation in favour of the victim of that experience. Fourthly, 'witnessing' provides a useful conceptual framework for connecting the burden of care towards victims of atrocious events and the imperative of social responsibility contract exercised through visual reportage. --- III. METHODOLOGY OF STUDY The study drew on the benefits of 'mixed methods' used in a qualitative way. The dominant method was content analysis. The complementary method was oral interview. The latter was meant to enable the authors respond to some issues that could not be addressed through content analysis. Four respondents in all were selected for oral interviews: Two were chosen from The Punch Newspapersone a senior photojournalists and the other an acting news editor with The Sunday Punch ; the other two were senior academics from the University of Uyo and the University of Lagos . About thirtyminutes telephone conversation was held with each interviewee in January 2018. Their views were recorded, transcribed and analysed. The samples for content analysis were four different footages , about the killing and setting aflame of the boy in Badagry as well as the Odunfa-Okepopo street conflict in Lagos Island. The primary foci were the manifest contents of the videos circulated online and their accompanying text commentaries. The focus of the videos analysis was on frequency measurement, rather than on intensity. The text commentaries were drawn from only one online source . The measureable units were tied around predetermined thematic categories, such as news value, age value , social status , emotional moods , voice , presence of security, environment of perpetration, predetermined outcome , presence of care, determination of perpetrators, and the mobile technology type. The conceptual units of analysis chosen were meant to address different issues in relation to the four sets of research questions proposed at the beginning of the paper. Below is a representation of findings in tabular format and their frequency analysis : Here, we attempt to make sense of the data, using the four research questions as guides. Answers to some of the questions are footage-based; others are based on data from indepth interviews or both: --- A. The Moral Agency of Citizen Photojournalists The moral agency of citizen visual reporters and the cultural activity that defines and underlines that agency were tested by four elements, namely, the presence of citizen witness, the characteristics of the witnesses/perpetrators, the news value of the video reports, and the respect for ethical standards in the exercise of the moral agency of citizenship. Evidence indicates a high prominence of citizen witnesses in the two sets of videos. Citizenship reporting or what Allan [1] describes as "accidental journalism" or "first-person reporting", has become a central and crucial aspect of our media and communication landscapes. As a special genre of the communication practice distinguishable from the mainstream, citizenship witnessing, in its diversified forms and formats, constitutes "the spontaneous actions of ordinary people compelled to adopt the role of news reporter in order to bear witness to human suffering" [3], [3:4], particularly during a time of crisis or disaster when they happen to be present on the scene. This cultural practice has often been associated with the imperative of witnessing because of its orientation towards the provision of intervention to mediate in the sharp pull resulting from conflicts and atrocities. In all the visual cases, the citizen-photojournalists are part and parcel of the bystanders. They are personally present as eyewitnesses at the scenes of the crimes against their young victims. Allan's [1] tripartite witnessing modalities , however, reveals the complexity of responses possible in representing the different positioning of the witnesses who recorded the atrocities in Badagry and Lagos Island. One of this paper's greatest challenges, however, is that of being able to state in explicit terms which of the modalities characterize our videographers of interest. As a result of this difficulty, this paper prefers to refer to these anonymous videographers as 'bystanderparticipants' or 'citizen-witnesses'. In this regard, Allan's warning that the modalities represented should be seen, not just as social contingencies, but as imagined modalities for the purpose of analytical clarity should be seen to apply in this context; that is, if we must avoid the problematic that could arise if these categories were to be understood too literally [1: 175]. Again, it must be noted that all the footages analysed were taken by the 'bystanders' at the spur of the moment. While in the Badagry case, the person recording stands almost shoulder-by-shoulder with the perpetrators, in the case of the Lagos Island fighting, the shooting is made from the first or second floor of a storey building. But in all the videos, the citizen witness has made himself or herself conspicuously absent from the shots. His or her presence is confirmed only by camera presence and activity. The purposes for the generations of the footages are not explicit from the videos themselves. But circumstantial evidence tend to suggest witnessing for the purpose of information, activism, and mediation as the principal objectives. While the victims are by all standards minors , the perpetrators are largely young adults whose intellectual grasps of sociocultural and legal realities would have been expected to be higher. The perpetrators are also largely of 'low' social status; only a limited number could be considered to be of 'middle class' category. How the perpetrators' "low" social class could impact on their understanding of the need of their immediate sociocultural environments and their ability to control the floods of emotions rising from within them in response to the perceived criminal offences of the victims is not explicit in the video, but could only be 'read' from their actions or inactions in the videos. Also, visual evidence shows that the contents of all the videos have high news values. The contents are meant to tell stories and communicate information to the public. Reliance on Tulloch and Blood's [4] value specification could highlight other forces, outside news consideration, that could shape the integrity of these videos. For instance, there are their resourcefulness in demonstrating significant moments in history; their foregrounding of the dominant social ideology of a particular community; expressing the social conditions that possibly generated the events; underlining the kind of rhetoric that gave inspiration to its uprising; bearing witness to the wounds of atrocities; as well as the processes needed to attain liberation. The ideological dimension, in particular, may be tied with the acceptability of the value of jungle justice as a sociocultural practice in righting wrongs in a local community in Nigeria. However, the communicative, modelling and reformative potentials of these footages could not be far removed from the effects of their emotive powers and their ability to expose, interpret and correct what ordinarily would have been hidden from public discourses and public policy decision-making processes. One other significant value is that these videos could be depended upon by mainstream and other alternative news establishments to develop story ideas and for activism against all forms of atrocious public engagements by young people. Chouliaraki's [5] "visual politics of journalism" can also be extended to the understanding of certain issues about these visuals. The concept, according to the author, is crucial for thematizing visual images as "the symbolic definition of world events", whereby these images function to make visible and render intelligible "places and people otherwise not available to us through immediate experience" [5: 520]. It is this symbolization role that also defines what is at stake in citizen visual reflexivity; that is, the awareness of how the choices of certain visual images of conflicts and their associated languages could bear implications for the making of a community and the "dynamics of collective belonging" [5: 520-1]. As valid as Chouliaraki's arguments are, it must also be stated that the intrinsic characteristic of these citizenship visual reportage, as narratives of witnessing, is not anchored on their symbolization role alone, but also on the value of the citizen bystanders 'being there' on the ground. A number of scholars [1], [2], [6], have placed strong emphasis on this 'eye witnessing' aspect of the practice. Studies [1], [6], further confirm that it is precisely this eye witnessing imperative that could, in the main, be drawn upon to negotiate conceptually the distinctions between amateur images and the more professionally made images meant for relaying hard facts [1]. Though newsworthy in all cases and reflective of other values, it is worth noting that the way and manner by which the 'raw' scenes of the events in Badagry and Lagos Island are foregrounded in the two sets of the inter-related videos seriously breach the principles of 'good taste and decency' in the display of images of the suffering and death of others. The need to protect the public from any emotional traumas that could result from viewing these gruesome images of spectacles has always been at the baseline of the good taste and decency policy [5]. The representations, in their 'raw' states, of the continuous brutal cutting of the 19-year old boy with a machete in Lagos Island and the repeated stoning of the 16 year old boy in Badagry to make him unconscious are seriously offensive and disturbing. While the visual depictions of these spectacles in their raw states reveal what truly happened at the scenes as well as the evil tendencies of the perpetrators of these crimes, these 'raw' depictions of social realities does not show the citizen photojournalists' sensitivity to the emotional shock that could possibly result from watching them. Yet, drawing insight from Chouliaraki's [5: 522] tripartite regulatory regimes -the economy of taste and decency ; the economy of 'display' of gruesome images ; and the economy of 'witnessing' -in relation to visual videography, the only defence for such breaches of expected journalistic standards by the citizens photojournalists, as is often the case in other videos of conflicts and traumas, is the imperative of mediatory witnessing and testimony. It is the economy of witnessing, with its central drive towards advocacy and to prove the authenticity of the reported events that could be drawn upon to balance out the ethical limits inherent in the foregrounding and display of these events in their 'raw' states. While this paper disapproves of the resulting ethical breaches, it does however accept that the validity of such breaches be grounded on nothing other than the need to seek quick interventions or mediations in favour of would-be victims of similar events in the future. In this regard, the central values of the 'raw' images of the gruesome murders of the two boys in Badagry and Lagos Island, respectively, may be more meaningfully placed, not merely against their aesthetical and moral regimes, but more so against their testimonial and mediatory regulatory regimes. --- B. Technological Structure and Care for the Victims How the mobile technological structure enables and/or impede the exercise of moral agency by the citizen visual reporters, in terms of care for the victims of the atrocious events, was tested by identifying the type of recording device used; the nature of the recording process; the presence of care for the victims; the presence of security operatives; the environment of the crime; the voice of plea of the victims; and the mood of the perpetrators in response to the plea. Visual evidence suggests that the recordings of the spectacles were all made with mobile devices; possibly mobile phones. But it is not clear from all the videos the mobile phone types with which the footages were made. Evidence of lack of clarity of the mobile device types is high. Studies [7], [8], however, show that mobile phones and their accompanying recording devices have become a taken-for-granted part of contemporary Nigeria society. While the first sets of mobile phones emerged in Africa in the 1980s, the Nigerian government introduced mobile phones into the country as late as 1990s. However, mobile handset capable of doing a number of other functions outside voice calls and meant for average consumers gradually appeared only in 2002. By late 2005, mobile devices with different designs, functionalities, and affordability were fast becoming the norm. Adomi [8] admits that an increasing number of Nigerians are now taking advantage of the potentials of the mobile communication systems. Apart from increasing access to the use of telephone services, it has brought about plurality in citizenship news coverage. In particular, their usages now make the management of personal photographic and video archives possible as well as the sharing of some of these visual images across global digital platforms. The prioritization of technologies over and against the human factor has, however, remained a point of critical consideration in scholarship, with the increasing distancing of the debates from the trappings of technological determinism [9], [10]. The pull of the debates has changed to an increasing recognition that technologies by themselves are inert. Their power to communicate and enable change through witnessing is rooted in the way they have been taken up by users. This paper agrees with this scholarly position. Thus, if anything meaningful is be understood about the structural and witnessing potential of the mobile devices used in the recording of the atrocious events, that understanding should be driven the concept of interdependence between the technological and a plethora of human factors, such as human social cultures, human vision and imagination [11]. In this respect, the enabling and impeding qualities of the mobile technologies, used in the recording of the events in Badagry and Lagos Island, cannot be tied with the technology alone. The human judgements, the ways the frames are made, how the shots assist the sociopolitical movement of the time, competencies and other factors that impinged on the uses of the technologies, also count. Cammaerts's [12] exploration of "technologies of selfmediation", for example, offers other significant perspective for connecting the values of the mobile technologies with citizenship visual witnessing. His specific emphasis is on the "affordances" and "constraints" of these media types. The enabling potentials include disclosure , examination , and remembrance . The impeding elements, in contrast, are tied around the issues of "access" to the technologies and news contents and the "technical skills" in relation to their usages . The significance of these factors must, however, be evaluated both at the point of production and reception [12]. These specificities are particularly useful for understanding how the communication abilities of the mobile devices used in recording the events in Lagos-Nigeria could be seen to enable or impede the moral agency of the citizen photojournalists in relation to the narratives of the killing of the two boys. Again, while the problematic of "access" are now more than ever tied around the availability of technological innovations, telecommunication services, government's regulatory policies, as well as the willingness of mainstream media to report on the frames from citizen reporters, "competency" is not regarded as a prerogative of citizenship visual practice; it is only a defining quality of professionalism [13]; [19]; 20]. The unprofessional nature of the videography is clearly self-evident in all the visual cases from Lagos: the camera movements are generally unsteady; the shots, too, are very shaky in response to the quick swings of the mobile phones. The shots are also lacking in camera angle variations. Most of the shots are wide angle, long shots. There are some close-up and extreme close-up attempts in relation to the Badagry scenario. But only a few times do the cameraman attempts close-ups in relation to the Lagos Island conflict; thus, making it difficult for viewers to see at very close range the expression of pains and tears on the face of the 19-year old boy as his body is repeatedly cut with machetes. These technical elements clearly illustrative of the human elements and/or the unprofessional approaches associated with the making of these visuals. The lack of the expected technical skills in relation to the productions of these visuals, to a great extent, is illustrative of the experimental nature of the production processes themselves, which is often at the heart of citizenship amateurish videography. However, the potential of these footages in assisting the sociopolitical movement of the time is very much tied with the wrongful assumption that jungle justice could assist in the protection of local communities against the spread of violent crimes. This ideological reality is, however, not self-evident in any of the videos but can only be inferred as part of the self-reflexivity of their viewers. Again, evidence from the videos shows that care for the victims of the atrocities was absolutely lacking. There is no presence of the personnel of any of the security agencies. The foci of the bystander-videographers are explicitly on the events before them, not the need to rescue the victims. The only evidence of care is the lone voice of an elderly figure in some of the videos pleading for the life of the victims to be spared. For example, an elderly person, dressed in a white flowing gown, is seen in the first Lagos Island video trying to intervene; but was pushed aside and stones hauled on the head of the 19-year old victim who then lay still. Duty of care is a universal principle. This principle is also implicated in the ways technologies are used. That an individual may owe a duty of care to another to ensure that they do not suffer any unreasonable harm or loss is particularly a central requirement of divine revelation, moral philosophy, and even legal policies. For example, the Aristotelian ethical tradition, which is substantially grounded on the evaluation of human character and virtue, upholds that having a virtuous inner disposition will also involve being moved to act in accordance with that disposition [14]. Realizing, for example, that care is the appropriate response to a situation and feeling appropriately disposed towards caring will also lead to a corresponding act of kindness towards others. Also, the Nigerian legal systems [15], with their roots in acceptable philosophical principles and the English legal systems, in some areas, absorbed and expressed the requirement to care for all, particularly victims of potential crimes and civil offences. While the Constitution, on the one breath, places the duty of care towards all citizens in the hands of government , it recognizes, on the other breath, the sovereign right and power of the citizens to shape how government exercises this duty in a participatory manner . This oscillation between government and the citizens is a significant recognition of the kind of collaboration necessary for the actualization of the principle of care in its broadest sense. This sense of care is absolutely lacking in the videos about the treatment of the two young boys. Lack of care is further proved by the very uses of the mobile technologies themselves as well as the responses of the perpetrators to the 'voice' of plea by the victims. The uses of the technologies in both the Badagry and Lagos Island events are not in connection with voice calls or even SMSs to seek out help for the victims of the atrocities. Rather the uses are tied around visual image generations to bear witness to the atrocities. While the visual images themselves are iconic, at least for forensic purpose, the lack of specific focus on voice calls or even SMS could be taken as indicative of a possible presence of an unresolved moral dilemma in the judgement of the citizen-witnesses themselves. The citizen photojournalists appear themselves caught up in the heat of the crowd's passion. Also, in all the videos, there is no evidence of victim's explicit affirmation of committing a crime. However, a 'voice' of plea for mercy and pardon is moderately evident in the two videos about the Badagry spectacle. A speech conveyed through 'pleading-silence' is not completely absent either. In the Badagry videos, when speech eluded the boy, he mimed for clemency with the wave of a hand. This is foregrounded against the strong voice of assertions, anger and promise of extermination of the victim by the crowd: "Your life will end today" ; "Knack-am well" ; "Bring the fuel" . The voice of the victim of the Lagos Island fighting is completely muted, perhaps due to the distance of the mobile device from the spot of the mutilation. But the voice of the perpetrators, which also provides tints of their moods and intention, is highly audible. Apart from the materiality of 'voice' in enabling one give an account of himself, voice can also become expressive of a distinctive political perspective of an imagined community that needs to be addressed, particularly a community with entrenched structures of inequality and legal breaches. Seen from these perspectives, this paper agrees with Couldry's [16] position that 'voice', expressed through the instrumentality of citizenship media, is not just a political tool but can also become a force for overcoming the dominance of violence in a community and for bringing about a radical behavioural and cultural transformation within that community. In this regard, the verbal and non-verbal gestures of plea by the victims could be taken to speak in favour of greater need for empathy towards the victims of atrocious events in local communities, even when there is the burning need to redeem such communities from the stench of criminality. While the overall intention of the perpetrators is to rid the community of crimes, the killing of these young boys is generally lacking in ethical value. This position is more specifically captured in the accompanying text commentaries which either denounce as evil or approve as good the crowd's actions against the victims. There is also an evidence of neutral ethical positioning. For example, out of the 47 commentaries that accompanied the Badagry spectacle , a total of two are in affirmation of the atrocious events; 37 disapprove of them; only eight are neutral. While the affirmation frequency is medium, the denunciation range is high. This is a sign that majority in the community do not support such unethical conducts against any victim. It is also a statement of the ethical position of the broader global community against jungle justice. Chouliaraki [5], for example, argues that the foregrounding of such subtle approvals or denunciations of barbarism displayed as linguistic accompaniments to visual images, is valuable as an ethical strategy in journalism for managing people's unique mode of encounter with distant but potentially traumatic events. In view of these findings, what then is the connection between the mobile technologies used and the moral agency of the citizen-witnesses? The technologies as enablers have assisted, for example, in capturing the 'voices' of the victims challenging the community to a growing sense of pardon. The citizen reporters have also found their 'voice' through the recording of the misery and suffering of the victims, the contents of which could become catalysts for social actions. Evidence of the impeding factor lies more in the unprofessional ways the videos were made. Other enabling or impeding factors are external to the mobile devices. They are directly tied with the perception of the actions of the victims by the bystander-participants, as well as their 'hypothetical' conception of the state of the local communities. The technologies could also become enablers if their previously recorded contents either become motivating forces for future infringements on the rights of other citizens accused to have committed similar crimes or serve as catalysts to secure care for would-be victims or help local communities seek out and adopt new values that are ethically acceptable in their attempt to secure a crime-free society. --- C. The Ethical Tragedy in Citizenship Visual Practice Identifying what constitutes the ethical tragedy in the citizenship visual practice and locating where the tragedy resides are based on this paper's hypothetical assumption that there is a level of hypocrisy in the practice of citizenship visual reporting in contemporary times. The answers are sought, not only from within the videos themselves, but also from outside the footages through oral conversations. Though the word 'tragedy' is often used to represent a very sad situation, it is taken within the context of this paper to refer to the shocking condition whereby citizenship visual practice is increasingly being valued primarily in terms of its news values and less for its inherent humane requirement. From the visual perspective, none of the videos confirm the presence of security personnel or any attempt by the bystanders to seek out police help to protect the lives of the victims. In this regard, there is a clear evidence of a serious negligence of the requisite duty of care towards victims of tragic public situations by the bystander-participants. This negligent tendency, because it undermines some of the core values of citizenship reporting , it is argued, could also function to reveal the hypocrisy underlying citizenship photojournalistic practice today. Outside the visual contents, the occasional flawed nature of the techniques involved in the formation processes of these visuals, though often overlooked as consistent with amateur videographic practice, could also be viewed as tragic, especially when the visual claims made are carefully manipulated to tell only one side of the story, neglecting the side that could have favoured a more sympathetic association of viewers with the wrongful persecutions or destructions of victims of alleged crimes. From oral interviews, respondents disagree as to whether the contradiction involved in citizenship visual practices is tragic. Those that see it as tragic are of the view that, normally, the citizen reporter would be expected to do his or her work generating stories and the neighbours who also witness the events to do theirs. It is, therefore, not advisable for a reporter to put himself in a harm's way by attempting to intervene directly on behalf of a victim of public attacks, to avoid becoming a victim himself. That is because the ability of the perpetrators to engage in any logical thinking is always very low in a moment like that. Those who see the situation as not tragic admit that most citizen reporters who report on such events do actually fulfil a moral responsibility that is tied around their duty to inform the public about such events. In this regard, they cannot be held to be negligent even when their passion is first directed towards news reporting alone. Other interviewees, in contrast, maintain that "the whole situation" of reporting on atrocity is tragic. The tragedy is not only in relation to the activity of the citizen reporter, it relates also to the action of the perpetrators. In fact, "it is the tragedy of humanity", which is tied around a number of indices, including the image of the local communities that tolerate such barbaric activities and the loss of life often involved in such events. The action of the citizen reporters, though done at the spur of the moment, becomes tragic when their coverage of the events does not roll out enough facts to inform the public that what had taken place was tragic and should never be allowed to happen again. These interviewees argue that "as a people, we are very much sensitized to see and believe that some of the public killings of innocent citizens are necessary. People have come to accept them as normal. This is tragic. The news reports themselves do not always raise the level of anger necessary for the larger population to react against these atrocious events. This too is tragic because it allows impunity to continue". All the respondents, however, acknowledge that there is always a dilemma inherent in the practice and which needs to be resolved on the spot. They argue that the first instinct of any photojournalist is to generate a story in order to let people know what is happening. Often, there is little consideration of how help could be sought for the victim. They maintain that it is the positioning of the reporter in relation to this social function that often determines his response type. The view of respondents and footage evidence notwithstanding, the tragic, it is argued, resides both within and outside the practice of citizenship reporting itself. And because they breach some basic ethical expectations that are grounded on the principles of care and compassion, the contradictory tendencies in citizenship visual reportage, should be seen to constitute an ethical tragedy for the practice and requires public criticism, that is, if change must be brought to bear on the practice at all levels: ideological, political, and pragmatic. --- D. Duty of Care and the Drive to Bare Witness How citizens who use mobile technologies could reconcile the duty of care for victims of tragic public incidences and the immediate concern to capture and represent events in ways that are fresh and immediate was tested from conversational and argumentative points of view. The interviewees all recognize the complexity involved in recording any public atrocious events. For them, it is the nature of the event or the situation the citizen reporter finds himself or herself that, to a great extent, determines his or her line of action. The interviewees maintain that ability to reconcile the two dutiesthe duty to care and to report eventswould depend largely on what the citizen reporter sees on the ground. It might not even be possible for a citizen reporter to talk and calm down an angry mob. In some cases, rushing in to record an atrocious event might not even be wise, as the mobile camera could be seized and destroyed by the mob. In other cases, the reporter could help by finding a way to connect with nearby mainstream media houses or security agencies to intervene. But generally, there is always a 40-30% chance that even if a call is put across the victim would be rescued, as the mob reaction is always very swift and not properly thought through. But where it is a case of a motor accident, for instance, it would be easier to know what to do firsthelp rescue the victims and then photograph the event later or do both simultaneously. But outside this, interviewees argue, it might be difficult to know what to do first. Where one is dealing with a mob reaction that could become volatile the response would always have to be personal. One interviewee's response was tied to the way news contents are ordered. Citizen reporters, he argues, should work to balance their stories to bring out, not just the objectivity, but more importantly the moral aspect. In other words, their news contents should be ordered to sensitize us on the values that are dear to our common humanity, such as respect for human life. Non-adoption of recognized legal procedures should be made to be seen as unacceptable. Another respondent argues that "citizen reporters need to exercise restraints in the kind of pictures they put out for public consumption, to avoid institutionalizing what ordinarily is condemnable. Again, when dealing with online media that is populated by people who have little consideration for ethics, what are important are our individual moral dispositions and our general attitude as a people that could work to reduce impunity in the way we handle issues involving human life". All said, it is the position of this paper that witnessing to trauma through the foregrounding of spectacles so as to mediate in them is a worthy objective. It is an expression of the social responsibility of every citizen. But equally significant is the need for witnesses to position themselves, first and foremost, as people with a responsibility to love others and to preserve life. While the confused state of the bystander-participants in a scene of public strive itself could create difficulties in making quick choices as to what to do first in such momentswhether to call for a rescue or to document the event or to challenge the injustices verbally -, it is generally the failures of the citizen witnesses at 'conflation' in relation to their expected positions and roles that might, perhaps, offer the tipping line for locating where the dilemma and the tragic in citizenship visual reporting sometimes resides. The avoidance of this tragedy, it is argued, rests more significantly on the willingness of the bystanderwitnesses to do the needful in terms of carefully balancing their social responsibility expectation when recording atrocious events with the need to care for the victims of such events. When evaluated, therefore, from the points of view of the requirements of divine revelation, moral philosophies and of legal prescriptions outlined above, careful balancing between the responsibility to bear witness through reportage and the need to protect life appears to be the necessary pragmatic path to follow by every reasonable person. Thus, beyond the broader applicability of the doctrine of care within the business and medical worlds [17], [18], [19], it is important for all privatized practitioners to understand that they also owe a duty of care to their fellow citizens. Citizen photojournalists, in particular, would need to accept that they, too, cannot just wash their hands off it and say they have no duty of care. To do so will amount to serious act of negligence. Building the awareness and making a conscious decision to be doubly responsible could help citizen videographers in bypassing the dilemma sometimes inherent in the practice, as well as help in reducing the tragic implications indifferent attitudes and behaviours could bear on the integrity of citizenship photojournalism itself. --- V. CONCLUSION The study focused on citizens' uses of mobile technologies in reporting atrocious social events and the ethical tragedy that arises in the course of that practice, caused by unresolved dilemmas or indifference on the parts of the citizen-participants. Additionally, the paper sought to understand the cultural activity that define the moral agency of citizen photojournalists, the functionality of mobile devices in relation to the moral agency of every citizen videographer, as well as how citizen photojournalists could reconcile the requisite moral duty of care with the need to video and represent tragic public events. While findings in relation to the issues were specific to the Nigerian experience, replete of similar problematic situations could be found across a number of other countries, making the research problem a universal problem that requires global research responses. Findings reveal, firstly, that while witnessing in and through the media is at the heart of the moral agency of the citizen photojournalists at the moment of the tragic events, mediation through information provision and activism for the purpose of practical reforms remains the ultimate objective. While the ethical imperative of journalistic visual representation of spectacles require that photojournalists exercise reasonable care in displaying such images to avoid offending people's taste and decency, evidence also suggest that 'raw' images of the events were displayed without much regard for their possible emotional impacts on viewers and that these 'raw' visual images could function, regardless of the regulatory breaches, to elicit intervention in favour of victims of the conflicts. Secondly, mobile technology by itself is inert. The enabling or impeding functionalities of the technologies are tied more directly with the manner of usages of these technologies by citizen photojournalists and with other human factors such as imagination and subjective viewpoints. Directly, the technology has provided 'voice' for the victims, the perpetrators and the citizen reporters enabling social actions for change. The only context within which the technologies could be seen to have an indirect enabling impact on the moral agency of the citizen visual reporters is where their recorded contents could become catalysts for future infringements on people's rights by other young adults under the guise of jungle justice. Thirdly, the tragic in relation to the moral agency of citizen photojournalists could be tied directly with the inability of the citizen reporters to reconcile, at the spur of the moment, the demands of their social responsibility contract and the religio-legal-moral requirements to protect life and to seek out care for victims of atrocious social events, using the voice call or SMS capability of their mobile devices. The duty of care the practice of citizen videography requires towards victims of tragic events and the 'sites' of the moral dilemmas sometimes faced by those who report these events have been logically argued in this paper. The 'tragic', it is argued, is in the negligent tendencies of the citizen visual reporters, manifested in the increasing loss of empathy towards the victims and the confusion often encountered by bystanderparticipants in deciding which role ought to be predominant: the recording of the events or the seeking for rescue for the victims. This negligent tendency is well documented in the two sets of videos from Badagry and Lagos Island and other instances. In all cases, putting out a quick voice call or SMS to security operatives was neglected by the citizen-reporters, while the desire to visually record the horrible events took greater hold. This approach to reporting, it is argued, constitutes nothing but a tragedy to citizenship photojournalism. It is 'tragic' for the people's journalism whose core values have been underlined, not only by the ethical values of objectivity and neutrality or the aesthetics of empirical truthfulness, but more so by empathetic and moral considerations. These are some of the key values that could be realized when the logic of care is given prime importance in the practice. The question begging for answer is: How do we make sense of this development, reconciling it with the central values of citizenship reportage? How can this inherent dilemma be avoided in the future? The position of this paper is very clear. There must be a conscious, reasonable and careful balancing between the call to social responsibility through media witnessing and the prerequisite moral duty to care for and protect the lives of victims of traumatic events. The acceptance of this demand must, however, go beyond the mere imperatives of sociocultural regulations or policy formulations [20], to the internalization of sets of acceptable moral values [19] or a conscious actualization of the Aristotelian ethical dictumact in every situation as any virtuous man should [14]. While we need citizen reporters to spontaneously generate newsworthy stories, we also need, above all else, those whose sense of moral responsibility in the alternative journalistic field is limitless. The adoption of this alternative approach could, at the long run, help minimize brutal killings resulting from jungle justices as well as open up broader social spaces for the needed legal processes to take their reasonable courses, in response to any kind of crime or social upheaval committed within our communities.

The paper considers the reportage of tragic events in Nigeria by ordinary citizens, using mobile phones and other digital devices. The focus is on the moral agency of citizenphotojournalists, the dilemma inherent in the exercise of that agency, the technological structure that enable/impede such agency, and the resulting ethical tragedy for citizenship photojournalism. The questions addressed are: On what cultural activity rests the moral agency of citizen-photojournalists in Nigeria? How does mobile technology enable or impede the exercise of that moral agency? How do the citizens who own and use mobile technology reconcile the duty of care for victims of tragic public incidences and the immediate concern to represent events in ways that are fresh and immediate? Does the failure to provide care for victims of tragic events in the course of citizenship reporting constitute an ethical tragedy for the practice? Two cases of citizenship visual reporting in relation to tragic events in Lagos-Nigeria are studied (the Badagry boy's saga and the Odunfa-Okepopo conflict). Mixed methods approach (content analysis and oral interview) is used in a qualitative way to generate data. The theory of media witnessing is used to ground the study. Findings show that in each case of citizen visual reporting, there is a serious negligence of the requisite duty of care towards victims of tragic public situations, resulting in an unconscious or deliberate undermining of some of the core values of citizenship journalism. Such negligence is also a revelation of the hypocrisy underlying the practice of citizenship photojournalism. Beyond social regulatory needs, the paper recommends careful balancing in practice responsibilities.

BACKGROUND According to the United Nations Population Division , in 2005 there were 191 million international migrants-representing about 3% of the world's population. Previous studies have found an association between the high mobility of migrants and an increased risk for HIV infection . Several factors have linked migration with increased HIV risk, including: separation from wives, family, and other support networks ; increased number of casual sex partners ; more money or higher socioeconomic status ; lack of access to social and health services ; limited knowledge about HIV/AIDS transmission and prevention ; and hardship and hostility while in the receiving country . However, the particular social, economic, psychological, and cultural processes that might explain how labor migration leads to increased HIV risk have not been well studied. With 80% of the country living below the poverty line and a GDP per capita of less than $290, the Central Asian Republic of Tajikistan is one of the poorest countries in the world . Faced with difficult economic challenges resulting from a 5-year civil war that resulted in over 60,000 deaths and widespread unemployment, an estimated 700,000 men leave Tajikistan each year in search of work, especially in Russia . Most Tajik migrants in Russia are seasonal workers in Moscow, carrying goods in bazaars , performing manual labor in construction sites, sweeping streets, and working in food services . Tajik labor migrants may engage in high-risk sexual behaviors such as having multiple concurrent sexual relationships . The migrants in Moscow also often use sex workers, and these sex workers have been documented to have an estimated HIV prevalence rate 30 to 120 times higher than that of the city's general population . Eighty-five percent of migrants aged 20 to 49 years old are married . The wives of migrants are at risk of being exposed to any illnesses that the migrants carry, including sexually transmitted infections such as HIV. Thus, labor migrants could serve as a bridging population and thus contribute to the acceleration of an epidemic in Tajikistan that has risen dramatically since 2000 . In May 2011, 3051 HIV cases were registered in Tajikistan, of which 145 were migrant workers . An increased number of HIV cases has also been observed among wives of male Tajik migrants following their return home . The available statistics are widely believed to underestimate the true rate of HIV infection in Tajikistan. In the field of HIV prevention, social psychology theory has been used to inform the design, implementation, and evaluation of behavioral interventions seeking to reduce HIV risk . Theory can help explain key and often complex psychological and social determinants of risk and preventive behavior, such as attitudes, perceived norms, and self-efficacy, defined as the belief that one has the skills and abilities necessary to perform a behavior in a variety of circumstances . With regard to labor migrants, social psychology theory has been utilized to inform HIV prevention and behavior change educational programs for labor migration under adverse and complex social conditions. For example, Campbell used social identity theory to guide qualitative studies that described how the construction of masculine identities among South African male mineworkers rendered them especially vulnerable to HIV. Campbell found that masculine identities and macho sexuality helped mineworkers to deal with the harsh day-to-day living and working conditions within the mines. She defined macho sexuality as a repertoire of insatiable sexuality, need for multiple sexual partners, and a manly desire for the pleasure of flesh-to-flesh sexual contact . Specifically, she delineated how difficult living and working conditions brought on a sense of "powerlessness" in the mineworkers, which accounted for their decreased self-efficacy regarding HIV protection. Campbell wrote, "for many migrants, deprived of key markers of masculinity on a day-today basis, frequent assertion of what are regarded as healthy and manly sexual urges could arguably serve to compensate for reduced opportunities for assertion of masculine identities in other contexts" . Prior studies of women living in different contexts have examined the concept of powerlessness in sexual health promotion for HIV prevention. For example, Tawil, Verster, and O'Reilly argued that addressing the powerlessness of women to protect themselves against HIV requires an "enabling approach" that creates circumstances facilitating HIV risk reduction. For example, one strategy may be to focus on improving women's access to resources so that they may be more financially independent from their male partners . Another study examined the prospects for community-based interventions to reduce HIV risk among female sex workers in Tamil Nadu, India, and suggested that these approaches were unlikely to succeed given the high degree of repression, taboo, social stigma, and fragmentation surrounding sex work in the city of Madras . This study concluded that individual level empowerment strategies were inadequate to address the complex social dimensions, such as economic structures, cultural traditions, gender roles, and societal institutions underlying HIV risk for sex workers in other countries. These studies underscore the importance of social and cultural contexts in understanding powerlessness and its possible role in HIV prevention. They also suggest that it is necessary to consider both actual and perceived powerlessness. Thus, for the purposes of this study, we focused on whether actual powerlessness and/or perceived powerlessness was associated with sexual risk behaviors for HIV among labor migrants. The possible link between difficult conditions, powerlessness, and HIV risk could in part explain the elevated HIV risk behaviors among external Tajik male labor migrants . Tajik migrants working in Moscow often speak of being "unprotected" . Unprotected means being without legal status, subject to arrests, imprisonment, beatings, and bribery, and without adequate housing, safe working conditions, or access to healthcare. These harsh living and working conditions caused us to hypothesize that Tajik male migrants could be practicing macho sexuality or something like it to compensate for reduced masculine role functioning while in Moscow. Thus, this study was an opportunity to assess the applicability of Campbell's claims regarding powerlessness based on social psychological theory. --- RESEARCH QUESTIONS AND DESIGN In accordance with International Organization for Migration and United Nations Educational, Scientific and Cultural Organization definitions, Tajik citizens who travel to work in Tajikistan are internal migrants and those who travel to work in Moscow are external migrants. To investigate the possible role of powerlessness in HIV risk for labor migrants, we took advantage of an opportunity to compare Tajik internal and external migrants. Specifically, for the internal sample, we focused on Tajik men who traveled to the Tajik city of Regar, where one of Central Asia's largest aluminum factories supplies many well-paid jobs and dormitories for the men to live in. Our research design was to compare internal and external migrants using ethnographic data and qualitative methods . We hypothesized that external migrants in Moscow would experience powerlessness to a greater degree than internal migrants in Regar and that this would result in higher HIV risk behaviors. We sought to answer the following research questions: 1. How do social conditions differ between internal and external labor migrants? Do internal and external labor migrants differ in their experience of powerlessness? 2. Do internal and external labor migrants have similar sexual and HIV risk and protective behaviors? --- 3. If there are differences in HIV risk and protective behaviors, can those differences be explained by "powerlessness," or is there some better explanation? --- METHODS --- Procedures and Sample To respond to the public health risk of HIV in Tajikistan, we formed a collaboration between U. Minimally structured interviews and focused field observations were conducted with 30 married male internal migrants in Regar, Tajikistan, and 30 married male external migrants who worked in Moscow. All subjects were identified through purposive sampling at sites in Dushanbe or Regar. The subjects for interviews in Dushanbe were identified with the help of staff members from the Republican HIV/AIDS Prevention Center. This site was selected because it is the location where the majority of Tajik migrant workers go in order to undergo HIV testing and counseling prior to leaving the country. Although there were several regional sites in the provinces of Gorno-Badakshan, Khatlon, and Sugd where migrants could receive an HIV test, the majority went to Dushanbe for testing because it is the main point of departure for flights and trains to Moscow. It is further worth noting that while Russian and Tajik policies aimed to have all labor migrants traveling to Russia tested for HIV, in practice, many migrants bypassed this requirement by entering Russia on tourist or student visas or by paying for certificates without being tested. The interviews with external migrants took place during the season when they resided in Tajikistan, which was typically from December to April. These interviews were conducted in a private room at the HIV/AIDS Prevention Center. Only a small number of interviews were conducted each day so as to reduce the likelihood that subjects could be identified by others as research subjects. Nonetheless, we informed subjects that there was a possibility that other migrants may find out that they were participating in our research project. The subjects for interviews in Regar were identified with the help of a local physician who worked in the government health cabinet that provided health care services to the majority of the workers at the aluminum factory as well as to local sex workers. To protect confidentiality, most of the interviews in Regar took place in private rooms of the physician's clinic. Two interviews were conducted in a semiprivate tent of an outdoor restaurant near one of the factory worker's dormitories. External migrant subjects were only included if they gave verbal and/or written consent and met the following criteria: male migrant from Tajikistan between 18 and 59 years old; married or engaged to a woman in Tajikistan; seasonal or temporary migrant workers who return to Tajikistan at least once per year; and risks : recent unprotected sexual contacts in Moscow; present or past sexually transmitted disease ; present or past tuberculosis ; present or past intravenous drug use; and present or past alcohol abuse. The criteria for internal migrants were the same as those for external migrants with the exception of criterion "c" and criteria "d, I" . Of note, in the process of screening male migrants for risks, only three persons from the external migrant cohort of 33 potential participants, and one person from the internal migrant cohort of 31 were excluded for not meeting criteria. We found important demographic distinctions between internal and external migrants . Internal migrants were older and more highly educated and had more children. External migrants were younger and less educated and had fewer children. Many external migrants were not married because many Tajik men go to Moscow in order to earn enough money to start a family. All interviews were conducted in either Tajik or Russian. The reason for using Russian is that some of the internal migrants spoke Uzbek and Russian, but not Tajik. The interviewer in Dushanbe was a Tajik physician. The interviewers in Regar were either the Tajik physician or, for 14 of the 30 interviews, a trained American medical student assisted by a Tajik translator. The interviews lasted 30 to 90 minutes and were audio-taped and transcribed into English. The content of interviews focused on the following domains: work and living conditions in Regar or Moscow; family in Tajikistan; HIV risk and protective behaviors; HIV risk and protective knowledge and attitudes; spousal communication; and alcohol and drug use. For example, we asked: "What are your living arrangements like?" "What is your opinion about condoms?" "Do you talk to your wife about the girls you see here?" --- Methodological Approach Our study utilized a grounded theory approach to qualitative research . The initial study questions were refined through an iterative process of data collection and analysis that followed standardized qualitative methods and used ATLAS.ti, a qualitative data analysis program . After establishing coder reliability, labels were attached line-by-line to segments of transcripts in order to form codes. Codes were systematically compared with one another and similar codes were classified into concepts, categories, or processes . Coders also wrote memos or extended notes, which further explained key constructs or identified gaps in understandings. Through this approach of comparing codes, writing memos, and developing concepts, categories, and processes, a grounded theoretical model emerged . This report summarizes the key components of that model. To ensure the trustworthiness of our findings, codes and concepts were reviewed by the entire research team at multiple points during and after data collection and data analysis to enable checking for other explanations and contrary evidence. Additionally, as part of the iterative process of research, when gaps or questions emerged, we addressed them by collecting new data, conducting further data analyses, and consulting key informants. --- RESULTS The grounded theory model that was derived through this investigation of Tajik male labor migrants represents the findings that external migrants described being "unprotected" due to harsh living and working conditions in Moscow. However, both external and internal migrants reported regularly having unprotected sex with sex workers when away from home. Thus, the model demonstrated that powerlessness did not directly explain whether or not migrants regularly had unprotected sex with sex workers. Yet it did demonstrate how the difficult social environment in Moscow was associated with a somewhat different manifestation of HIV risk behaviors. In the face of harsh conditions, external migrants' actions were more highly determined by masculine group norms and behaviors . In contrast, internal migrants were more likely to individually solicit sex workers and to use condoms based on each man's own sense of a woman's cleanliness. Table 2 summarizes the differences between internal and external migrants, and Figure 1 presents the grounded theory model derived through qualitative analysis of ethnographic data to explain these differences. In the sections that follow, the elements of this model are explained in greater detail with illustrative quotations from both internal and external migrants. --- Living and Working Conditions The external migrants working in the bazaars lived in overcrowded, unsanitary, and poorly heated dormitories or apartments. One migrant said, "Fifty people live in one dormitory. The conditions are terrible, there is no ventilation system and it is very difficult to breathe. We do not even have a bed." External migrants lived together in order to save money and because they felt safer when in large numbers. Another said, "Ten people were living in one room as it was slightly bigger. We all manage to live together. As you know, it is difficult to find rooms for rent because migrants are not trusted. Also, it is expensive for one or two people to rent a place in Moscow." Most migrants were working in Moscow illegally, and they were subject to frequent arrests, imprisonment, bribes, beatings from police, and unprovoked attacks from skinheads and ultranationalists. One migrant said, "If Russians want, they can kill you, beat you and say whatever they want. A Tajik does not have the right to say a simple word. Tajiks are unprotected." Most of the Tajiks who worked in construction reported that they lived in a barak or wagon. A barak is an old building within the construction site that would eventually be demolished. While construction was going on, the firm owners offered these old buildings to house the Tajik migrants. Wagons were one-room units including both a kitchen and sleeping area that were also located within the construction sites. The Tajik men reported that there have been many cases of fires in the wagons where Tajik migrants were burned and killed. Those working in construction were also subject to workplace accidents at the construction site for which they could not get medically treated in Moscow. As a consequence, external migrants lived with the daily fear that they would get seriously injured or die. One said, "I saw myself in the skin of a slave in Moscow. We are unprotected." The external migrants were seasonal migrants who typically departed for Moscow in April and returned home in December. Most external migrants reported being part of social networks organized around their extended family or their village. In the harsh conditions of life in Moscow, the external migrants were even more dependent upon these networks than they would be if living in Tajikistan. For example, one migrant said, "We listen to our village leaders. They keep us united, which is very important for us today. We are far from our homes and need support from each other. Nobody can survive alone in this terrible society. You have to belong to a network." These networks helped migrants through sharing information, providing social support, and problem solving. For example, one workgroup leader said, "During the gatherings at work or when we have a party, I can talk to the migrants. If somebody gives information to people there, they can share this information with others." All the internal migrants interviewed in Regar worked at the aluminum factory and lived in a nearby dormitory or apartment. As part of their work contracts, most internal migrants were offered discounted rent at the factory-owned dormitories. One internal migrant described his living arrangements, "There are two people in each room. I live with my coworker and friend." Another migrant described the dorms as being affordable and acceptable: "We have two rooms for three people … if the dormitory belongs to the factory-it costs five or six somoni per month … many people who live here are working in the factory." In the city of Regar, the internal migrants felt at home and did not face any persecution from police or locals. The internal migrants spoke positively about the work in the factory. "There is no work in the place where I live. You are only able to work in the farm. Here at the factory, I have good work." Many migrants said they came to the factory in Regar because they had technical skills or degrees that could earn them higher wages than in their home communities. One said, "I decided to come to Regar. The payment at the factory is better than anywhere else. The factory is paying me good money and I am feeding my children." The work at the aluminum factory was done in shifts with brigades of 20 to 30 workers. For each shift, the work was divided among the different members of the work brigade: "My work consists of melting metal … some workers clean, some produce locks … there are different tasks within one shift." Typically, each shift lasted 15 days. Many internal migrants returned home for 15 days after working two shifts . Internal and external migrants also differed in their patterns of alcohol use. External migrants reported drinking alcohol in large parties held in their apartments or living quarters. One migrant explained, "Usually we invite girls on Sundays when we arrange good parties. We used to have parties occasionally on birthdays or during national holidays. We sit together and drink different alcoholic drinks, have fun and later dance with girls who we bring in from the street." In contrast, most internal migrants reported either not drinking alcohol or drinking in relative moderation. None of the internal migrants interviewed reported getting intoxicated in conjunction with or during sexual experiences. When one internal migrant was asked how often he consumes alcohol, he replied, "Two or three times per week, because I am bored." Another said, "I drink vodka with friends … but not so much, just for pleasure." Neither any external or internal migrants reported using any drugs. Both external and internal migrants expressed the same kinds of beliefs regarding the role of men in male-dominated and Muslim-influenced Tajik society. They spoke about respect for elders, honoring village traditions, and protecting their family. A strong man, they described, is someone who is able to be a breadwinner and decision-maker for his family. One notable difference reported by internal and external migrants with respect to beliefs about men is that external migrants had more difficulty embodying these beliefs about men's roles given their separation from their wives and children and their being a part of an impoverished minority group in an alien society. --- HIV/AIDS Knowledge Knowledge of HIV/AIDS was variable and vague among the external migrants. One said, "No I don't know [much] about HIV/AIDS. I have no good information about it, really. I know that it is transmitted through girls who are dealing with sex." Another said, "I read in a magazine that using condoms is very useful for preventing HIV/AIDS and it is an easy way to protect yourself when you having sex with sex workers or other women. We have a proverb in Tajik language that says: be careful of incidents and solve problems before they happen. Therefore having enough knowledge about AIDS and using condoms primarily are very good ways for migrants to prevent themselves from being infected by HIV." Like external migrants, internal migrants also had variable and vague knowledge about HIV/ AIDS. When asked how a person could become infected by HIV, one migrant said, "Truly speaking, I don't know." While some internal migrants were not sure of how HIV was transmitted, others demonstrated a greater degree of knowledge. One migrant explained, "If a migrant looks after himself, and takes care of himself, does not use dirty syringes, and refrains from sex … I think these are some of the ways to prevent HIV." Neither internal nor external migrants considered HIV/AIDS to be a major problem for Russia, Tajikistan, or other Central Asian countries. An external migrant reported, "AIDS is a big problem in Africa but not here." An internal migrant said, "I think there are some HIV positives but it is not a large problem. If it was problem there would be more news [about this issue]." Another internal migrant stated, " … but I was never infected by any of them, thank God. I heard that there is also a dangerous disease called AIDS and it is transferred by sexual contact if someone doesn't use condoms. But I think that this disease is not common in Tajikistan." Internal migrants had positive opinions about HIV testing but rarely reported ever getting an HIV test themselves. One migrant said, "I only know that there is a place in Dushanbe where one can get a HIV test, but I did not go there. I have never been tested. But I think that every migrant should have a HIV test." None of the migrants interviewed reported testing positive for HIV. --- Sexual Behavior Most external migrants in Moscow reported regularly having sex with sex workers. In order to reduce the risk of being caught by the police, a leader of a group of migrants often arranged for a sex worker to meet the group back at their living quarters. Sex workers may be women needing food, vodka, or drugs , or less commonly a woman from the tochkas or another sex worker operation . One migrant described how the cost of a sex worker was split between many migrants: "One girl cost $200, and there were 20 guys in our wagon, so she agreed to sleep with us for a night. Then each of us paid her $10." Another external migrant explained, "You know when a person is drunk and women are present, they obviously will have sex. During that night we had to exchange our women." The external migrants described how heavy alcohol use impaired their judgment and reduced condom use. One migrant said, "We drink vodka and want girls. When we are drunk we never use condoms." Another said, "When I am drunk, I can't tell the difference between A from B … then how can I use condoms?" In contrast, internal migrants drank alcohol in moderation and none reported drinking during sexual contacts, including with sex workers. Unlike their external counterparts who often had to share women, internal migrants demonstrated autonomy in choosing where and when to engage in sexual contacts with sex workers in Regar. One factory worker described clearly the different choices of sex workers that existed for migrants: "There are three types of sex workers-sex workers who stay in cafes and restaurants. These are cheap sex workers. Another group of sex workers are more expensive and work independently. You call them and negotiate the price over the phone. The third group are the sex workers who work through Mama Rosa [Mama Rosa is the name given to madams who negotiate with clients for the services of sex workers]. I always choose sex partners through Mama Rosa. I think this is safer because Mama Rosa is responsible for the health of the sex workers." Internal migrants in Regar often engaged sex workers through telephone networks, for example: "My friend has the phone number of a Mama Rosa and she provides us girls for 20 or 30 somoni for two or three hours of sex." Sometimes frequent clients were identified and were the recipient of calls from women soliciting sex work: "If you have money, they will offer to go with you somewhere. They call us many times." Both external and internal migrants expressed the belief that men were expected to have sex when away from their wives. One internal migrant said that sex with women when away from wives was natural: "Something is inside a man that he needs to be with a woman. It is natural!" Another migrant claimed that the lack of regular sex caused referred physical pain: "My wife is far away. That's why I need to have sex. If I haven't had sex for many days I can feel pain in my back." An external migrant explained that it was part of the Tajik tradition of masculinity to have sex with other women when away from home: "It is in our tradition that a man keeps his sexual activities secret and it is a sign of a real man, mardi haqiqi [masculinity]. My wife and other women know that when their husbands are away in Russia they obviously will have sex." --- Condom Use Both internal and external migrants reported that condoms reduced pleasure during sex. One external migrant said, "A person has to feel that he is having sex, but with condoms you do not feel anything." Similarly, one internal migrant thought that condoms reduced pleasure because they were not natural. He said, "I don't need them. You cannot take pleasure with condoms. It is not natural and isn't pleasurable for me." External migrants cited several additional reasons for their unwillingness to use condoms. Some reported being afraid of being caught by the police while out buying condoms: "I do not have time to go out and buy condoms. It is dangerous and the police might catch you." Another said, "We really could not go to a shop or drug store to buy condoms, because policemen would definitely catch us, take a lot of money, or deport us." Other external migrants said that they sometimes did not speak the same language as the sex workers, many of whom were migrants from other countries, including Vietnam and Korea. In contrast, internal migrants did not report any fear of police and did not have difficulties in obtaining condoms or communicating with the sex workers. One internal migrant commented on the wide availability of condoms in Regar, "In every kiosk there are condoms, and I always buy them myself. Everywhere you can find condoms in the drug stores." Though most external migrants could correctly identify condoms as protective against HIV and other sexually transmitted infections , many continued to have unprotected sex. One external migrant said, "Everybody says that sex without condoms is dangerous, but nobody changes their behavior. Even I have sex with prostitutes and don't use condoms. Knowing is different from doing." Internal migrants reported making more independent and deliberate choices regarding whom to use condoms with. One migrant spoke of being able to detect whether or not a woman could have an STI: "If she is young-and you are suspicious, you can use condoms. If you have no suspicion, or you know her very well, you don't need to use condoms." Many internal migrants used condoms based on their sense of a partner's cleanliness or style of dress. One migrant said, "Sometimes, I used condoms with sex workers but never with my girlfriend. I just look at the girl I am going to have sex with, and if she is clean I don't use condoms. If I see that she is not clean or not well dressed, I use condoms." Another said, "You can know that from their clothes or their cleanliness. I ask them if they have any diseases. But with sex workers I mostly use condoms." --- DISCUSSION This study aimed to explore the relationship between difficult living and working conditions, sense of powerlessness, and HIV risk and protective behaviors among Tajik external and internal labor migrants. The ethnographic findings revealed that due to differences in the social environments of the receiving country, external migrants experienced more harsh living and working conditions that internal migrants did not, including having no legal status; being outsiders in an unfamiliar city; arrests and beatings from police and attacks from skinheads and ultranationalists; overcrowded and unsanitary housing; and lack of access to health care. Tajik external migrants regarded themselves as "unprotected" and unable to change this situation, which is similar to what Campbell described as powerlessness among South African mine workers. In contrast, Tajik internal migrants described their social and work environments in more positive ways. However, external migrants' powerlessness did not appear to be associated with elevated HIV risk, as both external and internal migrants from Tajikistan reported engaging in HIV risk behaviors including concurrent sex partners through the use of sex workers when working away from home; vague and variable knowledge about HIV risk and protection; and reluctance to use condoms. Other possible explanations of the high-risk sexual behaviors for external and internal migrants alike should be considered, such as simply that pleasure seeking trumps self-protective behaviors. This study found other important differences in social environment that appeared were to impact how HIV risk behaviors were manifested among migrants . Most external migrants belonged to social networks defined by their extended family or village in Tajikistan. They lived in close quarters with other Tajik migrants who provided much needed support amid adverse conditions. These close living arrangements also reinforced group masculine norms and behaviors such as heavy alcohol use, concurrent sex with multiple partners, sharing of sex workers, and reluctance to use condoms. In contrast, internal migrants lived more independently, exercised more autonomy in selecting the type of sex worker , had little to moderate alcohol use, and used condoms based on their sense of a woman's cleanliness. However, our findings indicated that both internal and external migrants were at elevated risk for HIV because both groups had unprotected sex with multiple concurrent partners. Our study has important implications for building theory about migrants and HIV risk that involves powerlessness. Campbell's contextual formulation of powerlessness was helpful in explaining the interpretative repertoires by which South African mineworkers shaped their own social and sexual identities ). She described how the mine workers experienced powerlessness and a lack of control in a variety of contexts, from mine accidents to health problems . While their concept of powerlessness was an appropriate starting point for this investigation, and did resonate with the Tajik external migrants' sense of being "unprotected," it did not appear to be as useful in the context of Tajik labor migration as an explanation for HIV risk. We do not want to trivialize the experience of powerlessness among labor migrants in South Africa or Russia. However, our results point toward limitations in theorizing powerlessness as a psychological factor that could become removed from social context. Rather, we find advantages in viewing powerlessness through an ecological lens, both as being strongly linked to social context, and because powerlessness is one of several multilevel consequences of men living and working in groups amid difficult living and working conditions. Previous studies of migrants and marital HIV risk in rural Mexico also utilized an ecological approach . The authors argued that in the social spaces of migration, male migrants established and maintained relationships with other male migrants through demonstrating assertive, competent, and sexually independent masculinity with respect to one another . This model may help to explain why Tajik external migrants living in shared apartments engaged in the exchange of sexual partners during drunken parties. It was not only because it was less expensive, but also because this practice strengthened the men's relationships with one another. In contrast, internal migrants did not share women during sex because they had better defined roles in the Tajik society and did not need to resort to affirming their bonds to other men through such sexual behaviors. In addition to implications for theory building, our findings have practical implications for HIV prevention programs focused on labor migrants. First, it is worth noting that in response to the theory of powerlessness developed by Campbell in South African mines, the authors advocated for developing HIV prevention programs that focused on peer education, in part because of their ability to address group processes that may underlie risky sexual behaviors . However, a 2-year study of a comprehensive community intervention in the mining community of Carletonville, South Africa-which included peer education, condom distribution, syndromic management of STIs, and presumptive STI treatment for sex workers-found little evidence for significant behavior change, and the prevalence of curable STI increased . The authors attributed the lack of effectiveness to inadequate resources, inconsistent messaging about the role of HIV in AIDS, conflicts between stakeholders, and a lack in urgency . It could also be that as a prevention model, the conceptualization of powerlessness that underlay the intervention was in some way inadequate as a basis for HIV prevention. With respect to Tajik labor migrants, attempting to diminish their individual sense of powerlessness, which did not appear to determine their HIV risk-related behaviors, may not be the most productive approach. Rather, HIV preventive interventions should build awareness, knowledge, skills, relationships, supports, and confidence that would assist men in adjusting to the harsh living and working conditions of migrant life. In the Tajik external migrant context, this should include discouraging behaviors that increase risk such as such as unprotected sex with multiple partners and heavy alcohol use, because they are not consistent with the man's role as a protector and provider of his family. Interventions should also consider encouraging more positive aspects of men's social identity and social networks that may be found amongst male labor migrants. Given migrants' perceptions of the inaccessibility of condoms, HIV/AIDS prevention efforts should also focus on increasing condom availability, perhaps by distributing educational materials and condoms at points of contact between external migrants and sex workers in Moscow. Last, structural interventions are also called for, such as addressing migrants' legal status and access to health care services in Russia. There are several limitations of our study. First, this study used a purposive sampling approach and thus the sample was not representative of all Tajik labor migrants, internal or external. Second, recruitment was conducted from only one source each for internal and external migrants, which may have limited the range of subject available for possible recruitment. With respect to these concerns, future studies with labor migrants should consider using larger representative samples. Third, it is possible that due to concerns about confidentiality, some subjects may have tailored their statements about social norms to fit with what they believed we wanted to hear. Regarding this last limitation, we took many steps to encourage frank disclosure by assuring participants of confidentiality, exploring each topic at length and in-depth, and maintaining a nonjudgmental interview approach. Fourth, another possible limitation is that our analysis could have been biased toward some interpretations and away from others. To minimize this possibility, we explicitly considered alternative and discrepant explanations and sought feedback from key informants on study findings. Fifth, data collection was limited to one point in time and to one location on the trajectory of migrants' lives. Future studies should consider longitudinal and multisite methods so as to better examine psychosocial, cultural, and familial processes over time and in the multiple real-life contexts where HIV/ AIDS risk and protection occur. One final limitation was that our study was unable to elicit data regarding same-sex experiences among male migrants. We suspect that this was in part due to the highly stigmatized and hidden nature of such behaviors in Tajikistan. Low rates of same-sex behaviors have been reported from other HIV studies in Central Asia and Russia ; however, these figures may also underestimate the true prevalence as a result of the extreme social vulnerability and stigma that men who have sex with men experience in many Muslim countries. In light of these limitations in the field, a future study is needed to address awareness of men who have sex with men and HIV risk in a culturally sensitive way among countries in Central Asia with a Muslim majority population. Further research is needed to design and evaluate preventive interventions to change the risk behavior of male labor migrants. One implication of our findings is that HIV preventive interventions with migrants should pay attention to the different patterns of migration as these introduce different contextual factors that must be considered for the design and implementation of interventions. Another implication of our study is that exposure to traumatic events should be investigated as a possible factor in impacting HIV risk and protective behaviors . Our findings also suggest that successful HIV prevention programs with this population may also require interventions for the wives and other sexual contacts of migrants in Tajikistan and Moscow. Grounded theory model representing similarities and differences between internal and external migrants.

To investigate the role of powerlessness in elevated HIV risk among labor migrants, we compared internal verses external male migrant workers from Tajikistan using minimally structured interviews and focused field observations. The sample included 30 male labor migrants who traveled to work in Regar, Tajikistan (internal labor migrants), and 30 who traveled to work in Moscow, Russia (external labor migrants). Though powerlessness did not appear to account for whether labor migrants engaged in more HIV risk behaviors, the harsh living and working conditions of external labor migration impacted how the migrants manifested these HIV risks by amplifying group masculine norms and behaviors. Progress in preventing HIV infection amid the difficult social conditions of labor migration is contingent upon adequate conceptualization of how such conditions impact HIV risk behaviors

social group . At the same time, youth become more sensitive to perceptions of acceptance or exclusion based on their membership in these social groups . Several studies of sexual minority youth and youth of color have found that encountered bias uniquely contributes to adjustment, an important finding given the well-documented health and academic disparities among sexual minority youth and YoC . 1 Yet few studies have examined how the interaction between sexual orientation and race-ethnicity operates as an underlying mechanism of disparate health outcomes among adolescents . Further, despite discussions and critical reviews of intersections of sexual orientation and race-ethnicity among adults , the literature focused on sexual minority youth of color largely remains unintegrated. Given the considerable demographic shifts occurring in the Unites States-with the projection that by 2020, the population of racial and ethnic minority youth will surpass the non-Latino White youth population -it is increasingly important to understand what is known about the intersections between race-ethnicity and other marginalized identities, such as sexual orientation. This content analysis and critical review examined what is known about the intersection of sexual orientation and raceethnicity among youth in order to broaden understandings and conceptualizations of risk and resilience for SMYoC. Of note, our review is largely focused on the experiences of Black and Latino SMY given that the majority of the research has focused on these two groups. Throughout our review, we refer to the race or ethnicity of the population being studied consistent with the initial reports . An integrated understanding of how multiple marginalized identities contributes to adolescent development is a critical first step needed to inform future research and subsequent intervention strategies aimed at improving well-being among marginalized youth populations. Our content analysis and critical review of the literature included three aims: to examine what content areas are frequently studied among SMYoC, to examine who was represented in the extant literature on SMYoC, and to summarize and critique the extant research published in this area, identifying limitations and promising areas for future research. We expected that the literature would be focused on problematic outcomes, rather than normative developmental outcomes, given that the study of adolescence, in general, has largely been a science grounded in pathology . This pathologizing-normative disparity is further compounded when the focus is on marginalized populations and because NIH-funded studies are typically framed by a medical model that is aimed at understanding pathology and risk . --- Method Inclusion criteria For inclusion in the analysis, reports had to include original empirical results focused on SMYoC, participants aged 25 years and younger, and participants living in the United States. The upper age limit of 25 years was chosen in order to be consistent with theories of emerging adulthood that posit that the developmental experiences of youth in their early 20s are distinct from the experiences of individuals in their late 20s ; further, several reports would have been arbitrarily excluded from our analysis had our inclusion criteria been age 24 rather than 25. We limited the review to only include U.S.-focused samples given the unique sociocultural and historical salience of race-ethnicity in the United States . Finally, we included reports with either an explicit or implicit focus on the intersection between race-ethnicity and sexual orientation. --- Search and categorization strategies Search terms were truncated to allow for diversity of term usage across reports. Search terms included AND AND . Several comprehensive databases were used, including ERIC, Medline, PsycInfo, and Sociological Abstracts. Unpublished dissertations were included in this review in order to minimize the threat of publication bias, given that null or unexpected findings are less likely to be published in peer-reviewed journals . Two searches, one in 2013 and a follow-up in 2015, yielded 1,507 unique reports . The research team-two faculty members and three students-categorized the resulting 125 reports according to the following characteristics: design utilized , participant characteristics , and the constructs of interest examined in each report . The coding system for the constructs of interest examined in each report was developed inductively by focusing on the report's dependent and independent variables; eight content categories emerged and are discussed in the following sections. This strategy is consistent with the methodology used in other relevant content reviews . Reliability was assessed throughout the process , and any discrepancies in coding were discussed and reconciled as a team. --- Results --- Report design and participant characteristics Of the 125 reports , 66.4% were quantitative; 25.6% were qualitative; and 8% were mixed methods. Sample size varied widely across reports ; nearly half of the reports contained samples with at least 200 participants, and only 10 reports included samples with fewer than 10 participants. Few reports focused exclusively on participants aged 20 or younger . The number of reports addressing the intersection between race-ethnicity and sexual orientation has increased tremendously in the past 5 to 10 years ; over two-thirds of the reports included in this review were published after 2008. Regarding participant demographics, 50.4% of the reports included only young men; only 8% of samples included only young women. Less than 10% of reports explicitly included transgender youth when enumerating gender identity categories. Most reports included samples with multiple racial-ethnic groups ; monoethnic reports included exclusively African American or Black youth , Latino or Hispanic youth , and Asian or Pacific Islander youth . Only 34.4% of reports included heterosexual or straight youth; most were focused on the within-group experiences of young men who have sex with men , young women who have sex with women , or lesbian, gay, bisexual, queer/questioning identified youth. Notably, almost one-third of the reports focused on MSM populations. Finally, the majority of reports focused on health-related outcomes: 69 focused on sexual health , 42 on mental health , and 43 on substance use. Few reports examined intersectionality in terms of normative developmental processes, such as sexual orientation identity development or contextual and interpersonal relationships . --- Health-focused outcomes Sexual health-Most of the 69 reports focused on sexual health only sampled sexual minorities , and 37 of those 69 reports only sampled young MSM populations. Forty-one percent of reports focused exclusively on one racial-ethnic group; 22 of these reports included only African American or Black youth. Only a few of the reports examined their research questions or hypotheses using an explicit intersectional approach . For example, Hidalgo et al. examined how racial-ethnic-related discrimination and sexual identity-related discrimination were associated with HIV infection among sexual minority men. In another example, Stevens et al. focused on the "double minority status" of gay and transgender African Americans. Notably, these reports were published in the past 5 years, likely reflecting a growing recognition of the importance of within-group variability among sexual and gender minority populations . The results of the reports are presented in terms of whether the reports were comparative in nature or were studies of SMYoC . Results from comparative samples: Reports that examined sexual health focused primarily on HIV/AIDS . In general, these reports indicate that, compared to White, non-Latino participants, SMYoC are more likely to be HIV positive , and Black SMY appear to be at greatest risk . Reports also suggest that Latino and non-Latino SMYoC have riskier attitudes about sex and engage in risky sexual behaviors at higher levels compared to White, non-Latino gay and bisexual youth . It is important to note, however, that other reports found no differences in sexual risk behavior by racial-ethnic group or found riskier sexual behaviors among White youth compared to young MSM of color . Further, SMYoC shared that they experienced racial-ethnic discrimination or stereotypes in their sexual encounters , but these experiences were not associated with more risky behavior . The literature also suggests that SMYoC lack adequate access to information about HIV/ AIDS and are less likely to know their HIV status or follow up with appropriate medical care when compared to White, non-Latino SMY . Notably, research on sexual health largely focused on risk; few reports included in our review focused on protective factors for risky sexual behaviors . Yet there may also be racial-ethnic differences in effectiveness of interventions to reduce risky sexual behavior. For example, an intervention that combines case management, health care, and counseling with small group discussions was successful in temporarily decreasing unprotected anal intercourse among African American and White youth, but not Hispanic youth . This intervention also temporarily decreased unprotected oral intercourse for all groups, but this effect did not persist for African American youth after 6 months. Rotheram-Borus, Reid, and Rosario also found that an intervention consisting of one to 30 sessions about HIV facts, coping skills, health care access, barriers to safer sex, and prejudice against gays was effective at the 3-month assessment in decreasing unprotected sex acts among White, African American, and youth of other racial or ethnic groups, but not Hispanic youth. It is important to note that the number of sessions attended did not consistently decrease risky behavior for all groups, and the effectiveness differed at the time of assessment . Results from SMYoC samples: Reports that sampled only SMYoC frequently introduced and discussed culturally specific factors that are likely important in understanding health disparities across groups. For example, Black youth identified spirituality, family support, education, and LGBT organizations as factors that helped motivate and refocus their goals . Another study found that school enrollment might be an important place to intervene for HIV risk and drug use for Black youth involved in house and ball culture . Clearly, more research is needed to understand the nuances of how culture, race-ethnicity, sexual orientation, and other relevant demographic factors interact to inform sexual health and behavior. Mental health-Over half of the 42 reports focused on mental health exclusively included sexual minority participants. In comparison to sexual health-focused studies, only eight reports sampled young MSM populations; thus, most of these studies focused on sexual identity rather than sexual behavior. Notably, only nine reports focused exclusively on one racial-ethnic group, and seven of these studies were focused on the experiences of Black or African American youth. The results are presented in terms of whether the reports were comparative in nature or were studies of SMYoC . --- Results from comparative samples: The results for mental health in comparative reports across racial-ethnic groups were much more mixed than those for sexual health, suggesting that additional research is needed in this area in order to understand culturally relevant or demographic factors that may help to explain mental health disparities within SMY populations. Results from some reports indicated that Black and Latino youth experienced greater mental health problems compared to White youth . However, other reports found that White youth reported more mental health problems compared to youth of color . Similarly, sexual minority young men who reported bullying had greater odds of suicidality than sexual minority young women, and this finding was stronger among non-Latino youth than among Latino youth . Others found no differences by race-ethnicity in feelings of hopelessness or suicidality or externalizing and internalizing symptoms . Results from SMYoC samples: Reports with mono-racial/ethnic samples did find evidence for poor mental health outcomes among SMYoC . Gender role strain may affect mental health, particularly among young Black MSM, because they may feel the need to be hypermasculine or risk being perceived as less than feminine . Black youth also reported more cumulative risk than Latino youth, and cumulative risk is associated with increased suicide risk . Thus, it appears critical to understand what cultural factors or learned coping mechanisms related to racism, for example, may protect SMYoC. Further, given that only a single report was identified that exclusively focused on mental health outcomes among Latino and Asian youth, respectively, more research is needed to understand the culturally relevant factors for these populations of SMYoC. Substance use-Over half of the 42 reports focused on substance use exclusively included sexual minority participants, and 15 reports focused on the experiences of MSM. Only eight reports focused exclusively on one racial-ethnic group, and six of these reports focused on the experiences of African American or Black youth. The results described in this section largely come from the 29 reports of sexual minority youth that included or focused on race-ethnicity identification as a factor related to sexual health. Instead of discussing the reports in terms of racially-ethnically comparative versus homogenous sample designs, the results in this section are described by the type of substance investigated , given that discrepancies existed across type of substance, and prevention/intervention efforts likely differ depending on type of substance. Smoking: Two comparative reports found that SMY from all racial-ethnic backgrounds had higher cigarette smoking prevalence compared to their respective heterosexual peers . However, there were racial-ethnic differences in types of smoking behaviors . Among high school students, Corliss and colleagues found that Black and Asian American/Pacific Islander SMY smoked more than White adolescents, suggesting that developmental period may be an important modifier of racial-ethnic disparities in smoking. Another report found that Native Hawaiian youth who were unsure of their sexual orientation reported more tobacco use than those who reported other sexual orientations . In the same study, White bisexual youth reported more tobacco use than their White heterosexual peers. One report of Black and Latino gay and bisexual men found that marijuana use was associated with African American young men's racial-ethnic identity , suggesting the need for an intersectional approach in order to understand how culture interacts with sexual orientation to inform smoking behaviors. --- Alcohol use: Comparative studies suggest there is either no racial-ethnic difference in the prevalence of alcohol use among LGB youth or that SMYoC have a lower prevalence of alcohol use . Further, some studies found that African American young MSM were less likely than Whites to binge drink , more likely to have unintentionally done something sexual because of alcohol or drug use , and more likely to develop alcohol dependence or abuse . However, Latino YMSM demonstrated similar risks as Whites . Factors associated with alcohol use among SMYoC include racism and antigay discrimination and bullying and victimization . Bullying and victimization appear to be more strongly associated with alcohol use among Asian/Pacific Islander youth . Yet there is evidence that family support is associated with less alcohol use over time across racial-ethnic groups , underscoring the importance of social support. Finally, alcohol may also be tied to social norms within a racial-ethnic group; for example, heavy alcohol use was tied to Latino gay men's racial-ethnic identity , again suggesting the need for culturally informed studies of alcohol use among SMYoC. Illicit drug use: Notably, illicit drug use was often explored in the context of HIV risk or coping . Illicit drug use was rare or lower among Latino and Black youth compared to White youth . One report found that youth of color generally reported lower prevalence of illicit drug use, although this difference was less pronounced among bisexual and unsure Black teenagers . Few studies examined illicit drug use among SMYoC using monoethnic samples. For example, one study that focused on Asian American and Pacific Islander sexual minority youth found that the association between sexual minority status and substance use was not present in the teen years but emerged in the 20s . Further, AAPI sexual minority young women might be at greater risk given the high prevalence of substance use compared to other AAPI groups. Among African American populations, illicit drugs are readily available in house and ball culture events , and participation in house and ball culture was associated with greater use of illicit drugs . --- Identity development All but one of the 22 reports focused on mental health exclusively included sexual minority participants; four of the 22 studies focused on MSM populations. Over half of the reports focused exclusively on one racial-ethnic group . Findings from these studies are discussed in terms of the two most common themes: identity disclosure and development processes, and identity centrality. Identity development and disclosure processes-The processes of identity disclosure and labels used appeared to be consistent across racial and ethnic groups , with the exception of unique terms that Black lesbians used to discuss identity . Few reports examined the experience of dual or multiple identity development exploration or resolution processes . For example, one study found that White SMY were less likely than SMYoC to have a diffused sexual identity status . Further, there were small correlations between racial-ethnic identity and sexual identity moratorium , diffused, and foreclosed statuses , respectively, but not achieved statuses . Among Asian and Pacific Islander men, racial-ethnic identity affirmation was positively correlated with sexual identity affirmation . In addition, in one study, being committed to and feeling positive about one's racial identity was associated with decreased internalized homophobia . To date, only one study has examined how racial-ethnic and sexual identities develop simultaneously ; youth in their study appeared to develop their racialethnic and sexual identities concurrently. It is important to note, however, that additional research does suggest the contexts and predictors of identity development may differ across racial-ethnic and sexual identity development. For example, whereas youth identified families as important for racial-ethnic identity development, community-based organizations and the Internet were important for sexual identity development . Important cultural constructs, such as machismo and familism for Latino youth, emerged when discussing the process of identity development . One report found that Latino youths' sexual identity development was influenced by their exposure to machismo . Latino youth also indicated that the identity development process was inherently relational because of the central role of family relationships in Mexican families . This was also found among Black MSM who described the perceived need to meet hypermasculine expectations for Black young men as heightened compared to White young men . Identity centrality-Reports of identity centrality suggest that the salience of a particular identity largely depends on context . For example, Latino youth who lived in predominately Latino-populated areas viewed their sexual orientation as more salient than their race-ethnicity in their identity . Conversely, Vaught found that being Black did not allow for the salience of a gay identity because it was seen as only existent in White culture. Youth in another report, however, discussed the importance of understanding how the intersectionality of identities plays a role in lived experiences, like the overrepresentation of SMY and YoC in the juvenile justice system . --- Interpersonal relationships and contexts Family-The majority of the 19 reports focused on family relationships sampled only sexual minority participants, and only one study focused on the experiences of young MSM populations. However, only seven reports focused exclusively on one racial-ethnic group . The findings from these studies are discussed in terms of the two major themes that emerged: disclosure to family members and acceptance/rejection experienced from family related to sexual orientation. Disclosure to family members: Several reports examining family relationships of SMYoC focused on the process of disclosing one's sexual orientation to family members. One report found that Black and Latino youth were less likely than White youth to disclose their sexual orientation to their parents . Yet other reports have not found racial-ethnic differences in level of disclosure . Across all racial-ethnic groups, youths discussed fear of disclosure because of potential for rejection, and parental reactions to disclosure included both acceptance and rejection . Only one report found a significant difference in experienced parental acceptance after disclosure, in that Black participants reported less acceptance than Latino or White youth . Acceptance and rejection from family members: Eight reports examined parental support and rejection that were not specific to disclosure. Overwhelmingly, the results indicate few racial-ethnic differences in levels of perceived family support ; only one report found that White youth reported higher levels of parental support than SMYoC , and this difference was only significant for females. Other reports found that parental support was associated with more positive outcomes and reduced risk for all youth, regardless of race-ethnicity . School-The majority of the 10 reports focused on school experiences exclusively included sexual minority participants, and none of the reports were focused on the experiences of MSM population. One-half of the reports focused exclusively on one racial-ethnic group . Findings from these reports are discussed in terms of schoolrelated stressors and positive school experiences. School stressors: School was largely examined as a source of stress . Homophobic victimization was associated with lowered school belonging, regardless of race-ethnicity, among SMY . A recent study by Sterzing , however, found that Black SMY were more likely to experience disciplinary actions at school compared to White, Latino, or multiracial youth. Another report found that among Asian American SMY, greater emotional distress was associated with more negative perceived school climates , but only among students with low self-esteem. Positive school experiences: Many of the school-focused articles examined gay-straight alliances . These reports documented that GSAs are largely composed of White participants and may be less protective for YoC . Thus, SMYoC may need to seek out different contexts within the school for different aspects of their identity . To date, studies have not documented any racial-ethnic differences in attitudes about school, academic integration, or perceived academic or career barriers among SMY , but these studies are 10 to 15 years old and should be interpreted with caution. Community-All but one of the 12 reports focused on community experiences exclusively included sexual minority participants, and only two focused on MSM populations. One-half of the reports focused exclusively on one racial-ethnic group . In these reports, community-based organizations were viewed as a critical source of support for SMY . Findings from racially-ethnically homogenous samples, however, demonstrate that these experiences may vary across subgroups of youth. For example, one report found that attachment to the LGBQ community did not reduce risk behaviors among Latino MSM , whereas another report found that CBOs were helpful in mitigating the experiences of sexual harassment among Black lesbians . Another study of Black female queer youth found that CBOs that focused on both ethnicity and sexuality helped foster agency and identity exploration in this population . CBOs may not be perceived as accessible for all youth. For example, one study found that very few Black lesbian adolescents participated in CBOs , and another study found that Latino youth felt less comfortable at a medical clinic for HIV-seropositive MSM compared to Black youth . Further, if YoC were not represented in CBOs, then SMYoC might have felt less comfortable in those contexts . YoC also felt more comfortable in race-ethnicity-focused CBOs compared to CBOs focused on sexual orientation . Compared to other youth, Black youth were more likely to endorse wanting support from other LGBQ adults in the community , but felt less attached to the LGBQ community compared to White and Latino youth . Violence-The majority of the 20 reports focused on violence exclusively included sexual minority participants, and only two focused on MSM populations. Only four reports focused exclusively on one racial-ethnic group . The findings from these reports are discussed in terms of violence specific to only sexual orientation and both sexual orientation and race-ethnicity. Findings from reports focused on only sexual orientation-related bias: Across the contexts discussed in the preceding, a common theme that emerged was the experience of bias-related violence; however, the results are mixed in terms of whether youths' experiences with sexual orientation-related violence differ by race-ethnicity. Two comparative reports found no racial-ethnic-group differences in levels of LGBQ-related victimization , and no difference was found in a report that examined general bullying . Yet three reports found that Black youth experience higher levels of LGBQ-related victimization compared to White or Latino youth . Finally, one report found that Latino, Native American, and multiracial youth experienced higher levels of verbal and physical general bullying victimization compared to White and African American youth . Sexual orientation disparities in violence were also identified by Russell and colleagues for White, Hispanic, and Native American/Pacific Islander youth, but not for Black or Asian youth. Note that, in their report, Black youth had the highest rates of violence, while Asian youth had the lowest rates, regardless of sexual orientation. Whether and how race-ethnicity modifies the relationship between victimization and outcomes is less clear. Two reports did not find any moderation by race-ethnicity. Three reports found that homophobic victimization was only associated with suicidality for White SMY . Further, Rosario et al. noted the important nuanced differences in the mediation of sexual orientation disparities in cancer-related risk behaviors by experiences of peer victimization by race-ethnicity. Findings from reports focused on multiple types of violence: Findings were mixed in terms of whether youth reported more violence related to their sexual orientation or raceethnicity and how strongly these different forms of violence contributed to health and wellbeing. Within-group reports found that Latino youth tended to report more harassment based on their sexual identity than based on their race-ethnicity . A report of Black youth, however, found similar levels of harassment based on sexual identity and race . Further, among SMYoC, racial-ethnic oppression was experienced within the LGBQ community, whereas LGBQ-related oppression was experienced from the larger heterosexual community . In terms of how violence contributed to well-being, one report found that both LGBQrelated and racially motivated bullying were uniquely associated with depressive symptoms . Thoma and Huebner found that there was an additive effect of racist and antigay discrimination on depression; however, antigay discrimination was only associated with suicide ideation whereas racist discrimination was only associated with substance use. Similarly, Garnett and colleagues found that youth who experienced multiple forms of bias had the highest odds of suicide ideation. --- Discussion Our review suggests that the extant research on SMYoC populations is mostly focused on sexual risk, substance use, and mental health problems rather than on normative developmental processes or positive youth development . This finding is consistent with a recent study that comprehensively reviewed all National Institutes of Health funded studies from 1989 to 2011 and found that only 0.5% of all NIH-funded studies focused on LGBQ populations, and they were largely focused on sexual health. This finding is also consistent with a prior content analysis of the literature of LGB people of color, which primarily focused on adults, and found that most empirical studies were guided by a risk perspective . The funding-research literature link is clear and problematic: the NIH, as well as other federal and private funding organizations, fund research that is often driven by a medical, problem-solving paradigm. This model requires that researchers examine the individual-and community-level risk factors that explain health disparities without explicit attention to structural-level factors that perpetuate risk . Further, given a focus on pathology and risk, funding agencies rarely require that researchers examine resilience or normative, positive development. The lack of attention to normative developmental processes is problematic because it undermines a comprehensive understanding of the lives of SMYoC and limits our scientific understanding to outcomes rather than the processes that contribute to those outcomes . These results suggest a critical need for more culturally informed research on normative developmental processes and resilience among SMY, with attention to the unique and shared experiences of YoC. This implication is consistent with the substantial gap in the literature on prevention and protective factors for health outcomes of LGBQ populations, particularly among YoC , as well as the research focused on adolescence . Research on the normative developmental process of identity development and interpersonal relationships in various contexts yielded very few differences by race-ethnicity. Differences that did exist were largely specific to processes of ethnic-racial identity development or ethnic-racial socialization , such that these processes were salient for SMYoC and were associated with their sexual orientation development. Note that very few reports examined how these parallel processes developed simultaneously. Family relationships and reactions to disclosure of sexual orientation functioned similarly across race-ethnicity, suggesting that these processes are fairly universal. LGBQ spaces in schools or in the community also tended to promote well-being, regardless of race-ethnicity, for SMY. Further research, however, is needed to illuminate how to make those support spaces more welcoming to YoC, given that studies indicate low attendance by YoC. Further, given that bias related to race-ethnicity is frequently experienced from within the LGBQ community , schools and CBOs that serve SMY should be cognizant of and address racism or ethnocentrism that is present in their programs. Finally, our review identified several limitations and areas of research that were underrepresented in the extant literature. We discuss these limitations in terms of the sociodemographic profile of participants who were represented in the extant literature and the conceptual approaches that were used to understand the lives of SMYoC. --- Who is represented in studies of SMYoC? Race-ethnicity-Although this review was focused on YoC broadly, the extant literature is not representative of the growing diversity of racial-ethnic groups in the United States. Asian and Latino populations are two of the fastest growing ethnic groups in the United States . Yet only five reports in our review focused exclusively on the experiences of Asian/Pacific Islander SMY, and only seven focused exclusively on Latino SMY. Given the changing demographic makeup of the United States, it is critical that future studies use purposive sampling methods to understand the experiences of these youth. In addition, many of the reports in this review used pan-racial-ethnic labels to identify youth, rather than youths' country of origin. Given the importance of the broader political, economic, and social histories and experiences that various racial-ethnic groups hold in the United States, future research should examine potential within-group differences . We acknowledge that studying these withingroup differences is complicated and requires complex sampling and recruitment strategies. Nonetheless, these nuances are likely important for tailoring prevention programming aimed at reducing disparities and promoting positive development among these populations. Gender-The majority of reports included in our review only sampled men; thus, the experiences of women and trans persons are largely invisible in the research. The few studies that did include men and women found gender moderation of the intersection of raceethnicity and sexual minority status on adjustment . More research is needed to understand how gender identity and expression intersect with race-ethnicity and sexual orientation to inform health and normative developmental processes. Age-Most of the reports included in our review had participants who were in their 20s rather than in their teens. Given that youth are coming out at earlier ages , it is increasingly important that research capture the experiences of all SMY, not just those over the age of 18. Further, given that research documents that sexual orientation health disparities may be heightened in adolescence , studies of early adolescents can help understand risk and resilience processes, particularly in the context of longitudinal studies spanning from early adolescence through young adulthood. --- Intersectionality Intersectional approaches seek to understand disparities and marginalization processes by considering the strengths and challenges that individuals face because of their multiple social identities. Rather than examining one social category at a time, an intersectionality perspective examines how multiple social categories contribute collectively to the experiences of individuals . Research that is not approached from an intersectional lens ultimately contributes to intersectional invisibility and a biased understanding of social categories. Notably, several of the reports in our review-especially those that were comparative in nature-did not intentionally focus on intersectionality, which may have resulted in fewer studies that directly assessed risk and resilience factors that may explain outcomes among YoC or potential buffers of bias-based experiences . For example, a recent review clearly demonstrates that racial-ethnic identity developmental processes are associated with positive adjustment . Other research has demonstrated that sexual identity affirmation is meaningful for the well-being of sexual minorities . Yet only seven studies examined these co-occurring processes for SMYoC . Thus, research focused on SMYoC needs to attend to these normative co-occurring developmental processes, particularly during early adolescence . Similarly, while overt experiences of bias were studied, microaggressions-or subtle, everyday experiences of mistreatment-were rarely discussed in the reports included in this review . Microaggressions based on race-ethnicity and sexual orientation are argued to be pervasive and more harmful than overt acts of discrimination because they are ambiguous . Specifically, targets of microaggressions may become distressed while trying to determine whether the particular encounter really was discriminatory . Further, if the intent of the perpetrator is unclear , targets of microaggressions may question whether they are being oversensitive because of the ambiguous nature of microaggressions. Work that examines microaggressions among SMYoC would significantly contribute to current scholarship. --- Limitations and conclusions This analysis and review is not without limitations. First, the reports included in our review differed in terms of sample size, which makes comparisons across studies difficult at times. However, given that our goal was to be comprehensive, we did not want to value findings from large, quantitative studies over smaller, qualitative or mixed-methods studies by excluding reports with small sample sizes. Second, our review was limited to the published studies focused on SMYoC in the United States; thus, our findings are likely not generalizable to the broader population of SMYoC in the United States given the limitations of sampling, definitions of SMYoC, and methodologies used in the extant research. Future research is needed to understand how the intersection of sexual orientation and raceethnicity contributes to health, well-being, and development in other contexts. Finally, given the structure of the extant literature, our review focused on broad racial-ethnic categories; as noted in the preceding, future research is needed to understand within-and between-group differences and similarities. Given the relatively brief history of research on SMY, there are an impressive number of studies that have examined the experiences of SMYoC. Future research is needed to ensure that the diverse experiences of all SMYoC are captured in order to best inform future policy, program, and prevention and intervention efforts. In addition, while an examination of the intersection between race-ethnicity and sexual orientation is an important first step, it will be important for future studies to consider how other socially relevant characteristics explain risk and resilience among SMY populations. Plot of the years that reports included were published. I = 1 report. --- --- Table 1 Articles included in each topic area. --- Topic area N Articles

This study analyzed the content of 125 unique reports published since 1990 that have examined the health and well-being-as well as the interpersonal and contextual experiences-of sexual minority youth of color (SMYoC). One-half of reports sampled only young men, 73% were noncomparative samples of sexual minority youth, and 68% of samples included multiple racialethnic groups (i.e., 32% of samples were mono-racial/ethnic). Most reports focused on healthrelated outcomes (i.e., sexual and mental health, substance use), while substantially fewer attended to normative developmental processes (i.e., identity development) or contextual and interpersonal relationships (i.e., family, school, community, or violence). Few reports intentionally examined how intersecting oppressions and privileges related to sexual orientation and race-ethnicity contributed to outcomes of interest. Findings suggest that research with SMYoC has been framed by a lingering deficit perspective, rather than emphasizing normative developmental processes or cultural strengths. The findings highlight areas for future research focused on minority stress, coping, and resilience of SMYoC.

Introduction Intimate partner violence is the most common type of violence against women , affecting about 30% of women worldwide [1]. IPV refers to "any behaviour by a man or a woman, or a boy or a girl, within an intimate relationship, that causes physical, sexual, or psychological harm to the other person in the relationship" [1]. IPV represents a serious violation of women's human rights and an urgent public health priority; it is an important cause of injury, and is a risk factor for many physical and psychological health problems [1,2]. IPV negatively affects the behaviors and emotions of children and often causes severe problems in the family and society through the intergenerational transmission of violence [1,[3][4][5]. The distribution of partner violence shows exceptionally large differences, with the IPV prevalence experienced in the past 12 months in many high-income countries being under 4% and that in some low-income settings being at least 40% [6]. Little is known about the factors that explain the difference in the prevalence of partner violence across countries [6]. Cultural views toward VAW have become a major concern worldwide, yet information remains lacking [7]. Korean society has shown lenient attitudes toward partner violence owing to its patriarchal, male dominated and family-centered culture [8]. However, with the change in the role of women owing to rapid modernization and Westernization, the public attitude toward partner violence is changing [8][9][10]. In Korea, the Gross National Income per capita has increased to 30,600 USD in 2018 [11]. Meanwhile, the nationwide prevalence of physical, psychological, sexual, or economic violence against the female partner from a male partner in the last year has remained high, at 33.1% in 2007, 39.1% in 2010, and 29.8% in 2013 [12,13]. According to the Korean Women's Hotline, at least 887 women were victims of homicide by an intimate male partners in 2009-2018, and 727 women survived a murder attempt against them [14]. As there may be more unreported cases, it is likely that the number of victims may exceed official reports. It is estimated that over 75% of VAW is perpetrated by the male intimate partner [15]. Across a sample drawn from eight low-and middle-income countries, 31% of men report having perpetrated physical violence against a partner in their lifetime [16]. Although there are differences among the countries in the percentage of men who perpetrated physical and/or sexual violence in nine Asian and Pacific countries, the percentage ranges from 25.4% to 80%. When including psychological and/or economic violence, the range rises from 39.3% to 87.3% [17]. Men's perpetration of VAW results from a complex, interconnected ecology of psychological, economic, and sociological factors [18]. There are many risk factors associated with IPV, such as age, economic status, educational level, marital status, exposure to prior abuse, acceptance of violence and traditional gender roles, and community sanctions [1,5,19,20]. Among other things, gender inequality in the distribution of power and resources and discrimination against women are the main root causes of IPV [1,2,20]. Nevertheless, there are multiple risk factors associated with both perpetration and victimization at multiple levels, such as personal history, personal background, social norms. Risk factors related to the occurrence of IPV differ by country and by the victim and perpetrator of abuse, and the response to them may differ as well [5,16,17,19,20]. In Korea, research on the risk factors related to the occurrence of IPV is lacking and has focused on female victims. Therefore, concurrent research on male perpetrators is necessary and ecological approach that considers various factors related to the occurrence of IPV at the individual, family and social level is useful. This study aimed to identify the factors that affect the occurrence of IPV from the aspects of both female victim and male perpetrator, based on the socio-ecological theory that involves considering factors contributing to the problem at various levels. Until recently, VAW was largely invisible within national and international statistics and surveillance systems [20]. Research on IPV was mostly conducted by individual research teams in Korea. The Domestic Violence Survey is a triennial study conducted since 2007 by the Ministry of Gender Equality & Family according to the Act on the Prevention of Domestic Violence and Protection, etc. of Victims [13]. Analysis of national survey datasets can be carried out to describe phenomena, generate knowledge for nursing practice, or shed light on the present, the past, or trends over time [21]. Therefore, this study conducted a secondary data analysis using the 2016 DVS. The results are expected to provide data that are helpful for policies and interventions for preventing and responding to IPV. The aim of this study was to identify the factors associated with IPV occurrence in Korean adults. The specific aims were as follows: to identify the prevalence of IPV, analyze differences in women's victimization in IPV and men's perpetration of IPV according to multilevel factors, and identify the multilevel factors affecting women's victimization in IPV and men's perpetration of IPV. --- Conceptual framework An ecological model was used as the conceptual framework for this study. The ecological model supports a comprehensive public health approach that not only addresses an individual's risk of becoming a victim or perpetrator of violence but also the norms, beliefs, and social and economic factors that create the conditions for IPV to occur [5,18,22]. The occurrence of IPV is determined by a complex interplay among various factors at each of the ecological layers. This study included witnessing violence between parents or experiencing child abuse by a parent in one's personal history, age, and structure of decision-making with the partner in the microsystem; education level, occupation type, and household income in the exosystem; and attitude toward gender roles and IPV, awareness of neighborhood and community, awareness of IPV-related laws and policies, and awareness of support facilities in the macrosystem. The conceptual framework for this study is shown in Fig 1. --- Methods --- Design and sample Our study used a cross-sectional and correlational in design with SDA using the 2016 DVS. --- Description of primary data: The 2016 DVS The 2016 DVS data was used in this study because it is the only large-scale national survey of IPV in Korea. What is referred as IPV in this study was identified using domestic violence data. The DVS has been conducted to examine the DV experience of general adults. Probability proportion sampling and systematic sampling were applied, with one person selected for each household. After the training of surveyors and field preparation, the survey was conducted by visiting each household and distributing a self-reported questionnaire. From September 22 to December 8, 2016, a total 6,000 adults aged more than 19 years were surveyed [23]. --- Samples of SDA This study analyzed the data of 3,961 participants, composed of 1,272 men and 2,689 women with a partner. The eligibility criterion was that they reported being married or in a commonlaw marriage. The exclusion criterion was being without a partner . --- Variables Dependent variables. Female victimization in IPV and male perpetration of IPV were set as the dependent variables. IPV was categorized into four types: psychological, physical, economic, and sexual. The Revised Conflict Tactics Scale [24] was reviewed, revised, and complemented to create a list of 15 behaviors; female victim was considered to have occurred if the woman had been subjected to any one of these 15 behaviors in the past 12 months. Likewise, male perpetrator was categorized as having perpetrated violence if the man responded that he engaged in at least one of the 15 items in the past 12 months. Independent variables. Personal history. For witnessing inter-parental violence in childhood, if the participant witnessed one or more of the three items when they were 18 years old or younger, the response was classified as "yes." Items assessed whether the participant has witnessed the following behaviors between their parents: used harsh words, such as swearing or slighting words; struck with a hand or foot; hit or injured with a belt, club, or others. For childhood maltreatment by parents, the participant's response was classified as "yes" if they experienced one or more of six items related to physical and psychological abuse as well as child negligence by parents when they were aged 18 years or younger. Six items assessed whether the participant heard harsh words, such as swearing or slighting words; was hit on the hand or legs with a cane; was hit with a hand or foot; was hit or injured by a belt, club, or others; was not given food or taken to hospital when sick; and was alone when they should have been with an adult. Microsystem. The structure of decision-making between partners was asked for the person who made the main decision for four items on a five-point Likert scale: "1. Completely decided by me," "2. Mostly decided by me," "3. Decided together through a discussion," "4. Mostly decided by my spouse," and "5. Completely decided by my spouse." The mean of the four items was determined. A mean greater than 3 points signified that the decision-making power of the spouse was greater. Exosystem. Education level, household income, and occupation type were investigated. Occupation type was classified into "professional and manager," "office job," and "other." Other jobs included positions in services, sales, agriculture, forestry, and fisheries, as well as device and mechanical operation and assembly workers, simple labor workers, and soldiers. Macrosystem. Attitude toward gender roles was measured using a total of seven items on a four-point Likert scale, with higher average scores representing more patriarchal attitude toward gender roles. The seven items included four items on whether the role of the man was to be a leader in society, promote important work, lead sexual relationships, and hold decision-making power over the household finances, and three items on whether wives should mostly do housework, obey the husband's decision on whether they should have a job, and prioritize the opinion of the husband for important decisions regarding their children. Attitude toward IPV was measured using a total of 10 items on a four-point Likert scale, and higher mean scores indicated a more tolerant attitude toward IPV. The 10 items were composed of three items on whether one can be violent to one's family if one loses control in anger, if one is under too much stress, or if one drinks too much; two items on whether an abuser can be forgiven if they genuinely repent after the violence or if they were maltreated in childhood; and five items on whether intimate violence is a within-family problem, whether women should endure the violence to protect their family, whether women can leave the relationship if they really want, whether they find it difficult to understand why someone would not leave an abusive spouse, and whether it is correct for the violent spouse to leave home if they engaged in violent behavior. Awareness of IPV-related laws and policies and of support facilities were determined using of six and four items, respectively, and scores were assigned based on the participant's recognition of each item: 1 for recognition and 0 for non-recognition. Higher total scores represented higher recognition. Awareness of the neighborhood and community was measured using a total of eight items on a four-point Likert scale, and higher average scores represented a higher sense of community. --- Data analyses All data analyses were conducted with IBM SPSS 24.0. Weighted data were used in the analysis of the report of the 2016 DVS [23]. The weighted data were calculated through three processes, namely, design weight , adjustment to non-response, and adjustment of using population information. Descriptive statistics based on frequencies, weighted percentages, and means were generated to describe socio-demographic variables, types of IPV, and other IPV related data. Next, chi-squared tests and t-test based on weighted percentages and means were used to analyze the differences in women's victimization in IPV and men's perpetration of IPV according to the multi-level factors based on the ecologic model. Subsequently, multiple logistic regression analyses were conducted to identify the factors affecting women's victimization and men's perpetration. The results are presented as p-values, and 95% confidence intervals . --- Ethics statement Considering the nature of the survey, the raw data were collected based on WHO guidelines for ethics and safety related to the research and survey of VAW [23]. The investigators of the present study requested raw data from the MEGF based on the Enforcement Rules of the Act on Provision and Active Use of Public Data [25]. To perform the SDA, institutional review board exemption approval was obtained from the Institutional Review Board of the Kangwon National University . --- Results --- Prevalence of female victimization and male perpetration of IPV The prevalence of female victimization in IPV in the past 12 months as reported by women was 12.1%. By type, psychological violence comprised 10.5%, physical violence, 3.3%, sexual violence, 2.3%, and economic violence, 2.4%. The prevalence of male perpetration in the past 12 months as reported by men was 11.6%. By type, psychological violence comprised 10.5%, physical violence, 2.1%, sexual violence, 1.8%, and economic violence, 1.5% . --- Differences in IPV occurrence by multi-level factor Differences in female victimization in IPV by multi-level factor. Participants who witnessed their parents' violence in childhood and those who experienced childhood maltreatment by their parents had a higher occurrence rate of IPV victimization , compared with participants who did not have these experiences . The IPV victimization rates of participants who graduated middle school and high school were higher compared with those without IPV victimization . Compared with the comparison group , the IPV victimization rate of participants whose income was between 2 and 3 million KRW was higher . Moreover, compared with the comparison group , the IPV victimization rate of participants whose job was classified as "other" was higher . The mean score for attitude on gender roles among women who were victims of IPV was higher compared with women who were not victims , showing that the attitude toward traditional patriarchal gender roles was stronger among victimized women . The mean score for attitude to IPV among women who were victims of IPV was relatively higher compared with women who were not victims , showing that the former have greater tolerance toward and have a lower level of awareness of IPV . Differences in male perpetration of IPV by multi-level factor. Men who witnessed parental violence in childhood and men who experienced childhood maltreatment by their parents had a higher rate of IPV perpetration , compared with men who did not have these experiences . The mean score for attitude toward IPV was higher among men who had experiences with perpetrating violence against an intimate partner compared with men who did not , showing that the former had a more tolerant attitude toward and lower awareness of IPV . --- Factors associated with female victimization and male perpetration of IPV The results of the multiple regression analysis on the women participants showed that 16.5% of the variance was explained by the four variables of witnessing parents' violence in childhood, childhood maltreatment by parents, household income, and attitude to IPV. The IPV victimization rates among women who witnessed parental violence in childhood and childhood maltreatment by their parents were 2.92 times and 1.7 times higher than the rates among women who did not have these experiences. Compared with women whose income was above 4 million KRW, women whose income was between 2 and 3 million KRW had an IPV victimization rate that was 1.97 times higher. In terms of jobs, the IPV victimization rate among women who had office jobs was 0.47 times lower compared with women with "other" jobs. In terms of the attitudes toward IPV, each point increase on the mean attitude toward IPV was found to be associated with a 2.25-fold increase in IPV victimization rate . The results of the multiple regression analysis on the men participants showed that 10.4% of the variance was explained by witnessing parents' violence in childhood and attitude to IPV. The IPV perpetration rate among men who witnessed parental violence in childhood was 2.64 times higher compared with men who did not. In terms of attitudes toward IPV, each point increase on the mean attitude toward IPV was found to be associated with a 2.68-fold increase in IPV perpetration rate . --- Discussion This study identified the risk factors that affect the female victimization and male perpetration of IPV in Korea to provide data for interventions and policy formation for IPV prevention and response. The prevalence of female victimization in the past 12 months as reported by women in 2016 DVS was 12.1%, showing a decrease compared with 39.1% in 2010 DVS and 29.8% in 2013 DVS. In contrast, the prevalence of male perpetration as reported by men was 11.6% in 2016 DVS [12]. Although male perpetration data reported by men from 2010 and 2013 DVS cannot be found, we hypothesized these to be similar to the prevalence rates reported by women. Thus, the prevalence rates of female victimization of IPV is gradually decreasing in Korea. This decline is interpreted to be a result of the multi-dimensional efforts to prevent partner violence for more than 20 years since the enactment of the Act on the Prevention of Domestic Violence and Protection, etc. of Victims, along with Korea's economic growth and women's advancement in society [26]. The growing awareness of gender equality owing to the expansion of new family policies and institutional changes based on the emphasis on democratic and equal gender relations in family and society [27] may have also helped. Meanwhile, the reports of analyzing 2010 DVS and 2016 DVS showed that awareness of IPV-related polices and laws and the willingness to report IPV gradually increased, consistent with the decrease in female victim prevalence [12]. However, the attitude toward patriarchal gender roles still remained the same, particularly among men, showing the need for continued effort [12]. It is important to consider the possibility that the decrease in the prevalence of IPV in South Korea is the result of survey participants being less likely to disclose personal aspects of IPV. Societal stigma related to the victimization and perpetration of IPV has changed with the shift in societal attitudes and female victims' feelings of shame or embarrassment [28]. Our study found that the occurrence of female victimization and male perpetration were about three times as high in men and women who have witnessed their parents' violence. This finding is consistent with many previous studies; witnessing parents' violence in childhood is an important or the strongest risk factor of IPV occurrence in both victimization of women and perpetration by men [5,20,29,30]. In addition, research on the risk factors of men's lifetime perpetration [16,31] and on IPV victimization of women [32] has also showed that witnessing parents' violence is an important risk factor of IPV occurrence. The strength and significance of the correlation between witnessing of inter-parental violence and IPV perpetration and victimization suggest the intergenerational transmission of behaviors and gender norms [16,33]. Observation and imitation occur throughout the lifespan but can be particularly important for children and adolescents. Interrupting this cycle is critical to reducing perpetration [16]. Next, we found that women who experienced childhood maltreatment by their parents in childhood were twice as likely to become victims of IPV compared with women who did not. This finding is consistent with many previous studies, which reported that abuse experience in childhood is an important risk factor of IPV occurrence in abused women [1,20,29,30,32]. This finding highlights how childhood experiences influence the likelihood of people later becoming perpetrators or victims of IPV, as well as the need for early childhood interventions, especially for children growing up in families where there is abuse [2]. According to a systematic review of the predictors of DV perpetration and victimization, abuse and family of origin problems experienced in childhood and adolescence, such as witnessing inter-parental abuse and experiencing maltreatment from parents, are consistent predictors of DV perpetration and victimization for men and women [33]. An increasing number of findings show that parental programs, including home visits and education, can reduce child abuse and have an increased effect of reducing children's violent behavior in the future. Therefore, parental programs to prevent and intervene in partner violence between parents are necessary because such initiatives would address not only parents' partner violence but also maltreatment of children [34]. On the other hand, our results showed that men's experience with violence from parents in their childhood had a significant relation with IPV but was not a risk factor that affected IPV occurrence. This was not consistent with previous findings [1,4,20,30]. A study on 242 Korean men showed that men's experience with violence in their childhood is a risk factor that affects the occurrence of male perpetration [35]. Given the inconsistent results, additional research is necessary. Attitude toward IPV was analyzed as an important risk factor for women and men in this study. Our finding was consistent with many previous studies, which reported that beliefs and attitudes about IPV are related to the occurrence of IPV for men and women [1,5,20,29,36,37]. Men who believe wife-beating is acceptable are more than four times as likely to report recent violence against their wives [38,39]. According to men's lifetime IPV studies in eight low-and middle-income countries, tolerant attitudes toward IPV and inequitable gender attitudes are associated with a higher likelihood of ever perpetrating physical IPV [16]. A study of Korean adults found that men are more tolerant toward IPV compared with women, and perpetrators are likely to victim-blame for the cause of perpetration and have lower awareness of responsibility toward the violence. Further, in the same study, both men and women score low on the item, "Men who perpetrate violence need to take responsibility for their behaviors," and lack awareness that domestic violence is not right. This implies a possibility of continued female victim by partners in the future [40]. Patriarchal societies, in general, view men as entitled to greater power, privilege, and control of women and children, and consequently, the right to punish them for perceived misbehavior [41]. These results are interpreted to have been found because Korea's social system and family norms continue [42] to adhere to the long-maintained patriarchal culture, despite the recent change in attitudes toward gender roles. Overall, when surveying Korean adults regarding their attitudes toward IPV, participants presented a tolerant attitude toward IPV, with gender having the greatest influence on these attitudes. Men have more condoning and tolerant attitudes toward IPV than women. It is common among Korean men to blame female victims rather than to think that violence toward them is the real problem [40]. As women's tolerant attitude toward IPV is associated with their actual experience of being a victim, changes in the attitude toward partner violence are important [43]. Because the attitude and cultural norm of IPV being acceptable are the most significant factors related to the likelihood of violent behavior [44], efforts to change this tolerant attitude in men and women are necessary. In Korea, society upholds a mixture of long-held Confucian traditions and modern values. As such, differences in the roles of men and women are attributed to the superior male authority in society and at home. Thus, the gender role norms and violence-tolerant cultures of society affect the formation of individual violence-tolerant attitudes, and in turn, individuals' learned perception and attitude sustain the norms and cultures of society. In other words, while the perception about partner violence is individualistic, it is socially formed and maintained because the individual learns and embraces the gender role norms and tolerance toward the violence of society [12]. Further, witnessing or experiencing abuse during childhood can lead to a tolerant attitude toward IPV, and thus has an intergenerational transmission effect that leads to spousal abuse. Therefore, this link needs to be managed [4]. In interpreting our analysis in relation to social-ecological theory has led to the observation that female victims of IPV are affected by both their personal history and the attitudes toward IPV at a macrosystem level. Attitudes toward IPV are related to male-centered gender inequality and discrimination against women, which still persist in Korean homes and society. Korean society is both more conservative and patriarchal than that of Western, and even Japanese cultures [45]. As can see from the gender role attitude score observed in this study, women's traditional values are weakening; however, a significant number of men still exhibit and adhere to a patriarchal attitude [45]-an attitude which is a predictor of violence toward women [45]. Women's educational background and economic activities, in particular, have increased, but the role of men in the home has not changed considerably, presenting a double burden for women [46]. Although the sense of responsibility to support parents-in-laws at home has weakened, the uneven gender roles in marital relations have hardly changed, while the obligation to educate children is on the rise [47]. These results demonstrate that unequal roles persist between men and women within families, indicating that patriarchal norms persist among nuclear families. In our study, we analyzed "Attitudes toward gender roles" in DVS at the macro-system level and "The structure of decision-making between partners" at the micro-system level. However, these were not related to the occurrence of violence toward women. Therefore, additional research is needed on the socialization process that affects IPV attitudes. Shedding light on the predictors of IPV occurrence and performing effective interventions should not be separated from the cultural context of the community [48]. Therefore, awareness-enhancing education programs and interventions to prevent IPV need to be provided in various ways that reflect the gender and life cycle differences at the individual and societal levels based on these results. Finally, many previous studies have shown that low socioeconomic status was associated with high rate of IPV in both victimization of women and perpetration by men [1,5,20,49]. On the other hand, our study found that low SES affected the rate of female victimization but not male perpetration. This finding is consistent with previous studies, which reported that women from low socioeconomic backgrounds had higher victimization of IPV [1,5,20,49]. In Korea, as of 2014, men's participation in economic activities was 74%, compared to 51.3% among women. The proportion of non-regular workers among female wage workers was 39.9% , and the ratio of women's wage in contrast to that of men's in similar jobs was only 67%. The rate of career disconnection among married working women accounts for a high rate of 47% [50]. These results suggest that the economically unfavorable situation of women in Korea-whereby women have lower employment rates and incomes than men-may act as a contributing factor to withstand IPV if it occurs. To reduce IPV victimization, policy measures are needed to ensure women's stable jobs and income such as ensuring equal employment opportunities, salary and minimum wages. In sum, the common risk factors of IPV in women and men were witnessing inter-parent violence in childhood and tolerance toward IPV. Another risk factor was the experience of maltreatment from parents, but only for women. Korean women remain underprivileged compared with men both in society and in the family. Thus, the witnessing of parents, who have authority, being violent and the experience with maltreatment from parents in childhood link to the tolerance of the abuse of husbands, who have authority, when they become adults. In contrast, men's experience of maltreatment in childhood does not link to the perpetration of violence as husbands with authority; it is the tolerant attitude toward IPV that has a greater effect on IPV occurrence. In other words, it can be interpreted that both men and women are affected by witnessing their parents' relationships in their childhood and by the changes in the social attitudes toward IPV. after the enactment of the Act on the Prevention of Domestic Violence and Protection, etc. of Victims. However, compared with men, women also have risk factors of IPV victimization such as experience with maltreatment from parents in their childhood, low socioeconomic status, clearly showing women's vulnerability. In this study, personal history and macrosystem level factors had remarkable effects on the occurrence of IPV for both men and women. For women, exosystem level factors also affected IPV occurrence. The number of risk factors is greater for the victimized women compared with the perpetrators men. Thus, it is necessary to strengthen the interventions and policies comprehensively to reduce these factors at multiple levels. Primarily, focus should be on changing harmful social norms through the education and campaign for gender equality and non-violence for both men and women across the life cycle from a young age and on providing various services to parents and children by detecting family violence early [51]. Further, it is necessary to establish integrated policies that consider women's stable work and income. --- Limitations This study provides a comparison of the risk factors that affect the occurrence of IPV from the aspects of both female victims and male perpetrators, suggesting a comprehensive yet differentiated direction of intervention on both sides. As this study is an SDA research using DVS, there were limited items in each level. In particular, the microsystem level had only a few items, which could have led to the lack of significant results. Therefore, it is necessary to add the items common in other studies. Further, as this study used a cross-sectional design, it was not possible to derive causal relations. Moreover, the use of national data based on the Act on the Prevention of Domestic Violence and Protection, etc. of Victims limited the sample to only those in legal marital or de facto relationships; thus, the study did not include adults who were divorced or dating. Further, the participants in the DVS are divided into men and women instead of couples. Thus, we could only analyze the results by sex; an integrative analysis of the risk factors that affect the IPV occurrence among women and men could not be undertaken. Owing to the limited SDA research on IPV, general research findings were mostly used in the discussion. In particular, the observations from the IPV prevalence survey indicate that there is a possibility of intentional under-reporting of IPV occurrence due to the fear of retribution and feelings of shame or embarrassment felt by female victims. Although there are cases of over-reported IPV-such as to obtain secondary gains or favorable results from divorce [52], DVS surveyed at the national level is likely to be underreported; this, as such, should be considered when interpreting the reduction in prevalence of cases. Because longitudinal survey data capture is more likely to provide an accurate representation of the population prevalence of IPV [28], it would be beneficial if future cohort survey studies were conducted to identify the accuracy of IPV occurrence. --- Conclusions There are common and different areas between the risk factors that affect IPV occurrence among countries and between genders. Our findings contribute to the understanding of risk factors for IPV occurrence in Korean men and women. This study revealed that witnessing inter-parent violence in childhood and the attitude toward IPV were strong risk factors of IPV in both men and women. In contrast, the experience of maltreatment from parents in childhood was a risk factor only for women. For women, witnessing inter-parent abuse in childhood, experience with maltreatment from parents in childhood, tolerant attitude toward IPV, low income, and certain occupations were risk factors of IPV. In contrast, for men, only witnessing inter-parent abuse in childhood and tolerant attitude toward IPV were risk factors of IPV. These results support the previous findings that witnessing inter-parent violence and experiencing maltreatment from parents in childhood lead to the internalization of violence, leading to the tolerant attitude toward family violence, and in adult women, to accepting abuse from husbands. In contrast, men showed differences in these risk factors. Additional in-depth research may identify the causes for these differences. Further, because the trend of prevalence of female victimization in Korea is decreasing, it is necessary to explore the future direction of interventions and support through research that examines the specific contributors to the changes in these trends. It is also necessary to seek a differentiated approach from the perpetrator's perspective through the accumulation of DVS research data on men perpetrators. In the future, IPV prevention programs should increase focus on transforming gender norms and attitudes toward IPV and addressing childhood abuse. To change the social attitude toward IPV, education and campaigns should be strengthened throughout the life cycle from a young age for both men and women. IPV-related service providers and policy makers may need to consider childhood family violence history in both men and women in the context of IPV. Moreover, the importance of strategies and policies to empower women socio-economically and personally needs to be recognized. In particular, as women victims are affected by more risk factors than men perpetrators, it is necessary to establish comprehensive policies that consider the risk factors at the ecological levels. --- --- Data curation: Young-Ran Han.

This study aimed to identify factors affecting the occurrence of intimate partner violence (IPV) in Korean adults aged 19 years and older. Specifically, we identified the factors in women's victimization in and men's perpetration of IPV. This study adopted a cross-sectional and correlational design in conducting secondary data analysis of the 2016 Domestic Violence Survey in Korea. Men (N = 1,272) and women (N = 2,689) with partner were included in the analysis. The ecological model was used as a conceptual framework. Multiple logistic regression analyses were conducted to identify factors affecting women's victimization in and men's perpetration of IPV. The results showed that the prevalence of IPV against female partner was 12.1%. IPV occurrence was higher among women (Odds ratios (OR) = 2.92, Confidence intervals (CI): 1.84-4.63) and men (OR = 2.64, CI: 1.62-4.32) who experienced witnessing inter-parental violence in childhood, and among women (OR = 2.25, CI: 1.40-3.61) and men (OR = 2.68, CI: 1.59-4.52) with a tolerant attitude toward IPV. The occurrence was higher among women who experienced childhood maltreatment by the parents (OR = 1.70, CI: 1.03-2.82) and women whose income was 2 million Korean Won (KRW) to 3 million KRW compared with women whose income was above 4 million KRW (OR = 1.97, CI: 1.10-3.55). The occurrence was lower among women with office jobs compared with other jobs (OR = 0.47, CI: 0.26-0.84). Based on the results of this study, early intervention in abusive families to reduce the negative impact of abuse experiences and witnessing inter-parental violence in childhood, and education and publicity for changing attitudes toward IPV are necessary at the individual and societal levels. The formation of policies for the stable workplace and income of women are required.

INTRODUCTION Marketing strategies used by electronic cigarette companies have been associated with the uptick in e-cigarette use among youth. 1 Many studies have found that the promotion of e-cigarettes through various channels and social media influencers) has lent to increased positive perceptions of vaping and intentions to use vaping products and contributed to e-cigarette uptake among youth. [2][3][4] In the UK, the advertising of e-cigarettes is regulated through Article 20 of the revised European Union Tobacco Products Directive , 5 which was transposed into UK law by the Tobacco and Related Products Regulations 2016. 6 The TPD prohibited the advertising of nicotine-containing e-cigarettes in channels with potential cross-border impact , including TV, radio, newspapers, magazines and sponsorship. 7 Online advertising was also prohibited under the TPD, although the regulations left scope for marketers to retain websites containing factual information about e-cigarette products. As there are currently no medicinally licensed nicotine vaping products in the UK, the prohibitions apply to all nicotine-containing e-cigarette products on the market. The TRPR requirements for e-cigarette advertising were set out in 2017 and are enforced by the Committee of Advertising Practice -a self-regulatory body of organisations representing advertising, direct marketing, media businesses --- STRENGTHS AND LIMITATIONS OF THIS STUDY ⇒ This study is one of the first in-depth studies to explore youths' responses to and engagement with electronic cigarette marketing on social media with a focus on the rise in user-generated and influencer content. ⇒ This research is very timely as governments consider the need for further e-cigarette regulations. ⇒ A methodological strength is the rich data set from semistructured interviews with 82 youths aged 11-16, facilitating understandings of youths' exposure to and interaction with e-cigarette marketing on social media. ⇒ Qualitative thematic analysis of the data allows depth of opinions but cannot offer predictions about the frequency of specific opinions with a wider population. Open access and sales promotion endorsed and administered by the independent Advertising Standards Authority . 8 9 Although Rule 22.12 of the ASA CAP code prohibits advertising in online media, social media content for e-cigarettes is permitted in 'non-paid-for space online under the marketer's control, provided claims are factual rather than promotional'. 10 Promotional content includes descriptive language that goes beyond objective, factual claims, significant imagery that is not related to the product, and health claims . 10 Social media marketing includes promotions by means of paid digital advertising, via compensated 'influencers' , and on brand pages that companies can create for free and use to share content. 11 Most adolescents use visual-based social media daily. Usergenerated and influencer marketing content on social media represents a key influence on young people's understanding of products, including e-cigarettes. 12 E-cigarettes are promoted creatively through social media, with welldesigned features including colours, flavour variations, emotional appeal, incentives and even celebrity endorsements. 13 14 Research has shown that online marketing that leads to exposure to e-cigarette advertising, including where it is concealed as information or recommendations from peers, can increase the likelihood of vaping in individuals, including among youths. [15][16][17] In addition, a number of researchers have suggested that e-cigarette advertising could lead to increases in how socially acceptable and desirable cigarettes are perceived to be, and subsequently influence their continued use or possible uptake in smokers, e-cigarette users and dual-users. 18-21 E-cigarette companies have been promoting their products on social media in an attempt to expand their market. 22 23 In addition, e-cigarette companies use social media platforms to promote e-cigarettes and create a socially attractive vaping image. 24 Previous studies have explored youths' perceptions of e-cigarette marketing on social media, finding that social media adverts entice non-smokers to use e-cigarettes, emotional appeals are featured in adverts, use of vibrant colours in the packaging, appeal of flavours and advertisements portray e-cigarettes as less harmful than combustible cigarettes. Liu et al 28 examined youths; perception of tobacco and marijuana messaging on social media and reported similar findings . One limitation of the literature in this area is that it focuses on either smokers, non-smokers and/or vapers. This leaves out non-vapers and dual-users , understudied. It is important to establish whether exposure to e-cigarette adverts on social media influence attitudes towards e-cigarettes in non-vapers and smoking in dual-users. Gaining a better understanding of the nature, extent and impact of e-cigarette marketing including their possible effects on youth e-cigarette perceptions and use can help guide the development of policies and interventions to educate youths about e-cigarettes. 29 This study, therefore, aims to explore youths', including nonsmokers and non-vapers and dual-users exposure to, and engagement with e-cigarette advertising on social media, including user-generated and influencer content. --- METHODS This section closely follows the methodological approach detailed in reference. 30 The data set has been rigorously analysed to answer two distinct research questions which have been answered in two separate manuscripts. Smith et al 30 explored youth's perceptions and engagement with e-cigarettes focusing on one type/generation of e-cigarettes; disposables. This manuscript explores young people's exposure to, and engagement with all types/ generations of e-cigarettes but focuses on advertising on social media, including user-generated and influencer content. We conducted 20 focus groups in Scotland between March and May 2022. Focus groups included between three and five participants . Purposive sampling was used to recruit a diverse sample of youths in terms of age, sex, socioeconomic background, smoking status and vaping status. 31 Eleven groups were recruited through youth workers in local youth organisations. These gatekeepers handed out information sheets and helped achieve the sampling frame in terms of youth demographics and experiences concerning combustible cigarettes and e-cigarettes. The three organisations that helped with participant recruitment worked specifically with young people from disadvantaged backgrounds in urban areas. This recruitment strategy resulted in the inclusion of a range of participants from more and less deprived backgrounds and with experiences of smoking and vaping. Seven groups were recruited through the Schools Health and Wellbeing Improvement Research Network Newsletter which is distributed monthly to over 500 schools in Scotland. The remaining two groups were recruited via personal networks directly by MJS. Focus group discussions were facilitated to allow the research team to explore opinions on online advertising and marketing of e-cigarettes. Friendship groups of 3-5 participants were used to facilitate in-depth insights and promote participant interaction. Each participant was given a £20 shopping voucher as compensation for their time. Prior to the start of the focus groups, participants completed a short anonymous questionnaire about their age, biological sex, postcode, smoking and e-cigarette use status. For both combustible cigarettes and e-cigarettes, the questionnaire asked participants to specify whether they had tried or used them in the past or were using them at the time of the study. Based on a scoping review of user-generated content and influencer marketing involving e-cigarettes on social media, 14 a Open access topic guide was developed which covered three key areas, including use of social media, perceptions of social media influencers versus user-generated e-cigarette content on social media and views on the marketing of e-cigarettes on social media. Images of different types of e-cigarettes and e-liquids posted by social media influencers and general users were used as conversation starters found during the scoping review part of this research. 14 Group discussions were facilitated by MJS . Ten of the groups were conducted online using Microsoft Teams and 10 of the groups were conducted face-to-face. Of these, one of the groups was conducted on the youth organisation's premises, and the other nine were conducted at the school, with representatives of the youth organisation present. Groups lasted between 40 and 66 min. Field-notes reflecting on the focus group and individual issues discussed were written up for each group. All focus groups were audio recorded and transcribed verbatim. --- Analysis Following Braun and Clarke's six-phase framework for thematic analysis, 32 we conducted thematic analysis of the data from the focus groups. The research team read and reread the transcripts to become familiar with the data, and then iteratively constructed a coding frame to enable consistent organisation of relevant data. NVivo was used to organise categories on the basis of inductive themes that emerged from close reading of the, capture of both areas of agreement and less typical perspectives across a range of categories. Each transcript was imported into NVivo V.12, coded independently, cross-checked and analysed by MJS and SH . Contradictory cases and group dynamics were discussed, making use of transcripts and field notes. The researchers reflected on their role as researcher, remained constantly aware of their position and took care not to introduce bias throughout the research. To further reduce bias, the researcher recorded the focus groups and analysed them some time after they were completed ensuring a more reflective viewpoint of occurrences. --- Open access Patient and public involvement Patients and/or the public were not involved in the design of the study, development of the research questions, recruitment, outcome measures or conduct of the study. A summary of the results will be disseminated to study participants. --- RESULTS --- Participant characteristics Eighty-two youths aged 11-16 years participated and 35 males ), representing diversity in sociodemographic characteristics and smoking-related behaviours. Age distribution within the sample was skewed slightly towards 14-15-year-olds, with 14-year-olds making up the largest subgroup . While the majority of participants did not currently smoke or use e-cigarettes, the sample included 10 smokers and 18 youths who used e-cigarettes. Deprivation rank was assigned using the Scottish Index of Multiple Deprivation . 33 The area-level deprivation grouped using a binary deprivation variable in which the three most deprived quintiles were grouped into the most deprived category for the Central Belt of Scotland area. Table 1 describes the focus group composition and participants in more detail and table 2 summarises smoking and e-cigarette use among the sample. Perceptions of advertising tactics used by the e-cigarette industry to target youths Youths discussed three ways e-cigarette companies market their products to target young people. --- Influencer or celebrity endorsements During the focus groups, participants frequently discussed the prevalence of social media influencers advertising e-cigarette products. When discussing why e-cigarette and/or tobacco companies would want to use social media influencers to advertise their products, participants highlighted the popularity of social media and how this would increase the reach of the adverts. Participants reported seeing marketing and advertising of e-cigarettes on social media and that they were portrayed as 'cool' and 'fashionable' by social media influencers. I If an influencer posts it, a lot of people see it and be like, that's the trend now, it's a cool thing to do, so I want to take part in that. It's like they're showcasing them. It's not just like it's in the background or whatever, it's not a normal pose to have them in, it's like they're showcasing them and that they are a fashionable thing to carry or to have. Participants discussed that this would influence the viewer to try the product, thus increasing the prevalence of use. People get tempted easily by seeing the post. Like if they saw some influencers trying that stuff [ecigarettes], people get tempted quickly to try that stuff, because then that person is trying it so the like the kids would think this is something that I should also try that stuff like that, that's the point. When you see what an influencer is, most people think oh if they're doing that, then it's cool, so it like, it's making more people want to buy the vapes, and stuff. Several participants discussed how social media posts make it too easy for underage youths to purchase and hide the fact you are buying e-cigarette products from websites. The ones that stood out to me the most were the Ben and Jerry's and fruit ones compared to the tobacco ones. I feel like, if I was going to buy them, I would buy them 'cause they are attractive and are things I like. Several participants stated that the variety of flavourings was to appeal to youths. I feel like no one really wants to taste tobacco, to be honest. But I think putting out there with like grape and blueberry flavourings are going to attract younger people to the marketplace. I think it's an intentional design and I feel like that is probably what it does attract youths. --- USING VIBRANT COLOURS TO ATTRACT YOUTH ATTENTION In addition, participants spoke about the presentation of e-cigarette products on social media, for example, the use of vibrant colours and filters. You can tell by the picture they're promoting them. Because if it was not promoting them you would have like a red colour, it's like a stop, it's kind of like a no colour. But they are using really positive colours, like pink and green. --- Connecting emotionally with youths Participants discussed the emotional connections and positive portrayal of e-cigarettes on social media, stating that individuals in the social media posts 'looked happy' to be using the products. I think the person who posted this product, I think he or she was trying to emphasise the fact that these products are good because they look happy to be advertising it. So, you should try them and like they were not promoting vaping negatively or that it is bad at all. I think they were trying to sell the aesthetic idea of this product. It's the way they're all holding them, they looked quite passionate with a smile on their face. A big cheesy smile. Participants highlighted the use of emojis in e-cigarette social media posts. Emojis and emoji sequencing 34 are used to enhance the message meaning. 35 36 When asked if this would affect their interaction with social media posts that included emojis, participants stated that they would be more inclined to read the post, rather than scrolling past. Perceptions of what can be done to limit e-cigarette marketing on social media Participants had divergent views on the marketing and advertising of e-cigarettes on social media. Several participants stated that marketing and advertising should not be allowed as youths, including those of a similar age, could access this type of content and it may encourage them to try e-cigarettes. --- If you see it in colours --- Open access and that prohibiting e-cigarette advertising may limit their role as cessation devices by smokers. Those participants suggested instead of prohibiting the marketing and advertising of e-cigarettes, warnings or other restrictive measures could be imposed on these types of social media posts. I think there needs to be age restrictions on the posts or parents could control what their child sees. DISCUSSION E-cigarettes have become increasingly popular and visible in public life, particularly on social media. Our qualitative thematic analysis of youths' perceptions of and engagement with e-cigarette advertising on social media highlighted a public health concern. Similar to previous research, [37][38][39] youths in our study highlighted the tactics used by the e-cigarette industry to promote and advertise e-cigarettes to underage populations through influencer and celebrity endorsements, variety of flavours and using promotional endorsements. Youths discussed the positive portrayal of e-cigarettes by social media influencers, and they were enticing the viewer to buy the product. Youths highlighted that e-cigarette and/or tobacco companies use social media influencers to advertise their products on social media and by adverting on social media platforms, they will increase the reach of the adverts to a wider audience. We found that youths had divergent views on e-cigarette marketing on social media. Several participants stated that e-cigarette advertising on social media should be prohibited, citing the high prevalence of youths on social media who could access the content. Whereas others suggested a balanced approach to regulating advertising on social media so that they can still be advertised as cessation devices by smokers. Youths suggested that all adverts should contain warnings or other restrictive measures to protect youths and non-users. Previous studies which examined e-cigarette-related social media 25-27 40-44 found that the vast majority of the content depicted positive attitudes towards vaping, while negative characterisations were mostly absent. Our study found that youths had similar experiences when exposed to e-cigarette content on social media. Consequently, there is a risk that youths will be exposed to, and possibly engage with, content that promotes vaping while staying uninformed about the negative aspects, including potential health harms. Our study adds to existing concerns about youths' perceptions of e-cigarettes as cool, fashionable products, 30 45-48 by considering the advertising of the products by influencers on social media platforms, frequently accessed by youths. The use of social media influencers to promote products and celebrity endorsements encourage their followers to share and interact with their content, and ultimately purchase their products. Social media influencers often collaborate with multiple industries in addition to e-cigarette products/brands. These influencers could potentially expose their non-e-cigarette-focused audience to e-cigarette content. 49 Therefore, these influencers could be considered an even higher risk for youth compared with influencers who post exclusively about e-cigarettes. This is problematic and concerning as most e-cigarette brands claim that their advertising and promotional content is meant to target current cigarette smokers to help them switch to e-cigarettes, 50 51 rather than people who do not use nicotine at all. Previous literature has focused on either smokers, non-smokers and/or vapers, leaving out non-vapers and dual-users. Participants in our study were primarily non-smokers and non-vapers, we argue that it is important to explore non-smokers and non-vapers exposure to and engagement with e-cigarette adverts on social media. This is important for helping to determine the potential influence of e-cigarette adverts on social media in non-smokers and non-vapers. Before plain packaging, the tobacco industry successfully used colourful packaging to represent its brand and increase the appeal of smoking. [52][53][54] The marketing tactics used by the tobacco industry were so effective that several e-cigarette manufacturers have used the same trade secrets to advertise their products. [55][56][57] Research has shown the plethora of bright-coloured packaging, eyecatching designs and variety of flavourings available all appeal to youths and may result in experimentation of the products. [58][59][60][61][62][63] Previous research has illustrated YouTube viewers' exposure to e-cigarettes while watching videos on the platform. 64 Our study adds to this by illustrating the potential mechanisms by which exposure to e-cigarette advertising on social media platforms may influence perceptions and resultant trial/use of the products. While restrictions on mass media marketing of e-cigarettes are increasingly considered internationally, social media platforms are more difficult to manage because user-generated content will not be covered by incoming regulations. Our study findings resonate with previous research, which found that youths were easily able to purchase e-cigarettes from the internet due to the absence of age-verification measures used by internet vendors. 65 66 As with all research, our study has some limitations, several of which are similar to that detailed in reference 30 as this research closely followed the same methodological approach. First, and consistent with the qualitative design, the sample does not aim to be representative of UK youth, as our study focused on Scottish youths. However, we did have a diverse sample of both sexes. Second, the study's geographical remit has to be considered when interpreting the findings. The UK is considered as an international leader in tobacco control policy. It is possible that participants' views may have been influenced by Open access the UK's unique favourable policy approach to e-cigarettes and legal and sociocultural context, including low smoking prevalence. Third, using friendship groups may have influenced youth's honesty when answering questions and their responses. Youths may not have wanted to disagree with their peers, thus responded in a similar way to the rest of the group. Fourth, participants in this study were primarily non-smokers and non-vapers and it is possible that this may have impacted in their discussion of e-cigarette advertisements on social media. However, as we used predominately visual prompts based on data found during the scoping review of the research, 14 we did not find that non-smokers and non-vapers contributed less to discussions. If we had not used visual prompts and focused on participants own experiences of what they had seen on social media , this may have impacted non-smokers and non-vapers contribution to discussions. Fifth, the data were collected in different formats , and this may have influenced participants' responses. Two of the online groups were conducted in a classroom with a teacher present, and during seven face-to-face groups, a teacher/youth worker was present in the room. The presence of a teacher/youth worker may have influenced youth's willingness to answer questions and their responses. Finally, two of the groups were recruited through personal networks and this may have impacted the youth's responses. The results from this study should provide guidance for future research. More research is needed to determine the most effective means to counter the favourable/positive aspects of e-cigarettes to reduce youths' interest in product trial and use. Future research could explore if there are different perceptions of e-cigarette exposure in social media between females and males. As females are one of the target groups for e-cigarette use, 67 68 it can reflect the marketing situation of e-cigarettes for females on social media, which deserves further study. In addition, research on the prevalence of warning statements on e-cigarette-related social media posts and the impact this has on youth perceptions would be an important future direction to expand this work. Considering that most youth access multiple social media platforms multiple times per day 69 70 and that exposure to this marketing is related to use, 71 72 our findings indicate a significant public health concern. Given the findings presented by our study, there is a growing need for policymakers to develop comprehensive plans to build more robust measures to protect youths and to restrict the ability of marketers to reach youths with enticing social media content promoting e-cigarettes. In addition, social media platforms should consider implementing more robust measures, such as age restrictions and portraying the negative aspects of vaping, to ensure the prevention of vaping-related content targeted at underage users. --- CONCLUSION The tactics used on social media platforms for advertising and marketing e-cigarettes directly to youths raise concerns and a new generation of youths becoming addicted to nicotine. The results of this study highlight that the e-cigarette industry is using previously employed tactics similar to the tobacco industry to advertise and promote its products on social media. Most youths use social media daily and the depiction of products on social media represents a key influence on young people's understanding of products. These findings may highlight a priority for governmental policy to restrict the ability of marketers to reach youths with social media content promoting e-cigarettes. --- Competing interests None declared. Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research. --- Patient consent for publication Consent obtained directly from patient. Ethics approval This study involves human participants. Ethical approval for the study was obtained from the University of Glasgow's Medical and Veterinary Life Sciences Ethics Committee . Participants gave informed consent to participate in the study before taking part. Provenance and peer review Not commissioned; externally peer reviewed. Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information. Supplemental material This content has been supplied by the author. It has not been vetted by BMJ Publishing Group Limited and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations , and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Objective Electronic-cigarettes (e-cigarette) are promoted creatively through social media and considering the potential influence of social media marketing on young people, we explored young people's exposure to and engagement with social media marketing of e-cigarettes. Design Semistructured discussion groups. Subjects Twenty focus groups with 82 young people aged 11-16 living in the Central belt of Scotland. Methods Youths were asked about smoking and vaping behaviours, social media use, vaping advertisement exposure and were shown illustrative examples of social media content (eg, images and videos) about different messages, presentations and contextual features. Transcripts were imported into NVivo V.12, coded thematically and analysed. Results Youths highlighted a variety of tactics ecigarette companies use, including influencer or celebrity endorsement, attractive youth flavours, bright colours and emotional appeal to advertise and promote their products directly to young people. Social media influencers who advertise e-cigarettes were described as portraying ecigarettes as 'cool' and 'fashionable' to entice viewers to try the products. Youths considered that there is a need for more restrictions on social media content to protect youths while also still allowing smokers to purchase them as a cessation device. Conclusions Our study highlights that the e-cigarette industry is using previously employed tactics similar to the tobacco industry to advertise and promote its products on social media. These findings suggest the growing need for governments to work together to develop and implement policies to restrict the advertising and marketing of ecigarettes on social media.

Introduction Physical inactivity is a major contributing factor to the obesity epidemic, and adolescence is associated with a decline in physical activity [1]. Despite known health benefits, a large portion of the United States adolescent population does not engage in the recommended 60 min of accumulated moderate to vigorous physical activity daily [2][3][4]. Of particular concern is that physical activity levels decline during the transition from childhood to adolescence, a period crucial to the development and creation of habitual participation in physical activity throughout the life course [3]. One prominent explanation of inactivity is that adolescents are spending too little time playing outdoors due to social and physical neighborhood factors [5,6]. It is possible that if parents restrict outdoor play due to concerns about their neighborhood, their adolescent offspring will be less active overall, as time spent outdoors is positively associated with overall physical activity [1]. Generally, individuals in more disadvantaged neighborhoods have lower levels of physical activity and higher rates of obesity, even when controlling for individual-level socioeconomic status . These relationships may be due to safety concerns ; the physical environment ; or to differences in neighborhood social environment [5][6][7][8][9][10]. Emerging social environmental research identifies perceived collective efficacy as a potentially influential determinant of physical activity opportunities for both adults [11][12][13] and youth [6,10]. Collective efficacy is a form of social capital and is defined as a measure of perceived social cohesion and social control [14][15][16]. Numerous social capital indicators, including collective efficacy, have been identified and tested in relation to health behaviors and outcomes, yet have resulted in mixed conclusions [7]. For example, physical activity studies operationalizing the social environment as sense of belonging, social cohesion or norms of reciprocity illustrate no significance, whereas studies assessing collective efficacy report significant results [9,17]. Furthermore, the majority of studies relating the social environment and physical activity assessed the impact on active transport, rather than outdoor play in general and were performed mainly in older populations [10]. While a number of studies have assessed the impact of the social environment on children's physical activity, few studies have considered physical activity-related parenting behaviors , which may ultimately restrict or facilitate adolescents' physical activity [1]. Adolescents who are free to play outdoors and travel actively without adult supervision accumulate more physical activity than those who are not; therefore understanding whether parental perceptions of their neighborhood impact physical activity-related parenting behaviors may be crucial to improving overall activity among adolescents [18]. Constrained behavior is defined as the act of restricting offspring's physical activity and is categorized as "avoidance" or "defensive" behaviors depending on whether physical activity was avoided or modified, respectively. Examples of avoidance behavior include parents driving their children to school instead of allowing them to walk or forbidding unsupervised outdoor play. Examples of defensive behavior include parental accompaniment while walking to school or restriction of outdoor play to the backyard [1]. Carver and colleagues is the only study to examine environmental influence on constrained behaviors and found that perceived risk of harm to a child in the neighborhood was positively associated with constrained behavior resulting in lower levels of active transport and MVPA outside of school hours for children and adolescents [1]. However, no studies to our knowledge have explored the influence of parental perceived collective efficacy on constrained behaviors, nor have any further explored the impact of the physical environment on parental perceptions and behaviors impacting adolescents' activity. The physical environment may be the foundation for, or etiology of, perceived collective efficacy and may ultimately explain differences in perceptions of collective efficacy among parents living in low-versus high-incivility neighborhoods. The social disorganization theory [19] and the norms and collective efficacy model [20] posit that characteristics of the built environment influence the ability of community members to establish cohesive relationships and create a shared set of socially-accepted norms that promote the willingness to intervene on behalf of the common good [21]. Therefore, achieving the recommended 60 min of daily physical activity is difficult due to a complex interaction among the physical environment, parental perceptions of the social environment and their constraints on adolescent outdoor play. Additional studies are necessary to understand the influence of the social environment on parenting behaviors and offspring's physical activity, especially among populations who live in high-incivility neighborhoods. The current study examined relationships between parental perceived collective efficacy, constrained behaviors and adolescent offspring neighborhood physical activity in low-versus high-incivility neighborhoods. Conceptual path diagrams are presented in Figures 1 and2. of the built environment influence the ability of community members to establish cohesive relationships and create a shared set of socially-accepted norms that promote the willingness to intervene on behalf of the common good [21]. Therefore, achieving the recommended 60 min of daily physical activity is difficult due to a complex interaction among the physical environment, parental perceptions of the social environment and their constraints on adolescent outdoor play. Additional studies are necessary to understand the influence of the social environment on parenting behaviors and offspring's physical activity, especially among populations who live in high-incivility neighborhoods. The current study examined relationships between parental perceived collective efficacy, constrained behaviors and adolescent offspring neighborhood physical activity in low-versus high-incivility neighborhoods. Conceptual path diagrams are presented in Figures 1 and2. of the built environment influence the ability of community members to establish cohesive relationships and create a shared set of socially-accepted norms that promote the willingness to intervene on behalf of the common good [21]. Therefore, achieving the recommended 60 min of daily physical activity is difficult due to a complex interaction among the physical environment, parental perceptions of the social environment and their constraints on adolescent outdoor play. Additional studies are necessary to understand the influence of the social environment on parenting behaviors and offspring's physical activity, especially among populations who live in high-incivility neighborhoods. The current study examined relationships between parental perceived collective efficacy, constrained behaviors and adolescent offspring neighborhood physical activity in low-versus high-incivility neighborhoods. Conceptual path diagrams are presented in Figures 1 and2. --- Materials and Methods --- Study Participants Adolescents in the Molecular and Social Determinants of Obesity in Developing Youth study were accessed for the current study. Participants were recruited from an existing cohort of healthy, exclusively pre-pubertal children who previously participated in the Mechanisms for the Metabolic Syndrome in Prepubertal Youth study . All 149 MET participants were eligible to participate. Only those who were able to be contacted and who were willing to participate were included in the current study . The characteristics of the MET cohort have been described previously [23]. Participants were included if they did not meet exclusion criteria during their initial enrollment into the prior MET study. --- Procedures Parents and adolescents who expressed interest in the study and met eligibility criteria were invited to participate in one study visit at the Louisiana State University Health Sciences Center Clinical and Translational Research Center at University Medical Center in New Orleans. A parent/guardian provided informed consent, and each participant gave written assent prior to enrollment. The study was conducted in accordance with the Declaration of Helsinki and was approved by the LSUHSC and CTRC Institutional Review Boards . During the visit, adolescents participated in a physical exam by trained pediatric nurses. The adolescent and his/her parent/guardian completed self-administered questionnaires, and an accelerometer was sent home to be returned via a pre-paid envelope approximately one week after the visit. --- Measurements --- Perceived Collective Efficacy Perceived collective efficacy was assessed using a survey instrument modified from the 1995 Community Survey of the Project on Human Development in Chicago Neighborhoods [10]. Items assessed two constructs: perceptions of cohesion, the level of trust and attachment among neighbors; and social control, a belief in the capacity of neighborhood residents to intervene to reach a collective goal [11]. Parents/guardians were asked to indicate their level of agreement with 5 statements assessing how cohesive they felt their neighborhoods were, such as "people around here are willing to help their neighbors" and "this is a close-knit neighborhood". The parent/guardian chose from 5 response options with subsequent coding : strongly disagree ; disagree ; neither agree nor disagree ; agree ; or strongly agree . Cohesion was summarized as the mean across all 5 items that assessed the parent/guardian's perception of neighborhood cohesion . Six statements assessed the perception of social control based on how likely the parent/guardian thought that neighbors would intervene in certain situations . The parent/guardian chose from 5 response options with subsequent coding : very unlikely ; unlikely ; neither likely nor unlikely ; likely ; or very likely . Social control was summarized as the mean across all 6 items that assessed the parent/guardian's perception of neighborhood control . Collective efficacy was calculated as the mean of cohesion and control summary variables. A higher score indicates higher levels of perceived collective efficacy : 4.0 ; range: 2.3-5.0). This method was applied from Sampson and colleagues [24]. --- Constrained Outdoor Play Practices Constrained outdoor play practices were assessed using a questionnaire adapted from Ferraro's [25] indices of constrained behavior with regard to crime victimization and was further validated by Carver and colleagues [1]. To assess avoidance behavior, parents/guardians were asked to indicate their level of agreement with 9 statements about protecting their offspring's engagement in various outdoor activities in their neighborhood. Similarly, to assess defensive behavior, parents/guardians were asked to indicate their level of agreement with 6 statements about defensive measures taken regarding their offspring's engagement in various outdoor activities in their local neighborhood. Response options and subsequent coding were: strongly disagree ; disagree ; neither/don't know ; agree ; strongly agree . All responses were summed to compute overall avoidance and defensive behavior scores, with a higher number indicating more avoidance or defensive behavior, respectively. Carver et al. found this survey to have moderate to high internal consistency for each score and moderate test-retest reliability when used in children aged 15-17 years [1]. --- Neighborhood Physical Activity A self-administered questionnaire asked parents/guardians how often their child is physically active in a series of places in their neighborhood . These ordinal responses were rescaled to indicate the number of times per month each child was active in specific neighborhood locations. This number was then dichotomized to active versus not active in the neighborhood, based on the median value . The median value was approximated to the nearest whole number. Methods were adapted from Tappe and colleagues who reported high test-retest reliability and internal consistency [26]. Adolescent physical activity was also assessed objectively using GT3M ActiGraph triaxial accelerometers . Study participants wore the accelerometer around their waist for a minimum of 3 days and a maximum of 7 and were allowed to remove the accelerometer during sleep [27]. Data were treated using a 24-h protocol to determine nocturnal sleep period and a separate non-wear algorithm previously published by Tudor-Locke and colleagues [28]. Valid cases were defined as having ≥3 days with ≥8 h of waking wear time in a 24-h period. Only valid cases were used for the analysis of Evenson cut points into sedentary, light, moderate and vigorous physical activity [29]. --- Weight Status An electronically-calibrated scale and a calibrated stadiometer were used to obtain the weight and height of each study participant, respectively. BMI was calculated using the following formula: weight in kilograms/height in m 2 . BMI z-scores were calculated using the Centers for Disease Control and Prevention's program for calculating BMI percentiles and z-scores based on a child's sex and age for BMI, weight and height, based on the CDC 2000 growth charts. Weight status was categorized as obese , overweight and healthy weight using BMI z-scores. --- Incivilities Systematic social observation performed at the parcel-level using virtual applications was used to collect data on four major components of neighborhood incivilities: physical disorder, physical decay, perceived safety and street safety. This method was previously proven reliable compared to street-level SSO and direct observations performed by raters traveling to the neighborhood [30]. Observations were performed over 6 months in August 2015-January 2016. Google Street View images were dated from 2011-2015. Physical disorder: A principal component analysis was used to determine whether the following items denoted physical disorder: garbage/litter on the street, in residential yards, commercial, businesses, institutional, industrial lots ; graffiti or graffiti that had been painted over on buildings or signs , yes ); and residential porches that were cluttered with personal items , yes ). However, no clear factor emerged , and items were instead analyzed individually . Ultimately, garbage/litter was considered as a marker for physical disorder due to its having greater variation : 2.3 ; min, max: 0.0, 19.0), whereas there were few streets with graffiti/graffiti that had been painted over or cluttered porches . Physical decay: A principal component analysis was used to identify whether the following items denoted physical decay: condition of residential units; condition of residential yards; and the presence of abandoned cars , yes ). A physical decay scale was created by summing all of the items, which resulted in a greater value indicating worse physical decay : 3.1 ; min, max: 1, 12; Table 1). Safety: A principal component analysis was used to identify whether the following items denoted safety: raters' perceptions of whether the neighborhood was a "safe place to live"; raters' perceptions of whether they would feel "safe walking at night" ; and whether the rater felt they could live in the neighborhood . A safety scale was created by summing all items with a greater value, which indicated a less safe environment : 4.9 ; min, max: 2, 12; Table 1). Street safety: A principal component analysis was used to identify whether the following items accurately denoted street safety: the presence of speed limit signs , yes ); presence of a bike lane , yes ); lighting ; and traffic volume . However, no clear factor emerged , and items were instead analyzed individually . Ultimately, traffic and lighting were considered in markers of street safety. The presence of a bike lane was not included as an item indicating street safety because no streets contained bike lanes. Low-versus high-incivility neighborhoods: A principal component analysis was used to create an overall incivilities summary variable from the following variables: garbage/litter on the street; physical decay summary scale; safety summary scale; traffic volume; and lighting on the street. The factor included: garbage/litter on the street; physical decay summary scale; and safety summary scale with an eigenvalue of 2.2 . The factor was then dichotomized at the median value to create low-and high-incivility groups. graffiti or graffiti that had been painted over on buildings or signs , yes ) residential porches that were cluttered with personal items , yes ) Finding: Due to no clear factor and poor alpha levels, a physical disorder scale was not created. Items were analyzed individually. Physical Decay 6 3 condition of residential units 0.6 0.4 1, 10 condition of residential yards the presence of abandoned cars -yes ) Finding: A physical decay scale was created by summing all of the items resulting in a greater value indicating worse physical decay. Safety 4 3 raters' perceptions of whether the neighborhood was a "safe place to live" 0.8 4.9 2, 12 raters' perceptions of whether they would feel "safe walking at night" whether the rater felt they could live in the neighborhood Finding: A safety scale was created by summing all items resulting in a greater value indicating greater danger. --- Street Safety 4 No Scale the presence of speed limit signs , yes ) -0.3 4.5 3, 7 presence of a bike lane , yes ) lighting traffic volume Finding: Due to no clear factor and poor alpha levels, a street safety scale was not created. Items were analyzed individually. Overall Incivilities 5 3 garbage/litter on the street, in residential yards, commercial, businesses, institutional, industrial lots , physical decay summary scale, safety summary scale, lighting on the street, traffic volume . Finding: An overall incivilities scale was created and included physical decay, safety and garbage/litter. Not Applicable is included for the alpha value for neighborhood activity because a scale was not created and therefore, an alpha value is not applicable. --- Data Analysis Statistical analyses were performed using SAS Version 9.4 . A p-value < 0.05 was considered statistically significant. Descriptive statistics were calculated for demographic and anthropometric characteristics. Unpaired t-tests and chi-square tests were used to test differences between low-and high-incivility neighborhoods. Multilevel linear or logistic models, which included households as a random effect to account for adolescents living in the same household, were used to examine the relationship of perceived collective efficacy and the outcomes of interest . Analyses included both parent-reported neighborhood physical activity and objectively-measured MVPA as physical activity outcomes; however, due to the small sample size of objective measures , only parent-reported neighborhood physical activity is presented. First, univariate models were run for all potential correlates. Next, multivariable models were analyzed using multilevel general linear and logistic regression models, controlling for demographic variables . Furthermore, multilevel regression analyses, controlled for demographic variables and BMI z-score, were analyzed by incivilities to examine the relationship between perceived collective efficacy and all proposed outcomes within low-and high-incivility neighborhoods. The Kenward-Roger approximation was used to calculate the degrees of freedom [30]. The publically available Hayes PROCESS macro [31] was implemented to determine whether constrained outdoor play practices mediated the relationship between parental perceived collective efficacy and adolescent physical activity . Mediation models were run overall and within low-and high-incivility neighborhoods. A parallel multiple mediator model with 95% confidence intervals derived from 1000 bootstrap resamples was used in order to include both avoidance and defensive behavior as potential mediators. Mediation models controlled for gender, race and annual household income. --- Results --- Descriptive Statistics The study participants ranged in age from 11-18 years : 14.2 ). Approximately half of the participants were male . The majority of the participants were white and high-income . More than half of the participants were either obese or overweight , as follows: 40.9% of participants were obese; 15.5% of participants were overweight; and 43.6% of participants were healthy weight. On average, parents/guardians reported that their child was active in locations around their neighborhood 18 times per month . Valid accelerometry data were available in 47.9% of the sample. Accelerometry data indicated only 25.7 ) minutes of overall moderate-to-vigorous physical activity per day and 21.0 ) minutes of MVPA outside of school hours per day . Overall, parents reported high levels of perceived collective efficacy : 4.0 ) with a possible range of one to 5.0 . * p < 0.05; independent sample t-test and chi-squared tests determined significant differences between low and high income participants. a Incivilities data were collected using systematic social observation at the parcel-level using Google Street View. Factor analysis was used to create a summary variable, which included physical decay, safety and litter summary variables . The factor was then dichotomized at the median value to create a measure of high and low incivilities. b Low income is defined as <$40,000 annual household income. c Collective efficacy was measured using 6 questions on cohesion rated on a 5-point Likert scale to strongly agree ) and 5 questions on control rated on a 5-point Likert scale to very likely ). Summary variables for cohesion and control were developed by calculating the mean across questions for each. Collective efficacy was calculated as the mean of cohesion and control summary variables. d,e All responses were summed to compute an overall avoidance or defensive behavior score using a 5-point Likert scale to strongly agree ) with a higher number indicating more avoidance or defensive behavior, respectively. f Moderate to vigorous physical activity is defined as the mean minutes measured by accelerometry. g Neighborhood activity was measured by parent-reported times per week their child was active in locations in their neighborhood to 4 days/week or more ). Responses were coded to indicate the number of times per month and summed. The total was dichotomized at the mean to indicate physical activity in the neighborhood. h Weight status was categorized based on BMI percentile as obese , overweight or healthy weight . i BMIz scores were calculated using the Centers for Disease Control's statistical program for calculating BMI percentiles and z-scores using BMI, weight and height based on the child's sex and age [32]. --- Differences in Low-versus High-Incivility Neighborhoods Neither parent-reported nor objectively measured physical activity differed between adolescents living in low-versus high-incivility neighborhoods. However, participants who lived in high-incivility neighborhoods had a significantly higher BMIz score : 1.4 ) than participants from low incivility-neighborhoods : 0.9 ; p = 0.03; Table 2). Furthermore, parents who lived in high-incivility neighborhoods reported more avoidance : -3.3 ) and defensive behavior : -1.3 ) compared to parents who lived in low-incivility neighborhoods : -7.4 , -3.8 ; p = 0.01, p = 0.02, respectively; Table 2). Parents who lived in high-incivility neighborhoods also reported lower levels of perceived collective efficacy : 3.8 ) compared to parents who lived in low-incivility neighborhoods : 4.1 ; p = 0.03; Table 2). --- Parental Perceived Collective Efficacy and Constrained Outdoor Play Practices Perceived collective efficacy was inversely associated with avoidance behavior in the univariate model . However, in the final multivariable model controlling for demographic variables , this relationship did not reach the level of significance . Similarly, perceived collective efficacy was inversely related to defensive behavior , which suggests that parents exhibit less defensive behavior when they perceive a higher level of collective efficacy. While not statistically significant after controlling for demographic variables, perceived collective efficacy did vary in the expected direction with defensive behavior . Table 3. Regression models between perceived collective efficacy and: constrained behavior ; and c' neighborhood activity, controlled for demographic variables . * p < 0.05. a,b All responses were summed to compute an overall avoidance or defensive behavior score using a 5-point Likert scale to strongly agree ) with a higher number indicating more avoidance or defensive behavior, respectively. c Neighborhood activity was measured by parent-reported times per week their child was active in locations in their neighborhood to 4 days/week or more ). Responses were coded to indicate the number of times per month and summed. The total was dichotomized at the mean to indicate physical activity in the neighborhood. d Collective efficacy was measured using 6 questions on cohesion rated on a 5-point Likert scale to strongly agree ) and 5 questions on control rated on a 5 point Likert scale to very likely ). Summary variables for cohesion and control were developed by calculating the mean across questions for each. Collective efficacy was calculated as the mean of cohesion and control summary variables. e Low income is defined as <$40,000 annual household income. High income is defined as ≥$40,000 annual household income. In multilevel regression analyses stratified by neighborhood incivilities, relationships between perceived collective efficacy and avoidance behavior were not significant within either low- or high-incivility neighborhoods . Furthermore, relationships between perceived collective efficacy and defensive behavior were not significant within either low-incivility or high-incivility neighborhoods . 1). The factor was then dichotomized at the median value to create a measure of high and low incivilities. c,e All responses were summed to compute an overall avoidance or defensive behavior score using a 5-point Likert scale to strongly agree ) with a higher number indicating more avoidance or defensive behavior, respectively. d Collective efficacy was measured using 6 questions on cohesion rated on a 5-point Likert scale to strongly agree ) and 5 questions on control rated on a 5-point Likert scale to very likely ). Summary variables for cohesion and control were developed by calculating the mean across questions for each. Collective efficacy was calculated as the mean of cohesion and control summary variables. f Neighborhood activity was measured by parent-reported times per week their child was active in locations in their neighborhood to 4 days/week or more ). Responses were coded to indicate number of times per month and summed. The total was dichotomized at the mean to indicate physical activity in the neighborhood. --- Parental Perceived Collective Efficacy and Offspring's Neighborhood Activity Relationships between perceived collective efficacy and parent-reported offspring neighborhood activity were not significant . When controlled for demographic variables, relationships between perceived collective efficacy and neighborhood activity remained non-significant. Relationships to objectively measured physical activity by accelerometry were also not significant. Perceived collective efficacy was significantly related to neighborhood activity in high-incivility neighborhoods , but not in low-incivility neighborhoods . However, perceived collective efficacy was not related to objectively measure physical activity in either low-or high-incivility neighborhoods. --- Mediated Pathways Neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . Mediation results were similar when analyzed using objectively measured physical activity. As reported above, direct relationships between perceptions of collective efficacy and neighborhood physical activity were not significant . Furthermore, the direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . Within low-incivility neighborhoods, neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . The direct relationship between perceptions of collective efficacy and neighborhood physical activity were not statistically significant . The relationships between perceived collective efficacy and avoidance and defensive behavior were not significant with low effect sizes. Furthermore, the direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . Within high-incivility neighborhoods, neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . However, there was a statistically-significant direct relationship between perceived collective efficacy and neighborhood physical activity in high-incivility neighborhoods . Although not statistically significant, the relationship between perceived collective efficacy and avoidance and defensive behavior had large effect sizes relative to the same relationship in low-incivility neighborhoods. The direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . Within low-incivility neighborhoods, neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . The direct relationship between perceptions of collective efficacy and neighborhood physical activity were not statistically significant . The relationships between perceived collective efficacy and avoidance and defensive behavior were not significant with low effect sizes. Furthermore, the direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . Within low-incivility neighborhoods, neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . The direct relationship between perceptions of collective efficacy and neighborhood physical activity were not statistically significant . The relationships between perceived collective efficacy and avoidance and defensive behavior were not significant with low effect sizes. Furthermore, the direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . Within high-incivility neighborhoods, neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . However, there was a statistically-significant direct relationship between perceived collective efficacy and neighborhood physical activity in high-incivility neighborhoods . Although not statistically significant, the relationship between perceived collective efficacy and avoidance and defensive behavior had large effect sizes relative to the same relationship in low-incivility neighborhoods. The direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . Within high-incivility neighborhoods, neither avoidance nor defensive behavior mediated the relationship between perceived collective efficacy and neighborhood physical activity . However, there was a statistically-significant direct relationship between perceived collective efficacy and neighborhood physical activity in high-incivility neighborhoods . Although not statistically significant, the relationship between perceived collective efficacy and avoidance and defensive behavior had large effect sizes relative to the same relationship in low-incivility neighborhoods. The direct relationships between avoidance or defensive behavior and neighborhood physical activity were not statistically significant . --- Discussion This novel study explored associations among the social and physical environment and parenting behaviors to explain anticipated low levels of physical activity in a small sample of adolescents. This preliminary study was the first to report that parents who perceived low levels of collective efficacy were more likely to constrain their adolescents' outdoor play practices. Yet, constrained behaviors did not impact neighborhood physical activity in these offspring. Results from multilevel regression analyses by incivilities suggested that parental perceptions of collective efficacy might be more influential in high-versus low-incivility neighborhoods. Our findings are consistent with the ecological perspective and stress the importance of understanding how the social and physical environments interact to impact parenting behaviors and adolescents' neighborhood physical activity to reduce adverse health outcomes at early ages. Parents who perceived higher levels of collective efficacy exhibited less avoidance and defensive behavior; however, gender, race and annual household income diminished these findings. Other studies have reported similar inverse relationships between other characteristics of the social environment and parental constrained outdoor play practices [1,33]. Parents of children 15-17 years of age who perceived greater risk of their child being harmed in their neighborhood had higher levels of avoidance and defensive behavior [1]. In the same cohort, perceived personal safety was related to lower levels of avoidance and defensive behaviors [33]. In our sample, significant relationships between perceptions of collective efficacy and constrained outdoor play practices did not remain after controlling for demographic variables . This may be due to low-income families being disproportionately exposed to a host of environmental factors that may further restrict their physical activity, such as limited access to safe areas to play, that were not considered in the current study [34,35]. In our sample, a higher number of low-income and non-white participants lived in neighborhoods characterized by high incivilities compared to high-income and white participants, and collective efficacy may have been more influential in these high-incivility neighborhoods . Yet, our sample was predominantly white and high income , which was likely not an ideal sample in which to explore these relationships; moreover, the low variability in parental perceptions of collective efficacy may have limited our ability to detect significant relationships. Furthermore, parenting behaviors may be especially relevant in low-income populations, where children have less exposure to physically-active parents, fewer parental role models, less ability for parental supervision of outdoor play and less joint activity with parents [35]. Thus, a more diverse sample that better represented the racial and socio-economic composition of the urban New Orleans area may have revealed stronger correlations between parents' perceptions --- Discussion This novel study explored associations among the social and physical environment and parenting behaviors to explain anticipated low levels of physical activity in a small sample of adolescents. This preliminary study was the first to report that parents who perceived low levels of collective efficacy were more likely to constrain their adolescents' outdoor play practices. Yet, constrained behaviors did not impact neighborhood physical activity in these offspring. Results from multilevel regression analyses by incivilities suggested that parental perceptions of collective efficacy might be more influential in high-versus low-incivility neighborhoods. Our findings are consistent with the ecological perspective and stress the importance of understanding how the social and physical environments interact to impact parenting behaviors and adolescents' neighborhood physical activity to reduce adverse health outcomes at early ages. Parents who perceived higher levels of collective efficacy exhibited less avoidance and defensive behavior; however, gender, race and annual household income diminished these findings. Other studies have reported similar inverse relationships between other characteristics of the social environment and parental constrained outdoor play practices [1,33]. Parents of children 15-17 years of age who perceived greater risk of their child being harmed in their neighborhood had higher levels of avoidance and defensive behavior [1]. In the same cohort, perceived personal safety was related to lower levels of avoidance and defensive behaviors [33]. In our sample, significant relationships between perceptions of collective efficacy and constrained outdoor play practices did not remain after controlling for demographic variables . This may be due to low-income families being disproportionately exposed to a host of environmental factors that may further restrict their physical activity, such as limited access to safe areas to play, that were not considered in the current study [34,35]. In our sample, a higher number of low-income and non-white participants lived in neighborhoods characterized by high incivilities compared to high-income and white participants, and collective efficacy may have been more influential in these high-incivility neighborhoods . Yet, our sample was predominantly white and high income , which was likely not an ideal sample in which to explore these relationships; moreover, the low variability in parental perceptions of collective efficacy may have limited our ability to detect significant relationships. Furthermore, parenting behaviors may be especially relevant in low-income populations, where children have less exposure to physically-active parents, fewer parental role models, less ability for parental supervision of outdoor play and less joint activity with parents [35]. Thus, a more diverse sample that better represented the racial and socio-economic composition of the urban New Orleans area may have revealed stronger correlations between parents' perceptions of collective efficacy and constrained outdoor play practices, especially in high-incivility neighborhoods. Perceptions of collective efficacy were related to parenting behaviors, but direct relationships were not found between parents' perceptions of collective efficacy and their children's neighborhood physical activity. In contrast to our findings, numerous research studies in moderate to large child/adolescent populations reported positive associations between perceived collective efficacy and physical activity [5,6,10,34,36]. In particular, children whose mothers perceived higher levels of collective efficacy in their neighborhood played outside for longer periods of time, watched less television and visited the park or playground more frequently [10]. However, these studies were primarily performed in children under 11 years of age. Parental perceptions of collective efficacy may be less influential in older, more autonomous youth in the current study because parents/guardians may modify their actions to align with their belief that older offspring possess better self-regulatory skills. The present study, to our knowledge, is the first to explore how physical environmental factors measured objectively may alter the relationship between parents' perceptions of collective efficacy and: parental constrained outdoor play practices; and adolescents' physical activity. Individuals in our sample who lived in high-incivility neighborhoods reported lower perceived collective efficacy and more avoidance and defensive behaviors than those living in low-incivility neighborhoods . This finding is supported by the "broken windows theory", which states that the presence of disorder, such as graffiti, litter or abandoned homes, communicates a dangerous, unmonitored environment, which influences individual's attitudes and perceptions [10]. Although not statistically significant, regression and mediation analyses within low-and high-incivility neighborhoods suggest that relationships between perceptions of collective efficacy and behaviors are more meaningful in neighborhoods with high incivilities. Perceived collective efficacy may have a lower impact in neighborhoods with low incivilities, since neighborhoods with appealing physical environments have additional modes of social control, such as gated entrances, security that decreases graffiti, landscape services, well-kept sidewalks and roads for play [37]. Therefore, collective efficacy may not be the key process by which parents make decisions regarding children's outdoor play in low-incivility neighborhoods. Whereas, in neighborhoods with high incivilities, perceived collective efficacy may be a key process by which parents overcome the negative effects of incivilities, thereby increasing the importance of social cohesion and control among neighbors. This is supported by previous research indicating that the relationship between incivilities and fear of crime was mitigated by higher levels of perceived collective efficacy [37,38]. Several studies report relationships between the social environment and physical activity, but a paucity of research has investigated whether parenting behaviors are the mechanism by which the environment impacts children's physical activity and health outcomes [1,5,6,10,34,36]. The current study is novel because it investigated parental constraint of outdoor play practices and how this relates to neighborhood physical activity. Results indicated that constrained outdoor play practices was not a significant mediator. This can be explained by the non-significant direct relationship between perceived collective efficacy and adolescent's neighborhood physical activity, which ultimately precluded our ability to detect mediation. However, mediation results within high-incivility neighborhoods suggest that constrained outdoor play practices may be the mechanism by which perceived collective efficacy impacts neighborhood physical activity . Carver and colleagues reported that higher levels of parental constraint on outdoor play due to perceived risk of harm resulted in less active transport and lower total objectively measured MVPA outside of school hours for both children aged 10-11 years and adolescents aged 15-17 [1]. However, there were inconsistencies throughout their results. Constrained behavior influenced female, but not male adolescents' MVPA, and hypothesized associations were true only for younger children and adolescent girls. Our results found no gender differences in constrained outdoor play practices, which may indicate that the adolescent population in the current study may be more autonomous. As stated earlier, adolescents from the current study presumably were less restricted by parenting behaviors than those reported for younger children [1]. In Carver's study, parents of children ages 10-11 years reported more avoidance and defensive behaviors compared to those of children aged 15-17 years [1]. Adolescence is a time of increased autonomy when rules and boundaries regarding outdoor play may be re-negotiated [1,39,40]. Additionally, greater autonomy may increase the pertinence of the child's own perception of neighborhood collective efficacy on their physical activity behaviors, especially since adolescents may experience the neighborhood context in a manner distinct from their parents [1,41,42]. Furthermore, adolescents may lose interest in specific outdoor locations in their neighborhood, such as their yard, cul de sac or local park or playground, which may result in decreased physical activity [43]. Furthermore, the lack of significant relationships between constrained behavior and physical activity may also be explained by today's children belonging to the 'backseat generation' that are chauffeured to structured leisure-time activities outside of their neighborhood more than previous generations [1]. --- Limitations The low sample size limited several findings. Furthermore, due to the small sample and cross-sectional nature of this study, our mediation analyses are exploratory in nature. Nevertheless, our findings were able to detect a significant association between perceived collective efficacy and parental constrained behaviors. Subsequent research should examine hypothesized relationships in a larger, more diverse cohort. Regrettably, the target population, and resulting small sample size, was limited by the availability of potential volunteers who participated in the parent study, the MET study . The selection of this convenience group of returning study participants inadvertently generated another limitation. The MET cohort included children recruited throughout southeastern Louisiana and included primarily white families from high income households, which limits the generalizability of the findings to more diverse populations. The inclusion of objectively measured physical activity data is a strength of this study; however, 47.8% of participants did not achieve valid wear time and, thus, were not included for analyses. Defining valid wear time is complex, and varying definitions have resulted in a range from 38%-84% of study samples that achieved usable accelerometry data [44]. Furthermore, accelerometers were not equipped with global positioning technologies to determine the location of physical activity. Therefore, the study relied on potentially biased parent-reported perceptions of their child's neighborhood physical activity. Additional studies with a larger sample of valid accelerometer data that incorporates Global Positioning Systems technology are warranted to increase the power to detect relationships between neighborhood characteristics and physical activity occurring in the neighborhood. Although numerous factors were considered, additional factors may explain the overall low levels of physical activity in our sample. For example, participation in organized sports teams and the types of services available to youths in a neighborhood have been shown to influence adolescent's physical activity [1]. Furthermore, individual factors or neighborhood-level factors that were not considered may have influenced activity levels. Additionally, future research should consider psychosocial factors that may be influenced by parenting practices and ultimately influence physical activity levels in our sample [45][46][47]. However, a major strength of the current study was the inclusion of a novel conceptualization of the physical environment as incivilities, which were objectively measured by systematic social observation at the parcel-level using Google Street View. Measures were aggregated to create summary measures for the street segment on which the child lived. Therefore, this unique application may not have truly captured the entire neighborhood-level incivilities. Further research is needed to determine if SSO performed on a larger area would provide a more complete assessment of the associations among perceived collective efficacy and parenting behaviors, physical activity and obesity-related outcomes. Regardless of this limitation, the SSO method provided a more accurate overall assessment of incivilities because factors were measured at the smallest geographic unit possible , which potentially provides a more detailed description of neighborhoods and captures incivilities that are within an individual's control [30]. This may be advantageous for micro-level studies of individual health outcomes, such as those included in the current study. The social environment is broad; no definition or single measure encompasses its effect on an individual's behaviors and health outcomes [7]. The way in which the social environment is measured and conceptualized produces inconsistent results when examining physical activity and health [7,48]. The current study was limited to the assessment of perceived collective efficacy. Thus, other social environmental factors , which may be associated with parental restriction of physical activity, were not considered. Furthermore, overall high perceived collective efficacy scores and, therefore, low variation in levels of perceived collective efficacy among parents of study participants may have reduced the ability to detect relationships between perceived collective efficacy and physical activity. Lastly, the present study did not consider participants' length of residence in their current neighborhood. Greater length of residence may have implied increased stability within a neighborhood and greater permanence of network affiliations, which may have increased collective efficacy among neighbors. --- Conclusions Despite several limitations, this study was the first to demonstrate that parents who perceived a poor social environment exhibited more constraint on their offsprings' outdoor play in their neighborhood. Yet, constrained behaviors did not alter adolescents' objectively measured or parent-reported physical activity. The role of the physical environment was not clear; yet results reveal that a better understanding of how the physical environment interacts with parental perceptions of collective efficacy and their constrained behaviors is needed. This exploratory study was the first step. Studies in larger, more diverse samples of children and adolescents are needed to understand further the role of, and the interdependence between, perceived collective efficacy and neighborhood incivilities. However, perceived collective efficacy influences parenting behaviors related to youth physical activity. Therefore, community-based programs that seek to facilitate social cohesion and control may be needed to increase adolescents' physical activity in their neighborhoods, especially among families living in high-incivility neighborhoods. Reducing parents' constraint of outdoor play may be achieved by targeting modifiable social environmental factors, which offers a unique opportunity to improve adolescents' physical activity to mitigate health disparities plaguing minority and low-income populations disproportionately exposed to poor environments. --- Author Contributions: Maura Kepper, Stephanie Broyles, Richard Scribner and Melinda Sothern conceived of and designed the experiments. Maura Kepper and Stephanie Broyles analyzed the data. Maura Kepper, Lauren Griffiths, Tung-Sung Tseng, Jovanny Zabaleta, Stephanie Broyles and Melinda Sothern wrote and edited the paper. ---

Background: The current study examined relationships between the neighborhood social environment (parental perceived collective efficacy (PCE)), constrained behaviors (e.g., avoidance or defensive behaviors) and adolescent offspring neighborhood physical activity in low-versus high-incivility neighborhoods. Methods: Adolescents (n = 71; 11-18 years (14.2, SD ± 1.6); male = 37 (52%); non-white = 24 (33.8%); low-income = 20 (29%); overweight/obese = 40 (56%)) and their parents/guardians enrolled in the Molecular and Social Determinants of Obesity in Developing Youth study were included in the current study. Questionnaires measured parents'/guardians' PCE, constrained outdoor play practices and offspring neighborhood physical activity. Systematic social observation performed at the parcel-level using Google Street View assessed neighborhood incivilities. t-tests and chi-square tests determined differences by incivilities. Multilevel regression models examined relationships between PCE and: (1) constrained behaviors; and (2) neighborhood physical activity. The Hayes (2013) macro determined the mediating role of constrained behaviors. Results: Parents who had higher PCE reported lower levels of avoidance (p = 0.04) and defensive (p = 0.05) behaviors. However, demographic variables (i.e., gender, race and annual household income) limited these results. The direct relationship between PCE and parent-reported neighborhood physical activity was statistically significant in high-incivility neighborhoods only. Neither avoidance nor defensive behavior mediated the relationship between PCE and neighborhood physical activity. Conclusions: PCE influences parenting behaviors related to youth physical activity. Community-based programs that seek to facilitate social cohesion and control may be needed to increase adolescents' physical activity.

Introduction In South Africa, as in other countries, structural factors like poverty, migrancy and gender inequality drive high rates of sexually transmitted infections, including HIV [1][2][3][4][5]. Poverty is likely to increase HIV risk [6], partly via the way it tends to frame choices around sexual partnerships, particularly for young South African women [7]. A history of migrant labour in South Africa leads to higher prevalence of casual partners [8], again associated increased HIV risk [9,10]. Gender inequalities and women's social and economic dependency can increase risk by reducing bargaining power around decisions like sexual debut, frequency of sexual encounters and condom use [11][12][13][14][15]. Improved knowledge of these structural determinants suggests the need for new models of HIV prevention aiming for collective action and social change [16][17][18][19]. Freire's [20] notion of 'critical consciousness'-or the critical perception of reality that causes a group to take action-has the potential to inform such approaches. While critical consciousness has long been seen as key in empowerment education [21,22], social work [23,24], and community building [25], its application to HIV prevention is somewhat new. Critical consciousness can influence HIV prevention in a number of ways. Critical consciousness is best viewed as a process which is analytical, constructive and mobilizing [24]. It is 'analytical' because it encourages participants to question the everyday realities of their lives and re-examine how health risk and other problems relate to wider social forces [26,27]. For example, an HIV project in Massachusetts encouraged dialogue about gender, race and class among partners of injecting drug users and was an opportunity to examine how environment contributed to HIV risk [28]. Critical consciousness helps to build a 'constructive' understanding that change is possible, as participants reinterpret their situation as negotiable rather than fixed [11] and to develop strategies for improving it [29,30]. A project involving women in Mexico discussed 'foto-novella' drawings to stimulate new ways of addressing HIV vulnerability [31]. Lastly, critical consciousness is 'mobilizing' when it leads to collective change [32,33]. An intervention in Nigeria, for example, engaged sex workers in formal negotiations with brothel owners to raise prices, so that they could afford to refuse clients who would not use condoms [34]. Critical consciousness can also stimulate collective action to change the context of HIV risk: through development of local leaders [28,35], changes in government policy [35], improvements to local resource management [29] and the creation of healthy peer norms [36,37]. Despite renewed interest in critical consciousness within health education, few Freirian-inspired programmes have conducted research on the processes that promote critical consciousness [38]. The Intervention with Microfinance for AIDS and Gender Equity has been reported to reduce rates of gender-based violence and sexual risk behaviour [39,40]. It aims to foster critical consciousness among participants in order to build collective action addressing structural determinants of HIV and gender-based violence. We aim to provide lessons for future HIV prevention programmes employing critical consciousness. Based in a densely populated rural area in Limpopo Province, South Africa, IMAGE combines microfinance with participatory education and community mobilization. Small Enterprise Foundation , a poverty-focused microfinance agency, partners with University of Witswatersrand's Rural AIDS and Development Action Research programme to facilitate education and community mobilization alongside group microfinance loans. Once a village is identified to participate in IM-AGE, community members map the relative wealth of households [41]. SEF then invites the poorest one-third of women to participate as a centre in its loan programme. Around 40 women participate in each centre, and women self-select into groups of five to act as guarantors for each other's loans. The loan centre meets fortnightly to share business advice and jointly decide when to increase funding to other groups [42]. In addition, specialized facilitators deliver a curriculum of gender and HIV education known as 'Sisters for Life'. Staff recruited from the local area were trained by a gender activist to conduct 10 sessions on gender roles, sexual norms, partner communication, HIV prevention and gender-based violence, employing a mixture of information giving, whole-group discussions and role play. A key aim of the curriculum is to encourage dialogue that builds critical consciousness [43]. Following Sisters for Life, women from each loan centre collectively choose who among them will attend a week-long 'natural leaders' training course aiming to engage them in additional reflection and leadership strategies. Upon returning to Promoting critical consciousness and social mobilization their loan centres, natural leaders work alongside fellow IMAGE participants in developing action plans to address HIV and gender-based violence in their communities via collective action. IMAGE drew upon critical consciousness as one of several guiding theories. We then applied a critical consciousness lens to assess delivery of the 'conscientizing' elements of IMAGE. Though IMAGE was not a purely Freirian programme, we intend that our examination of its use of a critical consciousness approach may guide future interventions. --- Methods We conducted a process evaluation using a mixedmethod approach to exploring the implementation of IMAGE during a trial and subsequent scale-up. During the IMAGE trial , we collected prospective qualitative data: researcher notes on 134 hours of observation of intervention delivery, reflective diaries of four staff delivering the intervention, focus group discussions with clients , in-depth interviews with clients and clients who dropped out of the programme after completing a loan cycle . During subsequent scale-up of the intervention , we collected both retrospective reflections on the trial and prospective data, and therefore, many informants were interviewed more than once. Thus, we interviewed all providers with significant contact with the intervention: 7 interviews with RADAR managers , 33 interviews with RADAR staff , 16 interviews with SEF managers and 14 interviews with SEF staff . We also collected prospective qualitative data on IMAGE during scale-up through in-depth interviews with 24 clients randomly selected from client lists covering both branches. Qualitative data were transcribed verbatim from digital recordings or other formats and, where necessary, translated independently by two researchers from the local language to English. Interview transcripts were analysed by developing a text coding structure in NVivo . Two researchers coded the transcripts, developing codes inductively from the data, ensuring that our analysis incorporated the 'grounded' voices of informants. The quotes cited in this paper illustrate the emerging themes and are provided for transparency and to ease the reader in assessing the validity of our interpretation. Ethics approval was granted by the University of Witswatersrand and London School of Hygiene and Tropical Medicine. Participation in research was sought on the basis of informed consent and anonymity of informants was protected in all research outputs. --- Results We found that IMAGE was successful in building critical consciousness through a number of programme tools and methods. The training of facilitators was designed to bring about personal reflection and a sense of collegiality between facilitators and future participants. Distinct elements, discussed below were implemented to develop analytic, constructive, and mobilizing aspects of critical consciousness. Figure 1 summarises the components of IMAGE that emerged as key aspects of the intervention during our analysis. --- Training facilitators for conscientization IMAGE facilitators had a key role to play in building critical consciousness among participants. In particular, they needed to approach participants from a co-learner stance, rather than as an expert teacher: Obviously, though, the role of the facilitator is critical-how you talk to them, the manner of approach, understanding the language . knowing the fact that you are not necessarily the teacher, and how to humble yourself to their knowledge so that they are open and willing to share it with you. To develop the skills and approach needed for building critical consciousness, facilitators received We would have feedback sessions and always we would try and push them to go beyond the superficial-so not just do the training activities by rote, but reflect on what were the examples that you used? How did you make it meaningful? What were the stories that came out? Importantly, the role of mentoring was gradually handed over to the facilitators themselves: And as time went on we were able to mature. They critiqued us, and we were able to critique ourselves, the four of us-this is what we are supposed to do. The training seemed to cultivate a sense of collegiality among the facilitators. They explained that the purpose of facilitation was not to 'teach' the Participants but to change their own lives at the same time: We don't just go there and teach these women, we have to see it ourselves are changing with Promoting critical consciousness and social mobilization related to whatever we are doing in the work. All the facilitators we interviewed described how the training was intensely personal and required personal reflection: The training was so intense. It had everything to do with the sessions and everything to do with yourself, and everything to do with every faculty of your life. You had to be aware of yourself before you could actually educate somebody else. --- Analysing everyday realities of health Since IMAGE planned to tackle deeply rooted health issues of sexuality, HIV/AIDS and domestic violence, its curriculum encouraged participants to examine 'normal' cultural practices in a new light through critical reflection. To help participants analyse the roots of health in their communities, Sisters for Life sessions were 'rooted' in broader issues like cultural beliefs, gender roles, relationships and communication and domestic violence: They were not really aware of all the roots. And how the roots are vulnerable to those infections. So when I got to those sessions I really did my best to work very hard to make sure that they understand these issues. Participants were sometimes confused about why facilitators brought up topics that they thought were unrelated to health. However, as the curriculum continued, many participants seemed to appreciate these broader discussions as interesting and relevant: I used think that . health education meant diseases and ailments like flu and others. I did not know that health education also include people's lifestyles, culture and the way we see ourselves. I have learnt many things. Several facilitation techniques helped promote critical consciousness. By probing, or asking questions of participants, the facilitators encouraged critical analysis within Sisters for Life discussions: So the whole process of designing the Sisters for Life training was really about helping people to ask really good questions so that they could find their own answers. Probing helped participants question cultural traditions that had previously seemed unchangeable: I have realised how easy it is for people to say 'it is our culture that I should beat my wife.' I thought it was natural that it happens that way. I thought men were strong and women were weak. After we did a session about culture and roles, I realised that men suppress women and we use culture to justify it. Role-plays also served as an important tool for cultivating critical consciousness. One planner explained that role-plays were envisaged as a way to encourage participants to tell stories about difficult issues: It seemed a natural fit to think about using roleplays and that kind of participation as a tool for opening up critical consciousness . getting people to tell stories around issues that are quite difficult to talk about as a way to kind of open up that consciousness. Role-plays and other participatory activities were used as confidence building tools, giving women an opportunity to share their views in a public setting: Many of these women never thought that we could talk about violence like this. It was nice because it was held in our church but even better [than church] because everyone had a chance to have a say. IMAGE discussions were 'grounded in daily realities' of womens lives. For example, exercises were drawn from locally relevant scenarios like fetching water or singing songs at a wedding. --- A. Hatcher et al. Participants tended to accept Sisters for Life more, when they saw that the issues were relevant to their own lives: I got very interested because these were the things that were happening in our homes. I thought wow we are going to talk about issues that trouble our homes. They do happen and they are everywhere and nobody talks about them. Constructing a new understanding of health IMAGE was also designed to help participants construct their own solutions to local problems: It's really important for people to find their own solutions, cause that's the only way that solutions will be appropriate, first of all, and that they'll be implemented and sustainable in that framework. To build the constructive side of critical consciousness, IMAGE created a space where women felt that they 'share common problems' and find useful solutions: Women have more responsibilities than before. We are living in an era where husbands are either dead or retrenched. We carry a lot of burdens on our shoulders. And along came SEF and says 'you are not alone, there are other women like you'. Women come together and share their problems and success stories. Participants emphasized how important it was to feel supported by other women who were struggling with similar problems: Coming to SEF made me realise that it is not healthy to keep things bottled up inside me. We have to share our problems. We can only find solutions when we support one another. IMAGE also aimed to go 'beyond information giving' by packaging discussions of HIV inside deeper issues: We knew early on that we didn't want to just do the basic information giving that HIV stands for this and AIDS for this. We wanted to be much more couched. However, comments by facilitators and participants characterized the intervention as centred upon information giving. Some facilitators felt that participants already had a wealth of knowledge, but others described their role as sharing knowledge with women who were 'blank': If you are empty there is no way you can you can facilitate or share your knowledge with these women, because those women are blank. They only know things in general but they don't know the facts. New information was seen as an important benefit of being an IMAGE member. Participants commented on their getting information that many of their fellow community members did not have access to: Today I am grateful to be fed with such knowledge because we can share it with our children. If you as a parent have your children's best interest in your heart you will tell them to protect them from mistakes that our parents did because of lack of information. --- Mobilizing for community change The mobilizing aspect of critical consciousness compels communities to take action to bring about concrete changes to their lives. Critical consciousness was seen as a catalyst for women to share ideas with the broader community: It was a sense of conscientising women and giving them skills and then letting them bring that message up in the community. By training a group of natural leaders, IMAGE planned to instil a sense of empowerment so that Promoting critical consciousness and social mobilization participants could drive the process of mobilization themselves. Natural leaders took part in a 5-day training session that personalized the curriculum and cultivated specific leadership skills for community mobilization. This additional training gave many participants a sense of confidence and power: The power that those women have after being identified as natural leaders; they went into training and I mean they were very different-that week changed them quite a lot. And I have seen them in action in the centre meetings after they return from the training. Yet, some participants were unable to attend a week-long training due to responsibilities within their households and businesses: Participant #1: I am taking care of my school going children. Participant #2: I am staying with my husband. Unfortunately I cannot go. Participant #3: I would love to but I would have to find someone who can help to sell my stuff so that when we get back I would be able to repay my loan. Our observations suggest that no group, such as the least educated or most poor, was systematically less empowered by the natural leaders training. After the 5-day training course, natural leaders were meant to take over the role of facilitation, and in some centres, they did so successfully. However, fieldnotes suggest that facilitators continued to exercise power at loan centre meetings, rather than handing leadership over to the natural leaders. In some centres, natural leaders already held an elite status, because of education or wealth, so that the training reinforced power relations that existed within the community. Following the training of natural leaders, several sessions on social mobilization were designed to foster 'collective action' among participants. Collective action placed women in new roles in their community through meetings with community leaders, workshops, partnerships with local organizations and public marches. When a rape prevention committee met with local leaders, it was the first time that women had ever addressed the neighbourhood's traditional council: SEF women have played an important role in the community. We have organised the all-women meeting, in which we told the chief, civic leader and the police about the crime in the area. It was the day in which the 'women against crime' initiative was formed. Likewise, IMAGE was the first time that women of one village had ever taken a leading role in a protest march: We organised a march against women abuse in our area. Many women attended it. It was even published in our local newspaper and many people knew about us. Collective actions took varied forms, such as couples counselling, a rape prevention association that works with local police to combat domestic violence and a sit-in at a hospital notorious for its poor care of patients. Collective action gave one participant confidence to engage with local structures and speak out against injustice: It was a long process because we consulted organizations in the community for advices and support. It is difficult because people have different views and some are criticizing us . . The important lesson is that women need to speak with one voice. It is only then that people take us seriously and listen. Collective action also encountered challenges. Some participants felt frustrated in having to implement collective action on their own time, an important dynamic since the women were expected to run their own business as participants in IMAGE: Women do not have time to leave their businesses and concentrate on community activities A. Hatcher et al. because SEF does not want to know whether you have spent most of your time helping the community. It wants its money when the repayment time comes. So many women sacrifice such activities for their businesses. When local leadership was supportive of initiatives, women were successful in implementing collective action. However, local leadership sometimes discouraged activities that did not fall in line with their views, as when one centre planned to march on the local clinic and the chief discouraged it because he feared a march would cause clinic staff to quit. Lastly, participants felt frustrated that they were not given financial resources to take collective action forward: She does not understand why they were told that they should identify community problems if they were not going to be helped to solve them. In her case, she has a project in mind that she would like to start to benefit her community and she wonders what kind of financial assistance IMAGE would give her. In lieu of collective action, many participants built confidence and shared lessons from IMAGE through 'individual action'. Participants described speaking within the household to partners and children about taboo subjects like sexuality: They brought health talks to us and lives have been completely transformed. Many parents have difficulties in talking to their children about relationships and other intimate issues. But health talk made us challenge our fears. Participants also shared information with the broader community by organizing talks with neighbourhood children and speaking with friends and relatives. Participants played a central role in mitigating family conflict in their neighbourhoods. They also shared the information they had learned with people at their work, in savings groups and at their churches: I shared health education with other health care members, particularly issues on HIV/AIDS and gender-based violence. My colleagues say I am more enlightened and useful and I attribute most of it to SEF education. A common narrative was that many participants felt compelled to share the knowledge they had acquired through IMAGE: I am now able to talk to people at home. I am not saying it was easy but it is the information you cannot sit with. It burns you to talk about it. Participants described how individual information sharing cultivated a sense of empowerment and self-efficacy: I felt so proud that I managed to say something which made a change in someone's life. --- Discussion Our findings suggest that IMAGE promoted critical consciousness through a number of concrete programme activities but that several elements met challenges and may be worth re-envisioning in future programmes. We aim here to contribute to the health promotion field by highlighting promising aspects of IMAGE as potential tools for encouraging critical consciousness. We then critically assess the challenges of building critical consciousness within IMAGE and offer suggestions for future health interventions and research efforts. --- Curricular tools for critical consciousness Whereas traditional models of HIV prevention give people information, with the assumption that increased knowledge will lead to behaviour change [44,45], alternative models recognize that broader societal and cultural issues influence HIV risk [5]. The IMAGE curriculum was successful in sparking critical analysis of cultural and gender norms Promoting critical consciousness and social mobilization around HIV. IMAGE emphasized the process of 'rooting' Sisters for Life in a broader discussion about culture, a process that reflects new 'best practice' ideas around HIV prevention [4]. To foster discussions that were rooted in the local context, IMAGE hired and trained staff from the local area who were culturally and linguistically similar to women participants. This decision aligns with research that suggests programmes with oppressed women should be led by facilitators who are part of the local community [31]. The facilitation technique of 'probing' aligns with Freire's theory of problem-posing education: people can only identify viable solutions once they have been able to critically reflect on the problems that they face [46]. For IMAGE participants, the feeling that loan centres provided a safe space for 'sharing common problems' was central to their acceptance of the intervention. Like other consciousness-raising programmes, this process helped them to view difficulties not as 'personal failures and shortcomings, but as being rooted in structures affecting the life of every woman alike' [47]. Encouraging critical analysis is particularly important in South Africa, where studies have shown that men tend to control sexual activity [7,13] and that women feel a need to 'play dumb' when talking about sex, so that they are not perceived as loose [48]. A culture of silence around sex and sexuality inhibits HIV prevention efforts [49]. Relating discussions about personal problems to broader societal issues can be difficult, as facilitators sometimes lack the nuanced social insights necessary to promote critical discussions [27]. Our findings suggest that intensive training and ongoing mentoring of facilitators can encourage the social insights required to promote critical discussion. This process IMAGE took a great deal of time and drew greatly from the daily experiences of facilitators themselves. It involved an intensive sessions that focussed on emotional aspects of Sisters for Life, rather than facilitation skills. This time-intensive process aligns with Freirian pedagogy [22,50,51] but may be difficult to replicate in future HIV interventions which receive shortterm funding. --- Information giving as a key aspect of health programming While some facilitators emphasized the importance of 'learning from' rather than 'teaching to' participants, others found it difficult to go beyond information giving and help participants generate their own knowledge. This dynamic may emerge from the didactic style of education that is the norm within South Africa [27] and health programmes more broadly [52]. The tendency to give information may have also resulted from a weakness in the development of 'collegiality' between facilitators and participants. Collegiality is central to the constructive aspect of critical consciousness because it ensures that new knowledge is generated by participants, rather than being handed to them by teachers [51,53], but this process is often difficult and time consuming [50], particularly within a programme that aims to provide information as well as enable critical consciousness. At the same time, participants valued new information and saw it as a tangible benefit of taking part in IMAGE. Despite an effort by the programme to go beyond information giving, we argue that providing health information remains an important and beneficial component of HIV interventions. What is crucial is to involve programme facilitators in a process of building collegiality, so that information is not seen as an end point but an initial tool for spurring behaviour change and social action. --- Strategies for individual and collective action Community mobilization is a sustainable and impactful part of IMAGE, insofar as it prioritises local context over external solutions and is tailored to the unique needs of communities [18,35,53,54]. Some collective action by IMAGE participants will likely create long-term change , while other action may remain once-off activities . Despite many successes, IMAGE also faced important challenges in encouraging collective action among participants. The women participants were poorer than A. Hatcher et al. fellow community members, a pre-condition for joining the programme, and therefore faced competing demands on their time and energy. In a programme combining microfinance and community mobilization, there were sometimes tensions between the individual imperative to work to repay loans and the collective imperative to mobilize. Participants in other health projects have faced competing priorities [51] and therefore receive incentives for participating [31]. We suggest incorporating a 'seed grant' process that gives resources to participants for implementing mobilization plans. While these women clearly had tremendous resourcefulness in managing their own lives and support for their families, the natural leaders training was not always adequate in providing women with necessary resources to engage in mobilization activities, something previously documented in other projects working with marginalized populations [51,55]. As with other community development interventions, we also note the challenges in some participants having more 'voice' than others. We suggest linking not only a few selected natural leaders but the broader group of participants to resources and continued opportunities for leadership roles in future interventions. Involving marginalized groups in social action is also difficult because people with more power tend to dominate decision making [16,29,56]. We would argue that long-term partnership with local communities is essential for programmes that promote critical consciousness, particularly to ensure that groups have access to the support and resources needed to engender real systemic change. Indeed, community partnership has long been viewed as central to structural health programming [18,35,57], but it should also be incorporated into the research process itself by drawing from emerging fields of community-based and emancipatory action research [58][59][60]. While our own research was not informed by such approaches, our use of qualitative methods alongside the quantitative methods employed in the trial did at least enable an examination of the intervention building on the perspectives of those involved, rather than merely gauging participants' responses to constructs developed by the researchers. Our research has demonstrated that individual action can be a successful strategy for engendering change within the short timeframe of intensive health programmes. The recognition of the 'individual' component of community mobilization may help programmes to be more realistic in their approach to mobilizing communities. IMAGE participants preferred individual strategies, like speaking with friends and children, because they could occur alongside normal day-to-day activities, did not require much preparation and were inexpensive to implement. Other empowering health programmes have prioritized individual strategies, such as ingroup presentations [22], mutual support around individual changes [50] and harm reduction plans [31,61]. Likewise, microfinance programmes [62] and activism projects [63] often view community mobilization as a culmination of many individual actions. At the same time, it is important to note that most Freirian scholars see conscientization as necessarily a group process [50,[64][65][66]. Adding up individual actions may not adequately address the goal of achieving community mobilization that addresses power relations between disadvantaged groups and the larger society [18]. There is a tension between achieving outcomes within a programmatic timeframe and working towards real structural change. On its own, community mobilization within marginalized populations is unlikely to produce the structural change necessary to reduce HIV and other health risks [5,11], particularly in the light of the limitations of donor-funded development interventions. Our findings suggest a twopronged approach involving both community mobilization as well as broader structural interventions addressing institutional, legal and other structural drivers of vulnerability. --- Measuring the process of conscientization Our framework delineates critical consciousness into three distinct elements for ease of conceptualization: analytical, constructive and mobilizing. Freire, on the other hand, viewed conscientization as a fluid process, without a finite end point [24,67]. While Promoting critical consciousness and social mobilization conscientization in IMAGE was not linear, it did seem to build towards mobilization as the high point of the intervention, with important feedback loops throughout the process. For example, natural leaders training returned to the analytical techniques of the Sisters for Life curriculum. Likewise, individual and collective action seemed to draw participants back to sharing common problems with one another and translating the 'information' they learned into meaningful conversations with family and friends. This is consistent with Freirian pedagogy, in that social action should naturally loop back to analysis and dialogue [38]. Freire called this concept 'praxis' or the blend of reflection and social action that causes a group to move from reflection to action and back to reflection [46,66]. Figure 2 shows the circular pathways of praxis that emerged during critical consciousness building in IMAGE. Measuring a process of change is challenging, particularly within the lifetime of health programmes, which are often brief and intensive. While we have not attempted to link the health impact of IMAGE directly to measures of critical consciousness, it is likely that health outcomes were influenced by the development of critical consciousness among participants. IMAGE outcomes like the halving of intimate partner violence [68], a reduction in unprotected sex among young women [40] and improved communication between participants and their children around sex [69] can all be theoretically linked to the development of critical consciousness in participants. Further research would be needed to demonstrate how significantly critical consciousness influenced these outcomes and to describe the process of change within individual participants. --- Advancing health through critical consciousness In the context of traditional health promotion, IM-AGE is an innovative community-based approach to HIV prevention that reduces gender-based violence and sexual risk behaviour [39,40]. At the same time, the programme falls somewhat short of the Freirian ideal for building critical consciousness in a collegial and collective way. For example, the microfinance component of IMAGE was crucial in attracting women to participate in the programme [42], but the expectation that women mobilize their community while running a small business may have created undue pressure for participants. Nevertheless, our findings suggest that distinct programme tools were successful at moving IMAGE participants towards critical consciousness. As alternative models for HIV and violence prevention are tested, practitioners and researchers now have the crucial task of teasing out the programme elements that lead to health outcomes. By examining the concrete processes of IMAGE, we have taken a first step towards translating critical consciousness into practical lessons for the health promotion field. We have much to learn from the resourcefulness of women engaged in conscientizing programmes such as IMAGE, and further work should unpack precisely how critical consciousness can influence behaviour change, group dynamics and broader structural drivers of health. --- Conflict of interest statement None declared.

The development of critical consciousness is seen as a key stage in communities increasing levels of dialogue about priority problems and effecting structural change for health. However, relatively little research identifies concrete methods for programmes to build critical consciousness. We examined how a South African structural intervention used critical consciousness as a tool for prevention of intimate partner violence and HIV infection. We collected qualitative data in the form of in-depth interviews with managers, trainers, and participants of the Intervention with Microfinance for AIDS and Gender Equity intervention (IMAGE) in rural South Africa. The data were analysed through a coding structure developed in QSR NVivo. We draw practical lessons from IMAGE to guide other HIV programmes aiming to promoting critical consciousness and social mobilization. This research suggests that specific curricular tools can work towards critical consciousness and that mobilization efforts in future programmes can be strengthened by including individual and collective efforts by participants

Background With the pronounced burden of prostate cancer among men of all races and the disparity of incidence and mortality between African American and European American men [1,2], there is a critical need to develop technological interventions that can assist men with informed decision making [3]. In 2015, 220,800 men were diagnosed with PrCA and 27,540 are expected die from the disease [2]. However, PrCA incidence is 60 % higher in AAs and this racial group is two and a half times more likely to die from PrCA [2]. Informed decision making is described by the CDC as: when an individual understands the disease, is familiar with the risks, benefits, and uncertainties of a screening or treatment, actively participates in the decision-making process at the level he or she desires, and makes a decision at the time of service or defers the decision to a later date [4]. Informed decision making is recommended by the American Cancer Society as a solution for reducing the PrCA mortality rate because of the unclear findings regarding the efficacy of prostate specific antigen screening, a blood test used to detect PrCA [3,5,6]. The two most recent, longitudinal studies on PrCA screening, titled "The European Randomized Study of Screening for PrCA" and "Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial on Prostate Cancer Mortality," , concluded that the prostate specific antigen test was either not effective or led to over-detection of PrCA [5,6]. Over-detection is a serious concern because it can lead to the treatment of indolent forms of PrCA and in some cases treatment can lead to life-long side effects such as incontinence and/or impotence [7]. In addition to possessing a thorough knowledge of PrCA and its screenings/ treatments, an individual must also believe that he possesses the capacity to engage in the informed decision-making process with a doctor or other healthcare provider [8,9]. Multiple past studies have demonstrated that preparation for the informed decision-making process can be facilitated by computer-based education programs [10][11][12], but most of these and other studies on technology design do not report on the involvement of the target population in the intervention/technology design process. Applying community-based participatory research principles to systems design can potentially enhance the impact of interventions by identifying the specific needs of the user and any foreseeable barriers to implementation [13][14][15]. This paper uses the Nielsen's Usability Engineering Lifecycle [16] as a framework for discussing the design of a computer-based PrCA education program, but focuses on how CBPR principles can enhance this framework. CBPR strategies are a promising way to address cancer disparities because they leverage community involvement in each phase of the research process to assist with making the most optimal decisions regarding everything from conceptualization to intervention [17][18][19]. Through the development of authentic partnerships with the target audience and stakeholders, cultural and contextual relevance of interventions is increased [18,20]. Thus, the likelihood of improvement in knowledge and preventive behavior through an intervention is maximized, resulting in better health outcomes [18,21,22]. --- Community-Based Participatory Research Principles There are eight CBPR principles [14,15,[23][24][25]. These principles, created by Israel et al., include: CBPR approaches emerge as a critical strategy to engage stakeholders and identify culturally and geographically appropriate methods to overcome health and cancer disparities [17,18,26]. The key to the success of designing a PrCA education program was operationalization of the CBPR principles in our research. We were able to operationalize all of the principles with the exception of principle 7 and only partial operationalization of principle 3. This success stemmed from the multiple interactive and iterative forums where AA men in the targeted community were provided with an opportunity to actively collaborate with researchers to develop a resource for enhancing their ability to make informed decisions about PrCA screening. --- Usability Engineering Lifecycle The Usability Engineering Lifecycle is an approach to systems design that emphasizes nine core principles that, when followed chronologically, can lead to an interface that has maximum usabilty [16]. Usability is defined by Nielsen as learnability, efficiency of use once the system has been learned, ability of frequent users to return to the system without having to relearn the system, frequency of error, and subjective user satsifaction [16]. The UEL has been applied to projects such as desigining systems to allow gesture controlled interaction with virtual 3D content [27]. Some of the UEL's principles are similar in nature to CBPR principles, but do not emphasize the importance of community involvement throughout the entire design process. However, when the aforementioned CBPR principles are employed within the UEL, the conjunctional use of these principles may lead to a more optimal interface for any community-specific, digital interventions such as a computer-based PrCA education program. The discussion below is structured using UEL design principles as a chrononlogical framework while highlighting how CBPR principles can be employed within a UEL guided intervention development process. --- Design Process --- Know the User Within the UEL, Nielsen suggests that the developers should study the users to assess their individual characteristics and the environment in which the product will be used [16]. He also describes the process for implementing a competitive analysis where current products that are similar to a potential future product are empirically tested among members of the target population [16]. In CBPR it is customary, following an in-depth literature review of a problem, to recognize the target community for an intervention as a unit of identity. Recognizing the community as a unit of identity extends beyond the demographics suggested by Nielsen and can be characterized by norms, values, customs, language, sexual orientation, etc. [14]. For example, our literature review on PrCA revealed that South Carolina has one of highest PrCA mortality disparity rates between AA and EA men in the country [28]. In addition, American Cancer Society recommends that men make an informed decision about PrCA screening beginning at the age of 45 for high risk groups [3]. Therefore, a developer seeking to create an ideal computer-based education program for preparing AA men to make an informed PrCA screening decision must identify a defined community or subset of AA men who can help determine the best inclusions for the system and the environment in which the system should be housed. In a CBPR process, the researcher will investigate those cultural practices, shared needs, and self-constructed and social representations of identity. Becoming familiar with the community's identity, which can be separated into multiple social and geographic subgroups, can contribute to an end-product systems design that is customized to meet the needs of the target community. In addition, the formative nature of a CBPR approach essentially allows the community to have more involvement in and control over the product development. The prominent community of identity beyond the race of the men in our study was the faith community . Churches were targets for the study because AAs' spiritual needs in addition to other socio-cultural and psychological necessities can influence their participation and trust in health research [29,30]. Churches in AA communities have also been influential in partnering with universities to offer health-related programming [31][32][33][34], which includes PrCA prevention [35][36][37]. Recruitment for Study Participation. AAs are significantly less likely than other racial groups to participate in health-related research [38], which could also pose an issue for someone solely using UEL processes for design. There are multiple barriers to AA participation, including factors such as mistrust and time constraints [38,39]. Our recruitment was guided by Vesey's framework on the recruitment and retention of minority groups, which involves a series of strategies such as leveraging partnerships in the community to assist researchers throughout the planning and implementation process . These strategies are congruent with CBPR principles, particularly principle 3 which involves facilitating collaborative partnerships in all phases of the research. The specific strategies from Vesey's framework used for this study were: conceptualization, planning, and development of the recruitment plan and promotional materials in collaboration with community partners , recruitment of study sample with partners/stakeholders, and reporting findings to the community at various stages in the research process. Furthermore, knowing someone who has established relationships in the community of interest and allowing some flexibility in your recruitment/research implementation plan can be paramount to reaching a recruitment goal. Knowing Someone Who Knows Someone. Reaching out to a colleague or an existing community partner can be effective for recruiting in minority communities. For example, churches connected to your academic colleagues have a higher likelihood of being open to working with researchers than a church that doesn't have a history of partnering with university researchers. In the course of recruitment for our study to develop a computer-based PrCA education program, there were three academic colleagues who provided the research team with names of churches with whom they had relationships. These churches not only helped to recruit their members for our research study , but also scheduled dates and times when focus groups could be conducted. Recruitment efforts lasted two months in duration and resulted in 39 of the 40 men desired for the study. Almost all of these men were recruited through word of mouth within churches. Many of these churches were recommended by colleagues. Other Important Things to Know When Approaching Communities of Identity With Your Research. During the process of approaching communities of identity to gain support for your research, it should be noted that the timeliness and relevance of the research or system aren't always consistent with the priorities of the community. For example, our research team approached a church that questioned the impact of the proposed PrCA education program and elected to forgo participation in our study. In addition, some communities are already conscious about a specific problem such as PrCA/informed decision making and are capable of providing their members with solutions . Therefore, they may underestimate the benefit of your research to enhance their current goals. In this scenario you must make the decision whether to sell the importance of your research and how it can further enhance their current efforts or simply make contact with another community. Be flexible and prepared to work around the community's schedule . --- Competitive Analysis Competitive analysis is not a key component of CBPR, but it is necessary to determine if there are products that exist that may be appropriate for your user. In our study, an analysis of competing products was accomplished through an Internet search and literature review for computer-based PrCA education programs, but most available products had either not been empirically tested among AA men or were not available for customization based on our formative research findings. These findings revealed specific PrCA information necessary for AA men in the study population to make an informed decision about prostate cancer screening and the essential functions/aesthetics of an ideal computer-based PrCA program. Therefore, it was decided by the research team to develop an original PrCA intervention. --- Setting Usability Goals, Participatory/Coordinated Design Nielsen recommends setting usability goals based on five constructs: learnability, efficiency of use once the system has been learned, ability of frequent users to return to the system without having to relearn the system, frequency of error, and subjective user satisfaction. He also explains that the priority of each usability goal and additional important attributes of the system will be dependent upon the targeted user. He then recommends a participatory design process where users have input on a specific prototype and that all aspects associated with the interface contain consistent elements . In CBPR the product usability goals are partly determined by the community because the community is considered a collaborative partner in all phases of the research. These usability goals should also build on the strengths and resources within the community. For example, men in our study attend church often, so considering how the product could fit into the church environment could be advantageous because the church could be a place where the system could be more accessible than the home environment. Furthermore, the support of the system by social networks within churches may lead to a higher likelihood of system use [40]. In our formative approach, the research team examined the literature relevant to PrCA, informed decision-making behavior among AAs, and technology use/acceptance . We then convened multiple focus groups to determine AA men's current PrCA risk and screening knowledge, decision-making processes for PrCA screening, usage of, attitudes toward, and access to interactive communication technologies , and preferences for and characteristics of a computer-based PrCA education program. These discussions helped the research team determine what information should be included in a PrCA education program based on knowledge and decision-making needs such as facts about PrCA screening. Also, gaining information about technologies that AA men currently use on a frequent basis and what aspects of these technologies made them easy to use, helped the research team identify usability elements that could be incorporated into a digital PrCA education program. Finally, we were able to gain any additional input specifically on their openness to using a computer for PrCA decision making and create goals for designing a PrCA education program. All of the input gained through these groups was used to create a paper prototype in PowerPoint and a short animated clip displaying an avatar that could potentially be used in the program. A more detailed description of results can be found elsewhere [41]. The community was invited to participate in a second phase of focus groups to conduct a thorough review of the storyboard, the accompanying character script, and the clip of the AA male avatar who would be providing users with PrCA information throughout the course of the 12-minute education module. Prior to the focus group and consistent with CBPR principle 8, findings and knowledge gained through the first focus groups were disseminated to study participants. It was also explained how these findings had informed the development of a storyboard/script that captured the intended content and functionality of the future intervention. The community was then invited to ask questions prior to participating in our second phase focus groups. In the second phase of focus groups, participants were specifically asked to critique the content of the script to ensure that it was appropriate for users with diverse literacy levels and provide their thoughts on navigation elements for users who may have lower levels of computer fluency. Participants also provided input on the appearance and expected acceptance of the use of the AA male avatar. The focus groups provided a forum for co-learning and empowerment because the participants and the researchers' exchange led to both parties leaving the focus groups with useful information. For example, while the participants gained additional knowledge about PrCA and the development process, the research team gained knowledge about decision-making behavior among AA men. In addition, the research team learned about participants' specific technology needs, while the participants learned more about what was technologically possible. Ultimately, the feedback received was then used to revise the storyboard/script and develop a full prototype of a PrCA education program. The design of this prototype represented the integration of knowledge and action for the mutual benefit of our partners . --- Prototyping/Heuristic Analysis In the UEL, prototyping is suggested after a heuristic analysis is performed, but our research team developed a prototype prior to the heuristic analysis using a series of usability guidelines [42] and significant community feedback. Developing the prototype prior to the heuristic analysis allowed us to receive optimal expert feedback early in the process. Waiting until after the heuristic analysis could be more costly in circumstances where the product is difficult to develop or modify. The research team solicited assistance from an animator to help translate the storyboard into a full prototype. Co-learning and empowerment was also applicable to the relationship between the developer and the research team because both parties were actively involved in the development process, which translated into an exchange of information and skills . The prototyping process consisted of using software that facilitated motion capture through a Microsoft Kinect camera, which could be applied to a custom designed avatar. These rendered video clips were then uploaded to a learning software that was capable of playing clips based on user decision and administering quizzes throughout the user's educational experience. The resulting PrCA prototype was designed to be operated on any computer, but the preference of the community was that the final product be administered on a large touch screen monitor to accommodate aging users who may also have lower levels of computer fluency. The PrCA education program was then mailed along with a heuristic evaluation instrument to six experts with experience and knowledge of digital health intervention design. The evaluation instrument was based on Nielsen's 10 usability heuristics for interface design [43]. Most of the changes recommended focused on aesthetics as opposed to issues related to the usability of the education program. The PrCA education program was then moved to empirical testing through 10 in-depth interviews. The empirical testing was a means to validate the usability of the prototype through system use observation and follow-up interviews with the community. The interviews with the community focused on similar system design constructs relevant to our expert heuristic evaluation . These interviews exemplified involving the community in all phases of the research process. Details regarding the prototyping process and results from the second phase of focus groups and in-depth interviews can be found elsewhere [44]. --- Iterative Design CBPR, much like UEL, supports the development of interventions/systems through a cyclical and iterative process where any stage of the design process is revisited in order to produce the most usable system. In the development of the PrCA education program, the design from conception to prototype was iteratively orchestrated through multiple focus groups, in-depth interviews, and a heuristic evaluation with the community involved in each of these phases of research. In addition, prior to the research team's future field testing of the system, the PrCA education program will be further revised based on changes recommended during the heuristic evaluation and in-depth interviews described above. Furthermore, following this future field testing , the research team will revisit the design again to make salient changes that could improve the intervention. --- Collect Feedback from Field Use Based on findings from the prior heuristic evaluations and empirical testing, the research and development team will make changes necessary to increase the usability and professional value of the PrCA education program. The research team will then pilot the PrCA education program among AA men who did not participate in the development of the system. For the pilot, the PrCA education program will be administered on both tablets and all-in-one, touch screen computers. Men from the design phase will, however, be invited to participate in this phase of research by helping recruit other participants for involvement in the pilot. By implementing pre-and post-surveys, the research team can gather information about the system's effect on the research team's posited knowledge and behavioral outcomes and the usability of the system based on both general heuristics and overall satisfaction regarding user experience. At the conclusion of our study, we will disseminate findings and knowledge gained to all partners through local forums with study participants/stakeholders who will be invited to engage in further discussion regarding where the system would be most accessible to AA men within and beyond the AA faith-based community . --- Summary/Conclusions There are multiple strengths and similarities in UEL and CBPR principles. However, using UEL and CBPR principles in concert could lead to stronger computer-based intervention designs for minority populations who may be far less likely to participate in a non-targeted effort to solicit feedback on a product or system design. CBPR emphasizes an equitable partnership between the developer and the community, which is not central to UEL. Conversely, CBPR has not been used extensively in studies focused on the development of computer-based education interventions and cannot be used unaccompanied by a set of usability guidelines. Further research is warranted to assess the impact of the conjunctive use of UEL and CBPR principles to develop technologies for diverse populations to address the prevention of varying diseases.

To reduce disparities related to prostate cancer among African American men, the American Cancer Society recommends that men make an informed decision with their healthcare provider about whether prostate cancer screening is right for them. The informed decision-making process can be facilitated through technology by teaching men about prostate cancer and providing them with activities to build their self-efficacy. However, these tools may be most effective when they are developed using a set of validated design principles, such as the Usability Engineering Lifecycle, in conjunction with a community-based participatory research (CBPR) process. Using CBPR can be especially useful in designing tools for minority communities, where men have the highest prostate cancer incidence and mortality. This paper describes the author's process for using CBPR principles to develop a prostate cancer education program for African American men and also discusses the value of using these principles within an existing usability framework.

Introduction There are massive and continuous streams of digital chatter being generated by mainstream news, blogs, social networks, and other online content. A significant portion of this chatter is driven by news-worthy, real-world events. This paper focuses on capturing such events in online media, and studying the network properties and dynamics among different events involving hundreds of thousands of online users. Many real-world events quickly spread worldwide, some immediately affect the political, economic and social lives of millions , some bear long-term cultural and ideological influence , and some show a significant role in both . Large collections of real-time content have only recently become available, enabling event tracking at scale. The significance of event tracking can been seen at three different levels: to quantify the extent and evolution of real-world events; to reveal the connections between different events; and to anticipate the short-term effects and long-term changes they will incur. Existing solutions to online diffusion tracking have taken several angles. Information diffusion can be defined based on shared keywords or similar text between documents , on hyperlinks , on network-specific actions such as retweets or hashtags for microblogs , or on shared quotes called meme phrases . The operational definition of an event has included hashtags manually grouped into broad topical themes , or reflections of events in meme phrases . More recently, designed a two-step approach to first cluster the input Twitter stream and then perform event versus non-event classification on clusters. None of these event definitions is about a particular real-world event; an instance of an event is usually known soon after its onset on the microblog stream which may or may not contain meme phrases, and it can have a complex and evolving relationship with hashtags. We propose a novel method for finding all event-related documents across diverse media sources. This method aims at capturing online discussions about a particular real-world event, and it achieves this goal by harnessing Wikipedia for a registry of important events, representing events and documents with journalism-inspired features. This operational definition of real-world events allows us to quantify the volume, dynamics, and interactions among events. We use document hyperlinks in the main content to generate an accurate citation network. Such an event representation allows us to observe two different overlays on top of the linked documents -a network of users and another network of events . We analyze the ICWSM 2011 Spinn3r dataset, with over 60 million English documents covering a one-month period in early 2011. We observe that hyperlinks across different event-related documents account for the majority of the total links, and that such cross-event links some- times reveal surprising yet reasonable connections among events that co-evolve over time . We also study the user network structure with over 350K nodes. It turns out that 1% of the most productive users contribute over 40% of the content. The strongly connected core in the largest user network is much smaller than that of the wellknown bow-tie model of the web. Such observations, to the best of our knowledge, is seen for the first time. Weblog Mainstream News Social Media U 1 U 2 U 3 U 4 U 5 E 1 E 2 E 3 E 4 Document User Event Hyperlink In the rest of this paper, Section 2 proposes a method for identifying real-world events, Section 3 presents our data processing method and network statistics, Section 4 analyzes the diffusion patterns and interconnections between events, Section 5 observes the user networks by layering them with topological properties, Section 6 concludes this study. --- Document Labeling with Real-world Events A common conceptual model for describing hyperlink cascades among time-sensitive documents is to look at twolevel networks among documents and users . As shown in Figure 1, links in the document network are aggregated by their authorship, and these user-touser links are interpreted as cascade interactions. We propose an event-based dual representation of such a hyperlink cascade, shown in Figure 1. Hyperlinks among documents are aggregated by the event E each document belongs to. The resulting event-to-event links reflect the interrelationships between events as events evolve over time. There are two challenges in extracting event-to-event networks from the underlying document citation network: to identify real-world events, and to associate each document in our collection with the events that it describes. Our approach starts by learning event models from crowd-sourced online registries. We design document features motivated by journalistic practice to make it possible to compute the similarity between a document and an event. This is notably different from modeling the dynamics of meme phrases or approximating events using text clustering . --- Real-world Event Identification Choosing relevant data source of noteworthy real-world events requires some caution. Traditional news outlets such as NYTimes or the BBC are subjected to institutional and geographical biases; a social news reader such as Digg is presented by articles and not by events; news aggregation sites such as Google News have better coverage but retroactive crawling is not easy. The Wikipedia Current Events Portal1 is a chronologically organized event registry of public interest, continuously updated and discussed by volunteers. This seems to be the best event source, despite potential selection bias of users who self-select to be editors. Event Hierarchy from Wikipedia Event Registry We define an event hierarchy at three increasing levels of generality : Event i = {Episode j , j = 1..n}, Category i = {Event j , j = 1..m}. An Episode is the smallest unit of daily new happenings , and it is possibly categorized into one general Category . We argue that it is necessary to group subsequent episodes with a common subject, location, and/or real-world cause-and-effect into one Event as people collectively think of them as one event. For instance, some episodes are ongoing with variations for a short or long term, either continually or occasionally, and people think of the news as one event . As Figure 2 shows, an episode is mapped into each bullet point which describes a short summary of the news happening for that day. Every episode comes along with hyperlinks to one or more Reference Documents for the detailed description of the episode. We crawled all reference documents for training each episode model. Titles are referred to as categories, and we choose five major categories that are representative across professional news agencies, and have a sufficient number of episodes over time. They are Politics, Business & Economy, Disaster, Arts & Culture, and Sports. Target Real-world Events By parsing the Wikipedia event page, we built a real-world episode registry for the ICWSM '11 dataset . We target episodes that occurred between 6th January and 30th January, 2011 . Finally, we identified 284 episodes for this period. For the tracking of semantically meaningful events, we manually grouped the 284 episodes into 161 events according to their subject, location, and proximity in time. We found that the manual approach is feasible and unambiguous for less than 200 events in this investigation. Automatic grouping and inter-user agreement can be topics for further investigation. Episode Representation with Entities An episode, a unit of an event, is well defined by the "5W1H", i.e. Who, What, Where, When, Why and How, of journalistic practice. Note that among the five Ws, at least three directly correspond to entities, such as person name, organization, location, date and time indicators. Moreover, the rest of the 5W1H often contain entities to make statements precise and credible. Therefore, we propose an episode representation using an entity vector ) whose elements consist of the TFIDF score of each entity extracted from the episode reference documents . Entity recognition in documents has been an active research area of natural language processing for over a decade. There are still challenges being actively tackled by the community, and efforts to date have produced high-quality tools. We conducted entity recognition by using the OpenCalais API which provides up to 116 types of entities . We extracted 4,411 unique entities for 284 episodes from the crawled reference pages, and generated both 4,411-dimensional entity vectors and their centroid for each episode. --- Document Labeling with Identified Episodes We also represent each web document as an entity vector with the same dimensions as the episode vectors. We then use the vector-space model for classifying documents into episodes by calculating the similarity between document vector and episode class vector as shown in Figure 3; the most similar centroid vector specifies the most similar episode class. Each document can potentially be labeled with none, one, or more episode classes. For simplicity, we assign up to one episode class per document only in this work. Entity Resolution An entity name can occur in many different ways among web documents, resulting in multiple dimensions for the same entity. For example, in our data, collection we identify nine name variations for Tunisia's former president Zine El Abidine Ben Ali, including Zine Al-Abedine Ben, Zine Al-Abedine Ben Ali, and Zine Al-Abdine Ben Ali. To alleviate this problem, we employ approximate string matching techniques to cluster similar entity names. Such techniques are commonly used in entity resolution and data matching to identify similar strings that refer to the same entity. We investigate four techniques that have been found to be effective for matching names : edit distance, which counts the number of character edits to convert one string into another; Jaccard distance based on bigrams, which counts how many bigrams two strings have in common; Winkler, which is a comparison technique specifically for matching surnames; and the longest common sub-string approach which recursively extracts the longest sub-strings two strings have in common and then counts the number of characters in these common sub-strings. Each of these comparison functions returns a normalized similarity value between 0.0 and 1.0 . Evaluation of Document Labeling For the evaluation of document labeling, we use the crawled reference documents of 284 episodes as ground truth. We divided them into 538 documents for training 284 episodes and 115 documents for testing. The baseline classification accuracy based on cosine similarity between document and episode class vectors is 68%. The entity resolution brought the improvement of the accuracy up to 74% using the best-performing Winkler string matching technique. Since not all documents in our collection are related to the identified episodes, we need to choose a similarity threshold value that filters out documents corresponding to none of the known episodes. We empirically determine the threshold --- Basic Statistics of Document and User Networks Our analysis and observations are based on the ICWSM 2011 Spinn3r dataset. This dataset consists of over 386 million blog posts, news articles, classifieds, forum posts, and social media contents collected between January 13th and February 14th, 2011. Each document includes a timestamp, author information, language, estimated spam probability, and the HTML body with hyperlinks. Target Dataset We focus on analyzing weblog, news articles and social media, since these are most relevant to external news events, and constitute 98.37% of the data provided. We choose to only keep posts written in English to avoid the need for translation. We filter out duplicate documents and disregard the posts with a non-zero spam probability. Nearly 6 million documents are left after the selection. --- Main Content Extraction The HTML bodies of documents contain large spam links and lengthy header/footer information, which can lead to wrong interpretations of document linkage. Thus, our data processing first needs to remove such boilerplates and keep the main document content. We use the effective boilerpipe library to achieve this goal. --- Document Network Construction Link Extraction We follow out-links in the main content of a post for diffusion tracking. Hyperlinks in the document text, however, often contain shortened URLs, masking the true identity of link destination. There are over 300 URL shortening services, and this makes it infeasible to query their distinct APIs to recover the original links. Moreover, some links have been shortened more than once, further complicating the recovery. To tackle this issue, we extract the original location from the HTTP header. After extracting out-links, we remove self-links and outof-scope links that connect to posts outside of the dataset. We also disregard links that point to documents created after the referring document. Non-isolate Documents We filter out isolated documents which have no links to, and are not linked from other documents in the collection. The non-isolated portion has 4,138,283 documents, whose cascade size distribution is illustrated in Figure 4, showing a heavy tailed distribution. This accounts for 6.9% of the original 60 million documents. This is due in part to the majority of documents isolated. Note however that this is not a low percentage compared with the literature, where only 2% of 2.2 million blog posts are not isolated . In fact, our higher percentage results from links between three different types of document sources based upon a wide range of content types of the ICWSM'11 dataset. Figure 5 shows the daily number of non-isolated posts over the 1 month period and there is a clear seven-day periodicity: less documents are posted on weekend than on a weekday . --- User Network Extraction To obtain underlying user networks as illustrated in Figure 1, the identity of users and their created document information are essential. User Identity There is no universally valid user information, due to the diverse sources that the ICWSM'11 Spinn3r dataset draws from. We chose five major blog domains and four major social media domains as we can write regular patterns for extracting user identity from each. We also regard the second-level domain of news sites as the unique identifier of a mainstream news agency. There are 2.2 million documents with both episode and user ids . This method generates a significantly large set of users, and is consistent with prior blog user extraction methods . User Network We make a directional link from user U 1 to U 2 if there is a hyperlink from at least one document that U 1 wrote pointing to another document of U 2 . The link weight is set as the number of accumulated links. This yields the underlying user networks. As a result, we identified over 350K users and their largest connected network consists of about 310K users. Details are explained in Section 5. --- Event Diffusion Analysis We extracted non-isolated documents and labeled them with the identified episodes in Section 2.2. Based on the labeled non-isolated documents, we observe the linkage patterns of episode-to-episode network and aggregate the network into an event-to-event network as is shown in Figure 1. Event-related Document Network Topology We build document networks separately for each category and also generate networks of all categories combined. Figure 6 We observe that the largest cascade of all categories combined is about twice as large as the sum of the largest cascades of each category. Its linkage volume is about three times larger than the total linkage volume of the five largest cascades. This means that about 50% of documents of the all-category network are linked to documents of a different category, and there are about twice as many cross-category citations. In other words, the interactions between different categories is dominant in the largest event diffusion. Thus, we look into the linkage patterns between events across all categories in detail. Event Linkage Patterns Figure 7 shows the normalized linkage patterns between events where the value of element of the matrix is defined as the number of linkages from Event i to Event j divided by the total number of linkages of Event i . The matrix is rearranged to group events based on the category and expressed in the gray-scale map where black is the maximum value and white is the minimum one . Note that some dark-gray vertical lines are found in the normalized linkage matrix. These refer to some particular events that get linkages equally from most of the categories, and this causes event diffusion across categories. From our results, those events are the US banking crisis , Somalia pirates , 2011 AFC Asian Cup , From the observation of inter-category linkages in Figure 7, we investigate how closely events in different categories are connected by applying a clustering algorithm, the normalized cut on the network . Table 3 shows the distribution of events in the clustering results by normalized cut. It is difficult to find one-to-one matching between clusters and categories, which reconfirms that events are tightly connected across categories. The events in cluster 1 are about the 2011 Australian Open, the flood in Australia, a cancellation of a cricket game, a victory of England in the 2010-2011 cricket series in Australia, and England hooligans on trial. Also, interestingly, Cluster 1 has events mostly related to Australia and England, Cluster 3 to Israel and Iran, and Cluster 4 to Pakistan and India. It shows that the contexts of events are correlated with each other across categories and clusters seem to reveal general geographical relationship among events. Figure 8 shows a scatter plot of document in-degrees versus out-degrees for each of the five categories, and across all categories. As the figure shows there are no strong correlations between in-degree and out-degree size, which means that a document which is largely cited does not necessarily cite a lot, and vice versa. This section discussed the interconnection patterns among events and event categories. The finding can serve as parameter estimates for the diffusion rates between events or cat- --- Underlying User Network We analyze the user network at two different levels. The first one is a layered approach which looks into the largest connected network first, and extracts the next largest and more strongly connected network. Consequently, we extract the three major layers, namely, LWCC , LSCC , and LRC . Note that there is a hierarchy among the three layers: LRC is the subset of LSCC which is the subset of LWCC . The second approach is to analyze the network at a macroscopic level in order to obtain global structures and linkages within a network by breaking it into six meaningful pieces based upon the Bow-tie model . Three Layers of User Network In order to discover the connections between document and user networks, we extract different levels of user networks in terms of strength of citation relationships. Figure 9 shows that the users in LRC, whose size is only 1% of LWCC , create almost half of all documents in the largest cascade of non-isolated documents, while the remaining users of LWCC produce a similar proportion of documents . The users who are outside the LWCC create only 0.39% of the documents in it. Although LWCC consists of a large proportion of blogs and SNS users , LRC is mostly news media which contributes to half of the largest event diffusion. One possible interpretation is that with regard to public social events, individuals tend to refer to authoritative media even though there are easily accessible web posts created by micro/macro blog users. 1% vs. 99% of Users in LWCC We look into the structural patterns of document networks created by each user group. As Figure 10 shows, both the in and out degrees of the largest cascade documents created by 1% user group of LWCC are much larger than that of the remaining 99% of user group, and there is a tendency that 1% of users receive more in-links than the other 99% users. Also, both cases tell us that documents are more likely cited than to cite. Figure 11 shows three citation plots corresponding to each disjoint user group . As the figure shows, LRC users make the highest number of citations to different event-related documents compared with the other groups. This fact can be thought as LRC users discover connections between different events, and these citations are observed to other users, which may contribute to the wide diffusion of events. Structural Properties by Bow-tie Model The user network is based upon the citation relationships, so a user citation network itself reflects document citation structure at an abstract level. For the observation of global structures of event diffusion made by user networks, we use the bow-tie model which represents the macroscopic structure of linkages. The bow-tie model breaks the whole network into six parts. The first part is a central core, LSCC . All of its nodes can reach one another along directed links at the heart of the network. Secondly, IN consists of nodes which can reach the LSCC, but cannot be reached from it. The third part OUT contains posts which are accessible from the LSCC without citations back to it. Then there are TENDRILS which consists of nodes that can reach either IN or OUT, and TUBES that link between IN and OUT, but both components have no directly links to and from LSCC. The last part consist of disconnected components outside of the largest connected component. Figure 12 contains the number and percentage of users in each component. Compared to web structure , we can see that the size of the central core is small compared to that of the web structure . Our user citation network has a much larger IN component than the OUT component , whereas both are around 21% for the web. Such a surprising small core and the imbalance of IN and OUT is partially explained by Figure 12. The small LSCC consists mostly of news media and blogs . Note that IN, accounting for more than half of all users itself, is occupied by a majority of SNS users . The TENDRILS also have a significant SNS presence . Blog users are positioned at all over the part with minimum 20% of proportions. The majority of SNS users tend to cite to news media, but they are not cited back from it, while blog users have bidirectional interactions with news media. Also, notable is the dominance of blog users in DISC, suggesting that blogs still are popular forums for things out of main-stream interest, or of non-news items. Overall, news media seem to play a central role to diffuse events over the whole web. --- Conclusion We present event diffusion patterns across different types of social media in terms of document network, user network, and the connections between the two networks. First, events are interrelated with each other across all five categories . This tells us that all categories need to be examined for tracking event diffusion, and also that widely spreading events have influence on wide areas of our society, namely politics, business, arts, and sports. Second, there is a small proportion of reciprocal citation relationships between social media users which is only 1% of the largest weakly connected component . This 1% of user group creates half of the largest cascade of documents while 99% users of LWCC produce almost the same size of documents in the cascade. In addition, one-way relationships are more widely found between SNS users and news media than blog users and news media. Finally, the documents created by LRC users are both cited and cite much larger documents than those generated by the remaining users of LWCC. Also, LRC users make citations to a larger number of different events in a single document than other user groups, which possibly contributes to the wide-spread across categories by showing the unexpected relationships between events to other users, and consequently, to massive diffusion of events. Our analysis proposes approaches for real-world event tracking across different types of media. We expect that this work would shed light on an analysis of event diffusion patterns on the Web. As future works, one topic is to improve real-world event identification and model event diffusion based on this empirical study.

This study focuses on real-world events and their reflections on the continuous stream of online discussions. Studying event diffusion on social media is important, as this will tell us how a significant event (such as a natural disaster) spreads and evolves interacting with other events, and who has helped spreading the event. Tracking an ever-changing list of often unanticipated events is difficult, and most prior work has focused on specific event derivatives such as quotes or user-generated tags. In this paper, we propose a method for identifying real-world events on social media, and present observations about event diffusion patterns across diverse media types such as news, blogs, and social networking sites. We first construct an event registry based on the Wikipedia portal of global news events, and we represent each real-world event with entities that embody the 5W1H (e.g., organization, person name, place) used in news coverage. We then label each web document with the list of identified events based on entity similarity between them. We analyze the ICWSM'11 Spinn3r dataset containing over 60 million English documents. We observe surprising connections among the 161 events it covers, and that over half (55%) of users only link to a small fraction of prolific users (4%), a notable departure from the balanced traditional bow-tie model of web content.

Introduction The outbreak of Ebola virus disease in 2014 attracted worldwide attention. The outbreak originated in West Africa and swept through numerous African countries [1]. Because EVD is a highly infectious disease, the increasing traffic and interaction between West Africa and other countries because of globalisation meant that EVD had high potential to be transmitted to outside Africa [2]. EVD first emerged in Africa in 1976, and it typically occurs in the tropical regions of Central and West Africa [3]. Since EVD emerged, it has appeared in different parts of Africa [2]. The disease has a high mortality rate, with an average fatality rate of 50 % generally; fatality rates varied from 25 to 90 % in past outbreaks [3]. There is no licensed treatment for EVD, and early supportive care is the only measure and treatment that improves survival [2]. Disadvantaged social groups such as ethnic minorities are documented as being highly vulnerable to contracting infectious diseases and experiencing associated stigmatisation and discrimination [4]. In Hong Kong, ethnic minorities comprise 4 % of the populations [5], and approximately 1700 Africans were residing in Hong Kong in 2011, according to government statistics [6]. These African residents were mostly from Algeria, Nigeria, Ghana, and Cameroon, and they mostly worked in importing and exporting [7]. Because the EVD outbreak originated in West Africa, African residents in Hong Kong experienced stigmatisation and discrimination following the EVD outbreak. However, there is scant literature concerning the experiences of the migrant Africans overseas during the EVD outbreak. Infectious diseases do not merely endanger people's health; they also manifest a country's embedded social and cultural values. Infectious diseases such as HIV/ AIDS and tuberculosis often demonstrate and manifest social inequalities in terms of race, social hierarchy, and gender [8]. Amongst these social inequalities, stigmatisation of and discrimination against ethnic minorities are common [8]. Take HIV/AIDS as an example; social minorities such as homosexuals and Blacks are often stigmatised as high-risk groups and thus discriminated against [9]. Migrant Africans have been documented as one of the ethnic minorities most vulnerable to diseaseassociated stigmatisation and discrimination [10]. Because of the low position in the social hierarchy, low education levels, and difficulty to obtain social resources of ethnic minorities, they are often deprived of receiving proper treatment when they become ill [8]. The lack of social awareness of the difficulties encountered by ethnic minorities also creates difficulty for these minority groups to access health care resources for avoiding infectious diseases [8]. When outbreaks occur, ethnic minorities are thus easily infected and are scapegoated as a result [8]. EVD is an infectious disease with high mortality that mostly emerges in Africa, and, according to the literature, marginalised social groups are more vulnerable to stigmatisation when new diseases emerge [4]. Therefore, migrant Africans-as marginalised ethnic minorities in most communities-are assumed to be highly vulnerable to stigmatisation and discrimination associated with EVD. In Hong Kong, ethnic minorities, such as Africans, are often marginalised because of their ethnicity and low position in the social hierarchy. In addition, this highly vulnerable social group has low social visibility. This stigmatised position can put these Africans in a disadvantaged position when an EVD outbreak occurs, leading to their possible higher morbidity, which could pose a risk for the entire community's public health. Furthermore, stigmatisation can motivate patients to conceal their sickness [8,11,12], which is detrimental to the effective control of infectious diseases. Therefore, understanding the experiences of these migrant ethnic minorities during infectious disease outbreaks is crucial. However, there is a paucity of literature concerning the experiences of migrant Africans during the EVD outbreak. The social and cultural issues raised by EVD are not well documented, despite the fact that these issues can severely affect people's perceptions of the disease and their corresponding health behaviours, and thus affect the public health of the entire community. Hence, this study was conducted to fill this gap in the literature by providing a case study on the EVD-associated stigmatisation experienced by African residents of Hong Kong, how this stigma can be explained by the embedded social and cultural values regarding ethnic minorities in Hong Kong, and how the stigma shaped the Africans' perceptions of EVD, which in turn could have posed a risk to public health. Because many countries have become multicultural under the influence of globalisation, this article can guide health authorities in designing a culturally responsive infection control policy for these ethnic minority groups. --- Methods In this study, a qualitative design using in-depth individual semistructured interviews was adopted. A total of 30 African residents of Hong Kong were recruited from Chungking Mansions in Hong Kong by purposive sampling. Chungking Mansions, located in Tsim Sha Tsui-a major commercial and tourist area in Hong Kong-is a 17-storey building with numerous low-priced guest houses, and attracts many Africans staying in the city for work and business. According to a recent ethnographic study, a large majority of the people working in the building are Africans [13], and they mostly work in the trade or catering sector [13]. --- --- Data collection Thirty men from Africa were identified and sampled for this study by process of purposive sampling in Chungking Mansions in Hong Kong. In-depth semistructured interviews were conducted with the participants in July and August 2014, during which time the first cases of EVD were confirmed overseas. The participants were approached and recruited when the researcher was conducting observational fieldwork in Chungking Mansions. Most of the participants were interviewed immediately after consenting to be interviewed, whereas seven participants scheduled interviews at a later, more convenient date and time. Prior to the interviews, an interview question guide was developed. The guide was designed based on the literature about infectious diseases and ethnic minorities and the stigmatisation and discrimination of ethnic minorities, as well as on written and visual documentaries about ethnic minorities in Hong Kong. The question guide was used throughout the interview process to guide the discussion, and to ensure that the discussion stayed in relation to the research questions and followed a focused direction [15]. The guide contained a set of open-ended questions aimed to elicit participant experiences about being stigmatised in Hong Kong during the EVD outbreak, how these experiences were correlated with the embedded social and cultural values of Hong Kong, and how these experiences shaped their perceptions of EVD. The questions were as follows: The interviews were conducted with the participants on an individual basis. All of the interviews were conducted by the researcher to maintain consistency and ensure quality, and follow-up questions were asked during the interviews to collect more in-depth data from the participants. Interviews being conducted by a single researcher can minimise the risk of insufficient data collection that may be introduced by another interviewer. All the interviews were conducted in English, which was the common spoken language of the participants and researcher. The interviews were held in the workplaces of the participants for their convenience, and were audio-recorded with their consent. Each interview lasted 1-1.5 h. As an incentive, each participant was given a HK$100 supermarket cash coupon upon completion of the interview. 1. What --- Data analysis Quick data analysis was conducted during the interviews to ascertain what was known and what topics needed to be explored further [16]. The interviews recordings were transcribed verbatim by two student assistants. The researcher then crosschecked the transcribed interviews with the recordings to ensure accuracy. A phenomenological analysis of the data was conducted by the researcher to understand the meanings of the participants' experiences. Interview transcripts were analysed line by line, segmented into meaning units, and subsequently collapsed into categories. The major themes in the data were identified [17] through abstraction and constant comparison. A coding table was developed according to the inductive coding process, allowing the discovery of patterns of behaviours and thoughts [15]. The coding table identified themes, categories, and codes with supporting interview quotes. New thematic codes that emerged from the data were added to the coding table, and repetitive codes, categories, and themes were noted and highlighted to examine the frequency of their occurrence. Memos were used to record ideas and commentary during the coding process. The analytic procedures, codings, and findings were documented in the codebook to ensure the consistency and accuracy of the analysed data. Data saturation was achieved. Direct interview quotations from the participants were included in the analysis so that their ideas were clearly represented and for the analysis to achieve credibility. Neutrality was established and the findings were grounded in the interview data rather than the researcher's bias, motivation, or interest. Because this research was conducted by a single researcher, coding and recoding of the transcripts was performed to establish reliability and confirmability. The recoding was performed 1 month after the initial coding to ensure that the analysis was clear and free of ambiguity and overlaps. --- Ethical considerations The Committee on the Use of Human and Animal Subjects in Teaching and Research at Hong Kong Baptist University approved the study prior to the fieldwork. Participation in the study was voluntary. All the participants were provided with an information sheet in English explaining the purpose and nature of the study. The participants were also informed about the use of interview data, such as publication in academic journals with all personal identification removed. Verbal explanation and clarification was also offered to the participants before the interviews. Written consent was obtained from each participant to ensure that they were clear about the purpose of the interviews and use of data for academic publication. All the participants were assured of their freedom to withdraw from the study at any time. To protect the participants' privacy, no participant identifiers were mentioned during the interviews. Each participant was represented by a code in the data and interview transcripts to further protect their privacy. All data and field notes were stored in locked files and treated with strict confidentiality. The audio recording of the interviews were destroyed after the interviews had been transcribed and checked. --- Results The interview data showed that the participants commonly experienced EVD-associated stigmatisation in their workplaces and in the community following the EVD outbreak in West Africa. Stigmatisation of the participants became tangible and noticeable after the mass media in Hong Kong began widely reporting on the outbreak. Regarding the sources of stigmatisation of the participants, Chinese Hong Kong citizens were the key players, though stigmatisation from other ethnic minority groups was also common for the participants. The participants' experiences of stigmatisation were not all due to the EVD outbreak; rather, their experiences were correlated with the embedded social and cultural values regarding ethnic minorities in Hong Kong. The experiences of the participants, in turn, shaped their perceptions of EVD and their corresponding health behaviour, which had crucial implications for the public health of ethnic minorities. --- Stigmatisation in the workplace The workplace was a common site at which the participants experienced EVD-associated stigma. Most of the participants were working in the trade or catering business, and Chinese Hong Kong citizens were their main source of customers. During the outbreak, many participants noticed stigmatising behaviour by Chinese Hong Kong citizens. The following experience, as described by one participant, was common among the interviewees: The business has dropped a lot. No Hong Kong people come in to eat now. Even though few people might stop by, they felt hesitated in touching our things. Sometimes I even heard that they said "Are they Africans? Do they have Ebola?" They are afraid of us for having Ebola. I once tried to ask some Hong Kong people to come in to eat, but they were gossiping softly, and then they left. In addition to Chinese Hong Kong citizens, the colleagues of the participants were also players in the stigmatisation experiences. These colleagues were also ethnic minorities, and came from Pakistan, India, and Nepal. One participant recalled how he was stigmatised by his Pakistani colleagues: Our relationship was not good when I first worked here because we are in different [ethnic] groups and we are not friend with each other. My wages is lower than them so they think I am competing with them. When the Ebola comes, they keep saying bad words on me, and blaming me for making the business worse. To protect their own sales performance, they even tell the customers that I am from the "country of Ebola", which scares many customers off from me. I receive much lesser commission now, and my boss blames me now for making his business drops. --- Stigmatisation from the community All the participants also experienced EVD-associated stigmatisation from the community. They often experienced stigmatisation when in and around Chungking Mansions, which is famous among Chinese Hong Kong citizens for ethnic minority businesses. One participant recalled the changes he observed in Chinese Hong Kong Being stigmatised in public transport was also common for almost all the participants. One participant discussed the difference he noted when riding on buses before and after the outbreak: After the Ebola [outbreak], no one wants to sit close to me. When I sat down, the people who have been sitting close to me immediately left their seats. They moved to other seats which were far away from me. No one took the seat next to me even though the bus was full. It just appeared that they were really afraid of me. Several of the participants also experienced stigmatisation from neighbours: Because the participants had chronic conditions that required follow-up treatment at public hospitals, hospitals became another place at which they experienced stigmatisation. Stigmatisation from other patients was reported by more than half of the participants. One participant described how he was stigmatised while waiting for treatment: After the Ebola [outbreak], many Hong Kong people are scared of me. No one wants to sit close to me while we are waiting [for treatment]. I remember at one time, several people with the same skin colour as mine were waiting, and all [the other] people left their seats and sat far away from us, though only I am an African. I really feel doubtful if they really know how to distinguish Africans; probably they just think that all people with black skin are from Africa. I also overheard their conversation [which was] about treating Ebola, and they stared at me unfriendly. I am sure they were gossiping about me. A few of the participants mentioned that they experienced stigmatisation from health care providers while receiving follow-up treatment from public hospitals. One participant recalled how he was shocked by the nurses who treated him: The nurses were afraid of me. They wore full protective gear, with facemask, face shield, gown, hat, and gloves, when having [preconsultation] examination on me. They kept asking me if I am from West Africa, and if I had been to West Africa in these few months. Also, they asked me several times if I had contacted any patients with Ebola. I was not sure if they were just more cautious about Ebola these days; but they did not wear the same set of things and ask the same set of questions when they were seeing other Hong Kong patients. This made me feeling very bad, and made me feeling I am very different and dirty, just because I am an African. --- Correlation between the EVD-associated stigma and embedded social and cultural values regarding ethnic minorities in Hong Kong The EVD-associated stigma attached to the participants was not completely due to the disease itself. Embedded social and cultural values regarding ethnic minorities among Chinese Hong Kong citizens also played a considerable role in contributing to stigmatisation; indeed, experiencing stigmatisation had been common for all the participants since initially settling in Hong Kong. One participant recalled his experience upon first arriving in Hong Kong: Many Hong Kong people were unfriendly to me when I first came here. They called me "black ghost", and they kept a distance from me. They labelled us as AIDS carriers, and they were afraid of me. No matter when I went to a shop or a restaurant, I was not welcomed by them. Even now, still not many Hong Kong people can accept me. They still love calling me as "black ghost". Also, every time when I take public transport, they will not sit or stand next to me. I still cannot understand why, but definitely I feel bad with it. All the participants had worked in Chungking Mansions, and some of them recalled how this building was stigmatised by Chinese Hong Kong citizens because of the ethnicity of the people inside: For almost 20 years ago, I was living and working in Chungking Mansions. At that time, no Hong Kong people would go inside [the building]. Although there was a shopping mall, I could hardly see any Hong Kong people coming in. Almost all the people in this building were Indians, Pakistani, and Africans. Many Hong Kong people were afraid of us, and they were afraid of this building, too. Maybe we are poor in their [Hong Kong people] eyes, or maybe our black skin is just too horrifying to them, or maybe they thought that we would rob them. "Black ghost" was the first Chinese word that I learnt here, and even up till now, some Hong Kong people still like calling us as "black ghosts". A few participants wanted to work for major companies in Hong Kong. However, their marginalised social position in the city often served as a barrier preventing them from working in the mainstream community. The participant with a university education recounted his unpleasant experience job searching when he came to Hong Kong: I studied computer engineering in a university in my home country. With my university degree, I believe that it won't be too difficult for me to find a job in information technology in Hong Kong, especially [because] there was an IT boom all over the world some years ago. I believe Hong Kong should be a good place that can offer me a good career prospect. However, I am just too naïve since what I have thought about is obviously my false hope. When I finished my first contract here, I had to find another job in order to extend my work visa and stay in Hong Kong. However, most Hong Kong IT firms were unwilling to hire me. I had been to many IT firms for job interviews, and many employers looked shocked when they saw me. No one hired me, but I had to make a living; so I worked in a restaurant at Chungking Mansions for some years. It was not until recently that an international company's Hong Kong's office gave me an offer for an IT position. Now I realise that not many Hong Kong people can accept Africans. --- How stigmatisation experiences shaped perceptions of EVD The EVD-associated stigmatisation experienced by the participants both intensified the marginalisation caused by their ethnicity and shaped their perceptions of EVD. Most of the participants were unaware of the dangers of EVD before they experienced tangible stigmatisation in Hong Kong. Although the participants' family members were living in Africa, some in countries with EVD cases, the participants commonly perceived the disease as too remote to affect them. However, the EVD-associated stigma that they experienced in Hong Kong shaped their perceptions of the disease as being shameful and horrifying. However, some of the participants adopted their own cultural values and beliefs to understand EVD, interpreting the disease as a form of retribution that had been introduced by Westerners. --- A shameful disease Because of the participants' stigmatisation experiences, they perceived EVD as a shameful disease. As one participant stated: I can imagine that if I get Ebola, I will be discriminated against further in Hong Kong. Many Hong Kong people are afraid of us now, thinking that we all have Ebola. If any of us really gets Ebola, it will be ashamed for us, because we really have Ebola and will surely be discriminated against further. If I really get Ebola, certainly I will not go to hospital; otherwise people will know and discriminate against me definitely. The participants commonly perceived EVD as a disease of shame for all Africans. Because the disease originated in Africa and the 2014 outbreak occurred there, in addition to the stigmatisation from Chinese Hong Kong citizens, almost all the participants perceived the disease as being shameful for Africans. As one participant commented: Ebola is a shame to me, and it is a shame to Africa. We are just too poor to get the disease. This disease does not hit in wealthy and Western countries, but it just hits Africa. Just like AIDS, it also hit in Africa at the very beginning. Why always Africa? It seems to me that it is our fate to suffer from these deadly diseases, because we are just too poor, and our hospitals are just too backward. It is a shame for me to tell others that I am from Africa, because other people will just think about poor, backward, and disease when they heard "Africa". --- A horrifying disease Most of the participants were unaware about the dangers of EVD at the beginning, and had little knowledge about the disease. Their impressions of the disease were mainly shaped by the mass media and, most crucially, by the stigmatising behaviour that they experienced in Hong Kong. Because of their lack of knowledge of the disease and its high mortality rate, perceiving EVD as horrifying was a popular perception among the participants, though some of them believed that the mass media had been exaggerating the severity of EVD in depicting it as horrifying. As one participant commented: I do not know exactly what Ebola is. I just know that many people are died of it, and there is no treatment for it. That is all I know from the news. Many Hong Kong people are scared of me and they keep a distance from me, so I guess the disease should be transmittable. Some Hong Kong people stare at me and mumbling "Ebola, Ebola". Their reaction makes me feel that Ebola should be really very horrifying. However, when I ask my family members in Nigeria about Ebola, they do not know much about it. Ebola sounds horrifying in Hong Kong, but not in my home country. The mysterious image of EVD also contributed to the participants perceiving the disease as horrifying. Almost all the participants were unable to receive much information about EVD. Their lack of information about the disease was mainly due to their limited social network for accessing information about the disease, which contributed to the mysterious image of EVD as a result. As this participant indicated: I do not know what Ebola is, and do not know how one can get Ebola. The disease is just like a mystery, so it is even more horrifying to me. I do not know where I can obtain more information about Ebola. Although I live in Hong Kong, I do not have many friends here. I do not have television at my home so I cannot watch the news. I do not buy newspapers because English newspapers are expensive. I can just ask my African friends but none of them know much about Ebola, too. Probably the only source of information is from the Hong Kong customers, but they do not come to eat now. I have no way to get more information. --- A disease of retribution Because of the difficulties in obtaining reliable information about EVD, more than half of the participants used their own cultural beliefs and values to interpret the disease. One participant explained how he used his existing cultural belief to understand EVD: In my country, we believe bad disease will come to find bad guys, so I think Ebola only infects bad guys. Bad guys behave badly, and they will annoy the god. When the god is annoyed, bad guys will be died of a painful disease as a punishment. Ebola is a punishment for bad guys. If you behave well and are a good guy, the god will not punish you and you will never get infected. But if you behave badly, you will die of this painful disease as punishment no matter how careful you are. --- Such cultural beliefs affected if and how the participants took precautions against EVD: Ebola is a punishment to bad guys, and they will bleed to death painfully. I am a good guy, and I never do bad things. I am honest to my customers and never cheat people, so I am not much afraid of this disease. I will not get infected because I am a good guy. If you are a good guy, then no matter how dirty you are, you will never get the punishment. If you are a bad guy, then no matter how clean you are, you will still get this painful disease, because it is a punishment for you. --- A disease introduced by Westerners Nearly half of the participants said they believed that EVD had been introduced by Westerners. These participants widely believed that EVD was imported from Western countries through the "invasion" of Westerners travelling to Africa for business. In many cases the participants often used their cultural beliefs to connect the Westerners with the disease and their own social stigmatisation. One participant expressed his hatred towards Westerners for, as he believed, bringing EVD to his home country: Originally, my country is very beautiful, clean, and peaceful. We rarely got sick at all. However, after these Western people have come, we then get sick more often. It is them who bring along the bad spirits to us. These bad spirits make us sick. They buy all our lands and we have to work for them. We can no longer work in our own farms. They exploit our people to work for them. They cut down many trees as well, and this annoys the tree spirits. Because the tree spirits are annoyed, we get sick as a punishment. It is them who bring Ebola to us. It is their fault, but we have to suffer. Interlocking with their cultural belief of EVD as retribution, some participants considered EVD as punishment of the Westerners for their bad behaviour and exploitation of the Africans: The Western people come to different parts of Africa to exploit us, no matter [in terms of] work or health. They take many of our lands, and this annoys our god of land. We get sick more often, because the god is unhappy with their taking of our lands. Ebola is a punishment from the god, so it [the disease] is brought by the Western people. As we work for them, the god is annoyed with us, too, and so we also have to get the punishment. If it is not because of them, there will not be any Ebola. It is a punishment for them, but we have to suffer, because the god misunderstands us for helping them. --- Discussion The interview data demonstrated the intertwining relationship between EVD-associated stigma and the embedded social and cultural values regarding ethnic minorities in Hong Kong. The stigmatisation experienced by the participants in their workplaces and the community was not merely induced by the EVD outbreak; rather, it was a continuation of cultural stereotypes of Africans among the Chinese Hong Kong citizens, and made the existing stigma more tangible. Such disease-associated stigmatisation of the participants played a considerable role in shaping their perceptions of EVD and their corresponding health behaviour. Consistent with previous literature [4,18], the participants, who occupied a marginalised social position in Hong Kong because of their ethnicity, were vulnerable to disease-associated stigmatisation during an epidemic. Marginalised social groups, such as migrants belonging to ethnic minorities, in particular Africans, are more vulnerable to being scapegoated as high risk groups for deadly communicable diseases [19,20]. Some participants had been stigmatised as possibly having AIDS after first arriving in Hong Kong, and they were stigmatised as possibly having EVD during the EVD outbreak. In Hong Kong, ethnic minorities are often marginalised; they often encounter difficulties in study, work, and daily life [21]. Government statistics also shows that a high proportion of Chinese Hong Kong citizens reported unpleasant interactions with ethnic minorities [22]. Even people belonging to ethnic minorities who are born in Hong Kong are often not perceived as "Hong Kong citizens" [23]. Because of the language barrier, these people often have to go to government-designated schools for ethnic minorities [24], thus their marginalised social position and social isolation from the main Hong Kong community are created and reinforced from childhood. When they grow up, they often encounter substantial difficulties in finding employment, even though they may have a university education [23]. The situation may be even worse for Africans; in a ranking of Chinese Hong Kong citizens' acceptance of ethnic minorities, Africans placed in the bottom third [22]. The participants of the current study, who were not born in Hong Kong, were expected to encounter even more hardship in Hong Kong. As indicated, the participants were often labelled "black ghosts" and prevented from obtaining their career gaols. Many of the participants were unable to gain employment in major companies, forced to work instead in communities mostly comprising ethnic minorities. This further contributed to their social isolation and marginalised social position. They had very limited interaction with Chinese Hong Kong citizens in both their workplaces and in the community. Therefore, the participants were excluded from the main community and could interact only within their own social network with their own and/ or other ethnic minorities who encountered the similar situations. Such social isolation and marginalisation made obtaining reliable information about EVD difficult for the participants, rendering them more vulnerable to being infected by not merely EVD but also other communicable diseases. The participants' marginalised social position might also have served as a barrier to them seeking health care. It is not uncommon for ethnic minorities to encounter unpleasant experiences when interacting with health care providers in Hong Kong [25]. Because of their marginalised social position, they receive little health care support [25]. These unpleasant experiences in seeking medical treatment can undermine ethnic minorities' confidence and trust in Hong Kong's health care providers [25], which may result in unwillingness to seek treatment. This serves as a potential risk for Hong Kong's public health, especially because some participants came into contact with family members living in countries with EVD cases. The health care providers' adoption of infection control measures also embarrassed the participants. Wearing full protective gear was advised by the health authorities for only when dealing with suspected EVD patients, with standard precautions considered sufficient in general outpatient settings [26]. However, this official advice was not followed, according to the participants' recounting of receiving follow-up treatment for chronic conditions in public hospitals. The different protective gear used by the health care staff when dealing Hong Kong patients and the participants revealed the health care providers' cultural perceptions and stereotypes of the Africans; specifically, the Africans were widely stereotyped as a high-risk group and "dirty" patients. Such different treatment by health care providers of the participants was not exceptional; health care providers were already noted as lacking sensitivity in providing treatment to marginalised and stigmatised social groups [27]. Stigma and marginalisation of certain social groups can interfere with health care providers' decisions and behaviours towards marginalised patients [27]. Because of the ethnicity of the participants, health care providers behaved differently when providing treatment to them . This different treatment in medical encounters and the stereotyping by health care providers both intensified the participants' unpleasant experiences with health care providers and deepened their sense marginalisation and stigmatisation in Hong Kong during the EVD outbreak. Because Africa was the epicentre of EVD, it was perceived as a "dirty" place by most Chinese Hong Kong citizens following the EVD outbreak. The participants, hence, were discouraged from returning to their home countries during the EVD outbreak, and could have faced serious consequences from the community if they did. Feelings of insecurity towards marginalised social groups was noted to be common among ethnic majorities [28], because the social practices and cultural values of the marginalised groups clash with the existing social and cultural values of the community [28]. During epidemics, such insecurity towards ethnic minorities can be heightened. To overcome the sense of insecurity and ensure social stability, marginalised social groups are often oppressed [28]. The study participants were oppressed by being pressured to avoid connections with the "dirty" zones during the outbreak to ensure that their practices and behaviours were aligned with those of the ethnic majorities. A participant returning to his home country was perceived as an act against the majorities' social values. As a result, members of the ethnic majorities monitored the participants to ensure that they had any contact with a "dirty" zone. Even among different ethnic groups stigmatised each other, according to the participants. Indeed, the literature shows that it is common for ethnic groups in the same community to come into conflict with one another in attempts to protect their own rights [28]. The participants' experiences were no exception. Government statistics show that pleasant interactions with other ethnic groups is low among non-Chinese people in Hong Kong [22]. Because making a living is often difficult for most ethnic minorities in Hong Kong, the participants were perceived as competitors and intruders by the other ethnic groups. The marginalised positions of ethnic groups in Hong Kong made the competition in their shared job market even fiercer. In addition, because of the marginalisation of different ethnic minorities in Hong Kong, the intent of members of non-African minorities to distinguish themselves and their ethnic group from other members of other ethnic groups was stronger, with the aim of avoiding further stigmatisation. Distinguishing themselves from the "dirty" participants became a survival strategy for other ethnic minorities in Hong Kong during the EVD outbreak. The EVD-associated stigmatisation in Hong Kong shaped the participants' perceptions of the disease. Because of their social seclusion, the participants experienced substantial difficulty in accessing reliable information about EVD. Their lack of knowledge of EVD led them to hold a mysterious and horrifying impression of the disease, resulting in a sense of insecurity. To cope with this insecurity, the participants made sense of the disease through their cultural beliefs. Africa's historical background, particularly its colonial history, had created unpleasant social and cultural feelings in the participants. The new social system and cultural values introduced by Westerners clashed with the original environment, social systems, and cultural values. These social and cultural experiences led the participants to associate EVD with the unpleasant history and Westerners, and they thus perceived EVD as a disease of retribution introduced by Westerners. The attitudes and behaviours of the ethnic majorities in Hong Kong during the EVD outbreak also shaped the participants' perceptions of EVD, leading them to view it as a shameful and horrifying disease; EVD was not merely perceived as deadly-it also conveyed another stigma to the participants. Both the colonial history of Africa and the double burden of stigmas, in relation to ethnicity and disease, that the participants had experienced in Hong Kong greatly shaped their perceptions of EVD. The participants' perceptions of EVD influenced their corresponding health behaviour, which could have impaired the effective control of EVD. The participants perceived EVD as a shameful disease. As noted in earlier studies, a disease-associated stigma can motivate patients to conceal health conditions and delay seeking treatment [12]; the disadvantaged social position of marginalised populations can even influence how they access health services when seeking treatment [19]. Thus, the stigmaattaching nature of EVD might have prevented the participants from seeking timely treatment if they had contracted the disease. Cultural beliefs could also have served as a barrier for the effective control of EVD in the participants. The participants' cultural beliefs, which interpreted EVD not as a highly communicable disease but as a retribution for immoral behaviour and a disease introduced to Africa by Westerners, demotivated them to actively guard against the disease. Moreover, the participants' distrust and blame of Westerners could have prevented them from receiving Western public health or medical assistance. To the participants, the Westerners were evil "intruders" who had gone to Africa to take the resources there, and who had brought various diseases in doing so. Such cultural beliefs served as a remarkable obstacle to the participants trusting biomedicine, which would have made EVD difficult to control among them; this situation played out in reality in the infected countries in Africa. In addition to the participants' marginalised social position and the stigmaattaching nature of the disease, the participants could easily have become "hidden patients" had they been infected by EVD. Some of the participants perceived the dangers of EVD as having been exaggerated by the mass media. This might have further reduced their motivation to take preventive measures against the disease. The experiences of the participants demonstrated the extreme health inequality between the ethnic minorities and ethnic majorities in Hong Kong caused by the minorities' marginalised social position and narrow social network. Such health inequality resulted in many "blind spots" in health education and health care among ethnic minorities in Hong Kong, and this further marginalised these people, rendering them highly vulnerable to and at high risk from EVD. Although Hong Kong's health authorities had disseminated health education pamphlets about EVD prevention to African communities in the city [29], these had failed to reach the participants. Because there are other ethnic minorities in Hong Kong, and because the social positions of these ethnic minorities are similar to those of the participants, these other ethnic minorities may be assumed to also have encountered difficulty in accessing reliable infection prevention information [30]. Without reliable health information, the participants' cultural beliefs and social experiences were the predominant determinants of their perceptions of and ultimately health behaviour in response to EVD. Previous studies have noted that health inequality can results in poor and marginalised people being more vulnerable to various infectious diseases [31]. Hence, ethnic minorities require more socially and culturally responsive health promotion and education to provide them with accurate information about EVD and preventive measures against the disease. --- Limitations This study recruited only Africans who had obtained Hong Kong residency. Thus, this study was unable to investigate the experiences of migrant Africans with a shorter stay in Hong Kong. Because of the communication issue, sampling bias might occur since only the participants who are able to communicate in English were sampled. Hence, the findings might reflect the Africans who had a higher education level. In addition, other ethnic minorities were excluded from the sampling, which made understanding other ethnic groups' experiences more difficult. --- Conclusions This study investigated the stigmatisation experiences of Africans residing in Hong Kong during the 2014 West Africa EVD outbreak. Their stigmatisation was not solely due to EVD; rather, it was a continuation of the embedded social and cultural values towards ethnic minorities in Hong Kong. EVD was a trigger that intensified the existing stigma and made it more tangible for Africans living in Hong Kong. The experiences of being stigmatised shaped the participants' perceptions of EVD. In addition, because of the participants' marginalised social position and isolation from the main community, they had limited access to reliable information about EVD. As a result, they used their own cultural beliefs to understand EVD, which might ultimately have influenced their health behaviour towards the disease. The experiences of the participants showed that ethnic minorities in Hong Kong were in need of more culturally responsive social and health care support to enable them to obtain reliable information about EVD and preventive measures against it. --- Competing interests The author declares to have no competing interests in conducting this research. Author's contributions JYS was involved in the study conception and design, data collection, data analysis and interpretation, and the writing of the manuscript.

The outbreak of Ebola virus disease (EVD) in Africa in 2014 attracted worldwide attention. Because of the high mortality rate, marginalised social groups are vulnerable to disease-associated stigmatisation and discrimination, according to the literature. In Hong Kong, ethnic minorities such as Africans are often disadvantaged groups because of their low position in the social hierarchy. In 2011, approximately 1700 Africans were residing in Hong Kong. Their overseas experiences during the EVD outbreak were not well documented. Therefore, this study investigated the EVDassociated stigmatisation experiences of African residents of Hong Kong with chronic illnesses, and how these experiences shaped their perceptions of EVD. Methods: A qualitative design with 30 in-depth semistructured interviews was conducted with chronically ill African residents of Hong Kong. Results: The interview data showed that the sampled Africans often experienced stigmatisation in their workplaces and in the community during the EVD outbreak. Their experiences of EVD-associated stigma were correlated to the embedded social and cultural values regarding ethnic minorities in Hong Kong. These experiences of being stigmatised shaped the perceptions of the Africans of EVD, leading them to view EVD as shameful and horrifying. They also perceived EVD as retribution and was introduced by Westerners. The participants' perceptions of EVD influenced their responses to and behaviour towards EVD, which may have posed potential threats to Hong Kong's public health. Conclusions: The EVD outbreak was not the only cause of the participants' stigmatisation; rather, their EVDassociated experiences were a continuation and manifestation of the embedded social and cultural values regarding ethnic minorities in Hong Kong. The experiences of being stigmatised shaped the participants' perceptions of EVD. Because of their marginalised social position and isolation from the main community, the participants had extremely limited access to reliable information about EVD. As a result, they used their own cultural beliefs to understand EVD, which might have ultimately influenced their health behaviours. The experiences of the participants showed that ethnic minorities in Hong Kong were in need of more culturally responsive social and health care support to obtain reliable information about the nature of and preventive measures against EVD.

Introduction The study of informal roles in organizations emerged with the discussion of informal and formal organizational structures. Informal structures develop as "unwritten laws," which derive from the interpersonal relationships between organizational members, and operate in parallel with the formal social structure of organizations . One aspect of the study of informal organizational structures is informal leadership and leadership emergence. The acknowledgment of such phenomena has fueled inquiry into the identification of informal leaders and the organizational consequences of their behavior and effectiveness . The research presented in this article is conducted in a setting without formal structures operating in parallel to the informal structure. Absent from these considerations is an examination of the formal and informal structure in parallel, and the social processes that contribute to the development of leadership . The absence of informal considerations provides the opportunity for the current article to enter a complementary political perspective into the discussion of informal leadership. The political arena inherent in organizations results in part from the informal behavior of organizational members . We argue that the informal leader role is part of the informal structure, and as such, the nonprescribed nature of this role requires both political competency and ambition. Specifically, the likelihood of individuals flourishing in the space between the formal hierarchy and informal status requires both their willingness to invest and risk reputational capital, as well as the appropriate context within which to exercise their influence. Although a complete understanding of political considerations in leadership is still underdeveloped, recent work has begun to better articulate the processes through which political inclinations, abilities, and contexts affect the leadership experience . The present study incorporates and expands on this literature to develop a model of the positioning and performance of informal leaders. Moreover, this article integrates the roles of political will and political skill with a social network perspective, and in doing so extends the literature in several meaningful ways. Our study initially expands the domain of individual antecedents of informal leadership by considering contextually specific political will; that is, political will that is specific to the context of informal leadership. Political will is particularly relevant to the informal leadership process because personality and skill-related traits could not, by themselves, predict employees' acquisition of powerful positions . As such, it is necessary to consider the willingness of individuals to invest themselves in political processes. "Informal leaders emerge through a complex process of role taking and peer perceptual processes that determine who becomes leaders" . It is not surprising that research on the antecedents of informal leadership has focused on the personal characteristics of employees that make them more likely to emerge in these roles. For example, employees' general mental ability and their personality traits have been found to predict informal leader emergence . Absent from this work is consideration of whether individuals are politically motivated to obtain the influence inherent in informal leadership status. More important, are individuals motivated in the inevitable competition for social status that occurs in informal groups? We also add depth to the literature on political skill by investigating the role of political skill within informal leadership roles and, to some degree, within the broader context of social networks. Specifically, we investigate this ability to navigate the political arena is conveyed through the leaders' use of political skill . Politically skilled individuals more accurately comprehend the needs of organizations and the members within it, and they use that knowledge to influence others to help them map the informal communication pathways embedded within these organizations . Past research has linked political skill to both employee and manager performance . The present study broadens this understanding to include the role of informal leadership recognition in determining the success of politically skilled employees. Finally, much of the previous research on informal leadership has focused on the role of leader emergence in small groups or teams, and is viewed as a within-group phenomenon . On the other hand, leader effectiveness often is a basis of comparison between groups based on their performance. Based on previous research, we suggest that the job performance of informal leaders is determined by leaders' ability to adequately and appropriately evaluate organization dynamics, and leverage the resources and latitude afforded by their leadership status into elevated job performance . Pescosolido highlighted the role of emergent leader behavior in setting the norm for what is considered to be appropriate behavior. More specifically, it is the idiosyncrasy credits held by informal leaders that enabled them to influence the individuals in their immediate environment . It is indeed the resources and latitude afforded to these leaders that provide them with the appropriate context within which to exercise their influence. Overall, this study unites both the emergence of informal leaders and their subsequent effectiveness through an integrative social network-political perspective on the individual job performance of individuals recognized as informal leaders. --- Theoretical Foundations, Model, and Hypotheses In recent years, informal leadership has been a topic of growing interest. In the present study, an integrative socialpolitical conceptualization of informal leadership is proposed and tested. Specifically, the mediated moderation model argues that employees high in political will are more likely to be generally recognized as informal leaders, and the performance of these informal leaders should be contingent on their political skill. --- Informal Leadership Classic organizational theory suggests that the organization exists on two planes: the formal and the informal. Barnard defined the formal organization as a "system of consciously coordinated activities or forces of two or more persons" . The explicit and shared objectives unite employees to a common purpose, and the organization's success is dependent on their willingness to contribute to this purpose . As a fundamental component of every formal organization, the informal organization represents the aggregation of employee attitudes and personal connections. From these definitions, it is apparent that the informal organization is a response to employees' social needs and the subsequent organic grouping outside of the formal organizational structure. As central actors in the informal organization, it is necessary to recognize informal leaders, identify their distinctive characteristics, and understand how they can support organizational effectiveness through their individual performance . We conceptualize informal leadership status as a product of coworkers' social perceptions, reflected in the degree to which they are recognized as leaders. Indeed, an individual must be acknowledged and recognized in the aggregate perceptions of followers as a leader to exert influence over others . L. Lord, Jefferson, Klass, Nowak, and Thomas conducted in-depth interviews highlighting the specific nature of leadership in a nursing context, similar to the organization used in the current study. The authors specifically cited informal leadership as the element that keeps the organization running smoothly and provides motivation to other coworkers. Interviewees further elaborated that "the informal leadership . . . is the stronger leadership" and that informal leaders were viewed as sources of knowledge for others and more likely to be able to get important work done. The research summarized highlights the unique role of informal leaders for their peers as well as in the organizational setting as a whole. The current research contributes to an understanding of the motivations of individuals to become informal leaders and how they are able to maintain their performance in parallel to their informal leadership status. As such, we propose that leadership status may, in itself, be a goal, and as such, employees invest their social and reputational capital to obtain recognition as leaders by their peers when the organization does not formally confer such status on them. Because leadership status is viewed as a personal goal of employees, an appropriate measure of leadership effectiveness is an assessment of how this status relates to an increase in the informal leaders' personal performance in organizations. To capture these aspects of the informal leadership experience, the present study articulates a model incorporating political will and political skill. Political Will and Informal Leadership. Previous research on political will has identified need for achievement and intrinsic motivation as indicators for the degree to which individuals will engage in political behavior . Indeed, Treadway and colleagues found that these elements predicted the general propensity to engage in political behavior at work. Building on their findings, the current study seeks to explore a specific political context, namely, informal leadership. As such, given the context-specific nature of the question presented, this research specifically considers the need for power as an important element of political will in the recognition of informal leaders. McClelland's ) work on needs introduced the notion that individuals differ in their motivations. In order to better understand the role of motives, subsequent research on the impact of different need motivations shifted to understand how these motivational values translate into goal-directed behaviors . As discussed by McClelland and colleagues, the explicit nature of personal motives indicates a cognitive orientation that could lead to sustained behaviors congruent with those values. These explicit motives are likely to be activated by social contexts, such that it can be expected in situations that require leadership, a person with high self-ascribed power motives will cultivate a cognitive penchant for exerting influence and affecting others . On the other hand, implicit motives would be predictive of behaviors that ultimately could lead to managerial jobs, but are not limited to the context of the work environment . For the current research, we are interested in a particular context , as it is likely to trigger these behaviors in individuals high in power motivation. The motive to influence and exert control over others, as captured in power motivation, often has been a theme in organization politics research . Indeed, power motivation illustrates a particular likelihood to "expend energy in pursuit of political goals" , and is a likely precursor to informal leadership. Treadway et al. represent the first attempt to connect the willingness to engage in organizational politics and the ability to do so effectively. More specifically, these authors suggested that the motivation or will to engage in political behavior would predict actual political behavior, and their political ability or skill would translate the behavior into organizational outcomes. The present study builds on their findings to consider need for power as political will and informal leadership as the context-relevant political behavior that interacts with political skill to affect performance. Because power tends to be viewed as integral to effective leadership, power motivation is frequently examined in the leadership context. Individuals with a strong desire to influence others often find themselves in leadership positions . In fact, a high need for socialized power is a stronger predictor of the attainment of formal leadership status than is either a high need for affiliation or achievement . Early research on power motivation found support for the positive association between individual need for power and the management role . Individuals' achievement need motivates their personal development and desire to obtain results themselves, which may decrease potential efficiency as a leader. Similarly, the need to be liked, emblematic of a high need for affiliation, may impede performance on task-related objectives. In the interest of integrating research on leader characteristics and informal leadership, the present study seeks to understand the extent to which individuals with high power motivation are likely to be recognized as leaders. Individuals with a high power motivation are likely to display behaviors such as organizing the activity of others and taking advantage of leadership opportunities as they present themselves. These directive behaviors are consistent with our general expectation that leaders be decisive and dominant , whereby individuals who engage in these behaviors to a greater degree are more likely to gain the reputation, and be informally ascribed, as informal leaders. We then argue that the need for power of individuals should give impetus to their identification as leaders by their peers. Hence, we formulate the following hypothesis: Hypothesis 1: Need for power is positively related to informal leadership. Political Skill and Performance. Political skill is "the ability to effectively understand others at work and to use such knowledge to influence others in ways that enhance one's personal and/or organizational objectives" . Politically skilled employees are capable of understanding the social context and the motives of others in the workplace. Because of their effortless execution of these influence attempts and ability to mask self-serving intent, the targets of their influence behaviors generally see them as genuine and sincere in their actions. Ultimately, politically skilled individuals are able to construct broad and strong networks because of the favorable attitudes and reciprocities they build within the social structure . Employee effectiveness is a central tenant of the conceptualization of political skill in that highly skilled employees are expected to have greater success within organizations . Indeed, in the initial tests of the political skill construct and theory, increased political skill of managers was found to be associated with higher ratings of their job performance . Furthermore, the metatheoretical framework Ferris et al. presented suggests that political skill has implications for the self, surrounding others, and the group or organization. Nonetheless, the political behavior that characterizes politically skilled employees may be limited by a lack of social capital or, as in the current context, informal leadership status. Employees who are informal leaders tend to take a more active role in the organization and due to their salience and visibility, are able to operate with a wider range of acceptable behaviors -including political influence behavior. In other words, informal leadership status allows politically skilled employees to be more or less effective. More specifically, for individuals high in political skill, performance will be high for those who have been identified as informal leaders because the networking ability, capacity to read accurately and act on social situations, facilitate the work of others, and execute all of these behaviors in a sincere, genuine, and affable manner should be perceived and interpreted by superiors as effective performance from these informal leaders. That is, they will be perceived as demonstrating situationally appropriate behavior and executing it effectively. Similarly, for individuals who are low in political skill, there should be no relationship expected between informal leadership and performance. In other words, whether individuals are identified as informal leaders, due to their political will, the informal leadership status of lesser politically skilled employees' remains unrelated to performance as they lack the necessary political abilities to leverage their informal leader behaviors in ways that contributes to being evaluated more favorably. However, we expect that for individuals high in political skill, who have not been identified as informal leaders, their performance will be evaluated lower because they will be perceived to be engaging in situation-inappropriate behaviors. Individuals high in political skill are adaptable, but they still will try to utilize their networking ability, and opportunity recognition and capitalization behaviors, which fit in an informal leadership context, but not as well in a regular job context-as highlighted by their deflated performance. Indeed, politically skilled employees who are not recognized as informal leaders are expending "idiosyncrasy credits" that they simply do not possess by trying to influence others. By exercising political behavior without either formal or informal recognition, politically skilled employees experience a misfit in reputation and ability. This misfit results in a decreased ability to effectively manage the political arena of the organization and personal performance as well as their lesser skilled or informally endorsed counterparts. To that point, we specifically used the organization's criteria to capture individual job performance. This provides the most context-relevant assessment of criteria of which employees are aware and determine their success in the organization. Thus, the following hypothesis is formulated: Hypothesis 2: Political skill moderates the mediated relationship between need for power and individual performance. Specifically, for individuals high in political skill, greater recognition as an informal leader is associated with higher individual performance. For individuals who are low in political skill, greater informal leadership status is unrelated to variations in individual performance. Last, high levels of political skill without informal leadership status is related to lower individual performance. --- Method --- Sample and Procedure The sample was composed of employees from a mental health facility. The nature of the study required data to be collected both from the subordinates and the supervisors. The initial survey was targeted to a total of 88 professional employees that contained both self-report and social network measures. Respondents were provided with an informed consent document that explained to them the nature of the study, and that performance data would be obtained from their supervisors 3 months after they completed their surveys. Of the 88 employees evaluated at Time 1, some employees had left the organization; therefore, 76 employee evaluations were possible. The final response rate for Time 1 was 85.5% or 65 employees and of those, 40 provided complete leader networks to be used in the analyses. The respondents of a residential mental health facility in Northwestern United States completed surveys for this study. The employees were aides who lived in the residential facilities with the mental health patients. In this organization, of the sample that reported their ethnicity, 56.8% of the sample was Caucasian, 10.2% Hispanic, 1.1% Asian, and 7.9% other. The performance appraisals returned accounted for 65 of the original 76 employees . Of these final 65, 6 were supervisors and 59 were subordinates. --- Measures Political Will . Power motivation was assessed with a five-item measure from the Manifest Needs Questionnaire . Sample items were evaluated using a 7-point scale and included "I seek an active role in the leadership of my group" and "I find myself organizing and directing the activities of others." Valle and Perrewé captured power motivation using the need for dominance scale, and found it to be positively correlated with internal locus of control and proactive personality measures. These findings are consistent with previous leadership research . Political Skill . The Ferris et al. 18-item Political Skill Inventory was used to measure employee selfevaluations of political skill. Sample items include "I spend a lot of time and effort at work networking with others" and "I always seem to instinctively know the right things to say and do to influence others." Subordinates rated their agreement with the items using a 7-point scale . Previous research has shown that self-reports of employee political skill are correlated significantly with peer and supervisor assessments. Informal Leadership. In-degree centrality is an index of "aggregate prominence" and representation within a particular network. The results reported are based on the reports of the 65 employees who completed the network measure at Time 1 to provide a more comprehensive report from a larger portion of the organization. In order to calculate this index, each employee was asked to respond to a single question applied to a roster of all employees . Specifically, respondents nominated their coworkers as a leader and were asked to "put a check next to the names of people you consider to be leaders in your organization. These individuals may or may not be officially designated by your organization as leaders." This could range from no nominations to a nomination from every other coworker. This information was then used to calculate individual's network centrality using UCINET 6 . In this sample, informal leadership status ranged from 0 to 51.7 based on the Time 1 reports from 65 participants. In this context, high centrality demonstrates that individuals are readily identified as leaders in their organization by their coworkers. 1 Performance. We used the organizations own internal measure of performance and was gathered by the human resources department 3 months after the first survey. We were given the raw data from the organization similar to other work investigating job performance . This performance measure has been used within the company to regulate and provide performance feedback; thus, it provides context-specific information about individual performance. The company calculates the overall performance score based on different dimensions including employees' relationship with their coworkers and supervisors, their interaction with other staff and clients, professionalism, and punctuality. Each employee was rated from 1 to 4 on each dimension, and the composite score, generated by the company, was then used to determine overall performance . Control Variables. Of the number of demographic variables collected, age, gender, race, and management position were controlled for in these analyses, as age , gender , and race have been shown to have an effect on leadership and/or performance. In pretests, demographic factors such as gender and race were controlled for, but only age and management position were found to have an impact on informal leadership. This is not inconsistent with other research finding that age and performance are likely to be positively related as job performance could possible improve with age . By including management position, we are able to parcel out the informal leadership variance attributed solely to formal position and better capture the recognition of the individual as a leader. --- Results Table 1 presents descriptive statistics, correlations, and reliability coefficients for the variables used in this study. Since the data were not nested in teams and rather captures an entire organizational network, regression was chosen to test the proposed model. To test the mediated moderation model, Preacher and Hayes SPSS macro, with resampling or bootstrapping methods to guard against violations of the assumption of normality and Type I errors was used. The Preacher and Hayes method also tests the indirect and direct effects of the moderator through the proposed model allowing for the evaluation of alternative models. The results from the bootstrapping analysis are located in Table 2, and they indicate a significant mediated moderation analysis. Need for power is positively related to informal leadership . The model summary from the control variables and need for power predicting informal leadership was significant, R 2 = .90, F = 28.30, p < .00. Informal leadership was negatively related to performance, but political skill was not related to performance. Finally, the interaction of informal leadership and political skill is positively related to performance . The complete model summary was found to be significant, R 2 = .60, F = 2.18, p = .06. The direct effect of need for power on performance was not significant . The direct effect of informal leadership on performance was significant at high levels, or 1 SD above the mean, of political skill . This means that employees who were high in political skill and recognized as informal leaders also attained high performance ratings. The relationship between informal leadership and performance was not significant for employees who were low in political skill , or who possessed average levels of political skill . The graph in Figure 2 plots the mediated effects of informal leadership at ±1 SD around the mean of individual levels of political skill on performance. Therefore, all hypotheses were supported. --- Discussion --- Contributions to Theory and Research Scholars have called for research to expand our understanding of the role of informal leaders . Confusion in the concept originates from the tendency of scholars to interchange the terms "informal" and "emergent" to describe employees who are identified as leaders by their coworkers . Outside of the team context, in conjunction with formal structured interactions, patterns organically emerge with particular individuals at the center of the action-these are the informal leaders . The very nature of this process suggests an overlap between the concept of informal and emergent leadership. Often, scholars evenly exchange the terms informal and emergent with regard to the discretionary networks that evolve outside of the formal or prescribed network . In other words, the informal content of the relationships in the organization can involve both a task focus in addition to the social element and it functions outside of the team context. Furthermore, the present study explored a political explanation, involving both political will and skill, for an employee's rise to, and leveraging of, informal leadership positions in organizations. That is, an integrative social network-political conceptualization of informal leadership was proposed and tested, and the results provided support for the model and hypotheses. The results demonstrated that employees high in political will are more likely to be recognized as informal leaders, and the performance of informal leaders is found to be contingent on their political skill. As such, the present study extends the social networks, leadership, and politics literatures by providing an empirical evaluation of individual differences as they relate to informal leadership emergence and subsequent performance. The results indicated that politically willed individuals were more likely to be recognized as informal leaders , yet recognition by coworkers as an informal leader was not related to higher performance ratings. This suggests that informal leaders' performance was dependent on their ability to manage the demands of their positions through their elevated levels of political skill. The findings presented here suggest that employees high in political skill who lack the status of informal leader actually suffer in their individual performance. In other words, the informal leadership status insulates politically skilled employees from the negative effects of engaging in political behaviors without the expected status. The present results contribute several points of depth to the political interpretation of leadership. The first of these is its integration of both motivational and political approaches to understanding informal leadership. Whereas previous research has linked leader political skill to subordinates' perceptions of organizational support, team performance, and ratings of managerial job performance within the formal structure of organizations , scholars have ignored the informal context of organizations. This oversight is striking given that the political perspective acknowledges that leaders are successful to the degree they are able to traverse both the formal and informal structures . The interplay of political motivation and skill has been at the foundation of political views of organizations , yet few studies have fully articulated and tested the framework or systematically applied it to the leadership context. As such, our research not only empirically justifies the direct inclusion of political considerations in leadership research but also identifies the sequencing of motivation and skill through which leadership performance and effectiveness is achieved. Specifically, individuals' motivation to achieve influence over others is an important precursor to them risking social capital to achieve a goal. This furthers the positioning of political skill as a personal resource that can be viewed as functional or dysfunctional for organizations and individuals to the degree individuals' motivations align with those of the organizations , or with the expectations created by the context. Our conceptualization of informal leadership offers a complementary theoretical viewpoint to the structural and dispositional approaches applied in previous research. Despite a recognition that informal leaders operate outside the formal structure of organizations, no previous work has attempted to offer or test a model that explicitly evaluates personal characteristics that would allow employees to thrive in the informal organization. As such, the present study represents a context-specific test of political motivations or political will and ability in the informal structure of the organization. In doing so, this model further establishes the relationship between political skill and political will , suggesting that political will serves as a driver of employees' willingness to risk social capital to achieve particular goals. Whereas the Treadway et al. work focused on the role of political skill as a resource in reducing emotional strain as a consequence of engaging in political behavior, the present study concurs with the role of political skill as a resource, but one that allows informal leaders to leverage their status to achieve personal gain . These findings substantiate that the motivation to engage in a behavior and the skill to leverage that behavior are distinct entities, but both are important for job performance. Furthermore, they provide empirical support for the theoretical positioning of political skill as a personal resource . --- Strengths and Limitations The present study has several strengths that are worthy of note, most notably the use of the social networks methodology to study informal leadership as an organization-wide phenomenon. By collecting and aggregating employee perceptions of their coworkers as leaders, this study provides an inclusive perspective on informal leadership. Coupling the social network approach with the consideration of individual differences as predictors of perceptions of informal leaders is consistent with similar research conducted previously in the social networks literature . By using both self-report and social network methodology, common method variance demonstrated limited or no impact on the results presented in this study. Other strengths include the collection of data at two different points in time and from multiple sources. The time lag effectively diminishes the likelihood of bias due to prior knowledge of the questionnaire. In addition, the use of the social networks methods requires the collection of data from every employee. The response rate for this study was very high and increases the likelihood of capturing the aggregate perceptions. Finally, Preacher and Hayes's analysis of mediated moderation relationships offers a more robust test of the model and hypotheses than the Baron and Kenny approach. Together, the strengths of the current study improve the interpretability and generalizability of the results. This study is not without limitations, however. There is potential for ambiguity as it relates to the theoretical conceptualization of informal leadership as it may be confounded with other concepts such as emergent leadership and shared leadership. The present study chose consensus measure using a social network methodology to assess informal leadership. Given the need for participants to rate every other organizational member, the sample size may indeed reflect participant fatigue. It is important to note that previous research using this methodology has limited the network to that of the team and not the entire organization . Furthermore, the generalizability of these results to the informal leadership literature may be affected by the context as well as the unique operationalization used here which includes the formal leaders. While the surveyed organization provided specialized services in a specific context, generalizability may be called into question. However, the results found in the current study parallel findings in other studies which support the validity of the current results. Nonetheless, there is support for the use of the social network approach to capture socially constructed positions of leadership , and as such, this method seems appropriate in this context. --- Directions for Future Research If the present research demonstrates that political will or motivation is critical to becoming an informal leader, it may be useful to expand the number and type of motivational variables that are considered in framing political will as an indicator of individuals' motivation. Mintzberg's discussion of both the definition and mechanics of political will were insightful, yet few studies have attempted to operationalize his thoughts. The present study demonstrates that political will can be seen as reflecting concerns for both internal needs satisfaction and observable goal attainment. A more refined view of political will is needed to advance our understanding of how motivation affects political processes in organizations . This treatment should encompass not only internal and external motivation but also more strongly integrate context as a motivation and as a barrier to action. Contextual considerations are of particular concern for leadership researchers , yet informal leadership scholars have provided little discussion of these issues. In the intrapreneurship context, "champions" are considered informal leaders in that they actively promote innovation within their firms . These informal leaders champion the causes of their organizations, thereby encouraging movement and support among their coworkers by utilizing their ability to articulate a vision and inspiring others to follow it. --- Conclusion The instance of informal leadership fuels the inquiry into who are perceived to be leaders and what determines their effectiveness. Brass suggested that leader effectiveness is based less on leadership traits, and more on the "ability to accomplish work through others" . This study proposed that the political will or power motivation of individual employees propels them into the role of informal leaders. Their subsequent individual job performance was then contingent on the match between political skill and the informal leadership context. These arguments have been made through an integrative intersubjective-political conceptualization that highlights leader characteristics, informal leadership status, and their individual performance. In sum, the empirical support we found for the proposed conceptualization provides a basis for understanding the identification, role, and success of informal leaders in organizations. We hope these results stimulate further research interest in this area of work. --- Note 1. For the purposes of this study, the scores were not symmetrized; that is, when person i nominates person j as a leader in his or her network, it is not necessary for Person j to nominate person i in order to calculate person i's centrality score. Conceptually, this approach also makes sense in that formal and informal leaders need not recognize other employees as leaders to be considered leaders themselves. Finally, there is usually a discrepancy between individuals' ratings of leaders in organizations in that high-status leaders are unlikely to reciprocate nomination of lower status individuals as leaders . --- Declaration of Conflicting Interests The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. --- Author Biographies Brooke A. Shaughnessy is an assistant professor at the Ludwig-Maximilians-Universität München in the Institute for Leadership and Organization, Munich, Germany.

Informal leadership has been a topic of growing interest in recent years, with the recognition that much remains to be known about this phenomenon. In the present study, an integrative social-political conceptualization of informal leadership is proposed and tested. The research question was tested through individual self-report survey questions, a network-based consensus informal leadership measure whereby each employee identified informal leaders in their network, and individual performance provided by the organization. Specifically, the mediated moderation test demonstrated that employees high in political will, as operationalized by power motivation, were more likely to be collectively recognized as informal leaders than those low in political will, and the performance of these informal leaders was found to be contingent on their political skill. By capturing informal leadership using a consensus measure, the results of this study provide a first look at informal leadership in an organizational setting, not team or group. Furthermore, the current research offers a social networkpolitical conceptualization of informal leadership in organizations that contributes to theory, research, and practice.

Introduction Foremost among the justifications offered for self-driving cars is that they will offer dramatic improvements in road safety. The promise is based on an assumption that the automation of driving, an activity prone to numerous human failings, will be possible in the short term thanks to rapid developments in artificial intelligence. If computers can take over the tasks of sensing and interpreting the world, predicting the behaviours of objects within it, planning a safe path and controlling a car's speed and direction, the idea is that human performance can be rapidly matched and then exceeded. A well-known public health catastrophemore than a million global road deaths each year, a hundred per day in the US alone-provides a strong motivation for radical improvement, with technology offering powerful options. However, new technologies raise questions about safety as well as offering answers. If self-driving cars are to earn public trust, we should ask, at an early stage, how safety can be assured, demonstrated and improved over time. This challenge stretches beyond engineering . Questions of regulation and safety assurance have been given insufficient attention as self-driving car developers focus on demonstrating the technology's potential. There has been little research to find out what the public thinks about self-driving car safety. Postponing debates about safety presents hazards for the public and reputational risks for developers who may be undone by their own or others' recklessness. In 2016, a Tesla that was in Autopilot mode crashed in Florida, killing its sole occupant. This offered a stark reminder that technologies that attempt to automate at least part of the job of driving were less safe than their proponents claimed. In their crash investigation report, the National Transportation Safety Board were eager to point out that this vehicle was not a self-driving car, even though the data extracted from the vehicle suggests that its owner was behaving as though it was one . The NTSB went on to investigate other Tesla Autopilot crashes as well as a crash in Tempe, Arizona in March 2018 in which a self-driving car operated by Uber hit and killed a woman who was walking her bicycle across the road. These collisions, and the investigations that have followed, have revealed not just a carelessness among some developers, but also a lack of consensus about how to assess risk, and an absence of clear regulation or standards to govern the testing or approval of new self-driving technologies. These incidents remind us not just of a technology's limits, but also of the flaws of a mode of governance that leaves technology developers to their own devices. The 2020 independent expert report for the European Commission on the ethics of connected and automated vehicles recommended that the technology should reduce overall risk, be designed to prevent unsafe use and have clear standards for testing on public roads. These principles offer a strong regulatory ideal, but any approach to governance must engage with a political and economic reality. Self-driving car companies talk about a race to develop the technology. Governments' enthusiasm for innovation has seen them buy into this story, which has meant lax governance regimes. The question of how we can know a self-driving car is safe is complicated. It depends on assumptions about how safe is safe enough, who needs to be persuaded and what constitutes a self-driving car. This paper explores these qualitative aspects of risk using qualitative data from interviews and workshops with members of the public. My team and I conducted 50 interviews with self-driving car developers, researchers and policymakers in the UK, US and Europe as part of the "Driverless Futures?" project. The interviews lasted between 30 and 90 minutes and took place between 2019 and 2021. The aim of the interviews was to go beneath public accounts of the benefits and risks of self-driving cars, the hypothesis being that the people closest to research and development would have a clearer sense of the uncertainties, complexities and contingencies of the technology and would be able to articulate these during long interviews. Interview quotes are anonymised here using numbers. In addition, I draw on transcribed conversations from a large public dialogue exercise that took place in 2018 and 2019, commissioned by the UK Department for Transport and Sciencewise. The CAV public acceptability dialogue was the world's first substantial attempt at deliberation designed to inform policy for CAVs. It involved 150 public participants in five locations, over three weekends, informed by expert visitors. Participants were recruited to reflect the diversity of the UK population. I was part of the team designing, facilitating and reporting on the process.1 This research reveals the diversity of understandings from people inside and outside the community of innovators. While the discussion is currently dominated by engineers, there is a clear need to include perspectives from other stakeholders and members of the public. --- How safe is safe enough? In 1969, Chauncey Starr asked how, in weighing the benefits and costs of new technologies, we might consider variations in the acceptance of different types of risk. His question, "How safe is safe enough?", prompted consideration of the dimensions of safety that couldn't be captured by a calculus of probabilities and outcomes . The 1970s and 80s saw growing interest in research on risk perceptions that examined the importance of psychological biases and heuristics in explaining individuals' attitudes to risk . Risks that were seen as new, uncontrolled, catastrophic and artificial were found to be consistently exaggerated. For transport, we see markedly different societal assessments of acceptable risk. The risks of many transport systems are by now well known, but the evolution of law, regulation, technology and culture suggests that people are much less willing to accept risks from modes of transport they regard as highly centralised and out of their control. UK train operators are willing to spend, according to one estimate, tens of millions of pounds per life saved on the railways ; meanwhile, there is chronic underinvestment in affordable and available technologies for car safety , even though cars are far more dangerous. Revealed risk preferences are observable in hindsight. However, formal risk assessment demands the calculation of outcomes and probabilities, both of which will be uncertain for new technologies. Alvin Weinberg, a Cold War physicist and nuclear energy enthusiast, described the difficulty of risk assessment for rare, catastrophic events in complex technologies, such as a nuclear accidents: Because the probability is so small, there is no practical possibility of determining this failure rate directly -i.e., by building, let us say, 1000 reactors, operating them for 10,000 years and tabulating their operating histories. . Social scientists have analysed the limits of conventional modes of science and policymaking in conditions of uncertainty or ignorance and the tendency to treat incalculable uncertainties as controllable risks. To use Hansson's analogy, the tools of risk assessment trick us into believing we are in an environment like a casino, where risks are known, when we are actually surrounded by a thick jungle of unknown and possibly unknowable hazards. The development of new technologies, from this view, is a form of experiment whose variables and metrics cannot be well-defined in advance . --- 3 Starr's challenge and subsequent research into risk perception are limited by a one-dimensional view of risk. An alternative research agenda that symmetrically problematises new technologies as well as public responses to them has seen risk as multidimensional. The politics of new technologies mean that questions of risk may be unavoidably bound up in questions of equity or freedom. The assessment of risk, even though it is often discussed as a value-neutral activity, is political and ethical as well as scientific . The language of technology governance tends to presume a separation and a sequencing between scientific risk assessment and risk management. Risk assessment is assumed to be scientific and risk management is where questions of trust, acceptability, uncertainty and politics come in. The public, it is assumed, only have an interest in the management of risks. The rise of research on risk perception has hardened rather than blurred this boundary. Engineers' appreciation for public views on risk cemented the view that theirs was the correct assessment, and that it was the divergences from this that needed social science explanation. If we pay attention to the framing assumptions of risk assessment, we can however see the limits of such a model both empirically and politically . The assumptions behind risk assessment are revealed to be brittle when tested in terms of public credibility. In the face of social and technological uncertainties, different groups will seek to draw parallels and precedents that either problematise or downplay novelty. Proponents of genetically modified crops, for example, sought to argue that the technology was 'substantially equivalent' to its conventionally bred counterparts, while advocates for precautionary regulation highlighted novelties and uncertainties . A successful innovator must finely balance claims of novel benefits with reassurance that their technologies do not require radical regulatory attention . --- Performance, assurance and reassurance As sociotechnical systems have become more complex, more dependent on digital and automated technologies and more privatised , regulators have sought to make their rules less prescriptive and more 'performance-based' . The idea is to give innovators an end goal rather than tell them how to get there. The hope is that this approach encourages innovation and allows for more focussed regulation. The model presumes, first, that we are clear on how to judge performance and, second, that the public has no interest in what is going on behind the scenes. It represents a way of knowing as well as a way of governing: a presumption that innovators know best and will be able to account for the public interest. According to one study of the recent crashes of Boeing 737 Max aeroplanes, if a performance-based approach is going to encourage safety, it should resist simple metrics and have independent verification . Without external scrutiny, performance-based approaches look less like genuine safety assurance and more like naïve attempts at public reassurance. Unlike some technologies, developed in a laboratory before being released into the world, self-driving cars are being developed in public. Their developers have therefore been compelled to build public stories of their safety that they hope will be sufficient to secure a social license to operate on public roads. When the technology was brand new, these stories reflected a 'technological sublime' . As the technology has become entangled in real-world complexity, the stories have been modulated in response to others' questions and concerns . The stories provide first drafts for risk assessment that could become hugely consequential. From my interviews and publicly available sources, I have extracted two competing narratives. The first, safety-in-numbers, starts by presuming self-driving cars are a solution to a perennial safety problem and looks for metrics to show progress. The second, safetyby-design, starts with the question of safety assurance and problematises self-driving innovation. --- Safety in numbers In April 2020, during a conference presentation on 'AI for full self-driving' Tesla's senior director of artificial intelligence announced that the company's cars had driven 3 billion miles on Autopilot. Autopilot is a limited automated system, but the number is meant to reinforce the impression that a self-driving Tesla is just around the corner. The claim is twofold: first, that Tesla are harvesting more data than their competitors and, second, that their system has a track record of safety. The safety-in-numbers narrative starts with a simple calculation: we know the risks of human driving; self-driving aims to eradicate that risk; therefore, as long as the technology works and there are numbers to show it works, there will be safety improvements. The stated justification for developing self-driving systems is to solve a problem of safety; the system's adequate performance is therefore a demonstration of safety. This approach ignores the question of who needs to be convinced; the developers' own assessment of performance is the relevant criterion. It's an approach that has defined self-driving car development since Google first funded its self-driving car project in 2009. Google's engineers were given a target: if they were able to clock up 101,000 self-driving miles hiding in plain sight on California's roads, they would receive large financial bonuses. By the time these secret tests were publicised, they were almost complete. The New Yorker later revealed that the company's self-driving cars had been involved in multiple incidents during this time, but there were no laws compelling the company to report them. The response from the company, which had by then been spun out from Google and renamed Waymo, to the New Yorker revelations is telling: The Google self-driving car project was founded with a mission to improve road safety, and that's the standard we hold ourselves to in everything we do. Over the past near-decade, we've carefully developed a comprehensive testing program that includes more than 10 million miles on public roads. 2Waymo's claim rests on a statistic of number of miles driven without a death or serious injury. This superficial metrics demands further analysis. A series of reports from RAND have taken on the question of measuring safety. The first concludes that, if self-driving cars are to demonstrate improved average safety over human driving, they would have to rack up 275 million miles without a mistake. Their conclusion is that "developers of this technology and third-party testers cannot drive their way to safety" . However, the assumption behind the RAND reports is that average improvements in safety still justify the rapid deployment of self-driving cars . In the search for measures that might be both useful and publicly persuasive, RAND popularised a distinction between leading and lagging metrics in a report that was commissioned by Uber . The argument is that lagging metrics of outcomes might be easy to measure, but leading metrics, including the number and type of minor transgressions a self-driving car makes, might allow for the prediction of performance. Judging self-driving performance by number of fatalities per million miles driven might be possible after substantial experience of the technology, but this information is useless in regulatory terms and a poor indicator of performance. Humans move on from the embarrassing everyday near misses that characterise our imperfect driving. Self-driving cars and their regulators can and should learn from the crashes that don't quite happen as well as those that do. 3From the safety-in-numbers viewpoint, the technology's safety benefits are jeopardised by irrational public risk perceptions that mean we underestimate the safety of modes of transport that we presume to control, such as driving, while overestimating the risks of systems that are out of our control and seem uncanny, such as flying. Some early research suggests that a sizeable proportion of the public wants self-driving cars to be at least a hundred times safer than conventional cars. Interviews with self-driving technology developers and researchers provide an opportunity to get beneath the superficial story of safety and probe some of the claims being made. The public narrative of how self-driving cars 'work' hides a broad range of views even among those trying to get the technology to 'work'. Companies are adopting diverse strategies, with some emphasising safety and responsibility while others, particularly smaller start-ups, find it hard to divert core engineering resources to address safety assurance. The most optimistic enthusiasts for self-driving see self-evident safety benefits, meaning that public persuasion becomes just an extension of engineering. One of my interviewees, a leading artificial intelligence researcher, argued that the statistics would inevitably force the hands of regulators: At some point in the near future, it's hard to predict when… you will have [self-driving] cars that are maybe, on average, ten times safer than humans. It will be three times, then five times, then ten times safer. It's a matter of statistics… I'm not sure whether the factor of ten is sufficient, maybe you need a factor of 100, but at some point they're going to be mandatory. Some interviewees entertained a consequentialist argument that there might be short term hazards from a technology under development, but the long-term safety benefits would justify the means. Other researchers engaged with the reality of public risk perceptions. One concluded that, even if average safety improvements were unarguable, "That's a hard one to deal with when it's your child that got run over by the vehicle" . The recognition here is that the technology would change the qualitative as well as the quantitative aspects of safety, making questions of responsibility inescapable. Other interviewees referred to "algorithm aversion" , a hypothesis that members of the public might exaggerate the hazards of automated systems. 4While constructing numbers that they hope will offer reassurance, self-driving companies have sought to normalise the technology with public displays of flawless driving. YouTube is replete with companies' demonstrations of the technology working, but these videos only report on success, offering little assurance on the technology's limits. Even though self-driving car developers, as one interviewee put it, "need to be out on the road… racking up the miles" , they know that this costs money and that not all miles are equal. One interviewee from a large car company criticised self-driving car start-ups who drive for millions of miles to prove a point… it's end-less… We have to let go of that. Three or four years ago that used to be the criterion: How many miles have you driven? That's not the issue any more… if I'm driving a hundred miles between one intersection and another one -a straight, simple road -the fact that I drove a hundred miles is insignificant. . Another researcher argued "You could drive up and down the Nevada desert. A hundred million miles. It doesn't help me if I'm going to use it in London" . When Google's engineers were given their target, the company recognised the variability of driving: 100,000 miles could be ticked off on North California's easy roads, but there were also 10 predefined routes on more challenging terrain, including Lombard Street, known as "the crookedest street in the world". While engineers recognise the qualitative variation, the pure numbers remain seductive. Waymo continues to announce milestones of incident-free distance. Another company's CEO has argued that progress could be measured out by increases in the number of "miles per disengagement" .5 One interviewee, an investor in self-driving companies, ran with this idea of a "Moore's law for self-driving vehicles", starting with the rough calculation that human driving in the US produces a fatal crash every hundred million miles: How long will it take to get to one disengagement every hundred million miles?... 15, 16 years, something like that… We're not going to tolerate machines killing people at the rate of 40,000 a year in the United States. So they've got to be maybe an order of magnitude more safe. Add another order of magnitude to that? That's sort of the timeline. One British self-driving company, seeking to emphasise its responsible approach to safety assurance, has taken issue with the "disengagement myth": It's now clear to everyone that simply measuring progress as improvements in miles between disengagements hides many failures that might not bubble up to the level of disengagement, whilst at the same time enforcing an extremely slow development cycle. That's not to mention the need to physically drive hundreds of millions of miles to be statistically confident. 6The continued presence of simple statistics in the public debate even while engineers agree that they are flawed is an echo of self-driving's origin myth. The technology's feasibility, according to this story, is enabled by recent and rapid advances in artificial intelligence. The technology therefore takes its inspiration not from other mobility technologies, but from technologies like machine translation, which requires little linguistic expertise, relying instead on what the researchers behind Google Translate called "the Unreasonable Effectiveness of Data" . Machine learning, at its root, is statistical. The hope is that, with enough data, performance can become superhuman even if the way machine learning works is utterly unlike human learning and is usually opaque . The approach to self-driving that prioritises data for machine learning is sometimes called "brute force", but the question of safety has proven hard to force. Some interviewees talked about getting the technology to work safely in terms of percentages: I think we're doing a pretty good job with technology. It is really close to actually working. it's always tough to get the last one or 2% out of these things, it's easy to do 80%... 20 is hard. The last 2 is really hard. The last 0.2 is really, really hard. An interviewee who once ran a self-driving start-up concluded that "scaling safety is going to be so hard and take so long" . As discussed below, an alliance between probabilistic machine learning and probabilistic risk assessment will also struggle to achieve public credibility. Self-driving car developers, some of whom have switched over a decade from regarding the technology as impossible to seeing it as inevitable, now find themselves asymptoting towards an ideal of safety that may always be out of reach. More data and more miles produce better systems, but they also reveal more 'edge cases'-circumstances that the model cannot account for. Engineers recount the unusual things-balloons, ducks, wheelchairs, kangaroos-that their sensors have seen but which their software has struggled to make sense of. The sheer complexity of their challenge can lead to frustration. Interviewees often expressed disdain for unruly pedestrians or poorly maintained roads. Predicting the movements of pedestrians is impossible with certainty, but pedestrians are unavoidable. And yet, obviously, they must be avoided. Developers' usual answer to this challenge is that it will be met with more data from which the system can learn. As they confront the challenge of safety, some admit the impossibility of perfection. ). But all developers will remain troubled by the normal abnormalities that exist in a world designed by and for humans, whose autonomy and mobility in the environments that self-driving cars seek to occupy is, for now, relatively unconstrained. As one safety engineer explained, complex systems are impossible to describe with one probabilistic risk assessment: If you don't understand the design of this system, how do you know that number's right?… We're trying to solve problems that are actually impossible. We throw numbers at them, we almost make them up, but they don't apply to that design. This engineer points to a gap between the numbers that are available and the numbers that regulators might need to assess a system. The data that are of interest to developers as they seek to get their technology to work may not be relevant for safety assurance. Others may have a very different sense of what it means for the technology to 'work'. The safety-in-numbers narrative is superficially impressive, especially when weighed against the known risks of driving and as long as self-driving cars aren't implicated in high-profile crashes. It fits a prevailing regulatory assumption that what a system does is more important than why it does it, and that we should trust innovators to show us what they can do. The limitations of the narrative become more apparent when, rather than taking safety as self-evident, we start with the challenge of designing safe systems. --- Safety by design While most self-driving car developers focus on improvements in AI and demonstrating safety through performance, there are engineers emphasising safety-by-design who are more likely to have had experience with hardware, requirements engineering and safety assurance of other complex systems. Looking at a prototype self-driving car, they see a potentially lethal safety-critical system; a heavy robot travelling at speed in an uncertain, uncontrolled environment. From this standpoint, the challenge of safety assurance looks daunting. This group draws attention to issues that they see as important but neglected in the simplistic safetyin-numbers story. These issues, discussed below, include human-machine interaction, system safety, redundancies, interpretability and simulations. The self-driving ideal removes the human from responsibility, if not always from the driving seat, but engineers with experience of humans in-the-loop know that people can never be completely automated out of sociotechnical systems . Safety engineers now have decades of experience with aeroplane autopilots and other systems involving human-machine interaction . They have warned about the risks of 'mode confusion', 'skill detriment' and handovers between human and machine responsibility, issues that selfdriving car developers are now coming to terms with, sometimes in reckless ways . Some regard such issues as temporary, worthy of attention while prototype systems are being developed, overseen by safety drivers who are expected to take control in the event of a technological failure. But other engineers have called for users to be a permanent part of a new approach to 'informed safety' . As Lisanne Bainbridge argued almost 40 years ago, "there will always be substantial human interaction and involvement with automated systems" . Even if the humans in control of a vehicle are completely reliable, interactions with humans outside the vehicle multiply the complexity, and cannot be easily engineered away. One software engineer interviewee said, "in the context of self-driving cars, something that we don't yet know how to do is handle the humans in-the-loop and interaction with human-driven cars" . These interactions, which, for a human, define driving, must be reinterpreted by engineers to become amenable to a technological fix. The assessment of a vehicle's risk necessarily involves more that the vehicle itself. The vehicle's context, in engineering terms, is sometimes called an 'operational design domain' . The ODD represents the conditions in which a self-driving car can reliably operate, and may include material features like road types, weather, other road users and infrastructure as well as digital systems like highdefinition maps and communication between vehicles and the outside world. Many of these bits will be outside the control of a self-driving car company. One interviewee discussed the necessity of narrowing the ODD… The idea that you'll be able to flip a switch in a Tesla and it'll drive you anywhere there's a road is in my mind fantasy… If you can restrict an ODD… you can characterize the types of interactions that the vehicle is more likely to encounter. You're narrowing this whole available pool of scenarios to something smaller. In practical terms, this might mean 'geofencing' a vehicle to prevent it from straying into spaces that are too unpredictable, or it might mean changing the outside world to make a particular domain operational by, for example, restricting the movements of other road users or upgrading infrastructure. One engineer, discussing so-called 'smart infrastructure' that would be able to communicate directly with a vehicle, concluded "there are some things where it's so difficult to be able to assure safety without the infrastructure helping" . Another said there would be a "need to instrument the environment, for example, for self-driving cars, so that they can read traffic signs maybe automatically" . An interviewee from a company trialling the technology wondered, "are we going to have to tell people not to walk out in front of cars? I think we might" . Even setting aside the political ramifications of reshaping the outside world to suit a new technology, one can see the complexity of a safety-first view that sees risk as a product of complex systems rather than individual machines. For a safety engineer, the challenge might seem intractable. Innovators' emphasis on getting their technology to work obscures consideration of what to do when it doesn't. Some interviewees defended their technologies by pointing to redundancies and fail-safes. One self-driving car developer claimed that their company maps the environment because "we want to know in advance to expect traffic lights to be in a certain position. That's what gives us the redundancy in our solution. That's what makes us confident that we're not missing things." But a safety engineer wondered "when the RSU ['roadside unit', for communication between vehicles and infrastructure] fails, what's my back up?". This interviewee said there was a need to "build in redundancy and diversity" . Some engineers were particularly troubled by a dependence on AI systems that they regard as opaque and brittle. Asked about the challenge of understanding why a machine learning system does what it does, one engineer responded, Companies out there who sell GPUs… [graphics processing units-a type of chip used for training neural networks] claim that you can do everything inside the car. They will never build a car. They will never take the responsibility of cars driving automatically outside in the world… end-to-end learning? Having a neural network which takes in camera data, Lidar data, radar data and then operates the brake and the steering wheel? That's a nice showcase… but it will never happen on public roads… a system which we bring to the road always needs to be 100% deterministic... if you say, 'Well, I don't know what happened, there's a deep neural network', that won't work… it needs to be completely deterministic. The end-to-end learning referred to by this interviewee is an AI approach in which one model-a neural network-learns how to turn inputs such as the images from a camera into outputs such as turning the steering wheel. This interviewee worried that such systems were usually black boxes, and "engineering is pure responsibility", which meant the need for "somebody inside the company who signs off the system and is personally responsible." This person would not be able to "sign off a deep neural network" whose decisions were uninterpretable and non-deterministic. One AI specialist in a self-driving startup offered a counter-argument: Vaccines are not deterministic. You don't understand them. There is no transparency. There is no determinism. You use it based on the statistical guarantee that they give more value than damage, so I don't think it's a good argument against machine learning. Another machine learning advocate compared the interpretability challenge with a massive software package like Microsoft Windows: Millions and millions of lines of code are in that software package. Is that really explainable?... You can't understand really why certain decisions will be made. It's the same with neural networks… We should really look at the tests that are used to assess the software, not making sure it's human readable. This interviewee did however see the value of explainable AI as a PR tool: Every accident that an AV [autonomous vehicle] is going to be involved in, there's going to be press articles about it. Anything we can use to debunk any incorrect information I think would be very beneficial. The indeterminacy of both the worlds in which self-driving cars operate and the software that drives their decisionmaking persuades some engineers of the need for a fundamental rethink of the approach to self-driving that seeks a like-for-like replacement between a human driver and a computer: We shouldn't be replicating a human, who has such a lousy ability to perceive time and space and speed… We can actually define how the vehicle should respond so it will be safe. I don't want to mimic a human's faulty decision making. I can do it safer than it's done today. By this, we can save lives. This interviewee's design view encompasses the whole system-roads, other road users and more-rather than just a vehicle . Others agreed that the safety challenge could only be met by rethinking whole systems. Interviewees involved in self-driving start-ups were more likely to argue that if new technologies were unable to meet established standards of safety assurance, those standards would need updating. Some companies have proposed rules that they hope will guarantee safe driving while reassuring a sceptical public.7 These rules are egocentric; they are about protecting the vehicle and, crucially, protecting its makers from blame, rather than designing for collective safety. However, as with other self-driving innovations, these rules still require real-world testing. Human drivers must, in most jurisdictions, pass a driving test that certifies a certain level of aptitude, the assumption being that skills will be transferable across different driving conditions. In addition, their vehicles need to be certified as roadworthy. A self-driving test would blend elements of both tests, and would be confronted with some of the challenges discussed above: How could we test for a potentially infinite variety of edge cases? Would a self-driving car's capabilities be as adaptable as a human's? How could we avoid technology developers 'teaching to the test'? Would a licence be localised or portable to other ODDs? Would certification still count after a software upgrade? A self-driving test could, like its commonplace counterpart, combine a set of scenarios.8 These could be examined at a test track, on public roads or in computer simulation. Some developers would see such tests as straightforward extensions of what they are doing anyway to verify their driving algorithms. Simulation has quickly become a vital tool for risk averse engineers. ). But a test might only postpone the question of assurance. A safety engineer wondered: You've closed the world to build the simulation… When you go from the simulation to the real world, do any of the things that you've left out in building the simulation really matter? One developer argued that this meant a need to understand the internal processes of a vehicle's decisionmaking rather than just its performance: There is no world in which, as we develop this technology, we can actually test all of the possible permutations of things, so we have to understand at some logical level how it's understanding, interpreting the environment and making decisions so that we know that the methods that we are using to test it have appropriate coverage An AI researcher speculated on what a test for AI driving the real world would look like: Maybe you could have a simulated test… a billion miles in simulation? If it has fewer than ten accidents, you say 'you're good to go'. And it could be a completely uninterpretable one, as long as we have confidence in the test, which we don't yet, but maybe we could someday. Yeah, maybe that'd be fine. But one developer took issue with "people trying to introduce digital driving tests and rules about having to test autonomy in somebody else's simulator" before the technology 'works': "test companies and test facilities, test processes, they're talking about validation, certification, verification, digital driving tests, digital MOTs. We don't have anything for them to test yet" . As self-driving car excitement, investment, testing and development have expanded, the initially straightforward narrative of safety has been troubled by the perspectives of others, some of whom have become enrolled in the technology and some who are watching from the sidelines. It is unclear whether these perspectives can constructively mesh or whether they are incompatible. --- Safety first; safety last In discussions surrounding standards, tests and possible regulations, there is a clash of cultures. For one group, safety is a self-evident property of the technology. The challenge is therefore one of public reassurance. For the other, safety is a vital design criterion, supported by extensive domain expertise. The former would be sceptical of the latter, suspecting they are conservatively defending their own incumbency. The latter group's response would be that the upstarts are creating real and reputational risks through irresponsibility. One safety engineer was critical of "start-ups who don't have to lose anything. They only have to win venture capital" and a former self-driving CEO argued. Very few Silicon Valley companies have ever had to ship a safety critical thing… Google seem to think that if they code a program well enough, no one will ever die. That's just not how the world works. That's not how you build a safety protocol system… The lack of maturity about that has really hurt the industry overall. One interviewee criticised "a lax attitude… within the industry to safety" . Others were more diplomatic: "I'm not sure I'd be rude to the AI people but often all of them working in this area don't understand a lot of standard safety engineering" . On the question of whether AI can explain its actions, one AI researcher was critical of colleagues for encouraging "a culture of people building uninterpretable models and getting paid a lot of money for that" . Viewed optimistically, this antagonism within and between disciplines might be constructive, destabilising assumptions and leading to more robust systems. One software researcher who works on self-driving systems admitted that, after taking his code into the world, I now understand the safety question in more detail… There are a lot of easy things that people say in machine learning that aren't really true…. Speak to any machine learning researcher today and they say, well, you just have to get more and more data and everything is done… There is an implicit assumption in everything that actually getting more data is easy… And then people say, 'Well, we can do it in simulation'. But this is a chicken and egg problem, because then how do you make the simulation good enough for it to be useful? Notwithstanding this interviewee's politeness, the two cultures and their approaches to safety need work to improve their compatibility, and there is a danger that the momentum of the safety-in-numbers approach railroads the other. Some interviewees' reflections on culture were prompted by other companies' early missteps. The NTSB's report on the Uber crash points not just to technological flaws, but also to a woeful safety culture . In this case, engineers' desire to demonstrate their success led to what Diane Vaughan calls a 'normalization of deviance'. The risks were a product as much of economics as of technology. The Uber crash also revealed that the governance of trials was, in some places, threadbare. As technologies are being tested, gaps and disagreements between the safetyin-numbers and a safety-by-design approaches are papered over by safety cases aimed at reassuring local authorities that uncertainties are under control. Safety case are documents put together by an organisation that attempt to persuade a regulatory authority that their system is acceptably safe. 9 The argument for safety cases is that they should inculcate safety by making a company redesign its operations from the bottom up. Most safety case approaches are performance-based rather than looking to open technological black boxes . The presumption is that the organisations developing the technology are best placed to identify issues and that they are willing and able to regulate themselves. Early attempts at safety cases for self-driving are criticised by some safety engineers because, as one interviewee put it, "they're very focused on the goal. They're showing all of the arguments why they believe their goal is met. That is a positive-oriented argument." . An alternative would be to "start by assuming this thing is going to lead to a loss of life… regardless of what you think, let's start by assuming there's something wrong with it. We're going to find everything that's wrong with this" . Such a philosophy would represent a dramatic shift in the burden of proof. In the short term, many safety cases rest on a safety driver, a human on-the-loop who is behind the wheel and, in principle, able to compensate for the technology's shortcomings. The safety driver acts as scaffolding while the software is under construction. The danger is that, without external safety assurance, the scaffolding could be removed by a self-confident technology developer and the safety cases that govern testing could become de facto rules of the road. Although technologies are still experimental and legal frameworks unclear, a consensus is crystallising around safety cases and other performance-based approaches. However, as I will discuss in the final section, a premature lock-in to this mode of regulation would foreclose more prescriptive approaches that may be more publicly credible. The claims of engineers need exposure to public attitudes in order to appreciate the multidimensionality of self-driving safety. --- Safety and the public As engineers from different standpoints attempt to persuade themselves and each other that their approaches to safety are good enough, they are imagining who the public are, what the public think and what the public want. Engineers' private effort to know about safety is extrapolated into a public project of reassurance. The role imagined for the public is, at present, an exceptionally narrow one. Some engineers recognise that people may disagree on levels of acceptable risk, but the presumption is that the public should be kept out of risk assessment: The complexity that you need to go into is not going to be of any benefit to the public… people aren't really interested in that. I think with the public, you can only really prove safety through experience. The widespread assumption is that non-experts should not know or care how a system works or why it does what it does. Interviewees described non-experts as prone to biases, including "algorithm aversion", that would skew their risk perceptions. But the more pragmatic safety engineers saw such perceptions as unignorable. Ultimately, regulation 1 3 would be a public matter. One software developer said "what I as an engineer give society is the knob [to balance safety against efficiency]… and I also clearly explain the trade-off" . The question is whether society will be content with just dialling up or down risk, or whether there will also be a legitimate public interest in the process through which selfdriving cars are developed and assessed. Our workshops with members of the public revealed a set of complexities that suggest a simple narrative of the technology as a safety solution will lack credibility. People are sceptical that a technology as novel as a self-driving car on public roads can be guaranteed safe. During the public dialogue exercise, participants were quick to highlight complexities and identify what engineers would call 'edge cases'. . The discussions reflected an ambivalence typical of public attitudes to technology : excitement about the benefits coupled to a concern about the technology's limits and its governance. The workshop participants were not naïve about the hazards of human-driven cars, and many were optimistic about possible benefits of self-driving, particularly for disabled people, but they were not convinced that weighing these qualities, many of which were highly uncertain, was straightforward. For some, the question of safety was far broader than just road safety. One woman said, "I'm concerned about travelling as a woman in shared [driverless] rides. How can I guarantee that I'll be safe late at night?" Another participant concluded "There will always be vulnerability in technological systems… I wouldn't get in a plane without a pilot." They were used to their computer software crashing, but expected far more from other technologies. They were aware that cars were highly regulated and tested before they left the factory. After one participant expressed concern that self-driving could be regulated more lightly, like computer software, another participant sought to reassure her: Cars are different. No one is going to allow a car on the road till it's [proven] that nothing will happen…. whoever is designing these cars and is moving this technology forward has thought about all these things. They haven't put a car on the road and just hoped for the best. This participant's faith in automotive regulation suggests that self-driving car developers will need to work hard to earn similar levels of public trust. Many were concerned about what Wynne calls 'social risks', to jobs, businesses and local communities from a rapid introduction of new technologies. Some were worried that the technology could displace public transport. Others worried that their freedom to drive would be at risk in the long term. In the short term, recognising that things would go wrong, they saw risks in the context of responsibilities, as these comments from participants indicate: There will be risks, we will learn from accidents, but I don't want my family to be those on the back of which the learning happens. If the system fails overall, then someone needs to be accountable for a backup system. The algorithms will be written by humans, so humans have some responsibility. Most of the participants agreed that it would be unwise to leave safety to the market. They saw a need for oversight. In the groups' final sessions, they discussed their messages for Government. Across the five locations, their support for the technology was broadly conditional on the following factors: These early insights from public workshops give a sense of the public credibility challenge. People see safety as highly contextualised, entangling questions of science and technology with those of values. The social assessment of safety is not just a question of how safe is safe enough. It is also linked to the question of what the technology is for and who is seen as benefitting from its development. People engage with safety issues from multiple perspectives: as potential users of the technology, drivers, pedestrians, public transport users, parents and citizens. The safety-in-numbers story treats safety as an ex-post destination, while the safety-by-design approach sees safety as an ex-ante starting point. In discussing the technology's uncertainties and contingencies with experts and members of the public, it becomes clear that safety will actually be a journey, a collective experiment whose questions and metrics are not set in advance. One interviewee, discussing the various approaches to self-driving safety, said "the answer is it's going to be all of that. There won't be any single thing that says this vehicle is safe", before arguing that "we will learn as we go along" . When considering questions of trust and public credibility, we should ask who the "we" is in this response. Can we trust that the technology developers' questions and metrics are sufficient? Or should the process of learning be opened up? There is a clear case, given that much of the experimentation is happening in plain sight, using publicly-owned roads as a laboratory, for the democratisation of learning. This would mean, at a minimum, sharing safety-critical data for overall safety improvement, enabled by what some authors have called an 'ethical black box' . It could also mean governing the technology in explicitly experimental terms, as clinical trials do for new medicines. This approach, proposed by London and Danks would see staged approval for tests that would be overseen by government before being scaled up. Once we recognise the limits of narrow approaches to the assessment and management of risk, we can ask what trustworthy governance might look like. --- Trustworthy self-driving cars?-Performance, process and purpose A multidimensional view of technological risk should change how we think about trust in autonomous vehicles. First, if we take public views on risk seriously, we should recognise that trust is out of the control of innovators and regulators. They can design for trustworthiness, but trust is a gift of the public, hard-won and easily lost. Second, rather than talking about trust in self-driving cars as technological artifacts, we should consider trust in the systems that govern the technology. Third, we should think about trust beyond just performance. The Cold War adage 'trust but verify' presupposes that we know what to measure. As we have seen in the debate on self-driving safety, the relevant numbers are at the centre of a controversy that is not just about what selfdriving performance, but also about the processes by which they function and the purposes of innovation. Lee and See's framework has trust in automation resting on three pillars-competence, integrity and benevolence . Their focus is on people using automated systems, such as pilots and train operators, but if we see trust as a concern for governance as well as human-machine interaction, we might see trust depending on issues of technological processes and purposes as well as those of performance. Members of the public are likely to pay attention not just to what self-driving cars do, but also to how they work, how they are developed and what the technology is used for. As it stands, the regulatory debate on self-driving safety emphasises performance-based metrics that rest on companies' own safety cases. If governance becomes locked in to this mode of safety assurance it could be socially brittle. Just as engineers would be concerned with a single point of failure in a technological system, so regulators should worry about balancing public trust on a single pillar. There is a strong case for socially-robust governance that is more pluralist, building on cultural theories of risk that look for the multiple ways in which people prioritise and make sense of hazards . The interviews analysed for this paper reveal innovators' emphasis on self-driving performance and neglect of questions of process and purpose. Public discussions suggest that people will want to scrutinise why a self-driving car does what it does, and who self-driving cars are seen as benefitting. Only one interviewee drew an explicit connection between risk assessment and the question of unequal benefits: Risk acceptance will also be different whether you talk about trucks driving on highways from A to B,… autonomous shuttles driving at 30 [kilometres per hour] in a city [or] luxury vehicles driving on motorways up to 130, where only rich people benefit . Given the potential politics of self-driving car innovation, what is the potential for governance to connect technological means and ends? For genetically-modified crops, competing political constitutions of the technology led to competing governance frameworks in Europe and the US. While American regulators focussed on crops as products and assessed their performance, European regulators saw genetic modification as a novel process, surrounded by uncertainties and creating social as well as physical risks . In both cases, regulation was framed by assumptions of what the technology was for: who benefitted and how. For self-driving cars, as standard-setting and other governance processes gear up, regulators should challenge emerging models of de facto governance and seek more deliberate, and more deliberative, alternatives. One interviewee, a researcher at a self-driving company, described a set of dilemmas that faced the company as it scaled up its operations. This researcher, unusual in their reflexivity, saw potential gaps between colleagues' motivations and the public interest: The verification/validation side?… From a safety standpoint… I have no doubt about the commitments and intentions of people who are doing that... But, you know, where and how? What does it mean to be validated and verified? What voices are in the room, on what grounds? What are the metrics that get considered and don't get considered?... Who's making those decisions? Those questions will not melt away; they will intensify as and when the technology scales up. One lesson emerging from our workshops with members of the public is that there will be a public interest in how self-driving cars work and how they are developed and governed. People are unlikely to be satisfied just by public displays of self-driving in action and statistics showing average improvements in safety. Early proposals such as ethical black boxes and publicising leading metrics such as disengagements, while they need work, at least offer ways to open up the governance of processes of self-driving and its testing. In addition, there will be a public interest in the perceived purposes of self-driving technologies and who they are likely benefit. Members of the public will have different expectations from engineers. The questions 'safe enough for whom?' and 'safe enough for what?' will be unavoidable. Seen in this light, the question of how we know whether a self-driving car is social as well as scientific. Some of the uncertainties will be hard if not impossible to resolve, but the project must bring together a wide range of disciplines and draw on public as well as expert insights. --- Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Self-driving cars promise solutions to some of the hazards of human driving but there are important questions about the safety of these new technologies. This paper takes a qualitative social science approach to the question 'how safe is safe enough?' Drawing on 50 interviews with people developing and researching self-driving cars, I describe two dominant narratives of safety. The first, safety-in-numbers, sees safety as a self-evident property of the technology and offers metrics in an attempt to reassure the public. The second approach, safety-by-design, starts with the challenge of safety assurance and sees the technology as intrinsically problematic. The first approach is concerned only with performance-what a self-driving system does. The second is also concerned with why systems do what they do and how they should be tested. Using insights from workshops with members of the public, I introduce a further concern that will define trustworthy self-driving cars: the intended and perceived purposes of a system. Engineers' safety assurances will have their credibility tested in public. 'How safe is safe enough?' prompts further questions: 'safe enough for what?' and 'safe enough for whom?' Keywords Autonomous vehicles • Self-driving cars • Risk assessment • Governance • Public dialogue * Jack Stilgoe

Introduction The Ngaruwat Bumi traditional ceremony or known as Ruwatan Bumi is an ancestral tradition that has been going on for more than a century. At that age, the sacredness of a tradition in agricultural communities is still felt today, especially in the Banceuy Traditional Village, Sanca Village, Sariater District, Subang Regency. Ngaruwat bumi is an expression of gratitude for the results obtained from the earth. Hope for the year ahead, as well as respect for ancestors. Ruat in Sundanese means collecting and caring for. Meanwhile, what is collected and cared for is the community and its agricultural products . Ruwatan bumi is also called hajat bumi, or a series of ceremonies that were held previously, such as the hajat solok, mapag cai, mitembiyan, netepkeun, nganyarkeun, hajat wawar, ngabangsar and kariaan ceremonies. The majority of them are related to agricultural processes, especially rice cultivation. In the Ruwatan Bumi tradition, rice has a special place. Paddy or rice, in local people's beliefs, is not only a food ingredient. Rice is believed to originate from the activities of goddesses so it is sacred and all processes of producing it are considered sacred. Based on pre-research interviews with traditional leaders, the Ngaruwat Bumi tradition, which remains sustainable in the Banceuy Traditional Village, originates from stories from the ancestors of the local community that occurred in the 1800s. Before it changed to Kampung Banceuy, this location used to be called Kampung Negla. Starting from an ancestral story that has been passed down from generation to generation, in the village there was a tornado destroying people's houses, livestock even died and vegetation and agricultural fields were destroyed by residents. After the disaster passed, the seven leaders of Ngabanceuy village, in the Sundanese dictionary, means deliberation to find ways to ward off natural disasters if they come again. Based on deliberations, an agreement finally emerged, namely that the seven figures tried to bring in psychics or smart people in the village. The trusted paranormal is Eyang Suhab who comes from Ciupih Village, Pasanggrahan Village, Kasomalang District . Then a deterrence event was carried out by numbal or slaughtering buffalo. Based on the Javanese calendar or wuku, the new name agreed upon was Kampung Banceuy as a replacement for the name Kampung Negla, because Negla was believed to be the name that caused disaster to the village and its residents. Apart from that, by changing the name of the village, it is hoped that the residents will live better and be blessed, as Banceuy said, which means deliberation. The village leaders hope that their area can be used as a gathering place and a place to exchange ideas at that time, and in the future. This event is then commemorated every year and is known as Ruwatan Bumi, or also known as Ngaruwat Bumi. . Based on the results of interviews with the elders of the Banceuy traditional village, Odang, the Ruwatan Bumi ceremony was held on Wednesday, at the end of the month of Rayagung or Dzulhijah . The purpose of this ceremony is to express gratitude to God Almighty, as an effort to ward off evil , and a form of expression of respect for the ancestors. So the Ngaruwat Bumi ritual, which is an annual tradition of the Banceuy people, is still being preserved to this day. Where in the ritual, the people of the Banceuy traditional village work together to bring the offerings to the center of the village and then store the offerings in the four cardinal points. The existence of Ruwatan Bumi in the traditional village of Banceuy cannot be separated from the history of its predecessors, one of which is from the descendants of Eyang Ato who now continues the annual traditional tradition and is held by the community. Eyang Ato's descendants are categorized as opinion leaders who cannot be separated from the communication actions of traditional elders , which is one of the topics that is related to organizational communication. According to Luthra , leadership communication is defined asinspiring and encouraging an individual or agroup by systematic and meaningful sharing ofinformation by using excellent communicationskills. Leadership communication is a complexprocess which starts from developing strategy forcommunicating, writing precisely and thenspeaking effectively to control difficult situations. Banceuy Village is a cultural organization under the village government, whose influence is categorized as sacred. Not only does it believe in the ancestral mandate of traditional leaders, there are even some who believe in it even today. The impact of leadership communication creates a pleasant atmosphere. Just as a leader strengthens his influence among his followers, the process of communication should be a process that can be maintained by all parties. This is because the communication process is also intended as a means of interacting with each other. That is the importance of communication in the world of leadership, especially in Banceuy village, Sanca Village, Sariater District. The existence of a local leader needs to improve communication skills, so that he can develop communication skills which in turn will increase the effectiveness of his own leadership. As stated by James Humes, "The art of communication is the language of leadership" or in the words of Nitin Nohria "Communication is the real work of leadership" . This research focuses on discussing: Local Leadership Communication in the Ngaruwat Bumi Tradition in the Banceuy Traditional Village, Subang Regency. From the problem limitations above, the specific problem formulation in this research is to see how the communication process is in the Ngaruwat Bumi tradition in the Banceuy Traditional Village? and what is the role and communication of traditional leaders as local leaders in the Ngaruwat Bumi tradition in the Banceuy Traditional Village? --- Method The method used in this research is a qualitative method with a qualitative descriptive approach, through observation, literature study and interviews with traditional leaders of the Banceuy Indigenous village using purposive sampling techniques. Data as supporting material for research is needed so that research results are more accurate in accordance with real social phenomena. The data source according to states that: "First, there is no similarity or deductive relationship between research questions and data collection methods." Data sources include the results of interviews, observations, document analysis, scientific articles, and others. The data sources in this research consist of: primary data, namely data sources consisting of words and actions observed or interviewed, so that they are obtained directly from research informants using in-depth interview guidelines; secondary data, namely additional data sources to complement primary data sources, including written sources divided into books and scientific articles, sources from archives and official documents, as well as observations of the physical condition of the research location. Based on the data sources described above, the type of data used in this research can be identified. This type of data is described based on problem identification and research concepts so that researchers are able to describe the problems being studied. The subjects in this research were informants from community leaders in the Banceuy traditional village and the general public in the Banceuy traditional village. The informants in this research are not subjects who will present their group, so the number of informants is not about whether or not there are many people who can represent a group. Discussion of participants and research locations can include four aspects stated by Miles and Huberman in Creswell , namely: Setting , actors , events , and process . Identification of locations or individuals deliberately chosen by the researcher, the idea behind qualitative research is to choose deliberately and with full planning , which means that the selection of settings, actors, as well as determining the events and processes that are the focus of the research are made on the basis of research interests and careful planning so that researchers used purposive sampling techniques. Purposive sampling according to Soehartono is: Purposive sampling or sampling based on objectives, namely that the informants taken as sample members are left to the consideration of the data collector who according to him is in accordance with the researcher's aims and objectives. Determination of the sample in the purposive sampling technique is very much determined by the aims and objectives of the research. In other words, informants are adjusted to certain criteria that are applied based on the research objectives. The informants were 5 people consisting community leaders from the Banceuy traditional village and the community in general in the Banceuy traditional village, starting from the village head, hamlet head, elders and traditional leaders. --- Result and Discussion Banceuy Traditional Village is located in Sanca Village, Ciater District, Subang Regency. According to information, the name Sanca comes from Chinese which consists of two words, namely the word San which means big and the word Ca which means snake. If interpreted clearly in Indonesian, it means Big Snake. According to the story teller, the founding of the python was the beginning of regional development in the surrounding area. The founders of the python were two very admired figures, namely: Eyang Ngabeui, nicknamed Eyang Abrul . With the title Eyang Joglo, whose real name is Eyang Abdul Puspadiwangsa. During the heyday of the Majapahit Kingdom, it was tasked with deliberately expanding the wings of the Majapahit kingdom, because at that time it needed to expand its power and followers. Since then, he began cultivating the area starting from Sanca. When translated, the word Sanca also comes from the Sangsekerta word, namely hanca, which means starting or first being created/built, then spread to Cipatat passing through the hamlet of Banceuy. Banceuy comes from the words Ban and Ceuy which means a place to stop temporarily to relieve fatigue. The two founding figures of Sanca were people who understood the teachings of the Islamic religion, until now Sanca is a center for the development and cultivation of Islamic religious teachings, including the cultivation of leading figures, so it is said that many people stay overnight/visit the two graves above to pray to become officials. Historically, originally Banceuy Village was Negla Village which was located to the northeast of Banceuy Village now. From Banceuy Village, only a few hundred meters, in Negla Village there are 7 families, namely Eyang Ito, Aki Leutik, Eyang Malim, Aki Alman, Eyang Ono, Aki Uti, and Aki Arsiam. It is called Kampung Negla, because the village is in a highland and open area . Around 1800, a tornado occurred in Negla Village which destroyed people's houses, including the houses of the 7 families, causing livestock and vegetation to be destroyed. After the natural disaster subsided, the seven Negla Village figures held a meeting or deliberation to ward off the natural disaster. In accordance with mutual agreement, the seven figures tried to bring in a paranormal or shaman. The paranormal who was trusted at that time was Eyang Suhab who came from Ciupih village, Pasanggrahan village, subdistrict. Kasomalang now. Then they carried out countermeasures using numbal methods. Based on the Javanese or Wuku calendar. The new name agreed was Banceuy Village as a replacement for Negla Village, because Negla was believed to be the name that caused disaster to the village and its residents. Apart from that, by changing the name of the village it is hoped that the residents will live better and be blessed like the word "BANCEUY". Banceuy means deliberation, village leaders hope that the village can be used as a gathering place and a place to exchange ideas at that time and for the future. So this event is commemorated at the end of every year of the Hijriah year and is known as "Ruwatan Bumi", or the Banceuy people more often call it "Ngaruwat Bumi" Banceuy Village is located at the coordinates 6°42'16"E -107°42'2"S on a plateau so the climate of this region is cooler than other areas around it. The relief shape of the Banceuy Village location is flat. Administratively, Banceuy Village is included in the government of Sanca Village, Ciater District, Subang Regency, West Java. Banceuy Village has an altitude of around 770m above sea level, a minimum temperature of 18˚C and a maximum temperature of 34˚C and an average temperature of 26˚C. Rainfall falls around 2,700mm3/year. Banceuy Village consists of 2 RWs, namely RW 05 and RW 06, as well as 7 RTs, namely RT 13, RT 26, RT 14, RT 15, RT 16, RT 17 and RT 18. The population of Banceuy Village is estimated to be 948 people in 2021, with the number of family heads is 334 families . The location of Banceuy Village in Sanca Village, bordering other villages, is as follows: to the north, it borders the Tegal Malaka rice fields; to the south it borders Solokan Cipadaringan; to the west it borders Solokan Cipatat; and to the east it borders the Cipunagara River. When you enter the Banceuy Village area you will see a large gate with the words 'Welcome to Banceuy Traditional Village'. Then enter the Babakan area of Banceuy Village and you will find the Saung Celempung which is opposite the mosque. After that, not far from the hut there is a bale sawala which is often used for banceuy community activities such as; deliberations, earth care events and so on. In Banceuy Village there are several sacred sites, namely: the Aki Leutik site, the grave of Eyang Ito, and the peak site of Eyang Nata Rangga Kusumah. Access to Kampung Banceuy can be reached, if you are heading from Subang City to Jalan Cagak it takes around 30 minutes. Then from Jalan Cagak enter the alley of Kasomalang District, Sanca Village, and arrive at Banceuy Village. The time required for this route is around 30 minutes using a two-wheeled vehicle. The route passes through residential areas, tea plantations, pineapple plantations and rice fields. The condition of the road traversed is a narrow road, uphill and downhill. The condition of the road to Banceuy Village is that most of the road has holes and gravel. Therefore, it takes quite a long time to reach the village, because the road access is still damaged. The area of Banceuy Village reaches 157 hectares, 47 hectares of this area is forest, 78 hectares are rice fields, 20 hectares are gardens, and 12 hectares are residential land. Forests, rice fields and gardens are natural resources in Banceuy Village . The land in Banceuy Village is used productively by the community. Paddy fields are used for farming, such as growing rice and secondary crops. The plantation land is used for pineapple and palm tree plantations. Apart from that, there is some land that is managed by the community as tourism for Curug Pandan and Leuwi Lawang. Based on education level, almost 70% of the people of Banceuy Village only have an education level up to elementary school. Meanwhile, the results of the interviews show the fact that the population's lack of interest in continuing their education at a higher level is the main factor causing the level of education to remain relatively low. Apart from that, the lack of educational facilities and too far road access to middle and high school level schools causes a lack of public awareness of the importance of education. Judging from ethnic and cultural diversity, the majority of the residents of Banceuy Village are Sundanese, and all of them are Muslim. Banceuy is indeed steeped in traditional rituals and arts which are still maintained to this day. Rituals in Banceuy include: Traditional Ceremony of Ngaruwat Bumi Ruwatan Bumi or Ngaruwat Bumi comes from the word care or ngarawat which means to collect or maintain, in general it has the meaning of gathering the whole community and collecting all the produce of the earth, both raw materials, semi-finished and ready-made. The purpose of Ngaruwat Bumi is to express gratitude to God Almighty for everything that has been obtained from the produce of the earth and also as a source of reinforcements and an expression of respect for karuhun . The reason for implementing Ngaruwat Bumi was because of a natural disaster that hit the Banceuy Village area in the past. After this incident occurred, the ancestors of Banceuy Village together performed the Ngaruwat Bumi ceremony. This was all done to anticipate that natural disasters would not occur again. Hajat solokan carried out in the middle of the paddy age with the purpose of expressing gratitude and wiping blessings so that the water channels run smoothly. This wish is implemented in 3 water channels , namely Eyang Ito sewer, Cipadaringan sewer, and Kolong Tembok sewer. Two of the three drains were plugged with goat's blood while the Cipadaringan drain was plugged with sheep's blood. The result of the slaughter is then served to be enjoyed together and the rest is taken home. --- Mitebeyan tandur This ritual is a ritual when planting rice with the aim that the rice planted will grow well and produce abundant results. Mitembeyan tandur is carried out simultaneously with the procedure of tandur. Naderan It is a religious ritual activity usually carried out by Banceuy residents with the aim of purifying young male and female children. The series of processions before the circumcision include the covering of nutu, mapag beas, nyelamkeun, mandi koneng, gusaran, nyembahkeun, procession and sawer pangten circumcision. Hajat Puput Puser It is a thanksgiving ritual usually carried out by Banceuy residents when a baby's umbilical cord is removed. This activity is usually carried out after the baby is 7 or 8 days old, using turmeric as an antiseptic for babies which is rubbed all over the baby's navel. Ngabangsar Thanksgiving activities are usually carried out by Banceuy residents when pregnant women enter their 4th or 7th month of pregnancy. This activity is carried out to pray for safety for the future baby, so that at birth it will be given smoothness and safety, and there will be no shortage of anything. Hajat Safaran Hajat safaran is a traditional ceremony that is performed every Safar month. This ceremony is performed by parents who have children born in the month of Safar. This ceremony is done with the aim of begging Allah SWT, so that the child is far away from danger and given safety, and this ceremony is also a repelling force for the child. Something that is typical for the purpose of the journey is that there must be a stew of tubers, starting from cassava, taro, sago, ganyol, peanuts , corn, sweet potatoes, etc. In addition to rituals, there are also typical arts in Kampung Banceuy, among others: celemung, gembyung, dogdog, rengkong, durug, tutunggulan. Other folk attractions include miruha, nyumpit, icikibung, kokoleceran, and lais. In traditional children's games, there are gusur upih, engklek, jajajungk, gatrik, and others. --- Communication Process in the Ngaruwat Bumi Tradition in the Banceuy Traditional Village The description of the life of the people of Banceuy Village seen from the religious and belief system basically adheres to the Tatali Karuhun value system as a legacy of previous ancestral traditions. Basically the local community adheres to Islam. This can be proven by the routine and practical activities of worship which are carried out at all times in accordance with the teachings and beliefs of their religion, including praying, reciting the Koran and other religious activities. However, looking at the facts and realities of local culture, it turns out that their belief in their ancestors is very closely tied to their lives by carrying out traditional ceremonial activities, one of which is the earth ruwatan which is still carried out every year. This reality is a condition of tradition that has been passed down from generation to generation as a system of inheritance of the customs and traditions of their ancestors. Even though they follow the Islamic religion, they have their own traditions which are based on beliefs and beliefs in the transcendent powers of the spirit or spiritual world. In this way, they have very strong polarization of their thinking power and belief in the power of ancestral myths and ritual forms which are still carried out today. They really believe in the power of their ancestral spirits, thinking that the power of their ancestral spirits can save them from various kinds of problems and prevent threats in the form of calamities or disasters. The traditions that they still carry out include belief in spirits and spirits who inhabit certain places, such as springs, rivers, forests and sacred trees. The community believes that the sacred tree is hundreds of years old and its inhabitants are genies. This place is considered keueung and every year local people always give offerings to the trees as they grow bigger. One of the things that characterizes the power of influence of belief in the ancestral spirits in which it is believed, is the myth of Dewi Sri, known as the spiritual figure Nyi Pohaci. As they believe that Nyi Pohaci is a holy spirit who incarnates into rice plants as the staple food of the local community. Therefore, in their view, rice is considered a holy and sacred plant. Thus, the existence of rice by the local community is treated with great nobility and respect in terms of its value and position. So when the harvest is about to take place, most of the Banceuy people still carry out rituals by making offerings and burning incense in the rice fields that will be harvested. Likewise, in these ways, the local community implements respect for their belief in the figure of Dewi Sri through the ruwatan bumi ritual. This ritual is basically a way of honoring the existence of Dewi Sri who has given blessings to the lives of local people by dancing and singing Dewi Sri. Ngarak and Nyawer Dewi Sri are two forms of ritual that cannot be separated and are implemented as one unit in the earth ruwatan ritual. The implementation of traditional and religious rituals which are based on the presence of supernatural powers which are still carried out by the Banceuy Village community group is implemented in death rituals, thanksgiving or slametan rituals, rituals of rejecting evil, including the earth ruwatan ritual. The Ngaruwat Bumi ceremony, which is believed by the majority of the people of Banceuy Village, is believed to be that the ancestors are able to protect and save them from threats in the form of major disasters as a source of reinforcements. As one of the local cultural traditions which is the result of humans being able to create beliefs that are so close that beliefs in local cultural traditions and beliefs in religion have an interrelated relationship with various aspects, local values and norms. The belief of the people of Banceuy traditional village in the spirits of their ancestors is so strong and embedded in their outlook on life, that this ritual is carried out every year. They are so confident that if it is not implemented there will be a major disaster, thus making the community feel uneasy about this. To carry out the Ngaruwat Bumi ceremony, the community through the kokolot can determine the time of day and other provisions to prepare for the ritual. The time for the ceremony is usually held on Wednesday, ending the Hijri month leading to the month of Muharram. The Ngaruwat Bumi ceremony basically has a standard structure that is agreed upon by custom every year. The Ngaruwat Bumi ceremony is carried out for two days, namely through rituals on the first day by carrying out dadahut, ngadiukeun, ijab kobul , ngalawar, sholawatan and the Buhun Gembyung art performance. On the second day, it started with Dewi Sri's numbal, dancing and singing, Ijab Rasul and a wayang golek performance. The Numbal ritual is seen as playing a very important role in the implementation of the Ngaruwat Bumi ceremony, and as the culmination of the ritual. The purpose of the numbal ritual is "Ngahurip Bumi Munar weak" meaning that everything done and produced from the land of Banceuy is useful and a blessing. "Ngahurip Bumi Munar weak" is a slogan of faith values based on local community beliefs which are the result of ancestral mandates that have been believed for generations. According to tradition and local community understanding, those who have the right to lead the ritual in this numbal activity must be a descendant of Aki Ito's ancestors. The numbal ritual is a form of representation of respect for ancestral spirits as well as asking for permission from the ancestors so that the event can run smoothly and be a blessing according to local community beliefs. From a cultural communication perspective, these rituals can essentially be seen as symbols of transcendent communication in the form of dialogue between the real world and the spirit world that connects humans, nature and their creators. As states that the Ngaruwat Bumi ceremony through the numbal ritual represents the reality of social institutions which are full of symbols, acting as a medium for communication between fellow humans and also as a link between the real world and the unseen world. Likewise, shows that the dynamics of the social and cultural life of the traditional community of Subang's Banceuy Village are studied historically, that the development of the life of the people of Banceuy Village still maintains the traditions or customs carried out by their ancestors. Likewise, Supriatna's research, Study of Cultural Values Regarding Myths and Environmental Conservation in the Banceuy Community, Subang Regency, explains that the people of Banceuy Village have a distinctive characteristic of life, namely having ancestral traditional figures who still believe in noble values and ceremonial traditions. The ngaruwat earth tradition is one of the traditional traditions or ceremonies carried out by the people of Banceuy Village, Sanca Village, Ciater District, Subang Regency. The communication process in the ngaruwat bumi tradition can include several aspects, including communication between traditional leaders, ceremony participants and the community present. The following are several stages in the communication process in the ngaruwat bumi tradition in Banceuy Village, Subang: Preparation: before the ceremony begins, initial communication occurs between traditional leaders, elders, and community leaders who are responsible for preparing this event. They may discuss necessary procedures, regulations, and logistics; Announcement: announcements of the implementation of the ngaruwat bumi tradition are usually made by traditional leaders or community figures through various communication channels, such as loudspeakers, leaflets, or verbally, to inform the wider community about the event; Instructions: on the day of the ceremony, traditional leaders and elders can give instructions to ceremony participants regarding what to do during the event. They also provide guidance on the procedures and etiquette that must be followed; Prayers and Rituals: during the ceremony, communication occurs through prayers and rituals carried out by traditional leaders or religious figures. They can communicate with spirits or spiritual entities that are believed to be present in this ceremony; Symbolic Communication: during the ceremony, various traditional symbols and signs are used to communicate with the ceremony participants and the community. This can include dance, music, traditional clothing, and other symbolic items that have meaning in the context of earthworks; Socialization and Learning: Apart from being a ceremonial event, the ngaruwat bumi tradition also functions as a socialization and learning process for the younger generation about the values, norms and cultural heritage of the people of Banceuy Village. It also involves communication between older and younger generations; Social Interaction: during and after the ceremony, people present can interact socially by chatting, exchanging stories and sharing experiences. This is another way in which communication plays a role in strengthening social relationships within a community. The communication process in the ngaruwat bumi tradition in Banceuy Village, Subang, includes various aspects, from verbal communication to symbolic communication and social interaction. This all aims to maintain and strengthen cultural identity and values passed down from generation to generation in that society. --- The Role and Communication of Traditional Leaders as Local Leaders in the Ngaruwat Bumi Tradition in the Banceuy Traditional Village Communication between traditional leaders as local leaders in the ngaruwat bumi tradition in the Banceuy Traditional Village is an important part of the culture and social system in the village. The ngaruwat bumi tradition is a ceremony or ritual carried out to respect and maintain balance with nature and the surrounding environment. In this tradition, traditional leaders have a major role in communicating and leading local communities. Aspects of communication in the Ngaruwat Bumi tradition include communication with nature through traditional leaders acting as intermediaries between society and nature. They communicate with nature through prayers, spells, or special ceremonies. This tradition is a way to ask permission and blessings from the surrounding nature before carrying out activities such as planting, harvesting, or other ceremonies. Communication with the community, in this case traditional leaders communicate with the local community to explain the importance of the ngaruwat bumi tradition, the purpose of the ceremony, and the role of the community in maintaining the balance of nature. They also coordinate community participation in this ceremony. These traditional figures and elders act as spiritual leaders in the ngaruwat bumi tradition. They can provide spiritual teachings to the people, remind them of traditional values, and guide them in carrying out ceremonies properly. As local leaders, traditional leaders also communicate with the community to make decisions related to environmental maintenance and other aspects of life in the traditional village. This involves discussing with the community, listening to their opinions, and reaching mutual agreement on the steps that should be taken. Apart from that, traditional leaders also play a role in teaching the younger generation about the ngaruwat bumi tradition, carrying out the ceremony correctly, and understanding the relationship between humans and nature. This is a way to maintain the continuity of this tradition from generation to generation. In some situations, traditional leaders can act as mediators in resolving conflicts between community members or with external parties. They use wisdom and traditional understanding to achieve peace and agreement, as well as communicate with external parties. Traditional leaders interact with external parties, such as local governments or environmental conservation organizations. They can communicate to explain the values and importance of the ngaruwat bumi tradition, as well as ask for support or cooperation in efforts to preserve the environment. Communication between traditional leaders in the tradition of nurturing the earth in the Banceuy Traditional Village is the core of maintaining culture, ecosystem balance and the social life of the community. They act as a link between humans, nature and the traditional values that underlie life in the traditional village. The Ngaruwat Bumi tradition is a traditional ceremony carried out by the people of Banceuy Village, Subang Regency. In this tradition, traditional figures have an important role and play a role in various aspects, including communication. The following are the roles and communication of traditional leaders in the Ngaruwat Bumi tradition: 1. Leading the Ceremony: Traditional figures are the main leaders in carrying out the Ngaruwat Bumi ceremony. They are responsible for guiding the entire procession and ritual, ensuring that all steps in the ceremony are followed correctly according to traditional traditions. 2. Preserving Culture and Traditions: Traditional leaders are responsible for preserving the culture and traditions of their people. They act as traditional stakeholders who ensure that the values, norms and procedures of Ngaruwat Bumi are maintained and passed on from generation to generation. 3. Provide Education and Instruction: Traditional leaders often serve as teachers or mentors for the younger generation in society. They teach about the meaning and purpose of Ngaruwat Bumi, as well as how to run it properly. 4. Communicating with the Outside World: Traditional leaders also play a role in establishing communication with outside parties, such as the government or other related parties. They can act as community spokespeople and explain the meaning and importance of Ngaruwat Bumi to outsiders who may not fully understand the tradition. 5. Reducing Conflict: If there is a conflict or difference of opinion within the community regarding the implementation of Ngaruwat Bumi, traditional leaders often act as mediators. They help resolve differences and ensure that the ceremony proceeds peacefully and in accordance with tradition. 6. Conveying Spiritual Messages: Traditional leaders also have a role as intermediaries between society and the spiritual world. They can convey spiritual messages and prayers to the community in the context of Ngaruwat Bumi. Traditional figures become important. They need to communicate clearly and effectively to all residents as supporters of the culture, explain the meaning and purpose of Ngaruwat Bumi, and ensure that this tradition is carried out with full understanding and obedience, and continues to be maintained from generation to generation in the Banceuy Traditional Village, not just maintaining the rules. disaster or disaster, but also more than that, namely as a form of maintaining gratitude and preserving the environment. --- Conclusion The research results show that the Ngaruwat Bumi Ritual in the Banceuy Traditional Village is a local tradition which aims to create harmony between humans and the environment. This ritual is performed every year and is considered an important event in society, and serves as a symbol of symbolic communication between humans and nature, connecting humans and animals. This ritual involves various aspects, including verbal, symbolic and social interactions. Based on this, it can be concluded that: 1) The communication process in the Ngaruwat Bumi tradition in Kampung Banceuy Subang as one of the local cultural traditions, is a belief in local cultural traditions with beliefs in religion having an interrelated relationship with various local aspects, values and norms. From a cultural communication perspective, these rituals can essentially be seen as symbols of transcendent communication in the form of dialogue between the real world and the spirit world that connects humans, nature and their creators. The communication process in the ngaruwat bumi tradition in Banceuy Village includes various aspects, from verbal communication to symbolic communication and social interaction, with the aim of maintaining and strengthening cultural identity and values passed down from generation to generation in the community. 2) The role and communication of traditional leaders as local leaders plays an important role in community identity and values. They communicate with the community to discuss the importance of rituals, the role of the community in overcoming environmental problems, and communicate with the community to make decisions regarding environmental conservation. They also act as mediators in conflicts between community members and external authorities, using their knowledge and skills to resolve conflicts and achieve cooperation. The role of traditional leaders in the Ngaruwat Bumi ritual is to preserve culture, ecosystem sustainability and social life in society. Through this research, it is suggested that methodologically it is necessary to carry out further research on cultural communication and local leadership through various approaches, both qualitative through ethnographic studies, or quantitative to measure how big the role and function of traditional leaders are as leaders and opinion leaders in maintaining traditional traditions and local culture. which can be axiologically beneficial to sustainable development with a focus on environmental and natural sustainability.

Village, Subang Regency. The description of the life of the Banceuy Village community as seen from local leadership based on social position and the social role of the actors in the Banceuy Traditional Village community is important in seeing the course of Ngaruwat Bumi rituals and traditions. This local leadership role is related to religion and belief which is centered on the Tatali Karuhun value system as a legacy of previous ancestral traditions. This ritual is a way of honoring the existence of Dewi Sri (Goddess of Rice) who has given blessings to the lives of the people of Banceuy Traditional Village by dancing and singing Dewi Sri. The problem of this research is the communication process in the Ngaruwat Bumi tradition in the Banceuy Traditional Village, the role and communication of traditional leaders as local leaders in the Ngaruwat Bumi tradition in the Banceuy Traditional Village. The method used in this research is a qualitative method with a qualitative descriptive approach, through observation, literature study and interviews with traditional leaders of the Banceuy Indigenous village using purposive sampling techniques. The research results show that communication and the role of local leadership through the Ngaruwat Bumi ritual in Banceuy Subang Village is a local tradition that emphasizes the importance of ritual in fostering dialogue between humans and nature. This involves various aspects, including verbal, symbolic and social interactions, which aim to maintain and strengthen cultural identity and community values. These rituals also function as mediators of conflict between local residents and external authorities, increasing cultural awareness, environmental conservation, and social life.

The Direct Physical Harm of the Substance to the Individual User The physical harm caused by a drug needs to be considered in terms of its acute toxicity, as well as its likelihood to produce long-term health problems. 1.1. Acute Toxicity. Acute toxicity is the adverse effect produced by a drug following either a single exposure or multiple exposures under 24 hours apart. It is assessed by measuring the ratio of lethal dose to therapeutic dose. Research conducted on 20 human and nonhuman lethal doses of abused substances, which are distributed widely in Europe and North America, identified intravenous heroin to have the greatest direct physiological toxicity. Hallucinogens in general appeared to have the least [1]. The clinical sequelae of exposure to toxic levels of a drug can be present in a specific set of symptoms referred to as toxidrome. Cocaine, for example, at low doses produces euphoria, reduced fatigue, and a perception of increased mental acuity. Higher doses may result in several undesirable side effects including irritability, paranoia, panic, repetitive stereotyped behaviour, diaphoresis, mydriasis, tachyarrhythmias, stroke, and seizure [2,3]. A thorough exploration of common and important toxidromes is beyond the scope of this paper but detailed information for UK healthcare professionals can be found on the clinical toxicology database [4]. --- Long-Term Health Problems. As well as acute physical harm, many drugs when used repeatedly over time have chronic physical consequences. The long-term health problem can either be directly related to the effect of the drug or due to the method of drug administration. Several drugs including marijuana are ingested by smoking and put abusers at increased risk of chronic cough, bronchitis, and lung and upper airway cancers. The longterm adverse effects of cigarette smoking are such that they reduce life expectancy by an average of ten years [6]. Drugs taken intravenously can lead to complications related to this route of delivery. These include thrombosed veins, bacterial endocarditis, abscesses, pneumonia, and liver disease [7]. In addition to the effects of the drug itself, street heroin may have additives that do not fully dissolve and result in further damage to blood vessels. Intravenous drug use and sharing of hypodermic needles open a gateway for potentially lethal blood-borne infections such as HIV and hepatitis. Statistics have shown that 21% of intravenous drug users are hepatitis B positive while 50% are hepatitis C positive [7]. In 2005, 28% of intravenous drug users reported directly sharing needles and syringes, which accounted for 5.6% of HIV diagnoses reported in the UK [8]. Heroin use carries a particularly significant mortality risk. Users are at twelve times greater risk of mortality than the general population and intravenous users, further 22 times greater risk, relative to noninjecting peers [9]. Chronic intranasal usage is also associated with specific problems such as degradation of the nasal septum [10]. A common but untrue belief is that the smoking of cocaine chemically breaks down tooth enamel and causes tooth decay. However, cocaine does often cause involuntary tooth grinding, known as bruxism, which can deteriorate tooth enamel and lead to gingivitis [11]. Long-term health problems also result from the psychoneurobiological impact of chronic use. Chronic marijuana abuse has been shown to result in depression, anxiety, and, in some individuals with a predisposing vulnerability, schizophreniform disorder [13]. Hallucinogens can result in flashbacks and hallucinogen persisting perception disorder. Methylenedioxymethamphetamine and cocaine can both result in impulsiveness, irritability, sleep disturbance, anxiety, and addiction [14]. Many drugs of abuse also increase susceptibility and risk of suffering other conditions. Cocaine, for example, increases the risk of developing rare diseases such as systemic lupus erythematosus , vasculitis, Goodpasture's disease, Stevens-Johnson syndrome, and an array of kidney diseases [15][16][17]. It also doubles both the risks of hemorrhagic and ischemic strokes as well as increases the risk of myocardial infarction [18,19]. --- The Tendency of the Drug to Induce Dependence 2.1. Drug Tolerance. To understand the nature of drug dependence it is first important to understand the concept of drug tolerance. Tolerance is commonly encountered with substance abuse and is described as a decrease in susceptibility to the effects of a given amount of drug as a result of previous exposure [4]. This means that increasingly larger doses are required to induce the desired effect. Whilst tolerance itself is not a particular problem to the health of the individual, it can have serious repercussions [21]. Firstly, there is usually little harm caused by most psychoactive drugs when taken in smaller quantities with sufficient time in between; however, once tolerance develops, larger and more frequent doses are administered that may be toxic and lead to direct harm from the substance itself. Secondly, tolerance to the different effects of a drug does not develop uniformly. This can be illustrated using alcohol; for example, tolerance often arises to the recreational effects of alcohol, but little tolerance develops to the lethal effects [22]. This results in a lowering of the therapeutic index. Finally, the mechanism that underlies tolerance contributes to the person's compulsion to take a drug resulting in socially unacceptable behaviour such as theft and fraud. This forms the basis of addiction [23]. --- Drug Dependence and Theories of Addiction. Dependence can be a physical or emotional adaptive state which results from the body's homeostatic response to repeated drug administration. Upon cessation of the drug, the homeostasis is lost and the dependence is unmasked, for example, cold turkey with heroin dependence [4]. The cessation of the drug and resulting chemical imbalance triggers negative emotions and behavioural disorder to be indicative of the disturbance. This is referred to as "reward deficiency syndrome. " Drug addiction may initially cause and then further proceed to exacerbate "reward deficiency syndrome" [24]. Another theory is the "drug for reward theory, " which states that addiction is the malfunctioning collision of both motivational systems , stimulating pursuit of a substance that most probably no longer provides pleasure and in fact may be pathogenic [25]. --- Neurological Pathways in Addiction. Reinforcers such as food and sex increase extracellular dopamine in the nucleus accumbens via the mesocorticolimbic pathway. This results in the sensation of reward. All known addictive drugs activate this pathway but at levels up to ten times greater than food [5]. As a result of the elevated levels of dopamine, compensatory changes take place that result in tolerance and dependence. This is due to long-term neuroadaptations in the dopaminergic system. There is an increase in transcription factors as well as differential expression of proteins involved in synaptic plasticity [26]. There are also alterations to neurotransmission, and at a cellular level the morphology is altered by increased dendritic branching and spine density in the NAcc and prefrontal cortex [26]. Another difference seen consistently in drug addicts is the decrease in dopamine D2 receptor binding in the NAcc . This is because of the reducing number of dopamine receptors present if dopamine levels are repeatedly elevated. It is hypothesized that this dulling of the responsiveness of the brain's reward pathways contributes to the inability to feel pleasure, known as anhedonia, often observed in drug dependence [26,27]. Addiction comes about through an associative learning process characterized by compulsive drug taking, craving, and relapse [11,21]. The pathway thought to be involved in addiction is the mesocorticolimbic pathway which consists of dopamine neurons projecting from ventral tegmental area to the NAcc and prefrontal cortex . As we have mentioned, this pathway plays a critical role in reward and reinforcement. Behaviours that activate this pathway tend to be repeated even if the experience was not considered pleasurable [28]. Most rewarding effects can be attenuated by blockade of dopamine in this region [29]. and reinforcement come about through disinhibition of the dopamine neurons in the VTA. These neurons are usually fire tonically but are inhibited by GABA interneurons. When heroin activates the 𝜇 opioid receptor on the GABA neurons, it inhibits them from firing thus removing the inhibition of the dopamine neurons. The result is an increase of dopamine release in the NAcc which is responsible for rewarding and reinforcing behaviour at that time [21]. --- Physical Dependence to Opioid receptors are found not only in the mesocorticolimbic system, but also in other systems such as the spinal cord and locus coeruleus . The LC contains noradrenergic nuclei that are implicated in attention and arousal as well as the "fight or flight" autonomic response. Tolerance and dependence arise through chronic activation of opiate receptors, which leads to homeostatic compensatory changes. Acutely, heroin inhibits firing of LC neurons. With chronic use, the LC neurons return to their normal firing rates. This results in increasing doses of heroin being required to achieve the desired effect [21,30]. Relapse is a very common problem during rehabilitation from opiate use [31,32]. Associative cues develop by longterm potentiation during the drug use. Reexposure to the drug triggers these cues, making abstinence increasingly difficult and relapse likely [21]. Upon cessation of heroin, the user may suffer withdrawal symptoms which are essentially opposite to the acute effects and are considered an overshoot rebound to the initial drug-induced state [12]. Whilst not life threatening, the symptoms are extremely unpleasant, often being flu-like in nature. The severity of withdrawal depends primarily on the intensity of the initial drug effects as well as the dose administered and frequency of use. The general health and personality of the patient are also thought to play a part [21]. --- Amphetamine and Cocaine. Amphetamine is an indirect agonist of the catecholaminergic system. It causes release of catecholamines from the presynaptic endings as well as blocking the reuptake. It also inhibits monoamine oxidase activity and hence metabolism of catecholamines [26]. These stimulants act at the dopamine transporter which is the main mechanism related to the reinforcing effects [21]. As with most drugs, tolerance is a problem as well as the reverse phenomenon known as sensitisation. This can occur spontaneously to some users on repeated intoxication with doses that previously only caused euphoria. If this happens, users may suffer acute behavioural changes that are virtually indistinguishable from paranoid schizophrenia [11]. After chronic use of cocaine, user's dopamine receptors become downregulated as they adapt to the constantly elevated levels of dopamine. Upon cessation, users suffer an increasingly Comparison subject --- Cocaine abuser Figure 4: Lower relative glucose metabolism in the prefrontal cortex of a cocaine user than in a normal comparison subject. Image from Carlson [5]. bothersome withdrawal syndrome [33]. Positron emission tomography studies in chronic cocaine users have shown that there is reduced glucose metabolism in some brain areas when compared with nonusers . It is hypothesised that the hypoactivity in these areas may underly some of the behaviors associated with cocaine addiction such as loss of motivation, impulsive drug taking, and the inability to abstain when exposed to drugs because of loss of inhibition [5,34]. Another study which also used PET imaging showed evidence of drug cue-induced DA release in the amygdala and hippocampus. The preferential induction of DA release among cocaine users further suggested that these aspects of the limbic reward network might contribute to drug-seeking behavior [35]. A recent study investigating changes in drug use severity and physical health-related quality of life among untreated stimulant users concluded that the severity of cocaine use is linked directly to physical health-related quality of life; strong evidence of the drugs long-term health implications [36]. --- The Effect of Drug Use on Families, Communities, and Society So far we have discussed the harm of addiction for the individual. However the implications of drug use extend far beyond the user, often damaging their relationships with their family, community, health workers, volunteers, and wider society [10]. --- Family and Children. One of the biggest impacts abuse has is upon the children or dependants of the abuser. The Hidden Harm report by the Advisory Council on the Misuse of Drugs estimated that there were between 250,000 and 350,000 children of problem drug misusers in the UK [37]. As with any drug, harmful effects on pregnancy cause much concern and cocaine is no exception with 90% of female users being of childbearing age [38]. The National Survey on Drug Use and Health data collected during 2002 and 2003 indicates over 4% of pregnant women reporting using illicit drugs in the past month [39]. Prenatal cocaine use has immediate, short-term and long-term health implications. Immediate complications include increase of the risk of miscarriage and preterm labour . As a result, cocaine-exposed babies are more likely than unexposed babies to be born prematurely and with low birth weight. Premature and low-birth weight babies are at increased risk of health problems during the neonatal period, lasting disabilities such as learning difficulties, cerebral palsy, or even death [40,41]. A recent study on school-aged children has demonstrated that prenatal cocaine exposure is reliably associated with impairments in attention and behavioural problems [42]. Children of individuals who use drugs are often abused or neglected as a direct result [43]. Studies have shown that the users need to find and fund their drug habit often taking precedence over the care of their --- Communities and Society. The cost to the community and society of drug abuse is colossal. Drug abuse has a significant impact on healthcare services, public services, and criminal justice system. Drugs that lead to intense intoxication such as alcohol, amphetamines, cocaine, and heroin are associated with huge costs in terms of damage to the user or the general public and property and loss of economic output. A large part of the health care budget is spent on treating drug users [46]. Alcohol alone is estimated to be involved in over half of all A&E department admissions [1]. Data published by the home office reports the estimated economic and social cost of Class A drugs alone to be around £15.4 billion per year [47]. Decreased social functioning related to drug abuse renders many dependant on state support with the home office estimating 80% of problem drug users claiming benefits at a cost of over £40 million a year [44]. Severe dependence and altered behaviour secondary to abuse are responsible for the most significant implications to communities and to society. Users requiring increasing amounts of drug to achieve the same high and to feed their on-going addiction often turn to crime to support their habit. Drug abusers are estimated to commit 36 million drug-motivated crimes each year which financially accounts for 90% of the total cost to society. --- Conclusion Drug abuse has many damaging consequences not only for the individual, but also for the society as a whole. Underneath the pleura of social problems is a phenomenally complex mechanism involving tolerance, dependence, and addiction. Multiple brain areas are implicated in many functions such as reward, motivation, learning, inhibitory control, and executive function. Addictive drugs hijack the reward pathway that is intended for natural reinforcers and in doing so cause harm both acutely and chronically to individuals and society. --- Appendix For more details see Table 2.

From a public health perspective, substance abuse has long been a source of major concern, both for the individual's health and for wider society as a whole. The UK has the highest rates of recorded illegal drug misuse in the western world. In particular, it has comparatively high rates of heroin and crack cocaine use. Substances that are considered harmful are strictly regulated according to a classification system that takes into account the harms and risks of taking each drug (see the tables) (Nutt et al. (2007)). The adverse effects of drug abuse can be thought of in three parts that together determine the overall harm in taking it: (1) the direct physical harm of the substance to the individual user, (2) the tendency of the drug to induce dependence, and (3) the effect of abuse of the drug on families, communities, and society (Gable (2004(Gable ( , 1993))). In this report, we discuss published evidence relating to the harm of substance misuse and consider the neuropsychopharmacological mechanisms behind addiction in an attempt to gain an improved picture of the potential devastation that abuse of these substances may evoke.

BACKGROUND Historically, the accessibility research community has primarily focused on designing assistive technologies and systems related to single categories of impairments. In other words, accessibility researchers often focus on evaluating technologies and addressing the needs of technology users who experience a single type of sensory [25,27], cognitive [16,18], or physical impairment [4,26] at a time. While this approach is justifable and contributes to the advancement of the accessibility feld, many researchers argue that designing accessible systems and assistive technologies for one impairment at a time contributes to an oversimplifcation of disability and does not refect real-world experiences of a signifcant number of users of technology who live with multiple impairments [5,14,23]. A 2016 survey reported that almost 75% of people with disabilities in England live with more than one type of long-term impairment [29]. According to the report, the combination of sensory and mobility impairments is the most common combination of impairments among English survey respondents. However, a combination of cognitive and physical impairments is also prevalent among older adults [17,31]. Furthermore, fndings also show that more than 20% of respondents live with at least three types of impairments [9]. Not surprisingly, 51% of those living with three or more impairments do not think current assistive technologies address their needs or support their daily activities in an adequate way. Similar American reports [10,20] showed that the US Department of Education has to provide special accommodations for at least 122,559 students with profound or multiple impairments each year. In order to be efective, educational programs need to incorporate a variety of components, including the use of personalized assistive technologies that meet the considerable needs of these individuals. These results indicate that people with disabilities do not always belong to fxed, clearly defned categories and assistive technologies should be able to accommodate these individualities. Many scholars question the oversimplifcation of disability among accessibility researchers. Hofmann et al argue that the accessibility feld frames disability as discrete and isolated blocks of impairments for practical reasons. According to the authors, these categorizations are often decontextualized [23]. Similar arguments related to the need for a multifaceted approach in the accessibility community are also made by Mankof et al [14], Bennett et al [6,7], Shinohara et al [19], and Frauenberger [5]. Yet, little research has explored the challenges and design considerations involved in addressing the multidimensional and often complex experiences of users of technology living with complex or multiple impairments. If we look at the work presented at ASSETS over the years [1,3,8], the majority of technical contributions still rely on relatively rigid categories of impairments or user groups and empirical research rarely includes users with more than one impairment. According to related work examining the personal experiences of people with disabilities, many individuals living with multiple impairments do not feel like they belong to one category or the other. Often, these individuals are rejected by both communities for being "diferent" [22]. This issue is also reported among people with deafblindness who do not think assistive technologies designed specifcally for people with "visual" or "hearing" impairments efectively address their needs [28]. Similar issues are faced by people in the neurodivergent community who may experience a combination of diferent sensorimotor and cognitive impairments [11]. We argue that the design of assistive technologies and accessible systems play an important role towards the participation of people with multidimensional accessibility needs in society. In accordance with Wobbrock et al [12], accessible systems should be able to accommodate the needs of diferent users, regardless of their abilities, extent, or complexity of their impairments. However, examples of successfully designed systems that accommodate the complex needs of users living with multiple impairments are limited in the accessibility community. Diferent interaction modalities, such as gaze [4], audio-visual [25,27], haptics [13,15,24], or multisensory approaches [2,30], are not always easily adaptable to the multilayered needs of users who may experience challenges in compensating one ability with another. A multidisciplinary approach to this problem could help the accessibility research community address the challenges involved in designing accessible systems that take into consideration a diversity of users living with complex accessibility needs. In this workshop, we invite researchers and practitioners from diferent disciplines to share their perspectives related to challenges and opportunities in designing accessible systems that consider the multidimensional needs of users with multiple impairments. This workshop intends to challenge current paradigms in the accessibility feld and draft a joint research agenda that actively involves users with multiple impairments in our practice. --- WORKSHOP PLANS The motivation for this workshop builds on our previous experience working with communities of people living with multisensory impairments and other complex disabilities. As accessibility researchers, we realized that most currently available technologies do not address their user needs and that well-established research methods do not necessarily facilitate conducting work with and for these communities. Therefore, we invite academics and practitioners from diferent disciplines to discuss issues related to designing accessible systems and assistive technologies with and for users with multiple impairments. Our goal in this workshop is to explore questions in three subdomains using a multidisciplinary approach: sharing perspectives and experiences from individuals with lived experience of disability, scholars in the humanities, disability studies, engineering, human-computer interaction and so on. To add focus and rigour to our discussions, 500-word abstracts will help us split workshop participants into diferent groups working on sub-domains depending on their interests and expertise. Furthermore, we plan to invite our extensive collaboration networks of organizations, international agencies, special needs colleges, and industry partners to be part of the discussions. The workshop intends to challenge current paradigms in the accessibility feld, devise new approaches for designing multifaceted accessible systems, and draft a joint research agenda that promotes the involvement of users with multiple impairments in our feld. Planned sub-domains for discussion with workshop participants: Theory: 1. How can the accessibility feld leverage theories and concepts from other disciplines to support more appropriate and inclusive technologies for users with complex needs or multiple impairments? 2. Are commonly used design frameworks appropriate in the context of multidimensional user needs? What new theoretical frameworks are needed in order to support the design of accessible systems for users with multiple impairments? Technology: 3. How can we mitigate potential harms and make current assistive technologies more accessible for users with multiple impairments? 4. How can future assistive technologies take into consideration a diverse combination of sensory, cognitive, and physical characteristics? For instance, should a multisensory approach be used? How can we adapt interfaces based on multi-layered user needs? Users: 5. Are commonly used participatory design and evaluation methods appropriate and inclusive? What methods and tools should be used to efectively involve users with multiple impairments and complex disabilities in our research and practice? 6. How should we actively involve and foster collaboration with diferent communities of people with multiple impairments in accessibility research? The workshop will be held virtually and will involve focused group discussions during the event. Table 1 shows a draft structure for the day, however, this schedule can be adapted based on the needs and interests of workshop participants. We anticipate using the Central European Summer Time to accommodate participants who will be attending the ASSETS 2022 conference in person, but this can be changed to better accommodate the fnal list of participants. Following the event, we plan to create a community of researchers and practitioners interested in working with communities of people living with multiple impairments or complex disabilities. We will also invite participants to submit an expanded version of accepted 500-word abstracts to a special journal issue. For that, we will discuss the best strategy for moving forward with workshop participants during the day. --- DIVERSITY AND INCLUSION CONSIDERATIONS Our goal is to bring participants of diferent disciplines and lived experiences of disability to discuss challenges and opportunities related to designing accessible systems for, with, and by users with multiple impairments or complex accessibility needs. Participants will be asked to follow the guidelines for accessible submissions as part of the selection process. Furthermore, all workshop participants will be expected to follow the ACM Code of Ethics and Professional Conduct during the course of the event. In order to increase accessibility and equitable participation during unprecedented times, the workshop will be fully virtual and will include all the necessary reasonable accommodations requested by participants. Furthermore, all workshop materials and accepted abstracts will be shared with participants via the workshop website and email ahead of the workshop day. The workshop timings will be decided based on the most appropriate times for participants in diferent geographic regions, including those who will be attending the in-person ASSETS conference in Athens, Greece. To encourage participation from researchers, practitioners, and individuals with lived experience of disability from underrepresented communities, we will actively share in the possibility to cover workshop registration fees through the SIGACCESS Diversity and Inclusion Scholarship. --- ORGANIZERS Each of the organizers has experience conducting multidisciplinary research related to critical disability studies, inclusive design and engineering, accessible computing, and human-computer interaction. We bring diverse methods, theoretical frameworks, and extensive experience working closely with disabled users of technology. Furthermore, our team is made of scholars at diferent points in their careers, including early career researchers and established faculty. More information about the workshop organizers can be found in the list below: Arthur Theil is a Lecturer in Human-Computer Interaction within the School of Computing and Digital Technology at Birmingham City University, England. His research focuses on the study of novel interaction techniques to support users with diverse sensory abilities. His current focus is on designing accessible interfaces for individuals with multisensory impairments . Arthur has also conducted accessibility research with older adults who experience age-related changes in sensory, cognitive, and motor abilities. In addition to conducting academic work, Arthur currently also serves on the ACM SIGCHI Accessibility Committee and is part of the Program Committee for the ACM SIGACCESS ASSETS Conference. --- Chris Creed is an Associate Professor in Human-Computer Interaction at Birmingham City University where he leads the HCI Research Group. His core research interest is around the design and development of assistive technology for disabled people . He is currently leading multiple funded research projects focused around accessibility such as investigating new interface techniques for facilitating creative work via gaze/speech interaction , exploring the development of inclusive AR/VR experiences , making coding more accessible for people with physical impairments , and investigating the potential of wearable technology to support young people with special needs within residential care . Dr Creed's research is multidisciplinary in nature and has been conducted in close partnership with national charities, disability and accessibility organisations, special needs colleges, large arts/cultural partners, and disabled people. Mohammed Shaqura is a Research Fellow in Haptic Communication and Navigation at the Institute of Design, Robotics, and Optimisation, University of Leeds. He received a BSc degree in Control Systems from King Fahd University of Petroleum and Minerals, and a PhD in Controls and Dynamics from King Abdullah University of Science and Technology. He held various engineering positions in robotics in the Middle East. He has been involved in multiple research projects focussing on the engineering and non-engineering aspects of assistive technologies, including. His current research focuses on developing haptic interface systems and the design of wearable haptics for navigation and localisation. His research interests are optimal control, indoor localisation and mapping, embedded robotic systems, and learning algorithms. Nasrine Olson is a Senior Lecturer in the feld of Library and Information Science at SSLIS, University of Borås, Sweden. Core research interests have involved issues of power and the relationship between day-to-day action, and broader societal structures. In the more recent years the focus has been on the societal implication of ICTs and information practices that enable or hinder the potential for equal opportunity for all. Towards this, Nasrine has adopted participatory practices in her research and has developed methodologies and technical innovations for haptic communication. She has also been instrumental in creating research environments that promote and lead to improved inclusive technologies and environments by coordinating projects such as SUITCEYES and a new Horizon Europe project called MuseIT which intends to create multisensory representation of cultural assets for broader accessibility. Nasrine is also the director of an interdisciplinary research centre called INCLUDE -Centre for Inclusive Studies, where through critical examination the ideology of normal, the unequal treatments of societal members will be explored. Raymond --- WEBSITE https://www.mpdas-assets22.org/ --- PRE-WORKSHOP PLANS We plan to invite diferent research communities to our workshop. This includes sharing the Call for Participation with members of the SIGCHI, SIGACCESS, and GATE communities. Similarly, we plan to invite practitioners working with the European Disability Forum, Deafblind International, Anne Sullivan Foundation, SENSE, SCOPE, AbilityNet, African Community on Assistive Technologies, and members of the South Asian and Latin American HCI communities. All workshop materials and accepted abstracts will be shared with participants via website and email ahead of the workshop day. Based on accepted abstracts, we will split participants into diferent multidisciplinary sub-domains related to their interests or expertise . --- CALL FOR PARTICIPATION Historically, the accessibility community has focused on designing assistive technologies and systems related to single categories of impairments. While this approach is justifable and contributes to the advancement of the feld, many researchers argue that designing technologies for one impairment at a time contributes to an oversimplifcation of disability and does not refect real-world experiences of a signifcant number of disabled users of technology. This workshop intends to challenge current paradigms in the accessibility feld and draft a joint research agenda that actively involves users with multiple impairments in our practice. We invite researchers and practitioners of diferent disciplines interested in designing assistive technologies with and for people with multiple impairments or complex disabilities to submit 500word abstracts stating their existing work, new ideas, or their critical position related to challenges in designing technologies for users with multidimensional accessibility needs. Abstracts should discuss the relevant academic literature, include short position statements or issues related to the workshop's theme. We also welcome submissions in the form of videos or posters. Authors must ensure the accessibility of their submission by following the SIGACCESS template and accessibility guidelines. Submissions can be made by September 9, 2022 by emailing the abstract to [email protected]. Authors will be notifed by September 26. If accepted, at least one author must attend the workshop at ASSETS 2022 on Sunday 23 October . Authors will be invited to expand their abstracts and contribute to a Special Issue Journal submission after the workshop.

Historically, the accessibility community has focused on designing assistive technologies and systems related to single categories of impairments. While this approach is justifable and contributes to the advancement of the feld, many researchers argue that designing technologies for one impairment at a time contributes to an oversimplifcation of disability and does not refect real-world experiences of a signifcant number of users who live with multiple impairments. How can the accessibility feld leverage theories and concepts from other disciplines (e.g. disability studies, medical humanities, social work, engineering, education) to support more appropriate and inclusive technologies for users with complex needs? How can future assistive technologies take into consideration a diverse combination of sensory, cognitive, and physical characteristics? In this workshop, we invite researchers and practitioners from diferent disciplines to share their perspectives related to challenges and opportunities in designing accessible systems that consider the multidimensional and often complex needs of users living with multiple impairments.• Human-centered computing → Accessibility; Accessibility theory, concepts and paradigms; • Social and professional topics → User characteristics; People with disabilities.

Introduction Each year, 8.2 million people worldwide die of cancer; most of these cancer deaths occur in Asia [1]. Cancer patients often require transfer to intensive care units after they undergo invasive procedures such as cardiopulmonary resuscitation and tracheotomy or when they suffer comorbidities such as acute respiratory failure, infection induced by disease progression or treatment, cardiovascular disorders, and neurological disorders [2,3]. The patient or the surrogate may inform the medical team that palliative care is acceptable. After signing a statement in accordance with the Hospice Palliative Care Act, the patient is transferred to a hospice ward or a palliative care ward according to the condition of the patient. Policies for ICU visits differ by country and by institution. In Taiwan, for example, most ICUs limit visits to 30 min two times a day and by no more than two visitors at a time. The rationale is to protect the privacy of the patients, control infection, and minimize interference with treatment. In comparison, visitor policies are relatively more flexible in the United States but relatively stricter in the UK and elsewhere in Western Europe [4]. For cancer patients who lack capacity to make decisions about their medical care while in ICU, healthcare decisions are often made by a surrogate or relatives. Nearly 95% of cancer patients cannot make their own medical decisions at the time of ICU admission and must rely on a surrogate [5]. In Taiwan, a patient under clear consciousness can sign a letter of appointment for a pre-determined surrogate. A surrogate is defined as a family member, friend, or other person who communicates the preferences of the patient regarding medical treatment and makes major medical decisions for the patient, such as consenting to life-sustaining treatments, end-of-life care, etc. Decision-making abilities include the abilities to understand and clearly express ideas, to make decisions about treatment, and to evaluate and debate the merits of available treatment options [6]. In Taiwan, a surrogate must be aged 20 years or older and have behavioral capacity to make healthcare-related decisions on behalf of the patient. Legally, surrogates appointed by medical institutions have the right to make medical decisions for the patient, even if family members disagree. If a patient has not signed a letter of appointment for a pre-determined surrogate, healthcare-related decisions can be made by a legal relative [7][8][9]. An ICU admission is a crisis situation for both the patient and the surrogate. For example, 14-81% of the family members of ICU patients have symptoms of posttraumatic stress disorder after they participate in making healthcare decisions for the patient [5]. In a study of stress experienced by family members of ICU patients treated at a Taiwan medical center, Li found that stress in family members peaked within 48 h after the ICU transfer [10]. Contributing factors include the unfamiliar environment of the ICU, fatigue, negative emotions , and limited understanding of the knowledge aspects of the treatment. This stress can cause strain in family relationships or friendships, disputes over medical decisions, and poor healthcare-related decisions by surrogates. Additionally, some populations of patients may not have a strong personal affinity with their legal surrogates. Surrogates for these populations may endure inner conflict during decision-making and during efforts to reach a family consensus. The combination of a complex medical condition and a complex family relationship further increases the possibility of decision-making conflicts. The decision-making process in ICU surrogates may differ by culture, and culture directly affects individual behaviors. For example, social harmony is the essence of Confucian theories of social interaction in Chinese culture. Chinese society is relational and collective. The social interactions and behaviors of all members of society are guided by Confucian social norms, which make social interaction more complex compared to that in western culture [11,12]. All relationships in a Confucian society can be classified into five types: superior and subordinate, father and son, husband and wife, elder brother and younger, and friends. Each person in a social relationship has a personal responsibility to maintain the relationship. Social goals also represent different types and levels of love. According to the theory of love proposed by the social psychologist Zick Rubin in 1970, love is the deepest and most meaningful of sentiments. Love is also a multifaceted attitude that a person has about another [13]. Since love is an abstract concept, however, the definition of love depends on the perspective of the individual. That is, the definition of love differs because individual feelings, behaviors, and attitudes differ [14]. Another theory of love is the Sternberg "triangular" theory. According to Sternberg, love has three components: intimacy, passion, and decision/commitment. Each type of love can occur in an interactive and dynamic intimate relationship. In western culture, contact between two people, including physical contact and even eye contact, indicates a strong relationship [13,14]. Different types of love are also recognized in Chinese culture. With the exception of friendships, all five of the Confucian relationships are vertical relationships, e.g., "respect" versus "humble" and "up" versus "down". Therefore, in Chinese culture, emotions are expressed in a roundabout and restrained manner in all relationships except friendships. Because of their love, ICU surrogates worry about losing patients. The roles and relationships of family members also change when they must make healthcare decisions that could determine whether or not the patient survives. In this situation, families experience anxiety and other negative emotions [15]. Research indicates that 20-46% of cancer patients meet the clinical criteria for depression and worry, and up to 75% of spouses experience negative emotions [10]. Chinese populations often have a sense of responsibility and obligation to the family or social group. That is, family and social relationships are prioritized, unlike Western culture, in which the independence of the individual and the freedom to make choices are valued and prioritized [11,12]. Therefore, family and social relationships and attempts to integrate with the family or a social group can cause psychological distress. Cultural attitudes about medical decision-making may be different. For example, Dionne-Odom et al. investigated decision-making experiences in 19 primary surrogate decision-makers of patients in critical care units in the rural Northeastern United States. In these surrogates, decision-making was impacted by gist impressions, distressing emotions, and moral intuitions [16]. In another study in the United States, Moss et al. interviewed seven surrogates and deduced three major themes of the decision-making process: "communication as key in decision making", "impact of past experiences", and "difficulties and coping" [17]. Asians traditionally have low completion rates for advance care planning and tend to rely on the family model of medical decision-making. Su et al. investigated the decision-making experience in 69 medical care surrogates and identified six subthemes, including two Communication subthemes , two Emotion subthemes , and two Conflict subthemes [18]. These studies indicate that cultural attitudes about medical decision-making and filial expectations may cause some surrogates to experience stress and family conflict. Until now, most studies of surrogate decision-making for cancer patients in ICUs have been performed in western countries. Seldom studies have considered how surrogate decision-making is affected by cultural aspects unique to Asian populations such as the reluctance to discuss the impending death of a family member, which is considered taboo in Chinese culture. Even surrogates who recognize that discussion with other family members would improve the decision-making process for loved ones with cancer may avoid discussing these matters until a poor outcome occurs, e.g., coma or sudden death. Although ICU medical teams routinely assist surrogates in making medical decisions, the decision-making process in surrogates of cancer patients in ICU are not well understood. One study explored the quality of care delivered to ICU patients at the end stage showing end-of-life signs and symptoms. The authors reported that, although ICU personnel can determine whether or not death is inevitable, the ICU environment limits opportunities for surrogates to communicate with healthcare professionals. Lack of advance care planning or lack of other documentation that expresses the will of the patient places the decision-making burden on the surrogate. Surrogates have insufficient time to bear the decision related to patient's bad outcome, leading to poor quality in the face of patient's death for surrogates confusing [19]. In Taiwan, ACP has not yet matured. The Hospice Palliative Care Act can only be signed for patients with terminal cancer, and the Patient Right to Autonomy Act was only recently passed in 2019 [6]. However, ACP has limitations. First, most patients in critical care units in Taiwan are highly reliant on their surrogates. For patients with complex clinical conditions, accurately predicting the outcomes of a medical treatment is often difficult. The preferred treatments indicated by patients during the ACP process may later prove impractical or inapplicable. Additionally, patients may not anticipate various outcomes when they indicate their preference for active treatment. For example, during ACP, the active treatment that a patient selects for progression of cancer may not be appropriate if the patient also has pneumonia. Preferences for active treatment indicated in ACP may not be a good choice for the patient and may later present a dilemma for the surrogate. For patients who do not sign an ACP or who do not sign a "do-not-resuscitate" order, family members, especially in Asia, may prefer that the patient continue to live for reasons such as their strong emotional bond with the patient, pending property issues, etc., even if it is not in the best interest of the patient. Although longevity is considered a blessing in Confucian society, family harmony is equally important in Asia. Regardless of their current relationship, all family members desire harmony and may make a decision inconsistent with the preferences of the patient. Differences in family relationship, filial piety, etc. result in a decision-making experience very different from that in western culture [20]. Studies of medical decision-making for acute care patients or for children have elucidated the highly complex and multifaceted process of surrogate decision-making. However, little is known about the experience, the feelings, and the needs of surrogates of cancer patients in ICUs in Asia. Therefore, the objective of this study was to explore the experience of surrogates of cancer patients in ICU during the medical decision-making process. --- Methods --- Design, Setting, and Participants According to Denzin and Lincoln , six respondents are sufficient to achieve data saturation when in-depth interviews are used to collect data in a phenomenological study [21]. This qualitative study interviewed surrogates as part of a descriptive, comparative study . The researchers performed face-to-face interviews of surrogates of adult cancer patients who were currently in ICU and lacked decision-making capacity. A phenomenological methodology was used to understand the experience of surrogates in the course of decision-making. The inclusion criteria were designation as surrogate authorized to make medical decisions for a cancer patient currently in ICU and age over 20 years old. Surrogates were excluded if they had any emotional or cognitive deficiency. The principles applied in the subject selection process were high variability and good narrative skills. Data collection continued until data saturation was reached, i.e., when data analysis revealed no new information. The study was performed in a medical center in southern Taiwan with 1200 hospital beds, including 130 internal medicine and surgical ICU beds. At this medical center, cancer patients comprise 15.4% of all patients admitted to the ICU annually. --- Data Collection Institutional review board approval was obtained before this study was performed . Surrogates who met the enrollment criteria were contacted, informed consent was obtained, and interviews were arranged. Before the interviews, the researchers explained the objective and procedure of the study to each subject. The subjects were informed that a researcher would record each interview and take notes during the interview. Surrogates were interviewed in conference rooms near the ICU. Each interview lasted 40-60 min. The subjects were ensured that all interview data would be anonymous, and all subjects gave written informed consent to participate in the study. An experienced researcher asked four questions in semi-structured interviews: How would you describe your experience of the healthcare decision-making process during this ICU stay? How would you describe your relationship with the patient during this ICU stay? How have your daily life activities changed during this ICU stay? How have you coped with emotions and experiences during this ICU stay? Interviews were performed during the first 3 days of the ICU stay. At the end of the interview, each patient was given a gift as a token of appreciation for participating in the study. --- Data Analysis Content analysis of interview data was performed in the following steps. First, data were encoded to ensure confidentiality of the interviewee. Each interview was recorded, and the researchers transcribed the audio recording of each interview within 48 h. Additionally, the researchers carefully read the resulting transcripts and extracted significant statements related to new meanings that emerged after transcription. The common characteristics of meanings were then clustered, and subthemes were grouped as themes [22]. The strategies used in this process included peer review, in which the researchers individually coded and then collectively negotiated mutually acceptable statements and definitions. Consistency in developing the major themes and subthemes was further enhanced by inviting an independent researcher to identify themes by reading transcripts of two randomly selected interviews. Content analysis was performed simultaneously with data collection. Data collection ended when theoretical saturation was achieved, i.e., when no new themes emerged in further data analysis. Three experts, each with 15-20 years of experience in qualitative research, reviewed and discussed randomly selected interview transcripts and the results of qualitative analysis performed by the principal investigator . Finally, the qualitative analysis results were verified by further interviews and by comparisons with medical records. --- Rigor The reliability and validity of the qualitative study were evaluated in terms of credibility, conformability, dependability, and transferability, which are well established standards for assessing the rigor of qualitative research [23][24][25]. Credibility: The PI had approximately 9 years of clinical experience in ICU and was in constant contact with the surrogates throughout the study. For an improved understanding of the phenomena investigated in this study, the PI completed an 8-month clinical practicum in palliative care wards, oncology outpatient departments, and surgical ICUs at the research sites to observe surrogate decision-making processes in various situations. Conformability: Interview data were collected by audio recordings and stenography. After the interviews, the PI confirmed the accuracy of the collected data by listening to the recordings of the interviews and comparing them with the written transcripts. The PI then asked each interviewee to clarify any unclear or ambiguous content. Dependability: To ensure high dependability of the research results, the qualitative experts reviewed selected interview transcripts and compared their analysis results with those obtained by the researcher. Transferability: The PI took handwritten notes regarding transient details of the interview process to provide a reference for further studies by other researchers. The PI and the qualitative experts then agreed on a coding structure. In the case of a difference of opinion on coding, the PI and qualitative experts continued discussion of the coding until a consensus was reached. --- Results The subjects of this study were recruited during a six-month period from June to December 2017. Eight surrogates were interviewed. Table 1 lists the basic information for the surrogates. Figure 1 shows the four themes, thirteen subthemes, and the meanings of each subtheme. --- Use Love to Resist: Internal Angst The internal angst theme relates to the reflexive self-blame, the inner conflict, and the reluctance to make decisions, which the surrogates experienced when they observed the patient suffering. Figure 1 shows the three subthemes of Use love to resist: internal angst. --- Use Love to Resist: Internal Angst The internal angst theme relates to the reflexive self-blame, the inner conflict, and the reluctance to make decisions, which the surrogates experienced when they observed the patient suffering. Figure 1 shows the three subthemes of Use love to resist: internal angst. --- Self-Binding Shackles When faced with the possible loss of a loved one, surrogates reflexively blamed themselves, as if hoping to reduce or compensate for the pain and suffering experienced by the patient. For example, Subject A stated that she usually traveled with her oldest daughter every summer. However, a sudden change in the condition of Subject A required a transfer to ICU for treatment. Therefore, Subject A and her daughter returned from their vacation in Japan on the day after they arrived. Subject A said, "I think to myself that, if I had not gone on vacation, none of this would have happened; I would not have been sent here . . . ." --- Pity for the Old Callus Despite their short visiting hours with the patients, the surrogates could clearly discern the declining health and changing appearance of the patients as various tubes and other life support equipment and devices were attached to their bodies. The increasing use of medical equipment drew attention to the changing physical appearance of the patient and caused the surrogates to experience feelings of pity. For example, some patients required high doses of vasopressors to treat peripheral arterial occlusive disease. Subject G became concerned when she noticed that the patient had a spot on his hand after vasopressor treatment and said, "Yesterday, this spot was not this dark and not this big. I thought this hand was already as bad as it could get, but you can see how dark it is now. It is even worse than before." 3.1.3. An Old Callus Trapped in a Spider Web When a patient is admitted to ICU, the surrogate may experience a stabbing heart pain caused by the perception that the patient is hanging on a cliff and is as fragile as an old callus. However, a surrogate must be willing to release the emotional attachment to a patient who is approaching death so that the patient can die peacefully with minimal pain and suffering. The decision to let go can cause immense stress for the surrogate. Subject B recalled, "I said in his ear that it was okay if he did not want to live. I told him I would be fine alone if he just followed the Buddha ." --- Allow an Angel to Spread Love Among Us: Memories and Emotional Entanglements The surrogates in this study discussed their good memories of the patient. The semi-private nature of the ICU environment limited physical contact between the surrogate and the patient. The surrogate was similar to an "angel" who can comfort and accompany the patient. However, their emotional entanglements with the patient varied from happiness to sadness and from cheerfulness to anger. This study identified five subthemes . --- Remembering the Yellowed Past While waiting outside the ICU, the surrogates often reflected on their past relationships with the patients. They contemplated the times they felt harmony as well as the times they had fights and arguments. Subject E said tearfully, "Whenever I go home at night to sleep, whenever I close my eyes, I can hear him calling me. I think about how we used to spend time together. Everyone envied us, you know . . . how am I supposed to fall asleep?" Most of the surrogates had participated in important moments in the lives of the patients. Therefore, surrogates tended to associate the current suffering with unhappy experiences of the patient before hospitalization. Subject A described experience, "He knew that the drug therapy would be very difficult to endure, but he never lost his temper, and he always cooperated with the medical personnel until now. We always knew that, together, we could overcome depression or negative emotions." --- 3/4 Happiness If "1" represents a complete family with caring and mutually supportive family members, then "3/4" represents a family in which one member is missing. When the patient is hospitalized, family members and the surrogate are busy traveling between the hospital and their disrupted home. Subject A said, "During this period , everyone has been running back and forth. I am thankful to have them now that patient is in ICU. Otherwise, I wouldn't know what to do." Subject E described a similar experience, "If we were to go home now, the thing we would fear most is the phone ringing. We would always be afraid that the hospital was calling." Surrogates expressed fear of losing the patient and the willingness to do anything, no matter how difficult, to help the patient get better and return home so that the family could be whole again. --- Smile! Anger For a patient whose body is wracked with the pain of a disease, emotions and moods can change. Surrogates hope for the best and continue smiling while caring for their loved ones, even when anger and conflicts arise. The surrogate considers the disease, pain, and low spirits of the patient and hopes for improvement in the condition of the patient. Subject H angrily said, "At home, I ask him what he wants to eat before I start cooking, and he does not respond! He does not like talking! Sometimes I get very angry serving him every day; I am also exhausted, but I am still willing to do anything for him ." Even when the patients tire and cry during the course of caring, the surrogates grudgingly tell themselves there is no other way and that caring for the patient is their responsibility. --- Left all Alone The formation of a traditional family starts with two individuals who know each other and who care for and love each other. They live together, rely on each other, and accompany each other. After the departure of a loved one, the surviving partner or family member must contemplate the prospect of continuing alone. Subject E lost his wife to recurrent breast cancer two years earlier. He and his son were the only remaining family members. Subject E said, "Two years ago my wife passed away, I have yet to walk out of that pain. Now my son is dying. There is no one left in the family." Family structures in Taiwan and throughout the world have rapidly changed in recent years. For example, the number of households with non-traditional family structures , singles, elders living alone, and gay couples) has increased. Changes in social welfare systems and legal systems may be too slow to accommodate these emerging populations. Current laws do not ensure comprehensive care for such populations on a daily basis. Surrogates are often under extreme stress, especially when they have no one to consult and no way to express their feelings. For example, Subject H, who had immigrated to Taiwan 20 years earlier, provided daily care for a patient. Subject H tearfully described the feelings she experienced when the patient was suddenly admitted to the ICU due to the disease. "Others had help, but I had no one to turn to. I felt so sad. Others had family members, but I was alone. Others had at least one or two people for support, but I was running in and out of ICU by myself. Do you know how desperate I felt?" --- Engraving the Hippocampus In the current era of ubiquitous media and communications technology, family members may have difficulty remembering the date that a loved one was hospitalized. Surrogates, however, can often remember the date of every important disease-related event in the life history of the patient. Subject A, for example, could remember the date the patient was intubated and the days she waited for the patient to awake from sedation. These dates were engraved on her hippocampus. During the interview, Subject A said, "Yes, it has been a week since . . . was sedated last Wednesday; it was also a week last time.". --- Dilemmas of Love: Anxiety About ICU Visitor Restrictions The medical team may have limited time to provide information, which may be highly technical, that the surrogate needs to make healthcare decisions for the patient. Additionally, the presence of other patients and visitors in the confined space of an ICU can limit the ability of surrogates to share their emotions and the ability of friends and family to provide emotional support. This study identified two subthemes . --- Persistent Smog A patient in critical condition often requires rapid and uninterrupted medical diagnosis, treatment, and decision-making. The medical team may have very little time to provide a surrogate with the information needed to make a decision about a patient who is in critical condition. Additionally, the medical team may use rough language and behavior that can hurt the feelings of the surrogate, especially surrogates who have relatively limited medical knowledge and information. For example, Subject H mistakenly purchased a large amount of protein supplement and was distressed to find that it was not required. Subject H said, "The nurse said the patient needed two cans a day, so I bought 30 cans. The nurse then said I should have thought about it before I bought so much. She said I couldn't get a refund. I told her not to worry and that I would not return it." Most people have difficulty understanding unfamiliar medical concepts. Surrogates who have low education levels and limited medical knowledge may not even know what questions to ask in the high-pressure environment of a hospital ICU. Subject B mentioned, "The doctor said that I would not understand because I do not have sufficient education. Anyway, even if I did understand, I would not know what to ask." Misunderstandings about the treatment plan can rapidly lead to critical changes in the relationship between the surrogate and the medical team. Subject G said, "My father-in-law underwent a full body examination every year. The first year, the image was blurry, but the doctor said it was nothing. My father-in-law is a rural laborer with no medical knowledge. After the doctor looked at the report, he did not even tell us to go to a bigger hospital for another image. He just kept saying many people have blurry images and they are fine. So, we just kept thinking this way. It was not until last year that we realized this was not right." Surrogates often recalled how the patient felt disappointed during the diagnostic process. --- Blockaded Longing Although all ICUs have visiting hours, the visiting hours may not be long enough to enable the patient and surrogate to form an emotional bond or to help each other understand the current condition of the patient. For one patient in this study, Subject C was the current surrogate and had been the caregiver for the patient almost 10 years earlier. The first time the patient was in critical status, endotracheal intubation and a transfer to ICU were required. Subject C said, "You guys tie him up when he is staying here. You don't let us accompany him. You can see it's hard for him to breathe when he gets emotional. We can help you take care of him, but you won't let us." When medical care can no longer maintain the health status of the patient, surrogates and the medical team must begin discussion of hospice care. Again, the medical team may not allow the surrogate to accompany the patient. When the medical team suggested transferring the patient to in-hospital hospice care, Subject G quietly asked, "Does hospice care need to be provided here? Can he be transferred elsewhere?" --- Suffocating Love: Entanglement in Decision-Making During the decision-making process, the surrogates and the medical team rarely interact as equals. Surrogates may lack the medical knowledge and communication skills needed to participate effectively in decision-making, especially in the emergent, fast-paced environment of the ICU. Family relationships unique to Chinese culture and emotional entanglements among family members may prevent surrogates from thinking logically when healthcare decisions must be made. This theme revealed three subthemes: chasm, unpredictability, and tangled limbs . --- Chasm A chasm between the thoughts of the surrogate and the thoughts of the medical team may appear during the decision-making process. For surrogates who have limited medical knowledge, self-doubt, and hesitancy can become major obstacles in the decision-making process. Subject A was concerned that the patient could not tolerate the procedure and refused further examination of the patient. Subject A said,"He has already done this 4-5 times, and he has done TUNEL [trans catheter arterial embolization] on two consecutive days. Sometimes I just want everything to go smoothly, but he received an injection of a very strong drug. I heard this drug should only be used once in a while. So I said no. But the doctor wanted to do it. This is a situation that we do not want." After making decisions for the patient in the acute phase of the disease, some surrogates experienced doubt about the decision and began to question the suggestions of the medical staff. Subject F complained that the medical staff did not provide a timely response to his questions about follow-up care and treatment, "After surgery, no one explained how to maintain oral hygiene. That is what I am most worried about. The doctor did not mention whether further chemotherapy or radiotherapy is needed. Why don't they explain what further treatment is needed?" 3.4.2. Unpredictability For surrogates, understanding the possible results of a treatment is similar to making a guess about something that is unpredictable. That is, even in the high technology era of today, surrogate decision-making is as difficult as speculating about whether a rainbow will occur after a rain, how it will look, and how long it will last. Making a healthcare decision is particularly difficult when the surrogate knows that a poor decision made under mental distress may cause the patient pain, injury, or even death. For example, when Subject E had to decide whether the patient should undergo surgery, Subject E said, "I fear that, if he does not get well after the surgery, I will not be able to care for him because I have to work. I would not be able to live on if anything happened to him. He is stable now, which is good. Surgery, of course, is very risky. I have seen too many bad outcomes of surgery.". --- Tangled Limbs A well-recognized feature of Chinese culture is the need to balance the concerns of the immediate family with the concerns of the extended family [26]. For surrogates who must make healthcare decisions, this tension can cause self-doubt and hesitancy. Surrogates may be tormented by this constant tension between the need to act rationally and the need to be sensitive to family members. Subject E described feelings of tension he felt when he had to decide whether the patient should undergo endotracheal intubation, "I told them not to place the tube because he would not get better. Since no one wanted to say it, I had to say it, even though I knew everyone might hate me later. I did not want my dad to suffer anymore." The interview results show that the feelings of family members were a major consideration during decision-making by surrogates. In some cases, the surrogate deferred to other family members, even if the surrogate disagreed. Conflicts were most likely to occur when the surrogate made a decision for a patient who was in critical condition and required an invasive procedure such as surgery, endotracheal intubation, gastroscopy, or cardiopulmonary resuscitation. --- Discussion Figure 1 is the qualitative model of the decision-making process of surrogates in this study. Decision-making by a surrogate was affected by love for the patient and the close personal relationship with the patient. Four effects of this close relationship were identified: Use love to resist: internal angst; Allow an angel to spread love among us: memories and emotional entanglements; Dilemmas of love: anxiety about ICU visitor restrictions; and Suffocating love: entanglement in decision-making. The surrogates in this study reflexively blamed themselves and hoped to reduce or compensate for the pain and suffering experienced by the patient. In some cases, the feeling of reluctance to make decisions further evolved into self-blame and then feelings of internal angst, which they described as a heartbreaking situation. Research shows that approximately 30% of surrogates experience anxiety and depression when a family member is in ICU, but the anxiety and depression taper off during the ICU stay [27]. Surrogates in this study experienced inner conflict and hesitance when they were required to make decisions that would result in physical disfigurement of the patient. The surrogates in this study also perceived that limited visiting hours weakened their emotional bonds with the ICU patients. Limited contact with patients induced mental distress in the surrogates and indirectly induced negative feelings during decision-making. Each surrogate considered the patient an angel that the family needed to be complete. The family members described how their memories of the patient, whether happy or sad, were bits and pieces of their lives. Their positive and negative emotions coincided with positive and negative changes in the condition of the patient, and they expressed fears of being alone if the patient left them. The Confucian value system evolved from the need to maintain complex social relationships and to guide behavior and interactions among family members. That is, Confucian guidelines for behavior and conduct are intended to maintain strong family relationships and good public perceptions. Compared to westerners, however, Chinese populations tend to be more reserved in their expressions of love and other emotions. The results revealed that surrogates informally tended to avoid discussing certain highly stressful matters such as a perceived lack of family support, a poor relationship with the patient, and progression of the disease from onset to ICU admission. Ji used a phenomenological methodology and performed in-depth interviews to explore the experience of nine surrogates of ICU patients. The theme of "Being in a constantly fearful and pressured state" and its subtheme of "Family member and medical activity assistance" emerged in their study. The authors described how insufficient family support and conflicts between the family and the medical staff can induce negative feelings such as fear and psychological distress [28]. Surrogates of certain populations such as DINKs or new immigrants may feel extreme mental and physical exhaustion when no other relatives or friends are available to provide support and assistance. For many populations, especially in Asia, children are expected to provide lifelong care and support for their parents. However, this responsibility can be a heavy emotional burden. The emotional longings and desires of surrogates of cancer patients are rarely considered by the medical team, and limited ICU visiting hours deprive surrogates of their emotional bonds with the patients. The ICU team can help them by referring them for psychotherapy or other social services or by encouraging them to seek emotional support from friends and relatives. The major sources of negative emotions for surrogates in this study were the limited visiting hours and the enclosed environment of the ICU, which deprived surrogates and patients of their emotional connection and caused a constant yearning for each other. This feeling of yearning, coupled with changes in the physical appearance of the patients related to disease or treatment, often caused the surrogates to experience tension and shock, which impaired their relationship with the medical team. Disrupted communication and connection with an ICU patient can cause family members to lose confidence in their ability to make healthcare decisions [15]. The needs and longings of surrogates are not prioritized by medical teams. For example, medical personnel who are focused on providing rapid and effective treatment for an ICU patient may not have time or energy to consider the psychological distress experienced by a surrogate who observes physical changes in the patient, such as large surgical wounds, severe bruises, or edema caused by procedures such as endotracheal intubation or repeated placement of other invasive devices. Surrogates who perceive a lack of concern about their distress may then experience feelings of resentment, which can negatively affect their decision-making capability and cause them to experience doubt and unease about the medical team. These changes can cause negative emotions such as sadness and regret as well as suspicion about the competency of the medical team [29,30]. This study revealed that conflicts between rationality and sentimentality can impair decision-making, not only when the surrogate perceives a gap between the knowledge of the medical team and the information the medical team actual conveys, but also when the surrogate experiences complex emotional entanglements with the patient. Dionne-Odom et al. explored and compared decision-making processes in surrogates for ICU patients in their last stages of life [16]. The discussed how surrogate decision-making is affected by painful affections uncertainty about the medical treatment due to a knowledge gap, e.g., a surrogate may be concerned that the patient cannot tolerate a painful medical procedure. Specifically, the author discussed how "painful affections" decrease the willingness of surrogates to authorize further treatment for the patient. For surrogates, these relationships and entanglements resemble a net with which the surrogate must disassociate in order to make a rational decision. The interview results also revealed that a struggle for power among family members often causes conflict during the decision-making process. Family demands are a well-recognized phenomenon in the "moral and emotional concepts" factor. Surrogates must consider the demands and expectations of all family members so that the treatment decision is consensual [16]. Su conducted focus group interviews with 69 surrogates and found that communication, emotions, and conflicts are the three most important factors in the surrogate decision-making process. The struggle for power among family members with different thoughts on medical treatment for the patient is a major cause of conflict during the decision-making process [5,18] Therefore, support from family members is essential for effective surrogate decision-making. However, displays of affection and emotions such as love and gratitude are much more restrained in Chinese culture in comparison with western culture [31]. Health professionals can improve the decision-making process and help families reach a consensus by referring them for counseling. A notable cultural difference in surrogate decision-making is that surrogates in Asia tend to consider the feelings and concerns of other family members and the need to reach a consensus. In contrast, surrogates in other countries tend to focus on choosing the best treatment option for the patient. For medical teams, the main concern is rapidly and efficiently diagnosing the patient and then providing treatment in the high-stress and fast-paced environment of an ICU. Communication with surrogates is not prioritized, which increases the risk of a decision-making conflict. The risk of conflict can be reduced by clear communication that gives surrogates the information and confidence they need to understand evolving conditions during the decision-making process. Medical teams in ICUs should share their professional opinions with surrogates and should invite surrogates to ask questions about treatment. However, information should be presented in layman terms whenever possible. Visual aids can also be helpful for discussing technical matters. For example, radiologists often use simple drawings on paper or whiteboards to explain radiograph findings [32]. The medical team should also schedule time to meet with surrogates and family members to discuss the medical plan, to address their concerns and needs, and to confirm their understanding [32]. Scheduled meetings would also assist the family members and the surrogate in making consensual decisions. --- Implications for Practice We have four suggestions for decreasing surrogate decision-making conflict in the ICU. First, the ICU team should help surrogates by clarifying whether changes in physical appearance are short-term or long-term changes, especially after the patient undergoes an invasive procedure that substantially changes the physical appearance of the patient. Second, the ICU medical team should, as early as possible, identify surrogates who have limited support systems and refer them to social workers. Third, visiting hours for surrogates should be flexible. Flexible visiting hours would enable surrogates to maintain emotional bonds with ICU patients by providing opportunities to assist with simple tasks, e.g., sending text messages, and would enable surrogates to spend time caring for patients. Last, hospitals should provide a comfortable rest area near the ICU where the surrogate and the family can discuss medical information and treatment options. A limitation of this study is that the sample was recruited from a single medical center, which limits the potential generalizability of the results. Factors that affect the decision-making process in healthcare surrogates may differ in other geographic regions due to differences in cultural characteristics and differences in ICU visitor policies and other hospital policies. Thus, this study should be replicated in samples of surrogates of cancer patients in ICUs in different hospital levels and in different geographic regions. All participants in this study were young or middle-aged adults, which is another limitation of this study. Additionally, the long duration of the interviews may have inconvenienced or distressed the surrogates, which could have affected their interview responses. Another limitation is that, even when interviews were scheduled in advance, the emotional impact of external conditions at the time of the interview may have caused interview bias. To minimize data collection bias, future studies should perform two separate interviews for each participant. Observing the emotional impact of decision-making in healthcare surrogates, particularly in Chinese populations, is very challenging in quantitative research. Further studies are needed to perform qualitative research in a larger and more diverse population of ICU patients. --- Conclusions This study of the healthcare decision-making process for ICU patients revealed that surrogates experienced four types of love: Use love to resist. The surrogates reflexively blamed themselves and hoped to reduce or compensate for the pain and suffering experienced by the patient. Allow an angel to spread love among us. The surrogate and the patient accompanied each other through their memories of the decision-making process. Dilemmas of love. The emotional longings and desires of surrogates of cancer patients were rarely considered by the medical team, and limited ICU visiting hours deprived surrogates of their emotional bonds with the patients. Suffocating love. Conflicts between rationality and sentimentality impaired decision-making. A struggle for power among family members often caused conflict during the decision-making process. Since ICU visitor restrictions prevented the surrogate from accompanying the patient at all times, surrogates were left alone with their worries. Surrogates could not always predict the outcome of a healthcare decision. Decision-making was negatively affected by the gap between the knowledge of the medical team and the knowledge of family members and also by negative emotions generated by poor relationships among family members. This study also found that decision-making was particularly difficult for surrogates of patients in certain populations that tend to have limited social support . Certain characteristics of Chinese culture, e.g., restrained and/or implicit expression of emotions, were another barrier to decision-making [31]. The results of this in-depth analysis of surrogate decision-making can help medical teams understand the shock and internal conflict experienced by surrogates during decision-making and provide a reference that medical staff can use to improve their communications with surrogates. ---

Background: Few studies in Asian countries have explored the emotional entanglements and conflicts that surrogates often experience during the medical decision-making process. This study was to explore decision-making processes in surrogates of cancer patients in a Taiwan intensive care unit (ICU). This qualitative study surveyed a purposive sample of surrogates (n = 8; average age, 48 years) of cancer patients in the ICU of a medical center in Taiwan. A phenomenological methodology was used, and a purposive sample of surrogates of cancer patients were recruited and interviewed during the first three days of the ICU stay. Results: Based on the interview results, four themes were generalized through text progression: (1) Use love to resist: internal angst. This theme was related to the reflexive self -blame, the feelings of inner conflict, and the reluctance to make healthcare decisions, which surrogates experienced when they perceived suffering by the patient. (2) Allow an angel to spread love among us: memories and emotional entanglements. Memories of the patient caused the surrogate to experience emotional entanglements ranging from happiness to sadness and from cheerfulness to anger. (3) Dilemmas of love: anxiety about ICU visitor restrictions. The confined space and restricted visiting hours of the ICU limited the ability of surrogates to provide emotional support and to share their emotions with the patient. (4) Suffocating love: entanglement in decision-making. Emotional entanglements among family members with different opinions on medical care and their struggles to influence decision-making often prevented surrogates from thinking logically. Conclusions: Expression of emotions by ICU surrogates is often restrained and implicit, particularly in Asian populations. These results can help health professionals understand the psychological shock and inner conflict experienced by surrogates and provide a useful reference for improving their communications with surrogates.

Many researchers have focused on Loneliness due to its negative impacts on an individual's physical and mental health across the lifespan and ethnicities . Loneliness is defined as the subjective feeling of being socially isolated due to a perceived discrepancy between one's quality and/or quantity of social relationships . According to a national survey by Cigna , 46% of the sampled 20,000 American adults aged 18 years and older reported feeling lonely, with the younger generation scoring the highest on loneliness. 43% of the same sampled adults were also found to sometimes or always feel socially isolated from others. With the current COVID-19 situation, the prevalence of loneliness may increase as a result of limited social interactions with others, which if not managed, may be detrimental to physical and mental health. As such, a rise in the prevalence of loneliness may potentially lead to a decrease in well-being, which, in turn, can impose heavy societal burdens. With that being said, well-being has also been a central research focus. Transformative service research is one of the research domains that looks towards making improvements in the well-being of individuals and their surrounding environment . However, the pandemic has placed tremendous pressure on individuals aged 18 to 64 years old to maintain their well-being while being socially isolated from the rest of the world . --- Gaps in current literature A systematic overview by Leigh-Hunt et al. noted that many have attempted to consolidate works surrounding loneliness and well-being. Most syntheses accentuate the measurement of loneliness as an indicator for well-being, as well as the development and effects of loneliness among the older individuals . Regardless, only a few have focused exclusively on the factors that drive loneliness in adolescents and young adults . The reviews also fail to explicate these factors in the context of parent-child relationships and interactions with friends. Put differently, limited works have documented the experience of loneliness among young people based on their early life experiences, and how this may have impacted their current well-being. Thus, mental health practitioners, educators, and even the general public may lack ample, evidence-based insights to deal with the loneliness epidemic among younger individuals especially so during the present times. There are some realities that urge greater attention to the experience of loneliness among adolescents and young adults. Firstly, although loneliness can affect an individual at any point in life, recent statistics from around the world report individuals aged between 16 to 25 years feeling higher intensities and more frequent feelings of loneliness . Secondly, very few studies have been conducted to study loneliness among the younger people. Evidence on the risk factors of loneliness and poor well-being that cluster in childhood and adolescence remain scarce. With the adolescence phase being the most risky in terms of the emergence of various mental health problems, experiencing loneliness may be stigmatising due to the adolescent's strong need to feel connected to others . This paper proposes that a deeper insight into the experience of loneliness among young people is needed to better direct prevention and treatment approaches tailored to their needs. This will be done by synthesising evidence on the effects of loneliness on well-being and understanding how the two factors are correlated. --- Review question To manage the rising prevalence of loneliness, Pitman et al. have advised investment into raising public awareness regarding loneliness among the younger individuals since much has already been done for the older population. Whilst such investments should include increasing support from mental health services, more effort in the early detection of at-risk individuals and the promotion of concomitant loneliness prevention policies and/or practices are needed. In response, the current paper utilised a systematic scoping review methodology to synthesise evidence pertaining to the impact of loneliness on well-being. Another related purpose was to give considerations to the review's results in providing a deeper understanding of the pathways from loneliness to well-being, and their potential to direct mental health practices. Two questions were derived from the above, which directed focus of the current review: What factors contribute to experience of loneliness among young people? Do existing works help us to construct a pathway between loneliness and well-being among young people? --- Systematic scoping review Scoping reviews broadly summarise current research findings and identify any gaps within a given research topic . Such reviews are particularly useful when researchers lack the time and resources to undertake a more rigorous systematic review . Systematic reviews on the other hand are narrowly focused on finding answers by identifying, evaluating and synthesising all relevant studies on a particular topic . The present paper merges the two procedures to find answers to specific questions from a broad body of peer reviewed literature. This paper follows the five components of the six-step approach outlined by Arksey and O'Malley to scope, identify and elaborate on the key concepts found in the extant literature regarding the association between loneliness and well-being. The steps include identifying the research topic, finding relevant studies, selecting a study, synthesising and interpreting qualitative data and synthesising and reporting the results. The sixth step 'consultation exercise' being an optional component of the scoping study framework , will not be included in the present review. This approach will provide a context to address the current review's queries. Subsequently, the search strategies will be detailed before presenting the results. --- An evolutionary understanding of loneliness For the purpose of this paper, loneliness is understood from an evolutionary perspective, as theorised by Cacioppo et al. . This perspective posits that loneliness heightens survival mechanisms towards social threats to reduce the damages that can be inflicted from poor social interactions. Lonely individuals would subsequently be motivated to search for and maintain social connections to ensure survival and security. Experiencing loneliness over time can result in poor physiological and psychological health and its effects can be mitigated by interpersonal connections that function similarly to that of other basic human needs such as hunger and thirst. An experimental study by Poerio et al. reported that lonely individuals, who participated in daydreaming about socialising with a close other, showed a significant improvement in feelings of connectedness and belonging. The outcome indicates that the sense of belonging resulting from engaging in a pleasant social activity perhaps provides a rewarding, emotional benefit that alleviates feelings of loneliness. --- Loneliness and well-being In line with the emphasis on the evolutionary perspective of loneliness, an epidemiological cohort study by Matthews et al. found that loneliness is an indicator for poor functioning in the various domains of well-being. In particular, loneliness has been reported to have strong negative associations with subjective well-being in both cross-sectional and longitudinal studies. SWB is defined as people's cognitive and emotional perceptions of their lives . It reflects a positive state that drives individuals to continue maintaining this state. Conversely, as described in the previous paragraph, loneliness is understood as an aversive state that acts as a signal to change behaviour. Hence, loneliness can affect SWB through the changes in emotions and moods. To illustrate, Hsu reported that among older adults the unfulfilled expectations from the family can result in feelings of loneliness and isolation, which in turn is resulted in depressive symptoms and lower life satisfaction. Shankar et al. also noted that older individuals aged 50 years and above who experienced initial high levels of loneliness consistently have lower SWB over a six-year period. Similar outcomes are reported for young adults where evidence points to direct link between loneliness and low well-being . These results confirm the notion that loneliness and SWB share a negative correlation. The perceptions of loneliness appear to influence an individual's SWB despite the presence of factors that aim to promote SWB, suggesting that the aversive effects of loneliness may extend beyond negatively affecting one's SWB to influence one's overall life satisfaction due to the social disconnection from others since they perceive the people around them as a threat . Pitman et al. underscored the importance of understanding how the quality and quantity of social relationships among young people affect the pathway between loneliness and well-being. A qualitative meta-analysis by Achterbergh et al. noted that the experience of loneliness in later stages of life is driven by relational and environmental factors faced in childhood and adolescence. Such factors include parent-child relationships and interactions with peers. The experience of loneliness varies across sociodemographic factors. Early experiences with poor parental bonds can lead to insecure attachments with external others such as friends. This is in accordance with the attachment theory, which posits that interactions with attachment figures build an infant's mental representation of social relationships . These representations subsequently influence the infant's behaviour, thought processes, and emotions throughout the lifespan. Additionally, parent-child relationships appear to be the most influential form of social bonds. Positive interactions with parents are necessary for individuals to effectively adapt to the various developmental periods. For example, Bostik and Everall reported that insecurely attached adolescents found difficulties forming emotionally close friendships as compared to their securely attached peers. This is in line with findings from Wiseman et al. which found that undergraduate students who felt that they lacked parental care and nurturance in their childhood had impaired self-concepts. This decreased their abilities to socialise effectively, which subsequently led to a dissatisfaction in interpersonal relationships and increased feelings of loneliness. Similarly, Karababa also noted that adolescents with more positive parental attachments have higher self-esteem, which in turn led them to perceiving themselves as less lonely. The abovementioned findings imply that the quality of early bonds with parents is essential to an individual's development, which eventually influences life satisfaction. --- Social relationships and loneliness In younger individuals, their lives revolve around forming social networks . These social networks can act as a buffer against any ill effects of stress . Peer rejection and decreased social support has been found significantly associated with loneliness since it might make one feel disconnected with the social reality . At a stage when young individuals are expected to form fulfilling social bonds outside of their family, deficient peer relationships may make them vulnerable to physical and psychological health issues . Thus, an individual's early life experiences with social relations contribute to the experience of loneliness and well-being in later stages of life. The experience of loneliness can be minimised or exacerbated from the use of specific coping strategies. It is thus important for the mental health sector to understand what drives loneliness in young people, which prompted the current review. Keeping this as focus, the systematic scoping review was undertaken to examine the extent and range of research activity specifically focused on the experience of loneliness among young people and its subsequent effect on their well-being. --- Methodology Utilizing the widely endorsed six-step framework described by Arksey and O'Malley , the researchers charted the below mentioned steps to identify relevant works to be included in the review. As mentioned in the earlier section, we followed five of the six steps in the framework, choosing not to include the sixth stage -'consultation exercise', which is viewed as an optional component of the scoping study framework . --- Identifying the research question Scoping reviews are undertaken to summarize what is known about a specific topic to date by maintaining a wide approach in order to generate breadth of coverage. . Keeping in mind the objectives of this systematic scoping review we formed a broad research question which was: What is known from the extant literature about the experience of loneliness among youth and its subsequent effect on their well-being? Through this process we aimed to identify factors that could potentially influence loneliness in youth and we were also keen to know if those works focused on studying the effect of loneliness on well-being in youth. --- Identifying relevant studies Since scoping reviews are comprehensive, the search for the relevant works was done via multiple sources. To guarantee adequate and efficient coverage, specialized and interdisciplinary electronic databases utilized in this review were APA PsycINFO, Scopus, Emerald Insight and OneSearch . Grey literature, which might not be well documented or easily obtainable , systematic and literature reviews and intervention studies were excluded. Additional searches included checking the reference lists of all included papers. We also carried out citation searches of key publications to identify subsequent publications. --- Study selection The team members independently searched for relevant works but the exploration of the available works yielded a large number of irrelevant studies. Discussions based on the initially reviewed articles were undertaken to draw conclusions and direct subsequent eligibility criteria. This criteria was refined as the review progressed and we learnt more about the existing body of works. --- Eligibility criteria This review utilised the following inclusion and exclusion criteria: in order to be included in the review, the paper had to report loneliness among youth, either as an independent variable , or as a dependent variable . The review followed the definition of "youth" as individuals aged 15 to 24 years old [World Health Organisation , 2021]. Eligibility was limited to full-text, qualitative or quantitative research published within the last 10 years in a peer-reviewed English language journal. A 10-year time frame was used since boom in social technologies in the last decade have been documented to increase loneliness in modern society . The ten year period would also ensure that the information extracted from the studies would have relevance in the current times . Since many works on loneliness have already focused on older population and there is a paucity of works examining the experience of loneliness among youth, the review focused on young people within the community. Papers utilising samples with pre-existing mental health issues were excluded as the researchers wanted to investigate loneliness and related factors within the general population. Though this review focused on youth, publications that included adult populations were retained if the results offered retrospective explanations on the risk factors of loneliness during the childhood or adolescence period . The review was conducted between 20 October 2020 and 30 March 2021. Based on the 10-year timeframe chosen as a criteria, initially papers published between 1 January 2011 and 28 February 2021 were included in the review. However, due to COVID-19, the research team was delayed in meeting time target for wrapping this paper. So another search for databases was done and the search date for related papers was extended to December 2021 to include more works. Search terms included Loneliness, Well-Being, Youth, Relationship, Social Support as keywords that were then combined using the Boolean operators "OR" and "AND" to merge search terms. For instance Social Support AND lonel* AND well-being; Youth OR Students AND lonel* AND well-being and; Youth OR Adolescents AND Lone*; Relationship AND lonel* AND Youth. We screened titles and abstracts to determine whether the study met the inclusion criteria for this review. --- Results --- Data extraction and collection Database search yielded a total of 1260 papers. The journal articles were manually exported from the databases and imported into Rayyan QCRI , a webbased application for database management of systematic reviews. After removing duplicates a total of 1040 records were left for screening. Before starting the screening process, a benchmarking exercise was done to establish reliability with the blind function on. The reviewers first gave their decision on 10% of the papers and the first reviewer then turned the blind function off to reveal the decisions to the team. This allowed the team to see how many papers were unanimously agreed or disagreed upon. This initial phase of quality check when searching and selecting relevant studies is a crucial aspect of conducting a systematic review since it ensures that expended time and effort is not wasted. It also standardizes the whole process and makes it more transparent and replicable . The benchmarking exercise showed 80% concurrence rate. The reviewers then discussed the reasons for their conflicting decisions. Following this, the inclusion and exclusion criteria were further refined and made more explicit and clearer. Once the benchmarking exercise was complete, reviewers proceeded with abstract screening. Application of the eligibility criteria identified 167 articles for full-text review. Any articles that were disagreed upon were re-read by both reviewers and discussed for relevance. Eventually, 20 papers were retained for the current review. The study selection process is summarised using a Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram . --- Chartering the data The data from the selected articles were organised into a table in Microsoft Word, with the authors' names arranged in alphabetical order . The table columns comprised of author/year, country in which the research was conducted, sample demographics , research aim, study design, measures, and relevant findings from the study. --- Geographical location The geographical range of the 20 studies included in this review was limited . The studies took place in 13 out of the 195 countries listed by the Global Health Security Index . Seven of the 20 studies were carried out in Europe and United Kingdom , while four studies were conducted in the United States of America . In Asia, a three studies were conducted in China , and one each in South Korea , Pakistan and Indonesia . Two studies were carried out in Turkey while one study did not indicate the geographical location of participants. --- Research designs and samples Two major methodological patterns emerged in the included articles . Firstly, majority utilised a quantitative research design and two used a longitudinal approach . Two studies utilized qualitative approach and two made use of mixed-methods research design . Secondly, we found that 11 of the 20 studies recruited a mix of adolescent and young adult samples while of the remaining nine studies, four focused on only adolescent samples , and remaining on university students. The sample sizes for the included works ranged from 166 to 5883 participants. All studies included both males and females in their samples. --- Tools used to measure variables The included studies in this systematic scoping review used self-report measures of loneliness and well-being via questionnaires, interviews, diaries, focus groups and online forum threads. Studies defined the concepts of loneliness and wellbeing in similar ways. Loneliness was generally characterised as an unpleasant experience from the lack of quality and/or quantity of social relations. Though the authors were in agreement on the definition of loneliness, different studies chose to focus on different aspects of loneliness. Such aspects included feelings of neglect or disparities between the past and expected social inclusion. Majority of the studies utilised the University of California, Los Angeles Loneliness Scale or one of its versions as a measure of loneliness. Most studies defined well-being as a multidimensional concept which included notions such as self-esteem and life satisfaction. Rosenberg-Self-Esteem-Scale was the most commonly used tool , followed by Warwick-Edinburgh Mental Well-being Scale , 2013), Subjective Happiness Scale and Satisfaction with Life Scale . --- Findings This section consolidates themes that emerged from this review. Multiple factors such as parental bonding and relationships, personality, social support and stress emerged as significant contributors to experience of loneliness among youth, which in turn impacted their well-being. --- Loneliness as multidimensional construct Though most of the existing works approached loneliness as a unidimensional experience, a few studied loneliness as a multidimensional construct. Teneva & Lemay in their study examined the different impact of trait and state loneliness on self-esteem and affect. Their results showed that both trait and state loneliness could distort a person's vision of future and lead them to expect and experience worse outcomes. von Soest et al. conceptualized loneliness as having social and emotional facets. While emotional loneliness increased across adolescents and young adulthood years, social facet of loneliness plateaued when individuals reached their mid-20 s. Özdogan examined emotional facet of loneliness further by breaking down into romantic and family loneliness. Their study found subjective well-being to be negatively correlated to romantic and family loneliness. Increase in social and emotional loneliness predicted meaninglessness of life. --- Parental bond as a protective barrier against loneliness Positive parental bonding was frequently reported as a protective barrier against loneliness among young people . Notably, these studies explicitly reported the strong influence of past memories of nurturing and loving parents on the individual's current state of loneliness and well-being. von Soest et al. noted that perceived parental support predicted higher autonomy and relatedness in adolescents and young adults, which, in turn, predicted lower loneliness. Weisskirch too argued that young adults' experience with parenting influenced their ability to achieve intimacy resulting in personal enhancement associated with well-being. Thus, close bonding with parents, albeit indirectly, in a way protected the young adults from negative psychological outcomes emerging out of loneliness. Young adults who perceived their parents as more supportive of their sense of relatedness reported higher levels of connectedness to others, which in turn negatively associated with loneliness . Their study highlighted the relevance of supportive parenting for psychological well-being in emerging adults. --- Social support as an inhibitor of loneliness Majority of the included studies gave evidence for the inhibiting effect of social support on loneliness . These studies frequently implied that social support from family and friends was effective in decreasing the experience of loneliness. According to Inguglia et al. , a sense of relatedness to other people acted as a protection against feelings of loneliness. Social support was a significant mediator between interpersonal relationship and loneliness, specifically in the school setting . Not having close friends in adolescence was related with higher experience of loneliness . According to them, this could influence the future social trajectory of these individuals, where they could potentially lack friends and support network during young adulthood stage. Presence of carings realtionships resulted in sense of worthiness in adolescents and youths and reduced their need to daydream as a way of coping with loneliness . Having more friends on Facebook resulted in lower levels of loneliness and higher subjective happiness . Bonding relation and social support were strongly and negatively related to the level of loneliness in college students . This was similar to findings by Yousaf et al. , Kong & You, , and Janta et al. , who too found that perception of social support was negatively correlated with loneliness among undergarduate students, with those low in social support experiencing more loneliness. Social support as a stress buffer against loneliness varied for different sources of support . For college aged youth, only support from friends and romantic partners negatively associated with loneliness. Family support was not associated with loneliness. Quality of realtionship with teachers and same sex friends too reduced the experience of loneliness in adolescents . --- Social media, loneliness and wellbeing Of particular interest, social media was found to influence the pathway between loneliness and well-being. However, there were mixed findings regarding its influence. For example, Park and Lee's study found that the use of smartphones for supportive communication among university students improved social bonds and psychological well-being. Janta et al. found that online social interactions could act as a form of social support. They reported that doctoral students utilised online social platforms as a means of escape from their experience with loneliness in school. Also having friends on social networking sites acted as a buffer against loneliness. Phu and Gow found that having greater number of Facebook friends positively correlated to subjective happiness and negative correlated to loneliness. However, their study found no associations between loneliness and the time spent on Facebook among young adults. These findings have implications, especially in the present times, when COVID-19 driven changes to social interactions encourages us to use social networking sites more often as compared to pre COVID-19 days. Not all time spent on such sites is fruitful and contributing to our well-being. Outcomes from the review showed that Internet and social networking sites could both be tools to forge and nurture online relationships as well as lead to risky, impulsive, or excessive usage in lonely users. Fulfilling engaging relationships not only reduce loneliness but also can 1 3 help control dysfunctional online behaviours like problematic or addictive internet usage that can result in social, emotional, physical, or functional impairment . --- Loneliness, well-being and mental health Feeling lonely can have debilitating consequences for well-being and mental health. In adolescents, where on one hand sense of belongingness acts as a buffer against emotional and behavioral problems, loneliness has adverse impact on psychosocial development. Relationships play a significant role in our life and this is true for the youth as well. Loneliness has shown a strong effect on well-being, by increasing the absence of well-being and decreasing positive well-being. Arslan found that adolescents who had low levels of inclusion at school reported greater feelings of loneliness and mental health problems. In their study, loneliness also had a negative association with subjective well-being. Similarly, being lonely was associated with lower life satisfaction in young adults as well . Loneliness was negatively related to affective well-being and psychological well-being . The COVID-19 pandemic isolation generated strong feelings of emotional distress among a sample of university undergraduates in Indonesia who reported confusion, anger, frustration and unceratinity about their future . Loneliness was highlighted as a significant risk factor for mental wellbeing and self-esteem in Nordic adolescents as well . Reviewed works also showed that loneliness is associated with an increased risk of certain mental health problems, including depression and anxiety , increased stress , diminished self-esteem , and less positive affect . --- Discussion The two questions that directed this systematic scoping review brought attention to the relationship between loneliness and well-being, and how this pathway is affected by certain key factors identified in the literature. In summary, this review of 20 articles reinforces understandings of the experience of loneliness as a complex, evolutionary process that is influenced by relational and environmental factors. It further suggests that there are strong similarities between the experience of loneliness in different parts of the world. Insights can be concluded from the current review about the experience of loneliness on well-being among young people. --- The importance of relational support It is noteworthy that the included studies reported a high number of loneliness-prevention processes at the relational level of young people. Acknowledging this fits with the argument that positive social resources, including friends and family, can play a crucial role in abating loneliness. In other words, the attempts by mental health practitioners to tackle the issue of loneliness among young people should consider the relational factors. To this end, loneliness interventions should not prioritise only relational processes over other resources , or vice versa. This could be tricky, given the fact that close relationships have been recognised as a protective factor . Nonetheless, tapping on the bidirectional nature of relationships, mental health practitioners could enhance the support given to young individuals and their social systems in learning ways of advancing loneliness-prevention exchanges. It should be noted that although the relational resources were relatively similar across the included studies, applying this notion to specific contexts should be done with caution. To illustrate, significant ethnic differences were found in the relationship between parental bonding and loneliness . This means that appreciation of intimacy between the parent and child may not resonate with all across cultures. The cultural orientations or ethnic differences may cause the child to prioritise other external relationships over parental relationship to deal with the loneliness. Thus, mental health practitioners should caution against the assumption that a specific relational resource is ubiquitously protective in all young people. Instead, they should target the protective relationship that is relevant to that particular young individual. In doing so, practitioners enhance collaboration between them and their clients in the therapeutic process, and increase their chance at success by exploring acceptable means of moderating the challenges that a specific context imposes on the individual. --- Impact of loneliness on well-being and mental health of youth This scoping review identified the adverse effect of loneliness on emotional health of young population. Loneliness resulted in not only poor mental health but also significantly lowered positive well-being among youth. Experiences of diminished life satisfaction, lower affective and psychological well-being, among lonely youth, thus necessitates more focused research on loneliness experiences among the young and adolescent populations. Reduced or limited access to friendship groups or peers have a psychological cost for youth as many participants in the included works showed symptoms of poor mental health including anxiety and depression. With world now emerging from the shadows of CoVID-19 pandemic, there is an urgent need to focus on the emotional well-being of the young population. The debilitating effects of loneliness among youths have to be acknowledged by the society in order to start conversations on this not-so-trivial issue. --- Contributions of social media in current times Apart from the importance of relational support for managing loneliness among youth, this review also identified the contributions of social media as a tool for social support. The fact that the loneliness of young people was intertwined with social media usage could be due to globalisation and boom in social technologies. The use of social networking sites was means of escape from reality but this was also linked to a reinstated sense of happiness and belonging . Despite the contributions of social media usage in decreasing loneliness, individuals may not reap equivalent benefits at par with face-to-face social support or interactions. Significantly, one study emphasised the importance of in-person social interactions. Kekkonen et al. noted that lower frequencies of physically meeting friends significantly decreased life satisfaction among young adults. These findings gain significance in light of the social distancing measures in place across many countries to prevent the spread of CoVID-19. In such circumstances, social media could offer a safe alternative and be a valuable platform for lonely individuals seeking social support and interactions. In present times when COVID-19 driven restrictions might force people to limit their face-to-face interactions, social media could be an effective tool to manage the consequences of loneliness. What stood out in the review was limited works examining the experience of loneliness in youths. The body of reviewed works showed that the youth too are vulnerable to experiencing loneliness and suffer its consequences. Some parts of the world are still focused on fighting COVID -19 by imposing restrictions on social gatherings and limiting social contact if need be. Since youths thrive on social interactions and engagements, the impact of these restrictions on their emotional health might be more severe than other demographics within the society. Youth are our future and we need to now focus on their experience of loneliness. Studies focused on examining and understanding their experiences are required so that solutions can be offered to mitigate the effect of loneliness in order to ensure that this valuable demographic does not suffer silently. --- Limitations This review is not without limitations. Firstly, logistical constraints limited the ability to consult with relevant stakeholders . Given the vast research conducted on loneliness and well-being, as well as the dynamic nature of social media and parent-child relationships , this review will require an update. Ideally, it should be done with the consultation of relevant stakeholders. Secondly, the search for articles excluded non-English language journals. This creates the possibility that the current review could have reduced relevant insights on the topic of interest in other languages. Thirdly, the methodology did not focus on quality assessment of the included research papers. This is reflective of the scoping review approach . Nonetheless, quality assessment should be considered in the future, especially if the findings from this review are going to help direct future interventions. Lastly, this review only focused on empirical works and did not include literature reviews, systematic reviews, interventions and grey literature. Due to not including all available data on this topic, a selection bias might have occurred. Since for this review only four databases were utilised, with searched articles limited to the last 10 years, these criteria may have limited the scope of findings. --- Conclusion There is a rising public health concern worldwide due to the social distancing measures that were brought in to mitigate the effect of COVID-19 pandemic. In this systematic scoping review, information was gathered to shed light on the current evidence related to the experience of loneliness in young people and how this affects their well-being. Essentially, their capacity to ameliorate loneliness is rooted in their early relationships with parents, their social support and the ways in which they choose to cope with the loneliness . Understanding that these factors feature in empirical accounts of adolescents' and young adults' loneliness is vital to the mental healthcare sector's efforts to tackle this problem. Emerging adulthood is an important transition period with focus on growth, gains, and planning for positive future. However, COVID-19-related stressors and disruptions led to social isolation, confinement, loss of freedom, and a sense of uncertainty that significantly impacted the mental health and psychosocial well-being of youth. This systematic scoping review did not find many works that focused on the experience of loneliness among youth in community. Studies that measure loneliness and wellbeing at different time points appear to be need of the hour. This will allow researchers to assess the temporal unfolding of loneliness at different intensities of the imposed social distancing measures. Studies need to be replicated and utilise more rigorous designs to validate the results from the included works in this review. By addressing these factors, researchers can help facilitate better mental health in the young population. --- Appendix B --- Data availability The datasets generated during and/or analysed during the current review are available from the corresponding author on reasonable request. --- Appendix A Figure 1 Figure 2 --- ---

A systematic scoping review was conducted to explore the current evidence on the experience of loneliness influencing well-being among youths. The electronic databases Scopus, APA PsycINFO, Emerald Insight and One Search were used to identify relevant studies, followed by an analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. Reference lists of all shortlisted articles were searched for additional studies. 20 studies (quantitative, qualitative and mixed) published in the English language were identified for inclusion. Findings illustrate that the experience of loneliness is a complex, evolutionary process influenced by relational and environmental factors. Results from the studies identified factors that promote lower experience of loneliness and better well-being in future life stages. Future research can substantiate the issues related to young individuals being socially isolated from others for a prolonged duration.

Introduction As government has widely adopted electronic government [14,39], a growing body of literature has paid attention to e-government as a means of fostering citizen participation in the government agency policy decision-making processes [27, 29,57]. All levels of government agencies have used various Web technologies to offer various forms of electronic participation applications ranging from simple online voting to online policy forums . However, little is known about the driving forces of citizens' use of e-participation. Why do some citizens more actively use e-participation than others? This is a crucial question because the full potential of eparticipation cannot be achieved unless it is actively used by citizens. As discussed later, for analytical purposes, this research examines the e-participation domain as citizen-initiated participation in the phase of policy agenda settings where citizens post their input and views as well as comment on the input of others and government responses. Citizen participation literature in public administration has offered meaningful insight on why citizens actively use e-participation to engage in policy decision making process. Some focused mainly on the effects demographics and psychological factors of citizen participation [56] while others examined the relationship between citizen participation management and citizens' satisfaction [20]. Although prior studies have developed normative arguments on the importance of citizen participation design principles such as fairness in the participation process [61,42], a few have systematically examined the role of design principles in affecting citizen participation. Moreover, as discussed later, given the fact that one feature of citizen-initiated e-participation is online community, this research argues that social capital literature can provide a useful theoretical framework to better understand e-participant's behavior. To fill the research gap, this study develops a model of active e-participation by focusing on the role of individual social capital and e-participation management. To test the research hypotheses, the study uses the 2009 E-Participation Survey data collected from 1,076 eparticipants of an e-participation program called Cheon Man Sang Sang Oasis run by the Seoul Metropolitan Government in South Korea since 2006. The Oasis provides various e-participation services including online polling, online surveying, and online policy forums. This research focuses on one particular type of e-participation service, the online policy forum, which has served as an online community for citizens to participate in policy decision-making processes, especially policy agenda setting processes. --- E-Participation: Definition and Scope Scholars in public administration have attempted to define citizen participation and classify it in order to systematically understand its characteristics [1,30,37]. First, this research views e-participation as a special type of e-government service in that e-participation is available at government websites or as part of e-government services run by government agencies. Second, Macintosh [37] refers to e-participation as the use of web technologies to provide information and to support "top-down" engagement, or to foster "ground-up" efforts to empower citizens to gain their support. In this research, we focus on "ground-up" e-participation, which emphasizes participation initiated by citizens, as opposed to participation initiated by government . The nature of citizen participation varies depending on who takes the initiative of selecting and suggesting a policy agenda discussed during citizen participation processes [67,57]. Oasis provides both government-initiated and citizen-initiated e-participation services. Government initiated e-participation includes, but is not limited to, online polls, online surveys, and agenda-specific online discussion boards . Meanwhile, citizen-initiated eparticipation services range from email contact [67] to online policy forums. This research focuses on citizeninitiated e-participation, specifically online policy forums designed to provide citizens with an opportunity to initiate inputs about SMG's public policy and day-to-day operations. Third, public administration literature has characterized citizen participation as part of policy decision-making processes including policy agenda setting, policy formation, policy implementation, and policy evaluation phases [1,30]. This research focuses on citizen participation in the policy agenda setting phase because it reflects authentic participation, arguing that participation should be sought, at least in the early stage of the decision-making process, before any decisions are finalized [30]. Lastly, White [62] refers to e-participation as "the use of information technologies to engage in discourse among citizens and between citizens and elected or appointed officials over public policy issues ." This definition broadens our understanding about eparticipation by incorporating discourse among citizens. That is, e-participation serves not only as a communication channel which e-participants express themselves, but also plays as an online community. Citizen-initiated e-participation often involves a deliberative communication mode in which citizens enjoy many-to-many communications. This implies that eparticipants not only communicate with public administrators, but that they also observe, make comments on the input of others as well as respond to the comments of others. Through ongoing and repeated interactions, e-participants, as members of an online community, can build online networks which serve as opportunities or constraints for e-participants to create social ties with others online. For analytical purposes, this study defines eparticipation as e-government applications designed to promote citizen-initiated participation in policy agenda setting and to build online community providing citizens with an opportunity to discuss policy agendas with others and with government agencies. The scope of this research is limited to the online policy forums of Oasis as eparticipation run by SMG in South Korea where citizens are allowed to initiate discussion about SMG policy agenda by posting their ideas and making comments to the ideas of others. --- Theoretical Framework and Hypotheses Citizen participation has been considered as mechanisms for creating democratic values and instrument values [41]. Considering the working definition of e-participation discussed above, this research reviews and discusses core components of citizen participation in order to develop a model of active eparticipation. One stream of citizen participation studies concerns itself with the relationship between citizens' socio-economic background and their participation in public administration [56,6]. The importance of individual demographics has received attention by early studies on citizen participation in urban planning and government budgeting process. Some scholars have paid more attention to the role of socio-economic status in citizeninitiated contact [56]. Recently, this stream of research has stretched its focus to understand how socio-economic variables affect e-government visiting as a form of citizen-initiated contact [57]. Another stream of research has focused on individual citizens' psychological factors [16]. The psychological factors include different types of self-efficacy , prosocial behavior, needs, and personality. For example, early research on citizen-initiated contact found perceived citizen needs, political efficacy, and social involvement influence citizens' contact of government bureaucrats [56]. In a similar vein, the Technology Acceptance Model [15] has been widely applied to understand individual's use of new technologies such as e-participation [27]. Since e-participation relies on web-based applications as a technological platform, the use of e-participation can be understood as a technology adoption by citizens. The essence of TAM is that individual adoption of IT application depends of his or her intention to use that application, their perceived usefulness, and the ease of use of that IT application [15]. The social capital stream of studies was also reviewed. Scholars in social science have mainly studied social capital in the context of local and urban communities [47]. Given the fact that e-participation is viewed as online community, social capital literature helps us identify what factors influence citizens to engage in e-participation. Social capital studies argue that the success of community often depends on the degree of community's social capital . Empirical studies have supported that social capital is a crucial asset for all levels of government to implement policies effectively, to provide high quality services, and to make governments' innovation efforts more feasible and legitimate [31,47,48,50]. This line of thought implies that certain characteristics of social capital are related to active e-participation. However, little is known about the role of individual social capital built in an offline community in promoting citizens' eparticipation as a means of shaping online community. Lastly, the design and management of citizen participation programs has long been discussed among scholars in public administration [42,61]. Several principles have been identified as critical factors in facilitating citizen participation and managing programs. For example, Webler and Tuler [61] applied Habermas's theory of discourse to understand two broader principles of citizen participation process -fairness and competence -in the context of citizen participation in public policy decisions on forest use in northeastern states. However, a few have systematically and empirically examined the role of design principles in affecting citizen participation. By focusing on the role of social capital and citizen participation design, this research suggests a model of citizens' active e-participation. Our study asserts that three dimensions of individual social capital are associated with their active use of e-participation. Also, we argue that three components of e-participation management are related to active e-participation. --- Individual Social Capital Scholars in public administration have often considered social capital as a collective concept. This research, however, discusses social capital at the individual level because it ultimately belongs to individuals [7] and because it views social capital as an antecedent of individual behaviors such as citizen participation [19]. Although there is no clear agreement on the definition of social capital in the literature, many scholars [12,48] agree that social capital consists of at least three key dimensions: trust, social networks, and civic norms. --- Trust in Government The definition of trust in government varies. Here, trust in government is broadly defined as the extent to which citizens believe that government works in their best interest [10]. When citizens do not trust in government, they are likely to perceive that government policies are harmful, to distance themselves from government, to resist government policies and programs, and to lower their expectations of how government will treat them in the future [28]. Such cynicism toward government tends to decrease citizens' interests in participation in public administration [28,4]. Meanwhile, citizen's trust in government signals that government will be responsive to their needs and care for their best interests. Also, trust in government reflects citizens' willingness to comply, cooperate, adopt, and support government policies and innovative programs [13,3]. For example, research found that citizens' trust in government increases the possibility of adopting innovative e-government services [3]. Moreover, when citizens trust government, they are likely to show greater interest in government. Thus, given the fact that citizeninitiated e-participation often requires citizens' commitment to participation in public affairs, their willingness and interest can be expressed as a form of active participation in policy decision-making processes. Competing arguments are possible. That is, it is likely that trust in government reduces the citizens' demands for monitoring government, which in turn, weakens the strong incentives of citizen participation. This perspective, however, may underestimate various motivations driving citizen participation. As discussed later, citizens are motivated by not only a sense of ownership, but also by social norms of cooperation and prosocial behavior. One may argue that there may be an inverse relationship between trust and participation. That is, it is likely that citizens who actively participate in government put greater trust in government [49]. However, citizen participation may not directly increase trust in government because active participation does not necessarily represent that citizens are supportive toward government. Rather, it is reported that citizen participation negatively affects trust in government [35]. Also, a recent empirical study reveals that there is no direct relationship between e-government use and trust in government, which implies that citizen participation affects trust through the management of citizen participation process [29]. Hypothesis1: E-participants' trust in government is positively related to their active e-participation. --- Strength of Social Ties Social network literature considers social networks as resources in that people can access information, gain social support, and receive recognition through their social networks [21]. In particular, the strength of ties has been discussed to understand the characteristics of social ties [21,33]. Strength of ties is a multidimensional concept [21]. This study defines tie strength as the extent to which individuals frequently interact with other social groups. Advocates of strong social ties argue that people embedded in strong social networks enjoy benefits in terms of accessing information, exchanging social support, and receiving recognition easily and promptly [33]. However, people connected through strong ties tend to share similar information, face higher dependency, and spend more resources to maintain strong ties [9]. Proponents of strength in weak ties emphasize that weak social ties provide people with an opportunity to access diverse social groups thereby helping them gain nonredundant and new information, to enjoy autonomy, and to manage them with a lower cost [21,9]. Considering that e-participation serves as online community, this research asserts that e-participants' offline social ties affect e-participation use because offline social ties act as incentives to build online social ties. E-participants are limited to developing and sustaining strong online ties with others unless they actively engage in online community. Meanwhile, active e-participation offers citizens the opportunity to increase the visibility of their contributions, recognition, reputation, and status [46]. That is, e-participants are seen as active when they post more ideas and comments to others. Eparticipants receive different forms of social rewards including attention, recognition, and support from peer eparticipants, but only when they are actively engaged in eparticipation. Frequent exchange of ideas, comments, and responses with other e-participants and government officials helps e-participants build strong online ties. How does the strength of offline ties affect citizens' use of e-participation? This study argues that strong offline social ties are negatively related to active eparticipation. In other words, weak offline ties are positively associated with active e-participation. When people are connected through strong ties in an offline setting, it is likely that they have a limited opportunity to actively use e-participation applications. Early studies found that people tend to spend less time socializing faceto-face when they spend more time online [34,43]. This finding implies that people who frequently meet in social groups face-to-face may not allocate extra time and energy to commit to e-participation. Meanwhile, weak social ties serve as an incentive to use e-participation actively because weak ties increase the possibility that people will spend less time socializing face-to-face, but more time socializing online. But, spending more time online does not necessarily motivate people to use eparticipation actively. It is likely that weak ties offer an incentive to actively use e-participation because active eparticipation enables e-participants to build strong online ties. Moreover, e-participants who enjoy the benefits derived from strong offline ties may consider eparticipation as a complementary means of gaining added value by building weak online ties. When people are weakly tied with social groups in an offline setting they are likely to reap benefits such as access to new and nonredundant information and lower maintenance cost [21,9]. Active e-participation, however, provides an opportunity to build strong online ties, which enables eparticipants to gain complementary resources . Hypothesis2: E-participants' strong offline ties are negatively related to their active e-participation. --- Civic norms of volunteering Civic norms can be broadly defined as group-held beliefs about how members in civic society should behave in public affairs. This research considers civic norms as socially cooperative behavior associated with a more general interest rather than a specific interest associated with a partisan group of people [31]. They can be characterized by a willingness or desire to help others and can be captured by the degree which individuals have affective motives such as volunteerism [16]. As a specific form of civic norms, this research focuses on citizens' volunteering experience. The literature reported a positive relationship between citizens' volunteer experience and political participation [5,63,66]. For example, Wilson [63] addresses positive impacts of volunteering on community participation, civic engagement, and opportunities for professional development. Youniss et al [66] also found a positive relation between youth participation in service programs in high school and their engagement in community organizations as adults. Flanagan et al [17] found that high school students who volunteer are more likely to be engaged in a political campaign. Furthermore, Smith [52] finds that participation in extracurricular activities in one's youth is one of significant predictors of greater political and civic involvement in young adulthood. Scholars address several factors that may facilitate this relationship including the sharing of information [32], the opportunity to develop "civic skills" such as the ability to organize a meeting [60] and the fostering of generalized trust [54]. Moreover, this study argues that citizens' volunteer experience often represents their trait of extraversion, a person's tendency to be social [16], which affects their engagement in online community including e-participation. Hypothesis3: E-participants' volunteering experience is positively associated with their active e-participation. --- Management of the E-participation Process Scholars in public administration have paid attention to the importance of design and management of the citizen participation process and asserted that poor design and management of e-participation processes obstructs citizen participation. [30,61,23]. For example, Halvorsen [23] found that participants who perceive high quality participation program management assess that the agency in charge of managing the participation program was responsive to public concerns. --- Fairness in E-participation Process Scholars address fairness as one of design criteria measuring the quality and effectiveness of citizen participation programs [61,42,11,24]. For example, based on Habermas's theory of communicative action [22], Webler and Tuler [61] propose fairness and competence as core dimensions of developing criteria of desirable process of public participation. Following their study, this study defines fairness as "the opportunity for all interested or affected parties to assume any legitimate role in decision-making process" [61; p.568]. They offered three dimensions of fairness in the process of citizen participation discourse including fair attendance, fair participation in agenda setting and rule making, and fairness in discussion and debate [61]. Scholars have also analyzed a positive relationship between process fairness and outcome satisfaction and acceptance [2,25,51]. Research findings show the positive impacts of procedural fairness on the institutional legitimacy of governmental authorities [2] as well as increased levels of trust in political systems [51]. Other scholars analyzed the impact of fairness of the citizen participation process on citizen support for government decisions [25] and the impact of the use of fair processes on public trust in public officials [58]. Herian et al [25] finds that the inclusion of public input by local governments can increase perceptions of fairness and that the perceptions of fairness have stronger relationships with overall governmental assessments for those who are relatively uncertain about a governmental institution. Van Ryzin [58] found that the use of fair processes by public servants increased the public's trust in those officials. The study proposes a positive relation between perceived fairness of e-participation process and citizens' active e-participation. In order to explore the relationship between perceived fairness of e-participation management and citizens' active e-participation, three aspects of fairness in e-participation process are developed in this study-including availability of diverse participation programs, the equal opportunity for stakeholders and citizens to e-participation, the fair process of eparticipation decision making. Hypothesis 4: E-participants' perceived fairness in eparticipation process is positively associated with active e-participation. --- Access to information One normative argument about the design of the citizen participation process highlights that limited access to government information and its interpretation prohibits citizens from understanding existing government activities such as public policies and day-to-day operations and thus, citizen participation should be designed to grant citizens access to relevant information and its interpretations about government activities in participation process [61,42]. In a similar vein, advocates of TAM and other scholars imply that the design of e-participation applications must be effective and easy-to-use because the design affects citizens' access to information about the participation procedure as well as government responsiveness [29,45], which motivates citizens' active engagement in e-participation. According to the principal-agent model, the relationship between citizens and government is one of information asymmetry [64]. As the principal, ordinary citizens are often less knowledgeable with regard to government activities than government employees as the citizens' agent. When an e-participation process is designed and managed to enhance citizens' ability to access information of government activities, it is likely that citizens are better informed of what and how government agencies perform. Thus, increased access to information minimizes information asymmetry, which reduces uncertainty and ambiguity about government policy and programs. The decreased information asymmetry can strengthen citizens' capability of understanding government agencies. Knowledgeable citizens are likely to offer useful and helpful suggestions for government agencies to make better informed policy decisions. Thus, it is likely that they make meaningful contributions including posting policy inputs and suggesting ideas about problem identification and solving, and/or innovative proposals. Also, informed citizens are better able to monitor government agencies increasing both the government's commitment to openness and honesty as well as the likelihood that any government deception will be uncovered [65]. Hypothesis 5: E-participants' perceived easier access to policy information is positively related to active eparticipation. --- Responsiveness During the public participation process, government responsiveness has played a crucial role in shaping citizens' perception and behavior toward participation [30,36]. For example, research found that citizens' satisfaction with participation programs is determined by government employees' responsiveness to their needs and the quality of feedback for their inputs [23,36]. As part of management quality, public officials' interpersonal, discourse and facilitation skills have been emphasized as a means of implementing authentic participation programs [30], which require citizens' active participation. Although e-participation is promising, in some ways, it limits the ability of both government and e-participants to interact with each other interpersonally, to engage in verbal communication, and to facilitate discussion in the e-participation process, compared to conventional citizen participation setting. In this regard, management of eparticipation processes plays a crucial role in shaping active e-participation. In the context of e-participation, government responsiveness can be captured by the extent to which public officials provide quality feedback to eparticipants' input and inquiries. As the nature of citizen participation does not bind government decisions, government bureaucrats have no strong incentives to respond to citizens' input and inquiries in a sincere manner. Insincere responses or no responses from government concerning e-participants' input is likely to decrease their interest in e-participation and their willingness to commit to the community through e-participation. As a result, this lack of interest and willingness to participate discourages e-participants from e-participation actively. Meanwhile, it is likely that sincere responses from eparticipation management reinforces e-participants' interests in e-participation and their willingness to engage in e-participation by facilitating their commitment. That is, e-participation management's quality responses promote e-participants' self-esteem by enhancing the sense of importance within and identification with the community [55]. Increased identification often creates a sense of civic duty by motivating the participant to take more interest in community issues. Also, e-participants who receive quality feedback from government officials are likely to perceive that they gain useful policy information that helps them better understand community issues and in turn, contribute to community building. Moreover, online community literature found that quality responsiveness often motivates e-participants to stay longer and to participate in the online community frequently [38]. Hypothesis 6: Perceived government responsiveness via e-participation programs is positively associated with eparticipants' active e-participation. --- Data and Measurement To test research hypotheses, this study used the 2009 E-participation Survey data collected from the citizen members of the Oasis. As of June 2009, 34,792 citizens had joined Oasis. From this population a sample frame of 10,136 citizen members of Oasis who have posted at least one suggestion over the last three years was created. A web-based survey was administered for four weeks in May and June in 2009. Of 10,136 members, 1,076 participants responded to the survey . Because of low response rates, non-response bias test was performed to see if there is difference in demographics between respondents and non-respondents [40]. The results show that the respondents and non-respondents were not significantly different in terms of age, gender, and education. --- Dependent variable Active e-participation. As a measure of active eparticipation, this research employed the number of suggestions posted on the Oasis as a measure of active eparticipation. The survey participants were asked to indicate the extent to which they posted their suggestions on Oasis using five ordered categories ranging from "1-2 suggestions" to "More than 10 suggestions" . --- Independent variables Trust in government. The measure of citizen trust in government was derived from prior research [29,26]. Trust in government is measured by a single survey item rated on a 5-point Likert scale ranging from strongly disagree to strongly agree ; "To what extent do you trust that SMG works in your best interests?" Although the single item is limited to capturing various dimensions of citizen trust in government, it allows us to broadly understand the respondents' level of trust in government. Tie strength. To capture e-participants' strength of social ties, this study used respondents' self-reporting on the frequency of going out with people for socialization. Respondents were asked to indicate how often they go out with five different groups of people for socialization . Five items were designed with a 7-point Likert-type scale ranging from "Every Day" to "Once a Year" . The average scores of the five items were used in the analysis . Social norms. The respondents' volunteer experience is used to measure social norms [8]. The respondents were requested to indicate how often, on average, they have been involved in volunteer work for the past three years. This item was rated on a 7-point Likert-type scale ranging from "Never" to "Every Day" . Fairness in e-participation process. This research used four survey items to measure respondents' perception of fairness in e-participation processes [61,25]. The participants were asked to indicate the extent to which they agree with the four survey items using a 5-point Likert scale ranging from strongly disagree to strongly agree . The mean scores of the four items were used in the analysis . --- Access to information: The measure of access to information was adapted from citizen participation literature [61,25]. To measure e-participants' perception of access to information, this study used four survey items with 5-point Likert scale . These items were summed and averaged into an index. Responsiveness. Responsiveness is measured using the three items representing the respondents' perception of quality feedback offered by SMG. The mean scores of the items were used in the analysis. --- Control variables TAM variables. Two TAM related variables -intention to use and perceived usefulness -were included as control variables [15]. The intention to use and perceived usefulness are found to be associated with acceptance of new IT applications [59]. By modifying Davis's TAM scale [15], we used one item to measure citizens' intention to use e-participation and seven items to capture respondents' perceived usefulness. The seven items for perceived usefulness were summed and averaged into an index . Psychological factors. To control the effect of eparticipants' psychological factors, three variables were included. Political efficacy refers to e-participants' perceptions of influence on governmental decision making. As a political reward, political efficacy serves as incentives for active participation in public affairs [30]. To measure political efficacy, we used the four items using a 5-point Likert scale ranging from strongly disagree to strongly agree . The average scores were used in the analysis . This research used the length of Oasis membership as a proxy for capturing Internet self-efficacy. It is assumed that eparticipants who joined the Oasis since 2006 might be equipped with greater Internet skills necessary to use the Oasis. The item was scaled from 4 years to less than 1 year . To capture respondents' needs of e-participation, this study used a frequency of visiting Oasis sites as a proxy. This item was scaled from less than every six month to more than five times per week . Political Participation. Citizen-initiated contact literature has suggested political participation as one key factor [56,67]. This research used two types of political participation. Respondents were asked to indicate whether or not they voted on four most recent elections. This research combined their responses to four elections and created an index. Also, to control the effects of respondents' involvement in interest group on e-participation, three proxies were added to political participation category. As proxies, respondents' volunteer activities sponsored by nongovernmental organizations , unions, and political parties were included to measure the extent to which citizens as volunteers were indirectly involved in interest groups. Socio-economic variables. Respondents' socio-economic status has been identified as a barrier for to citizen participation [30]. We included gender as a dummy variable . Age was measured on a continuous scale. As a dummy variable, education was included as a dummy variable . Income was measured by households' monthly income with six categories ranging from 1 to 6 . Six income categories were coded as a series of dummy variable where the lowest income level is used as the base dummy. --- Analysis and Findings Descriptive statistics and the correlation matrix show that five independent variables are significantly correlated with active e-participation. However, strength of social ties is not significantly correlated to active e-participation. The results of multicollinearity tests show the VIF did not exceed 3.0 in this model [44], which implies that multicollinearity is not a serious issue. Because the scale of a survey item for measuring active e-participation consists of five ordered categories, an ordered logistic regression model is employed to estimate the effects of independent variables. Table 1 shows the results of the regression analysis. Three social capital hypotheses are supported by the data. Consistent with H1, the results demonstrate a positive and significant association between trust in government and active e-participation . That is, eparticipants with greater trust in government tend to post a greater number of suggestions on Oasis. As expected, H2 is supported by the data . That is, eparticipants who maintain strong offline social ties tend to post a smaller number of suggestions on Oasis. In other words, weak offline social ties are positively associated with active e-participation. The data support H3 as well . Citizens are likely to post more suggestions when they have been frequently involved in volunteer activities. Unlike the significant effects of individual social capital, three e-participation management hypotheses are partially supported by the data. The data does not support H4 and H5. The findings may imply that fairness in the participation process and information access do not facilitate e-participants' motivation to actively engage in e-participation. However, as expected, H6 is supported by the data . That is, e-participants tend to post a greater number of policy and management suggestions on Oasis when they receive sincere and useful feedback or they observe other participants receiving quality feedback from SMG employees. Several control variables reveal significant relationships with active e-participation. Of the two TAM variables, intention to use is positively associated with active e-participation while perceived usefulness is not. Among the three psychological variables, Internet self-efficacy and need of e-participation are statistically significant and positive, but political efficacy is not significant. Of the two political participation variables, voting participation is significant , but the three dummy variables related to involvement in interest groups are not significant. Age , education , and most income variables are significant while gender is not significant. --- Discussion and Implications The study results suggest that all three dimensions of individual social capital play crucial roles in shaping active e-participation. As discussed earlier, one could argue that when citizens have greater trust in government, they are less likely to engage in government-initiated citizen participation because trust in government decreases their motivation to monitor government actively. However, these findings suggest that trust in government facilitates citizens to actively engage in citizen-initiated eparticipation because trust in government encourages citizens to have a sense of cooperation with government and in turn, take appropriate actions. Another important finding implies that weak offline social ties, rather than strong ones, promote active eparticipation. This finding implies that weak offline social networks can serve as an incentive because active eparticipation provides an opportunity to build online social networks as a complementary means for mobilizing resources. Thus, when e-participants embedded in weak offline social networks actively engage in e-participation, they are likely to gain the complementary benefits from online social networks , which serves their interests by reinforcing their resources. Results of the study also suggest that there is a positive relation between volunteering and citizens' active e-participation. The finding supports prior studies on a positive relationship between citizens' volunteer experience and political participation [5,63,66]. The result implies that citizens' volunteering experience matter for their active citizenship behavior of participation in local community concerns. Concerning the design and management of eparticipation programs, this study's findings show that fairness and access to information in participation process are not related to active e-participation, which is not consistent with citizen participation literature emphasizing the design principles of citizen participation [61]. The inconsistency might be related to the nature of the citizen participation initiation and channel. Citizen participation literature has mainly been constructed on the basis of government-initiated citizen participation in an offline setting, which creates, by its nature, some barriers to citizen participation. When citizen participation is initiated by government, it is likely that citizens face information asymmetric circumstances because the government, as an agent, has more information than the citizens who lack information related to key issues of the purpose of participation . E-participation has been touted as a means of lowering the physical and psychological barriers of conventional citizen participation [57]. Because of much lower cost for both entering and leaving e-participation sites, there is no strong economic incentive for eparticipants to remain with the site. It is much easier for e-participants to leave e-participation. For example, consider a town hall meeting as a type of offline citizen participation. When citizens participate in the town hall meeting, it is not cheap for them to attend, continually pay attention, and commit to the meeting. Because of high opportunity and transaction cost, they might be more concerned about how government fairly treats them and provides the necessary information. However, eparticipants may not be concerned about fairness in the participation process and information access because of the lower opportunity and transaction cost for them to engage in e-participation. Lastly, the findings imply that government responsiveness-quality feedback-matters for facilitating active e-participation. The importance of quality feedback as a facilitator of e-participation is consistent with both conventional citizen participation literature [30,36] and online community studies [38]. Also, this finding supports a normative argument of the role of "listening bureaucrat" in enhancing responsiveness in public administration [53]. --- Conclusion While web-based e-participation programs have been championed as a crucial tool for e-government to facilitate citizen participation, there have been limited efforts to analyze the driving forces of active eparticipation from e-participants' perspectives. Active eparticipation in local governance could matter for effective and transparent decision making and problem solving in local governance. This study proposed a model of active e-participation and tested the model using the survey data collected from the residents of Seoul who have hands-on experience with e-participation run by SMG. This exploratory study contributes to eparticipation literature by uncovering both social capital and e-participation management factors affecting citizens' active e-participation. Also, the study findings suggest that local governments pay more attention to the role of government in facilitating individual social capital as a facilitator of active e-participation and in building effective design and management systems of eparticipation enhancing government responsiveness to citizens' input. At least, three limitations should be noted: external validity, cross-sectional research design, and online social networks. The results of this study could be outcomes of unique citizen engagement evolution that are affected by South Korea's historical, political, and cultural contexts. Accordingly, more in-depth case studies in various regions and countries are needed to advance active eparticipation models in local governance. Also, we suggest longitudinal research in order to better understand the factors affecting the sustainability of active eparticipation. Since this study analyzed the role of offline tie strength on active e-participation, it is limited to understanding how online social networks affect eparticipation behavior. Therefore, future studies are needed to extend the study model by incorporating the role of online tie strength.

Oasis, an e-participation program administered by the Seoul Metropolitan Government in South Korea. Using ordered logistic regression analysis, the study finds that active e-participation is positively affected by citizens' trust in government, their volunteer experiences, weak offline social ties, and perceived quality responsiveness during the e-participation process.

INTRODUCTION Despite the importance of this topic, that is, the need to explore access to health for quilombola communities, especially during the pandemic, there is no major academic focus on the issue. Existing studies on access to goods and services by quilombola communities, carried out before the COVID-19 Pandemic, demonstrate that these groups live in very precarious contexts and in a total process of exclusion. As Gomes et al. , although the use of health services, the result of a broad and complex set of determinants that include factors related to the organization of supply, the sociodemographic characteristics of users, the epidemiological profile and aspects related to service providers, the use of These services still demonstrate extensive inequality, especially in relation to groups that, historically, experience complete exclusion from the provision of social services. Access to healthcare by the Brazilian population is a subject of concern and constant examination by managers. There are many reasons responsible for the lack of adequate provision, such as: lack of places for care, intense demand, lack of adequate infrastructure, lack of planning and strategies linked to preventive medicine, a process known as bureaucratic insulation, etc. As if that were not enough, the aforementioned difficulties are expanded in number and intensity when we are talking about access to health by Brazilian quilombola communities, since, added to the list of obstacles, is the existence of what is conventionally called "Structural Racism"., responsible for leading and keeping these communities apart, relegated to isolation from indispensable public policies, which ends up shamefully segregating even more a descendant group of people . In this sense, this article was produced through documentary and bibliographical research. The analysis of the material found was based on the "historical-dialectical materialist" method, with the perspective of analyzing the phenomena based on the history of the subjects involved. In addition to the Brazilian Federal Constitution, International Conventions brought the intention of eliminating any prejudice related to these groups, such as the 1965 Convention on the Elimination of all forms of Racial Discrimination which states, in its 3rd article: "The States Parties condemn the racial segregation and apartheid and undertake to prohibit and eliminate all practices of this nature in the territories under their jurisdiction" . Regarding the aforementioned instrument, Galhano adds that it "also seeks to eradicate all forms of discrimination based on race, color, descent or national or ethnic origin, to guarantee the full exercise of civil, political, social and economic, with the application of the principle of equality". Although we recognize the importance of advancing the debate on issues linked to discrimination and racism at the legislative level, social inequalities generated by the issue of race are still a major public health problem, especially when related to access to healthcare for historically marginalized populations, as is the case with quilombolas . --- THE DIFFICULTY OF ACCESSING HEALTHCARE FOR QUILOMBOLA COMMUNITIES IN BRAZIL: THROUGH THE BUREAUCRATIC MODEL OF THE CAPITALIST PRODUCTION SYSTEM A false idea persists in the scope of social relations: that the availability of fundamental rights and guarantees in the constitutional sphere and in ordinary legislation would be sufficient for the full reach and enjoyment of citizens. In Torrens' words, "public policies are a dynamic process, with negotiations, pressures, mobilizations, alliances or coalitions of interests" . Another author recalls that "Politics is conflict, opposition and contradiction of interests. It is also a power relationship; therefore, it often becomes an unbalanced game, between antagonistic social classes" . In turn, the State, in an attempt to mediate between the very diverse interests of these social classes , needs to propose measures that alleviate such gross inequalities, which occurs through public social policies as ways of guaranteeing social security , access to education and health, housing, food and nutritional security, and all sectoral policies linked to the elderly, children, adolescents and families . In the words of Santos , "The State [...] expresses social relations through confrontation, negotiation, co-optation, the search for consensus and agreement, but always under the hegemony of a certain societal project" . And this "societal project" today is the protection of the "capitalist system of production". It is not possible, under penalty of having an uncritical debate, to remove the figure of capitalism as the driving force of this antagonism of interests, this is because social policies emerge as a means of continuous intervention to repair the injustices and oppressions generated when only a tiny part of the The population has the means to produce and most of the financial resources. In the words of one author: Social policies emerged in the capitalist world, from the second industrial revolution, as a strategy of continuous, systematic and structured intervention by the State in the social area, a consequence of the refunctionalization undergone by the State to respond to the monopoly phase of capitalism . With the expansion of the concept of health promotion, now seen not only as what aims to reduce risk, but rather as the set of actions that seek to prevent, recover, and must be perceived in its physiological, psychological dimension, etc., you know-It is clear that all these policies need to be coordinated with each other, meeting all human and social needs as the only way to achieve what is conventionally called good health. According to Jaccould , with the promulgation of the 1988 Federal Constitution, there was a guarantee of universal access to services offered by the State as a way of operationalizing the very idea of equity. In the words of the same author: The third pillar of social security is health policy. With universal access and governed by the principles of equity -serving each person and community, according to their health needs -and integrity, encompassing all types of necessary services, health policy began to be implemented through the Unified Health System. Health -SUS . Even though the Federal Constitution has guaranteed universal access to health, this expansion process cannot be discussed without taking into consideration, all the social phenomena that accompany quilombola populations and which are an inseparable part of the exclusion of which they are victims when it comes to promotion of their rights. Exactly for this reason, Freitas et al. add that: Skin color can be seen as a biological manifestation in the human figure, but it can also mask itself as a racialized expression of biology, when exposed to segregating attitudes within society. The terms race and ethnicity are social categories, more than biological, referring to groups that have a common cultural heritage . What the authors intend, when they argue that "skin color" has a social connotation far beyond biology, is to demonstrate that it is no coincidence that these populations have precarious access to health services, both in basic units and in healthcare units. specialized, and that this marker is connected throughout history with other cultural heritages, fruits of violence and abuse with current consequences. Along the same lines: It is necessary to expand the discussion of the right to health, which is one of the basic premises of the SUS, taking into consideration, that access to it also depends on the social and economic conditions of the population and not just their ethnic condition. But without losing sight of the fact that the universality of the SUS, which would be full access to quality public health services for the entire Brazilian population, has not yet been implemented in practice . According to Rizzotti : "The set of these aspects can only be understood and constituted in a totality, if articulated from the points of view of criticism and history". In the author's words, it is right to mention that problems related to social issues in Brazil need to be seen and worked on in light of a conjunctural dimension, expressed through a concrete totality, in which the State has a primary role in acting in accordance with achieve the reduction of social inequalities imposed by capitalism. According to Rizzotti : It becomes essential to discuss economic development and the characteristics of the Brazilian State, having as a central element of the social issue in Brazil and the relationship between civil society and the State, the specificities of the national bourgeoisie outlined by the dependent economic structure. When we talk about public policies, it is known that social measures still need to be better structured in order to abandon this model of bureaucratization, seen and recognized as a real obstacle to access to public health care in our country . Therefore, it is true to say that the difficulties in accessing fundamental rights do not only occur among the working class, but also among communities that, historically, live on the margins of society, sunk in poverty and social exclusion , fatally wounding what is called universal rights. This is all due to a centralized policy aimed at serving the interests of the capitalist production system . In the words of certain authors: The Brazilian State guarantees by law the right to health of its population, but there is a difficulty in accessing these services which, by analogy to Foucault's theoretical framework, are treated as "ordeals", not as a legal -political form which showed the maintenance in the 16th and 17th centuries but with evidence of the torture as a political and disciplinary agent that aims to control the power exercised by the Brazilian elite and the forms of resistance exercised by quilombola populations in this network of power . We can highlight, for the purpose of exemplifying bureaucratic actions, the enormous difficulty in scheduling appointments by basic health units for reference care, the long queues, the registrations, which are increasingly rigorous, excluding those who are not in the best of luck. of documentation, the division of units into a wide variety of sectors, so that it is difficult to reach the specialist doctor, the long queues for surgical procedures, so that people often die before undergoing surgery. It is clear and necessary to explain that, as previously highlighted, when it comes to the difficulty in accessing quilombola communities to public health services offered by the SUS , the process of bureaucratic insulation is not the only barrier, and it must be highlighted that race and color itself they are also considered obstacles . In the words of the author Silva et al. : Prejudice and discrimination are factors that impact health conditions. In North American literature, it is already clear that inadequacies regarding living conditions, social support, employability, access to food, lifestyle and access to health services are strongly associated with race. According to Almeida et al. , the network of bureaucratization of health services is not a mere technicality or fatality of the system for implementing social policies, but rather a way of guaranteeing the status quo, that is, a way of maintaining control over less vulnerable populations. favored and the power exercised by the elite . In the same vein as Foucault pointed out, regarding torture as a political and disciplining agent that aims to control the power exercised by the elite, it is important to highlight that this population also became the target of disciplinary power . Bluntly, the same author states that: The health conditions and precariousness in which quilombola populations live, characterizing them as a poor population with poor health, refers to the thinking discussed by Foucault about the disciplines and forms of power exercised in the 17th, 18th and 19th centuries on formation of docile bodies that, simultaneously, are useful to the economic system, but are conditioned to less political participation . Following the Foucautian theory, the author still argues that the denial of access to health and other services that legislation has made the responsibility and obligation of the State, in a universal way, towards its citizens, imposes on quilombolas a new deprivation of their freedom, since there is suppression of acquired rights, which is a serious facet of racism in the health system and exposes social and racial inequality . --- THE WORSENING OF ACCESS TO HEALTHCARE BY QUILOMBOLA COMMUNITIES DURING THE PERIOD OF THE COVID-19 PANDEMIC AND THE INCREASE IN CASES AND DEATHS Although legislation has advanced to a certain extent in addressing the universal right to health, in recent years, access to this right by quilombola communities has been worsening and becoming more fragile. In this sense, it is important to remember the public spending freeze proposed and approved during Michel Temer's government, which completely nullified any prospect of future progress in the provision of health services . With regard to access to health services for the quilombola population, including the elderly, studies indicate that: Geographic isolation, service hours, long queues and waiting times have been identified as factors that explain, at least partially, the lower prevalence of use of health services. Likewise, the problems faced in the functioning of the PSF in these communities, such as the high turnover of higher education professionals in rural regions and the precarious infrastructure for providing care, prejudice and dehumanized care on the part of health professionals, can constitute barriers access and use of health services by this population group ). On this subject, Silva et al. brought into their studies recent research from ABRASCO that demonstrated a huge reduction in the action of Community Health Agents in municipalities during the Pandemic, and, for many quilombos, this contact with the aforementioned professional is the only health service they can access. have. The author goes on to point out that: In response to the pandemic, across the country, communities have opted for self-isolation and have adopted autonomous health barriers. People who need to get food, medicine or take relatives to hospitals are advised to follow World health organization guidelines. However, the situation has worsened since Primary Care has always been precarious, with minimal coverage by the ESF Quilombola and the presence of doctors in most communities, where there are many people with chronic diseases such as hypertension, diabetes and sickle cell disease, is sporadic., which require regular monitoring. These people are in the risk group for COVID-19, which increases their chance of dying when having to seek health services in urban areas . Isolation, as an essential trait in many of these communities, is the result of the context of escape, recognition and belonging of their peers who, together, formed true protection networks against the hunting operation institutionalized by the Brazilian State for many years. In the words of Freitas et al. : In quilombola communities, they raise a series of socioeconomic, spatial, legal and cultural issues that are part of the discussion about what contemporary quilombos represent today regarding effective citizenship insertion . In other words, it is necessary to recognize that the Pandemic increased the isolation of these communities, worsened their access to health services, caused a greater lack of professional turnover in family health strategy units, "which expanded the history of social vulnerability to which they have always been subjected, making them especially affected by the effects of the pandemic" . Although there is no official data on the real situation of quilombola communities during the period of the COVID-19 Pandemic in urban Brazil, which only confirms the state of institutional abandonment of these communities, it is already known that there was a large process of underreporting of cases and deaths . However, "despite the lack of knowledge about the real impacts of the COVID-19 Pandemic, a partnership was signed between the Articulation of Black and Quilombola Communities and the ``Instituto Sócio Ambiental`` in the creation of the COVID-19 platform" , brought important data, for example, that, in the month of July of the year two thousand and twenty, there was a period of sharp growth in identified notifications . This assertion is demonstrated in Graph 1, below. Despite all the health problems faced by these communities in non-pandemic times, according to the graph above, an increase in cases and deaths from COVID-19 in quilombola communities can be seen during the most critical periods of the pandemic. "Some studies indicate that the black population in Brazil dies more from COVID-19 than the white population" . According to Professor Unaí Tupinanbás, "the explanation for this difference is social and economic inequality. During the pandemic, inequality was exposed. Mortality among the black population is much higher, not only in Brazil, but also in Europe and the USA" . An example of these supports is shown in Graph 2, below, which has the State of São Paulo -SP as a reference, and provides a comparison between excess mortality by race/ color in SP. --- CONCLUSION Access to healthcare for quilombola communities continues to be a challenge for the Brazilian State. Racial discrimination, perceived as a multifaceted phenomenon, with complex and countless social consequences, forms the backdrop that explains how the power structure of the Brazilian State operates by excluding groups like these. Especially during the COVID-19 Pandemic, as is to be expected in times of crisis, the harmful consequences of racial issues became even more evident when we noticed a significant increase in deaths in Brazilian quilombola communities when compared to the general public. Numbers objectively show that this pandemic has affected more black people than white people and this is not the result of mere chance. Black Brazilians live in a situation of greater social vulnerability, which implies recognizing that they are the majority in Brazilian prisons awaiting trial, they are the biggest victims of police and State violence, they have a lower per capita income rate, less access to education, and they are more likely to die. Due to preventable causes, they have greater difficulties in accessing healthcare. Although we see advances in the treatment of the issue by legislation, with the Federal Constitution of 1988 bringing the precept of universal access to health, with the reception of ordinary legislation that reinforces it, we have not yet found effective official protocols to put this notion into practice, in order to make it possible and fully viable to serve communities historically disadvantaged by the Brazilian State, such as the quilombola community. It is necessary to take health as a broad meaning, without it being perceived as isolated from other dimensions of human care , it is urgent to focus on parallel and complementary public policies that guarantee better living conditions for the quilombola population. Regarding the preventive health model, there is an urgency to improve access, increase teams of professionals visiting these places, create new Family Health Strategy Units , new medium and high complexity care hospitals intended for these populations, increase the number of places available for all medical specialties in the outpatient clinics of public hospitals, in the three spheres of government . The pandemic period showed a reality ignored by many: that the black population is more subject to the adverse events of life. If we do not have serious public policies that address the social dimensions in which the concepts of race and class are entangled, we will continue contributing to a historical debt with no end date.

Nature and Crisis of Hegemony. Class Consciousness and Social Struggles in Overcoming Barbarism. Held from November 16th to 19th, 2021, em Ambiente Virtual. Disponível em: https:// www.joinpp.ufma.br/jornadas/joinpp2021/anais.htmland was also published, as a chapter of a book, in the book named:``Serviço Social e Política Social: desafios da intervenção profissional``. Organizer: Soraya Araujo Uchoa Cavalcante.Atena, 2023.

Background Recent biomedical and political advances in the treatment of HIV have contributed to the realistic possibility for an end to the HIV epidemic. Continued progress toward achieving an end to the epidemic will require concerted efforts and resource expenditures to address persistent health inequities that exist globally. HIV is a highly stigmatized disease that often affects marginalized peoples who historically have had limited access to social resources. Racial and ethnic minorities and gay and bisexual men are disproportionally affected by the HIV epidemic in the United States . The increased HIV burden among these groups necessitates a better understanding of what social resources are available to them and what effects these resources have on health outcomes. Efforts to address the challenges of persons living with HIV who are disenfranchised or marginalized require innovative approaches that integrate their daily context-specific challenges . Understanding the relationships between social resources and health promoting or risk behaviors, and how key demographic variables influence social resources may help us develop targeted interventions with marginalized groups. Among all people, including people living with HIV, the achievement of optimal health requires individual and collective resources. Individual and collective resources have been studied over the past several decades through studies of social resources , social belonging/friendship , social capital , and social networks . The purpose of this study was to describe associations between social belonging/friendship, social networks, social capital, and the health promotion behaviors and quality of life of persons living with HIV. Social resources are defined as any concrete or symbolic item that can be used as an object of exchange among people . They are constructs that have consistently been observed to improve and maintain health. Examples of social resources include both tangible items such as money, information, goods and services, and less tangible concepts such as love/affection and status within society. Over the last few decades, social resources have been documented to have direct and indirect benefits for members of society that possess them . Individuals and groups with adequate social resources tend to be healthier and are more able to manage health threats in their communities. When social resources are present, they mediate the deleterious effect of stress on health as well as directly affect physical and psychological health . Foa and Foa classified social resources into six categories: love/affection, status, information, services, goods, and money. This categorization remains a useful framework for analysis despite being debated, elaborated on, and synthesized by other scholars . In this analysis, we use Foa and Foa's Social Resource Theory to help describe the distinct contributions of social belonging, social networks, and social capital on select health promotion behaviors and quality of life of persons living with HIV. Social belonging or inclusion is a universal and basic human need for being accepted as a member of a group . Elements of social belonging include companionship, affiliation, and connectedness . Associations between social belonging and increased heath promoting behaviors have been documented, but remain an emerging area of health research . Social belonging encompasses all six of Foa and Foa's social resource categories with an emphasis on love/affection and status. The relationship between social belonging and other social resources, and their combined impact on the health promotion behaviors and quality of life of persons living with HIV, has yet to be explored. However, understanding this relationship may provide critical information on the nature of social resources in this population, leading to improved health promoting interventions. Social support network is a multidimensional construct that emphasizes the structure and function of a person's social relationships . The members of social support networks can help to convey a strong sense social belonging, or the lack thereof. Social support networks can be shaped by individual factors , interpersonal and sociostructural conditions . Social support networks are prerequisites for the development of social capital and are recognized as essential for good health by the World Health Organization . The effects of social support networks have figured prominently in health research, practice, and policy. Extensive research has demonstrated that social support networks decrease illness and mortality across populations . However, the mechanisms by which social support networks lead to improved health outcomes are elusive and expanding the distinctions between social support networks, social belonging and social capital may lead to a better understanding of these mechanisms. Social capital is the "aggregate of potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance or recognition" . While debated , it can be useful to think of individual social capital comprising reciprocity, trust, participation, norms, social networks, value of life . Many of the definitions of social resources are similar to those of capital -an ability to exchange object entities, capabilities for other objects, entities of value -helping to make social capital a popular area of study in health research. However, conceptualization and operationalization of social capital vary tremendously across studies rendering aggregation of findings difficult . In general, social capital is thought to be associated with positive health behaviors and outcomes, although rigorous prospective trials are lacking. Similar to social belonging, social capital encompasses all of Foa and Foa's social resource categories, however social capital can more easily illuminate the concrete and particular categories including services, goods, and money. Given the overlap and inter-relatedness of these constructs, scholars have called for conceptual clarity among social resources constructs . For example, Locher et al examined the relationship between social isolation, social support, social capital and nutritional risk among 1000 older adults and found that indicators of social isolation were associated with greater nutritional risk. However, they did not examine associations between social resource variables and did not use psychometrically sound scales to measure these variables, leaving many gaps in the literature . To our knowledge our study is the first to examine the relationships between social belonging/friendship, social support networks, social capital, health promoting behavior, and quality of life among people living with HIV. --- Objectives We hypothesized that the nature of various constructs that contribute to conceptualizations of social resources, particularly those most commonly used in the health literature, are different and each construct adds distinct information about the impact of social resources on the health of people living with HIV. In this paper, our objectives were to examine the associations among three variables, social belonging/friendship, social support networks, and social capital. Guided by previous research , we further examined the role of the demographic characteristics gender, age and economic position on these social resources. Finally, we described the associations between these social resource variables and two important health behaviors, HIV medication adherence and weekly physical activity, and quality of life among persons living with HIV. --- Methods This analysis was conducted as part of a larger study examining the relationships between age, gender and self-management behavior in adults living with HIV. A more full description of the inclusion criteria and data collection methods for this analysis are included in prior publications . --- Sample and Recruitment We examined the relationships between social belonging/friendship, social support networks, social capital, health promoting behavior, and quality of life among people living with HIV using a convenience cohort. Adults living with HIV were recruited from HIV clinics, HIV service organizations, and a northeast Ohio HIV research registry and organized into four groups that most approximated equal sized strata and equivalent age cut points to ensure representation of both males and females, and older and younger participants. Individuals with a documented HIV diagnosis who were 18 years of age or older and on antiretroviral therapy were recruited from November 2011 to June 2012 . The study was approved by the institutional review board of University Hospitals, Case Medical Center . Written informed consent was obtained from eligible adults living with HIV before completing study related activities. Participants completed the demographic survey, which was collected via computer-assisted delivery, to enhance selfreport accuracy. Participants were compensated for their time with a $50 cash gift card. --- Measures Demographic characteristics were obtained by self-report and clinical data were abstracted from participant's medical records. Clinical data included HIV disease variables and current health conditions. Social belonging was measured with the 6-item Hawthorne Friendship Scale . This scale was developed from the 2004 South Australian Health Omnibus Survey and was refined through administration to 829 older adults. We used the refined scale, which measures perceived social isolation and perceived emotional loneliness in the past four weeks. Items are scored on a 0-4 scale with a total scale range of 0 to 24. Higher scores indicate greater social connectedness and lower scores indicate more social isolation. Examples of items include "It has been easy to relate to others" and "I had someone to share my feelings with". The Hawthorne Friendship Scale has good psychometric properties and a Cronbach's alpha reliability coefficient of 0.72 in this study. Social support networks were measured using the Lubben Social Network Scale . This 6-item scale is a psychometrically sound measure of perceived social support provided by friends and family. It measures both the structure and function of social support networks. Each item is scored 0-5 with a total scale range from 0 to 30. Respondents are asked to report the number of friends or family who provide various types of support. Higher scores indicate greater social network support and lower scores indicate greater social isolation. Examples of items include "how many relatives do you see at least once a month" and "how many friends do you feel at ease with that you can talk about private matters". The Lubben Social Network Scale had a Cronbach's alpha reliability coefficient of 0.84. Social capital was measured with Bullen and Onyx's 36-item individual Social Capital Scale. This widely-used instrument measures eight subscales including: participation in the local community, social agency, feelings of trust and safety, neighborhood connections, friends and family connections, tolerance of diversity, value of life, and workplace connections; these items were used to create a total score. Participants were asked to rate items on a 1-4 Likert-type scale. Higher mean scores indicate more social capital. Examples of scale items include "if you were caring for a child and needed to go out for a while, would you ask a neighbor to help" and "are you on a management committee or organizing committee for any local group or organization". Reliability and validity of the social capital scale have been reported as acceptable and the Cronbach's alpha reliability for the Social Capital Scale in adults living with HIV is 0.88 in this study. This scale has the advantage of being developed a priori for the purpose of measuring social capital, not integrating various measures post-hoc, and integrates the various perspectives of social capital . Physical activity was measured using a pen and paper exercise diary. Hours of exercise participants completed per week was assessed using a daily exercise diary completed by each participant for seven consecutive days. The diary contained four questions about whether a participant exercised at all that day, if so, what times the participant started and ended exercise, the perceived intensity of exercise, and the type of exercise. Each question was completed at the end of the day. The cumulative duration of each exercise session was summed for a daily total and then the daily totals were summed to obtain hours of exercise per week. Quality of life was measured with the HIV/AIDS Targeted Quality of Life Instrument. This widely-used, 34-item instrument is an HIV-specific quality of life measure assessing nine dimensions: Overall function, life satisfaction, health worries, financial worries, medication worries, HIV mastery, disclosure worries, provider trust, and sexual function . Each dimension is scored and reported individually. We administered the overall functioning and life satisfaction subscales to indicate participants' quality of life. Examples of items on these scales include "in the past 4 weeks, I've been satisfied with my physical activity" and "in the past 4 weeks, I've felt in control of my life" . Reliability coefficients ranged from 0.83 to 0.88 for all dimensions . --- Statistical Analysis All data were directly imported from the web-based data collection system, REDCAP, into a data management program . Subjects' demographics, HIV disease characteristics, social resources and health outcomes were summarized using descriptive statistics, e.g., means and standard deviations for continuous variables and frequencies and percentages for categorical variables. The primary measures of interests are social belonging, social support networks, social capital, and health outcomes. Distributions of these variables are not symmetric. Therefore, for studying associations among these variables we used spearman rank correlations coefficients. Similarly, for finding association between a risk factor and the measures of interests we used median regression analysis. The median regression analysis is a robust regression approach. It does not assume normality of the response/dependent variable, as in the case of mean regression. A covariate with p-value <0.05 was considered statistically significantly associated with the underlying response variable. --- Results --- Demographic and clinical characteristics One hundred and two eligible participants completed the study procedures. Of these, there were 27 males <51yrs, 27 males >51yrs, 23 females <51yrs and 25 females >51yrs. The mean age of all participants was 48 years, most were African-American , single , and had a high school degree or less education . Almost all, female participants identified as heterosexual and more than half of male participants identified as gay . Participants had been living with HIV for an average of 13.6 years. Mean duration of HIV antiretroviral therapy was 9.0 years, and most participants had an undetectable viral load . Eighty percent had a co-occurring health condition . On average, participants reported a mean social belonging/ friendship score of 17.0 , a mean social support network score of 13.9 , and a mean social capital score of 2.64 . Participants had an average 3-day HIV medication adherence rate of 92.7% and exercised on average 4.6 hours per week. Overall functioning quality of life was 66.5 and their life satisfaction quality of life was 74.3 . Additional participant characteristics are in Table 1. --- Social resources, demographics, and health outcomes To achieve our first objective, we examined the associations between our social resources variables and found that there were moderate-strong and statistically significant bivariate relationships. The strongest association was between social capital and social support networks , followed by the association between social belonging and social capital , and then the association between social belonging and social support networks . These associations can be found in Table 2. Multivariate median regression analyses only found two sociodemographic variables that influenced social resources. Having a college education significantly influenced the strength of one's social support network and increasing age and having a college education influenced one's perceived social capital. Full multivariate median regression analyses results are in Table 3. To fulfill our final objective, we examined the associations between social resources and select health promotion behaviors and quality of life. Of our three social resource variables, social belonging was the most strongly associated with health outcomes. It was significantly associated with HIV medication adherence , overall functioning quality of life and life satisfaction quality of life . While less strong, social capital was also associated with HIV medication adherence and life satisfaction quality of life . Social support networks were not significantly associated with health outcomes. None of the social resource variables were associated with weekly exercise. Associations are in Table 3. --- Discussion We found that in this sample of adults living with HIV the association between social belonging/friendship, social support networks, and social capital was moderate and significant but not perfectly overlapping. This was an expected finding. In examining the associations with our health outcomes, it appears that among our sample of adults living with HIV, social belonging is most strongly associated with health outcomes; followed by social capital. Social support networks were not associated with health outcomes in our study. This evidence supports Foa's contention that social belonging/friendship comes before any sort of assets can be derived from social relationships . Additionally, several explanations should be considered in light of this evidence. Since the development of Foa and Foa's social resource theory, much work has highlighted the significance of place/location in a person's ability to access and use social resources. In the United States, a person's health status is often linked to his or her zip code .With the exception of an overrepresentation of African Americans, our sample characteristics are similar to those of the national population of adults living with HIV and would be considered a marginalized group . However, our sample reported high levels of HIV medication adherence, exercise and quality of life. These health outcomes were associated with social belonging/friendship and social capital. One reason for this may be that among our Midwestern population, our participants may be less transient than in other geographic areas and are able to develop a sense of identity in place. This stability may allow individuals to form egalitarian social relationships and then, over time, build stronger and perhaps more reliable friendship ties and social bonds . This may lead them to feel as though people care more about them, instilling a desire to actively participate in the social relationships, which facilitates easier and fuller access to the benefits of those social relationships and the development of social capital . One implication of this explanation is that a more nuanced understanding of the individual's social resources, and their inter-relationships, in both academic/research and clinical settings is necessary . It is our contention that social belonging and social capital are distinct and measureable elements of social support that influence health outcomes among marginalized populations. Simply reducing these constructs to social support would minimize important information necessary for the development of interventions to improve health behaviors and quality of life among people living with HIV and perhaps other marginalized populations. We also found that social support networks were not significantly associated with health behaviors. One explanation for this is that the quality of social relationships may matter more than the quantity or density of social relationships. This hypothesis has been proposed previously, but evidence supporting it has been limited . By distinguishing between the existence of social relationships , their quantity , and the content and consequences of these relationships , our study provides additional support for this hypothesis in a sample of adults living with HIV. We further report the associations of these constructs with health behaviors and health-related quality of life, which improves our understanding of how social resources are consequential for health. An implication of this finding in the field of HIV relates to the current initiatives to link HIV-positive individuals to formal health care. Substantial resources have been invested in identifying HIV-positive adults, linking them to, and retaining them in health care services. This strategy is proposed to improve HIV outcomes and to reduce HIV infection . However, although many people living with HIV have been identified and linked to care, our ability to retain people living with HIV in care and to help them take their HIV antiretroviral medications in a way that suppresses HIV viral replication has been less successful . Engagement with care is often defined in terms of numbers of appointments rather than the quality of the relationship between the patient and the rest of the health care team. However, when investigating the quality of the social relationships between patient and health care provider, improved quality of the relationships has been found to be highly predictive of HIV medication adherence . Additionally, a recent multi-site randomized clinical trial found that enhanced personal contact with a clinic staff member improved the number of primary care visits in HIV clinics . Although the outcome was defined as a quantitative measure of visits and not the quality of the social relationships with the clinic staff, that study provided evidence that consistent personal contact was the efficacious intervention component. The findings from that study further supports the notion that the quality of relationships and not simply quantity has a bigger impact on patients engaging with and being retained in care. Recently, Christopoulous and colleagues conducted a qualitative study with newly diagnosed HIV-positive patients that examined barriers and facilitators of engagement in HIV care . Their results illustrated the primacy of the quality of the relationship with health care team members for continued engagement in care, which was particularly important as the individual experienced health changes over time. Our findings, when taken in context of this evidence on the quality of the relationships between persons living with HIV and members of their health care team, suggests that the development of a social resource intervention to improve health outcomes in this population is appropriate. Finally, our findings lead us to recommend that investigators interested in examining the role of social resources on health outcomes in people living with HIV or other marginalized populations should include measures of social belonging and social capital. The information yielded by these separate instruments was distinct and significant allowing for a more nuanced understanding of the social resources necessary for optimal health outcomes in this population. An additional advantage of the Social Capital Scale is that social support networks are a subscale of the instrument and allows investigators who are also interested in social support networks to analyze the subscale separately. Appropriate psychometric testing must be done before analyzing the subscale in lieu of a separate measure of social network support. Clinicians who routinely assess social resources of their clients in order to develop a holistic care plan, should be aware that there are distinctions between social belonging, social support networks and social capital. These unique constructs have differential associations with health behaviors and quality of life that are often the goal of the treatment plan. It will be important to maintain a conceptual clarity and verbal precision when assessing patients' social resources. --- Limitations We have several limitations. The Lubben Social Network Scale does not measure network density or activation, is confined to the structure of social support networks comprised of friends and family, resulting in an inability to understand the type of social network and extent of its support . While the scales we chose to measure our social resource constructs are limited, the scales we chose were designed a priori to measure the construct each purports to measure and each scale has undergone extensive psychometric evaluation in other studies. Further, these scales are commonly used in the health literature and the Lubben Social Network Scale has been included in the PhenX toolkit by the U. S. National Institutes of Health to help pragmatically facilitate common data elements and cross-sample and cross-study analyses. In the future, investigators may want to consider an approach similar to factor analysis or item response theory to further disentangle these constructs and methods of measurement. A framework like the PhenX toolkit may be the most likely way to accomplish these analyses . Additionally, our crosssectional data were only collected at one site and may not be generalizable to other populations of people living with HIV. However, our single site design allowed us to consider the unique geographic attributes of the sample and their implications for assessment and use of social resources in health research. Future, multi-site studies examining social resources should analyze data both at individual sites and pooled across sites to further explore the role of social geography in relation to health outcomes. Finally, our behavioral data were collected using self-report measures . This may have led to an overestimation of both physical activity and medication adherence in our sample. In conclusion, we found that there are distinctions among various, widely-used social resource constructs. Social connectedness/friendship and social capital were associated with medication adherence and quality of life among our sample of PLWH. By describing these unique associations and distinctions, our study helps identify which social resources should be targeted in the development of interventions to improve the health of members of this marginalized population. Our findings reinforce the importance of understanding the nuanced ways in which complex social factors, such as social belonging/connectedness, social network support, and social capital influence health outcomes of persons from marginalized populations. The adage "it takes a village" to manage HIV as a chronic disease is evident in our study's findings and requires close attention to the social resources available to persons living with HIV. ---

Purpose-To describe the associations among three social resource variables (social belonging, social support networks, and social capital) and two health promotion behaviors, HIV medication adherence and physical activity, and quality of life among persons living with HIV (PLHIV). Method-We conducted a cross-sectional analysis in 102 adult PLHIV. Social resource variables and quality of life were assessed using validated and widely-used instruments. Physical activity was assessed using a daily physical activity diary and medication adherence was abstracted from the participant's medical record. Spearman correlations and descriptive statistics were used to analyze associations among variables. Results-Fifty-four participants (54%) were male and most were African American (84%), single (69%), and living in poverty (82%). Participants had been living with HIV for an average of 13.6 years (+/-7) and most were living with at least one non-AIDS comorbidity (80%). Social belonging was significantly associated with HIV medication adherence (ρ = 0.25, p = 0.02), overall functioning (ρ = 0.48, p < 0.01) and life satisfaction quality of life (ρ = 0.50, p < 0.01). Social capital was also associated with HIV medication adherence (ρ = 0.17, p = 0.10) and life satisfaction quality of life (ρ = 0.29, p < 0.01). Conclusions-We found that there are distinctions among various, widely-used social resource constructs. By describing these unique associations and distinctions, our study helps identify which social resources should be targeted in the development of interventions to improve health promotion and the quality of life of members of this marginalized population.

Introduction The World Health Organization declared the coronavirus 2019 outbreak an international public health emergency on January 30, 2020 and a pandemic on March 11, 2020 [1]. In an effort to contain the COVID-19 outbreak, the Israeli government announced a number of new restrictions aimed at reinforcing social distancing [2]. On March 12, Israel announced that all universities and schools would close. On March 16, all non-critical government and local authority workers were placed on paid leave, and private sector firms were required to reduce the staff presence in the workplace [3]. On March 19, Prime Minister Benjamin Netanyahu declared a national state of emergency [4], saying that existing restrictions would henceforth be legally enforceable, and violators would be fined. Israelis were not allowed to leave their homes unless absolutely necessary, putting a near-lockdown into effect. Essential services-including grocery stores, pharmacies, and banks-remained open, but people were prohibited from venturing more than 100 m from their homes, apart from under certain circumstances . Non-essential stores were required to close, and parks were to remain shut. People were required not to participate in any social gatherings and to limit face-to-face interactions with individuals outside the immediate household [5]. Self-rated health -also known as subjective health, self-assessed health, or perceived health-is a simple and straightforward construct that has a strong predictive capacity regarding various health-related outcomes, including mortality [6][7][8][9]. Numerous studies suggest that, among people with chronic illnesses, SRH encompasses many of the objective aspects of health including clinical, cognitive, and functional aspects [10][11][12][13]. For example, studies among individuals with Type 2 diabetes suggest that SRH is highly associated with clinical status [14,15], and provides additional information regarding mortality risk, independent of demographic, socioeconomic, and medical risk factors. In addition, studies about changes in SRH suggest that a decline in SRH can independently predict better long-term prognoses among people with chronic illnesses [16]. Therefore, and based on the extant literature highlighting the importance of SRH and changes in SRH in the assessment and understanding of the medical status of people with chronic illnesses, we aimed to investigate whether people with chronic illnesses experienced changes in SRH after one month of the social distancing reinforcement due to the coronavirus outbreak. Changes in health behaviors were also assessed, as these behaviors may have been affected by the lockdown situation [3]. It is important to assess changes in health behaviors given that these behaviors can have detrimental effects on health [17], especially among people with chronic medical conditions [18]. A recent study conducted in France and Switzerland among the general population assessed whether changes in physical activity and sedentary behaviors during the COVID-19 lockdown were associated with changes in mental and physical health [19]. Results of this study showed that the lockdown in these countries resulted in an increase in sedentary behaviors, which was associated with a decrease in physical health, mental health, and subjective vitality [19]. Although no research has yet been conducted regarding the effect of the COVID-19 crisis on health behaviors in people with chronic illnesses, it may be presumed that the disruption in usual activity would be even more harmful in this population. Loneliness is an important determinant of health [20], and studies suggest that loneliness and social isolation are major risk factors for morbidity and mortality, risk factors comparable in importance to obesity, sedentary lifestyles, and possibly even smoking [21]. For example, a longitudinal study conducted among a U.S. nationally representative sample reported that loneliness was associated with an increased mortality risk [22]. Specifically, among people with chronic illnesses, severity of illness can predict enhanced feelings of loneliness, and loneliness, in turn, can aggravate the medical condition [23]. Nevertheless, despite the understanding that lockdown measures can increase loneliness, specifically among vulnerable populations [24], studies examining the impact of loneliness on the health of people with chronic medical conditions during the COVID-19 outbreak are, as yet, scarce. We also aimed to examine factors contributing to changes in SRH. The literature suggests that various factors, including sociodemographic and medical-related factors, can contribute to changes in SRH [25]. For example, there is some evidence that lower socioeconomic status, as reflected for instance by lower education level, or higher number of people living in the household, is often correlated with worse SRH [26]. Furthermore, females may be more vulnerable to psychological distress than are males [27], and this psychological distress may be linked with a stronger decline in mental SRH during stressful occasions . Indeed, Mazza et al. [28] found that female gender was associated with higher levels of depression, anxiety, and stress during the COVID-19 lockdown in Italy. In addition, studies among patients with chronic illnesses suggest that those with higher illness severity, as represented for instance in higher BMI, more medical visits, and longer illness duration, are more vulnerable to a decline in SRH [10][11][12]. Therefore, in the current study we sought to find those characteristics of people with chronic illnesses that might be linked with a stronger decline in SRH. An online survey was conducted one month into the reinforcement of social distancing among people with various chronic medical conditions in order to address the following research questions: 1. Did people with chronic illnesses experience changes in their SRH, one month into the reinforcement of social distancing? 2. Did people with chronic illnesses report changes in loneliness and health behaviors, one month into the reinforcement of social distancing? 3. Did people with different chronic illnesses experience different changes in self-rated health, loneliness, and health behaviors one month into the reinforcement of social distancing? 4. Were there specific characteristics, such as sociodemographic and medical-related factors, that were linked with a decline in SRH one month into the reinforcement of social distancing among people with chronic illnesses? It is important to address these questions, as chronic medical conditions may increase the risk of having a fatal reaction to COVID-19 [29][30][31][32]. That is, these vulnerable individuals are under additional stresses and likely more confused about how to handle their ongoing medical issues in the context of COVID-19 [1]. Relatedly, the fact that non-emergency medical services were completely shut down or reduced to an absolute minimum during the examined period must have been particularly challenging for people with chronic illnesses, who have no choice but to make use of these services on a regular basis. Hence, findings from the current study may help us understand how the reinforcement of social distancing might have influenced the psychological and physical health [33][34][35][36] of people with chronic illnesses, in order to inform policymakers about the effect of social distancing on the health of vulnerable populations. --- Methods --- Procedure Information was collected using an online, self-report survey, which was advertised through Camoni . Camoni is the first Israeli medical social network to have been established, and it includes several virtual health communities for people who share similar medical conditions, such as diabetes, cancer, pain, depression, or eating disorders. These are moderated communities-maintained by a governmental organization, the Gertner Institute for Epidemiology and Health Policy Research -that are free of charge and accessible to everyone. Users can log into the communities to share in-the-moment feelings or to seek momentary support from moderators and other users. The communities do not present themselves or intend to be seen as alternatives to professional treatment, and the moderators of the communities encourage community users to receive professional treatment. The invitation to participate in the study was advertised by the maintainer of Camoni through a newsletter, thereby using a convenience-sampling method. On the first page of the online survey and prior to the start of the questionnaire, participants were asked to press "continue" if they were over the age of 18 and agreed to participate in the survey. Participants were also informed that they could choose not to participate in the study or to stop participating at any stage. In addition, they were told that the questionnaire was anonymous, and that no personal information would be collected . The contact details of the primary investigator were provided, as well as the name of the funding organization. Following consent, participants were provided with a link to a designated website which contained the anonymous survey. Participants who wished to participate in a lottery were invited to provide their contact information at the end of the survey. A 50-shekel cash gift was the prize, equivalent to about 15 U.S. dollars. The study was approved by the university's institutional review board IRB. Ethical approval number 042011. --- Statistical Analyses Data were analyzed using IBM SPSS-25. Chi 2 and one-way ANOVA with LSD post-hoc tests were conducted to examine differences among the five medical condition categories. Paired T-tests were conducted to examine changes in health behaviors. Ordinal logistic regressions were fitted to examine which factors contributed to perceived changes in physical, mental, and general health occurring during the first month of the outbreak. The orthopedic/pain-related condition was selected as a reference group in the regression models, given that univariate analyses suggested that people with this condition reported the greatest decline in SRH. The independent variable "crowded living" was dichotomized in the regression models to not at all versus somewhat , as its distribution was not normal. --- Measures Individual and Socioeconomic Variables: These variables consisted of gender , age group , marital status , education , work status during and before the COVID-19 crisis , economic status , religiosity , number of people in household , and perception of crowded living . Categories containing few responses were combined. --- Medical Condition Participants self-selected their one main medical condition from a list of predefined conditions, based on the different communities within the Camoni platform, with an option to add an additional condition not mentioned in the list. In addition, participants reported the duration of their medical condition , whether they generally received medical care for their medical condition , and their Body Mass Index . In addition, one item was included to examine frequency of medical appointments/visits: "How often did you meet with your medical team before the COVID-19 crisis ? --- Health Behaviors Participants were asked to report the frequency of their health-related behaviors, during and before the COVID-19 crisis via the following questions-"On average, how many times a week do you participate in any exercise/sports activity for half an hour or longer?" This item was based on the 36-item short-form Medical Outcomes Study [37]. "On average, how many times a day do you eat fresh fruit?" "On average, how many times a day do you eat fresh vegetables?"-with answers ranging from 0-10 times per day, based on the 2-item Serving Fruits and Vegetables Scale [38]. Participants were also asked to compare the amount of food they consumed during and before the COVID-19 crisis, with answers ranging from 1 to 5 . --- Disease Management Disease management was defined to participants as actions that patients need to take routinely in order to maintain their health, for example, going to medical appointments, adhering to nutrition recommendations, taking medications, and engaging in physical activity. Perceived disease management, during and before the COVID-19 crisis, was measured by a single item: "In my opinion, I manage my illness optimally," rated on a 5-point Likert scale with answers ranging from 1 to 5 . This item was adopted from previously validated disease-specific self-efficacy measures [39] and from the Challenges to Illness Management Scale [40]. --- Time Spent on Social Media Total hours per day spent on social media, during and before the COVID-19 crisis, were measured by two single items: "How many hours per day, on average, do you spend on social media in general?" and "How many hours per day, on average, do you spend on online health communities to receive or provide information about the medical condition you are dealing with?" Participants were asked to answer this question in regard to the week before the outbreak and the previous week. Hours per day were measured on a seven-point scale: . These items were adopted from the Technology Use Questionnaire [41,42]. --- Self-Rated Health Five items were included to examine SRH before and during the outbreak, revised from the SF-36 MOS [37] and adapted for the COVID-19 pandemic: "In your opinion, as of today, your physical health condition is?" with answers ranging from 1 to 5 . "In your opinion, as of today, your mental health condition is?" with answers ranging from 1 to 5 . "Has your medical condition improved or worsened in the last month?" with answers ranging from 1 to 5 . "How much do you feel that the COVID-19 crisis has affected your physical health?" with answers ranging from 1 to 5 . "How much do you feel that the COVID-19 crisis has affected your mental health?" with answers ranging from 1 to 5 . --- Loneliness The 3-item version of the Revised UCLA Loneliness Scale [43,44] was used to assess participants' sense of loneliness, during and before the COVID-19 crisis. The Hebrew-translated version of the scale was used [45]. Items are as follows: "How often do you feel that you lack companionship?" "How often do you feel left out?" and "How often do you feel isolated from others?" with answers ranging from 1 to 2 or 3 . The three items are summed to form a total score ranging from 3-9. Participants were asked to rate their current and past loneliness using these three items. --- --- Results Table 2 provides an answer to the first research question. About half of the participants reported a decline in mental and physical health during the first month of the COVID-19 outbreak. Regarding changes in physical SRH, 8% reported feeling greatly worsened, 39% reported feeling slightly worse, 44% reported feeling no change, 5% reported feeling slightly better, and 4% reported feeling greatly improved. Regarding changes in mental SRH, 10% reported feeling greatly worsened, 41% reported feeling slightly worse, 42% reported feeling no change, 4% reported feeling slightly better, and 3% reported feeling greatly improved. Regarding changes in general SRH, 7% reported feeling greatly worsened, 20% reported feeling slightly worse, 60% reported feeling no change, 8% reported feeling slightly better, and 5% reported feeling greatly improved. Table 3 provides an answer to the second research question by demonstrating changes in health behaviors and loneliness. One month into the reinforcement of social distancing, participants reported a decrease in their perceived disease management abilities , and a decline in physical activity . In addition, total time spent on social media and on online health communities increased significantly. Over 50% of the participants reported eating more than they did before the outbreak. Furthermore, one month into the reinforcement of social distancing, participants reported a higher sense of loneliness . Whereas only one-third of the participants reported that they experienced loneliness before the outbreak, as many as two-thirds of them reported feeling lonely one month after the reinforcement of social distancing. Specifically, in regard to the question "How often do you feel that you lack companionship?" only 6% of the participants reported almost always before the outbreak, relative to as many as 27% during the outbreak. Similarly, in regard to the question "How often do you feel left out?" only 7% of the participants reported almost always before the outbreak, relative to as many as 20% during the outbreak. Finally, in regard to the question "How often do you feel isolated from others?" only 7% of the participants reported almost always before the outbreak, relative to as many as 21% during the outbreak. % Much less than before 2.9 1 "In my opinion, I manage my illness optimally" with answers ranging from 1 to 5 . 2 Answers ranging from 1 to 7 3 Half an hour or more of physical activity. 4 Revised UCLA Loneliness Scale. 5 "Relative to the amount of food I consumed before the outbreak, I estimate that I currently consume" 1 to 5 . Table 4 provides an answer to the third research question by demonstrating differences in background variables and SRH measures between each of the five medical condition categories. There were no group differences in current physical health , but current mental SRH was lowest among participants with mental health conditions . Similarly, the feeling of loneliness was highest among participants with mental health conditions . 1 Items rated on a scale from 1 to 5 . 2 Total score; items rated on a scale from 1 to 3 . 3 Items rated on a scale from 1 to 5 . Table 5 provides an answer to the fourth research question by presenting three ordinal logistic regressions to predict changes in SRH. The first model shows that a decline in physical SRH was predicted by crowded housing conditions , higher BMI , and higher frequency of medical visits before the crisis . In addition, participants with orthopedic/pain conditions were more likely to experience a decline in their physical SRH relative to participants with mental , metabolic , and cardiovascular conditions. The second model shows that a decline in mental SRH was predicted by female gender , crowded housing conditions , and higher BMI . In addition, participants with orthopedic/pain conditions were more likely to experience a decline in their mental SRH relative to participants with metabolic conditions. The third model shows that a decline in general SRH was predicted by female gender , higher education , crowded housing conditions , and illness duration . SRH = Self-Rated Health. BMI = Body Mass Index. Significant values appear in bold. 1 "How much do you feel that the COVID-19 crisis has affected your physical health?" with answers ranging from 1 to 5 . 2 "How much do you feel that the COVID-19 crisis has affected your mental health?" with answers ranging from 1 to 5 . 3 "Has your medical condition improved or worsened in the last month?" with answers ranging from 1 to 5 . 4 Before the outbreak. 5 Participants with "other" medical condition were omitted from the analyses. --- Discussion The present study investigated how people with chronic medical conditions perceived their health status one month into the reinforcement of social distancing due to the COVID-19 outbreak. We conducted an online survey among more than 300 people with chronic illnesses exactly one month after the Israeli government reinforced severe social-distancing regulations, and two days before these regulations were reduced. The survey included self-report information about SRH, health behaviors, and loneliness. Several interesting findings were identified. First, as many as about half of the participants with a chronic medical condition reported a decline in their physical SRH or mental SRH during the first month of the social distancing reinforcement. Second, a significant deterioration in health behaviors was reported, including a decrease in vegetable consumption and physical activity, an increase in time spent on social media, and a perception among about half of the participants that they were eating more than they had been before. Third, whereas only one-third of the participants reported that they had felt lonely before the outbreak, as many as two-thirds of the participants reported feeling lonely one month after the reinforcement of social distancing. Last, a decline in general SRH was linked with female gender, lack of higher education, crowded housing conditions, longer illness duration, and loneliness. Data indicated that as many as about half of the participants in this study experienced a decline in their SRH after one month of social distancing reinforcement . To the best of our knowledge, the current study is the first to report changes in SRH during the coronavirus crisis among people with chronic illnesses [33][34][35][36]. Contrary to the current findings, in a study by Lei et al. , conducted during the lockdown among 1593 healthy Chinese individuals, only 2.4% perceived their SRH as bad or very bad, and over two-thirds perceived their SRH as good or very good [46]. This difference may be explained by the fact that our study exclusively included participants with chronic medical conditions, who likely ordinarily have poorer SRH than do healthy individuals [47], whereas the Chinese sample was recruited from the general population. Furthermore, this discrepancy between our findings and those of Lei et al. may imply that the outbreak and the reinforcement of social distancing may have had a stronger negative impact on people with a chronic illness. Relatedly, Wang et al. , who collected data at two timepoints during the COVID-19 outbreak [48], reported that poor SRH status, physical symptoms, and a history of chronic illness may have contributed to higher levels of stress, anxiety, and depression. It could be that there was a vicious cycle in which during the COVID-19 outbreak, chronic illness and poor SRH contributed to higher levels of stress, anxiety, and depression, which contributed to a decline in SRH. Findings suggest a deterioration in health behaviors, including less exercising, less consumption of fresh fruits, and more time spent on social media . These findings can be explained by the limitations imposed by the social distancing reinforcement: People were not allowed to venture more than 100 m from their homes, and health-clubs, gyms, and recreational parks, as well as many stores and markets, were closed [3]. Although supermarkets and pharmacies did remain open, and local shortages of fresh produce were unusual [49], people with chronic medical conditions were discouraged from leaving the house [50], making it difficult to purchase fresh, healthy food. The reported decline in people's ability to optimally manage their illnesses may have been related to the disruption in healthcare access, including the transition to online medical care and the absence of non-emergency medical surgeries due to the outbreak [51]. These disruptions may have posed specific difficulties for people with chronic medical conditions. Some people may also have had poor access to or difficulties in mastering the new technologies in remote care that were introduced by healthcare providers [52]. Several studies have already examined compliance to health behaviors directly related to COVID-19 [53,54], but limited information has been available regarding routine health behaviors such as nutrition, exercise, or pre-existing disease management during the pandemic among people with chronic illnesses. A single study identified health-related behaviors significantly associated with mental health among people in quarantine due to COVID-19 in Brazil: diet, tele-psychotherapy participation, and exercise level. The results from Brazil highlighted the role of health behaviors during the outbreak by showing that specific health behaviors, including balanced meals, exercising, and the use of tele-psychotherapy, may have impacted stress, depression, and anxiety levels [55]. Our findings demonstrated some similarities and some dissimilarities between people with different medical conditions in terms of SRH . For example, the adjusted model suggests that people with orthopedic/pain conditions experienced more deterioration in physical SRH relative to people with other medical conditions . However, medical condition did not contribute to the variability we detected in general SRH declines. In Israel and the rest of the world, pain treatment and physical therapy centers closed their doors following the outbreak of the COVID-19 pandemic. Although telemedicine is being used to treat patients with many different medical conditions, pain management providers face a challenge in delivering services through video or other eHealth methods [51], preventing people with orthopedic/pain conditions from participating in treatments such as physical therapy. Moreover, the decline in physical activity reported in the entire sample may have more negatively affected people with orthopedic/pain conditions than those without, as these individuals are more vulnerable to lessened muscular strength, muscular endurance, and joint flexibility [56,57]. As expected, we found that participants felt lonelier during the first month of the outbreak than they had previously, and that loneliness was an important contributor to a decline in SRH . This finding is not surprising given that previous studies have already highlighted the importance of loneliness in health [18,20]. For example, a longitudinal study conducted before the coronavirus outbreak among a U.S. nationally representative sample reported that loneliness was associated with an increased mortality risk over a six-year period, and that this association was explained by health outcomes [22]. Furthermore, a study conducted during the COVID-19 lockdown period in Spain found that higher reported loneliness was associated with higher distress [24]. In line with findings from the current study, Brodeur et al. used Google Trends to show a significant increase in searches for loneliness, worry, and sadness before and during the lockdown in Europe and the U.S. [58]. In the current study, the "crowded housing conditions" variable was a strong predictor of decline in both physical and mental SRH . Although a lower number of people living in the household is often correlated with worse SRH [26], even after adjusting for socioeconomic status, it could be that the social distancing regulations that forced people to stay in their homes for an extended period of time increased the negative effect of crowded housing conditions on participants' health. In line with the current findings, Wang et al. reported that respondents staying in a household with three or more people one month into the COVID-19 outbreak had significantly higher posttraumatic stress disorder scores compared to respondents who lived alone [49]. Furthermore, female gender, higher BMI, higher frequency of medical visits , and longer illness duration were also linked with a decline in SRH. These findings are not surprising given that females may be more vulnerable to psychological distress than are males [27], and this psychological distress may be linked with a stronger decline in mental SRH. Indeed, Mazza et al. [28] found that female gender was associated with higher levels of depression, anxiety, and stress during the COVID-19 lockdown in Italy [28]. Moreover, it has been reported that the COVID-19 pandemic has increased the care burden of women, negatively impacting women and their families [59]. In addition, higher BMI, more medical visits , and longer illness duration are all likely to be linked with illness severity [60,61]. Therefore, our findings may imply that among people with a chronic illness, poorer medical condition is linked with a stronger decline in SRH. Future studies are warranted to clarify the long-term effects of social distancing on the health of this vulnerable population. This study has several strengths, including the investigation of people with chronic illnesses exactly one month after the Israeli government decided to carry out an aggressive response to the COVID-19 outbreak. We used real-time data from a relatively large number of respondents at the peak of the COVID-19 epidemic and reinforcement of social distancing, thereby contributing preliminary information regarding the effect of such regulations on people suffering from chronic medical conditions. Several limitations should be noted. First, although participants were asked to report about their past and present situation, the cross-sectional design of the study does not allow us to determine causality. Second, participants were asked to provide retrospective data about their health status prior to the lockdown, as we did not have access to their pre-pandemic records. A recent study [62] examining the validity of a retrospective measurement of health, by using a test-retest design to measure reliability and agreement between prospective and retrospective patient-reported health, suggests that a retrospective measurement of health is a valid alternative to using reference data for the purpose of estimating past health status. Third, we relied on self-reported data and not on clinical records. For example, height and weight were assessed by self-report, which is likely less accurate than objective measurement. However, studies show that web responders usually provide accurate information about themselves [63]. Relatedly, we do not have information about participants' specific medical diagnoses, and it may be that such information could have provided important insights into the detected variability in SRH. In addition, SRH is a subjective measure and may not always reflect objective measures of health, although a strong predictive capacity regarding various health-related outcomes is well established [6][7][8]. Fourth, the sampling design is subject to biases of internet-based surveys. Participants voluntarily selected to take the survey online, and although the current sample was large and heterogeneous, it was not a representative sample. --- Conclusions Findings highlight the important role of loneliness in health [58], while demonstrating how the near-lockdown measures, announced by the Israeli government in response to the COVID-19 outbreak, may have had a negative impact on the health of people with chronic illnesses. Future longitudinal studies are warranted to clarify the long-term effects of loneliness on the health of people with chronic medical condition during the COVID-19 outbreak. ---

We sought to examine how the near-lockdown measures, announced by the Israeli government in an effort to contain the COVID-19 outbreak, impacted the self-rated health (SRH), health behaviors, and loneliness of people with chronic illnesses. An online cross-sectional survey was carried out about one month (20-22 April 2020) after the Israeli government reinforced the severe social distancing regulations, among a convenience sample of 315 participants (60% women) with chronic conditions (27% metabolic, 17% cardiovascular, 21% cancer/autoimmune, 18% orthopedic/pain, 12% mental-health). Results suggested that about half of the participants reported a decline in physical or mental SRH, and as many as two-thirds reported feeling lonely. A significant deterioration in health behaviors was reported, including a decrease in vegetable consumption (p = 0.008) and physical activity (p < 0.001), an increase in time spent on social media (p < 0.001), and a perception among about half of the participants that they were eating more than before. Ordinal regression suggested that a decline in general SRH was linked with female gender (p = 0.016), lack of higher education (p = 0.015), crowded housing conditions (p = 0.001), longer illness duration (p = 0.010), and loneliness (p = 0.008). Findings highlight the important role of loneliness in SRH during the COVID-19 lockdown period. Future studies are warranted to clarify the long-term effects of social-distancing and loneliness on people with chronic illnesses.

Background When the SARS-CoV-2 virus emerged at the end of 2019, it spread rapidly around the globe because of its ability to transmit pre-and asymptomatically and because the population was completely susceptible to this new virus. The only control option that could immediately mitigate the growing epidemic was the implementation of nonpharmaceutical interventions, including social distancing measures . These social distancing measures aim to reduce exposure by reducing contact rates of for instance face-to-face and physical contacts, as each of these contacts could be an at-risk event for transmission. Contact surveys showed that in the first wave of the COVID-19 pandemic, contact rates decreased by more than half in China [1], the UK [2], the Netherlands [3], and other countries worldwide [4][5][6][7]. The reduction of contact rates effectively decreased the daily number of cases [8][9][10]. Social distancing was especially important to reduce exposure of older adults and persons with frailty, as they have the highest probability of hospitalisation and death after SARS-CoV-2 infection [11][12][13]. Frailty as a health status is a generic concept that can be defined and measured in many ways [14]. In hospital or care settings, the focus is usually on physical frailty, as expressed, for instance, by the frailty index [15] or the Clinical Frailty Scale [16]. In community settings, frailty is often used in a broader sense that also comprises cognitive and psychosocial dimensions, as expressed, for instance, by the Tilburg Frailty Indicator [17] or the Groningen Frailty Indicator [18]. All frailty instruments indicate that frailty increases with age and is generally higher for females than males of the same age [19]. The variation in frailty among persons of the same age and gender is substantial, especially in community-dwelling older adults [20]. Two studies conducted in the Netherlands in 2006-2007 [21] and 2016-2017 [22] showed that participants aged 70 years and older contact on average 7.2 and 7.0 different individuals per day , respectively. Surveys among older adults during the first wave of the COVID-19 pandemic showed a decrease in in-person contacts [23], associated with a decrease in general well-being [24] and an increase in frailty [25]. Contact surveys in the general population also showed a significant reduction of contact rates in older age groups [1][2][3][4][5]. However, none of these contact surveys distinguished participants by frailty. The contact behaviour of older adults may well differ by frailty. On the one hand, older adults with frailty might have fewer contacts due to physical inabilities or social isolation. On the other hand, they could have more contacts when they require more medical or home care. In general, the number of contacts per person decreases by age [26,27] and can differ by gender [26], but it is unknown how this is affected by frailty. As contacts can be used as a proxy for at-risk events for transmission [28], quantifying these contacts by frailty may lead to a better understanding of exposure and hence disease burden in older adults with and without frailty. We conducted a contact survey among persons aged 70 years and older in the Netherlands to quantify the number of contacts per person by frailty status. By conducting the survey during two distinct survey periods with stringent and moderate COVID-19 measures, we aimed to gain insight into how older adults with and without frailty change their contact behaviour when faced with lockdown measures. --- Methods --- Study design Persons aged 70 years and older were invited to participate in the contact survey. For each survey period, invitees were randomly sampled from the Dutch Personal Records Database [29], yielding two independent crosssectional study populations. The invitees were randomly sampled per age group and gender. The numbers of invitees were chosen to obtain similar numbers of respondents in each stratum, assuming lower response rates in older age groups. The questionnaires were unmarked to minimise the collection of personal data. As a consequence, it was not possible to send reminders to non-responders to increase response. Each participant was requested to provide some general information and to fill out a contact diary each day during a full week. The general information consisted of participant age, gender, country of birth, education level, household size, and vaccination status for COVID-19, influenza, and invasive pneumococcal disease . To determine the frailty status of a participant, the Groningen Frailty Indicator [18] was used. This indicator consists of 15 questions on physical, cognitive, social, and psychological conditions. When an answer contributes to frailty, it would score 1, or 0 otherwise. A GFI score of 4 or higher indicates a frail status. The GFI was preferred over other frailty indicators because it contains few questions and is easy to fill out. The contact diary consisted of a contact page for each day of the week, where a participant could report all persons with whom they had face-to-face contact that day. A contact was defined as a conversation of at least a few sentences and/or a physical contact; contacts via telephone or the internet were explicitly excluded. For each contacted person, the participant reported the gender, age , whether this person was a household member, whether the contact was physical, such as shaking hands, whether the contact was protected, for example, by a face mask, whether the contact lasted longer than 15 min, whether a minimal distance of 1.5 m was kept, and the location. The location options included the home of the participant, the home of the contact person, the work place , transit, leisure, shop, outside, and other . Also, a name or description of the person could be reported, not only for ease of filling out but also to identify repeated contacts with the same person on different days of the week. Participants could indicate when they did not have any contacts on a given day, which was helpful to distinguish between participants without any contacts and participants that did not fill out the contact diary on that day. Returned questionnaires that lacked general information and/or any filled-out contact day were not included in further data processing. Questionnaires that were suitable for analysis were entered in a database with internal consistency checks, e.g., whether the participant is at least 70 years old and whether no duplicate participant id's are entered. After data cleaning, the data was reformatted to the standard format for contact surveys on socialcontactdata.org and published online [30]. A full description of all steps in the study design is provided in the supplement . All code for data cleaning and analysis is publicly available [31]. --- Survey periods The study encompassed two survey periods. The first survey period in April 2021 featured stringent COVID-19 control measures: education was online for most students of secondary schools and universities; working from home was required when possible; face masks were mandatory in indoor public spaces; bars and restaurants were closed, as were all other cultural and leisure activities; and an evening curfew was in place. Most older adults had already had the opportunity to be vaccinated, but mass vaccination of the remaining adult population had only just started. In the second survey period in October 2021, most control measures had been lifted. Schools and work places were fully open; face masks were only required on public transport; and social, cultural, and leisure venues were accessible with proof of vaccination or a negative test result. All persons aged 18 years and older were eligible for vaccination, and many had been fully vaccinated. --- Analysis Response rates were calculated by age group and gender for both survey periods. The study population was described by summarising the number of responses by survey period, age group, gender, country of birth, GFI score, household size, and education level. To assess the representativeness of the study population, we investigated the non-response biases of participants not born in the Netherlands and participants in long-term care facilities, as we expected these groups to have low response rates. To this end, we compared the number of participants not born in the Netherlands and the number of participants in long-term care facilities to their expected values, based on the size of these groups living in the Netherlands in 2021 [32,33]. The observed vaccination coverages were compared to the actual vaccination coverages [34][35][36]. We evaluated which characteristics of the study population determine frailty and checked whether this agrees with literature. To focus the further analyses on community-dwelling older adults, participants who live in a long-term care facility were excluded. The contact behaviour was described by how many people a participant contacted and in which age classes. For the analysis, only community contacts were included, i.e., contacts with nonhousehold members, as these are the contacts that are mostly affected by control measures. The community contacts were summed over the full week to eliminate any day-of-the-week effect. Participants with fewer than 5 completed days were excluded from the contact analysis. The missing data of participants with 5 or 6 days completed were imputed while taking the effects of the day of the week and fatigue into account . The weekly total of community contacts was assumed to follow a negative binomial distribution, where the log of the mean was modelled with full interactions between frailty , participant age , participant gender , and survey period . The COVID-19 vaccination status could also affect the number of community contacts, but this covariate was omitted because the data contains few unvaccinated participants. We assessed whether household size should have been included as an explanatory variable by comparing the likelihoods of the models with and without household size using a likelihood ratio chi-square test. In a similar way, the effect of frailty in a specific period and the effect of the period on a specific frailty status were determined using subsets of the data. For example, the effect of frailty in the first survey period was assessed by comparing the likelihood of the model without frailty as an explanatory variable to the likelihood of the model with frailty, using only the data collected in the first survey period. If the likelihoods were similar , the number of contacts did not significantly differ between people with and without frailty in the first survey period. The expected weekly total of community contacts was compared for an average person with and without frailty in the Dutch 70 + population by survey period. To study age-specific mixing patterns, contact matrices were constructed by dividing the total number of contacts per week between two age groups by the number of participants in the participant age group. The resulting matrix of average numbers of contacts per week per participant is asymmetrical due to the skewed population distribution. We assumed people with and without frailty mix proportionally and divided the matrix by the Dutch population distribution in 2021 [37]. Assuming reciprocal contacts, diagonally opposed matrix elements were averaged, yielding a symmetric contact matrix. The resulting contact rates can be interpreted as the average number of contacts per week per participant if the population were uniformly distributed over the age groups. We compared contact characteristics between participants with and without frailty and between survey periods 1 and 2. For each contact, it was reported whether it was unprotected , whether it lasted longer than 15 min, whether it was closer than 1.5 m, and whether the contact was physical. For each of these four high-risk types, the fraction of community contacts per participant was calculated. The differences between the four comparison groups were tested with a Mann-Whitney U test. In a similar way, the fraction of repeated community contacts per participant was analysed. Finally, the distribution over contact locations was compared visually between frailty status and the survey period. --- Results --- Study population In total, 4914 invitations were sent out, and 820 questionnaires were returned, of which 730 were suitable for analysis. The 90 questionnaires not suitable for analysis were either empty with sometimes an explanation why the invitee was not able or willing to participate , without general information , without any contact data , or incomprehensible . Response rates decreased with increasing age group and were higher for males than for females. The overall response rate for survey period 1 in April 2021 was higher than for survey period 2 in October 2021 . The two survey periods consisted of similar numbers of participants . The age group distribution of the study population ranges from 17% for the 90 + age group to 23% for the 85-89 age group, close to the aim of 20% for each of the five age groups. More males than females participated. A large majority of the participants were born in the Netherlands. Based on the number of Dutch citizens with a first-generation migration background by age group [32], we would expect 53 participants with another country of birth in the study population. Instead, only 34 participants were not born in the Netherlands, which cannot be explained by stochastic effects, indicating a lower response rate in this group. Around one-third of the participants had a frail status, according to the GFI. Frailty was higher for females than for males in the same age group, and it increased with age but faster for males than for females . These findings agree with the literature and results reported for the Tilburg Frailty Indicator of communitydwelling older adults [38], which is a frailty indicator similar to the GFI. Most participants lived with one or more household members. According to the number of Dutch older adults in long-term care facilities [33], we would expect 60 participants in care. Instead, only 10 participants indicated that they live in a nursing home or similar long-term care facility, emphasising that the study population is more representative of communitydwelling older adults. One-person households were more common for higher age groups and for females . The education level was, in general, higher for younger age groups and for males . Participants reported whether they had been vaccinated in the last 12 months against COVID-19, influenza, or IPD . The highest coverage was observed for COVID-19 in all age groups. Only the age group 70-74 fell slightly below other age groups in the first survey period in April 2021, as they had only just been invited for vaccination. The overall vaccination coverage in the study population in the second survey period in October 2021 was higher than the actual coverage reported for the 70 + population [34]. The overall vaccination coverage against influenza in the study population was higher than in the general 65 + population [35]. Nothing can be concluded from the IPD vaccination coverage because the IPD vaccination programme for older adults has only been implemented recently for specific age groups [36]. --- Contact behaviour For the contact behaviour analysis, 4 participants without frailty status, 10 participants in long-term care facilities, and 12 participants who participated for less than 5 days were excluded. Of the 704 included participants, 46 provided contact information for less than 7 days. The number of contacts on their missing days was imputed while taking reductions in the average number of contacts per day into account of 4% on each additional participation day and 20% on Sundays . The household size did not affect the number of community contacts per week and is not included in the full model. By comparing full and subset formulations, we found some evidence that the weekly number of community contacts of participants with and without frailty differed in survey period 1, though not statistically significant and differed significantly in survey period 2 . The number of community contacts per week between the two survey periods differed significantly for persons with and without frailty. According to the full model, the expected weekly number of community contacts in survey period 1 was 14 for an average person with frailty and 19 for an average person without frailty. In survey period 2, they were 21 for an average person with frailty and 26 for an average person without frailty. When plotting the full model results with all covariates , we noticed some remarkable patterns. The weekly number of community contacts in survey period 2-with few COVID-19 measures-only slightly decreased with age in any stratum. Females without frailty had more contacts than females with frailty of any age, and males with frailty had the same number of contacts as a 12-year older male with frailty. In survey period 1, persons with frailty of all ages decreased their contacts to a similar extent compared to survey period 2, as demonstrated by the almost parallel lines in Fig. 2. For persons without frailty, however, the lines diverge, showing that the number of contacts in survey period 2 clearly decreased by age. This trend is obvious in both males and females. The same effect is apparent in the contact matrices by frailty and survey period . In all contact matrices, the highest contact rates are on the diagonal, meaning that participants interact mainly with people in the same age group. Participants with frailty had lower contact rates than participants without frailty, and in survey period 1, lower contact rates were observed than in survey period 2. The largest differences between survey periods 1 and 2 are seen for older participants without frailty. In survey period 2, they interacted mainly with other people of similar age, from which they refrained in survey period 1. --- Contact characteristics Contact behaviour change is not only reflected in the number of community contacts but also in the type of community contacts. For all types of contacts, participants shifted to more risky behaviour in survey period 2 compared to survey period 1: the fraction of community contacts per participant that were without protection, closer than 1.5 m, lasting over 15 min, and involved physical contacts increased significantly . For the most part, differences between participants with and without frailty were not significant. One exception is that participants with frailty used protection such as face masks more often than participants without frailty in survey period 1. Another exception is that participants with frailty reported a significantly higher fraction of physical contacts than participants without frailty in both survey periods. The location where contacts take place also differs by frailty status and survey period. Participants with frailty had a similar fraction of community contacts at home in both survey periods and always more than participants without frailty . In survey period 2, participants without frailty had especially more contacts at leisure activities. The fraction of repeated community contacts, i.e., persons who are not members of the household that are contacted more than once a week, did not differ between persons with or without frailty. Persons who lived alone, however, did have relatively more repeated community contacts than persons who lived in a multiperson household . --- Discussion In this survey, we studied how the contact behaviour of persons of 70 years and older in the Netherlands changed during periods with stringent and moderate COVID-19 measures in 2021 and how these changes differed for persons with and without frailty. We found that people with and without frailty had more community contacts in October 2021 with moderate measures than in April 2021 with stringent measures. During both survey periods, people without frailty had more community contacts than people with frailty. When we consider people with and without frailty separately, the number of community contacts only slightly decreased with age in the survey period with moderate measures. In this survey period, the observed frailty status of a participant is a more informative predictor of the number of contacts than age. The two survey periods differed in both timing and in the intensity of COVID-19 measures . Of these two differences, we believe the latter is most essential in interpreting the observed results. Results of an earlier contact survey in 2016-2017 [22] showed little effect of timing within the year on the number of contacts [3]. Therefore, it is unlikely that the differences found between the two survey periods are due to a seasonal effect, and the differences between survey periods are most probably caused by the difference in intensity of COVID-19 measures and the risk perception at the respective times of the surveys. The results might have been affected by a poor representativeness of the study population. The study's overall response rate of 15% was low. One reason for the low response rate is that the returned questionnaires were unmarked by design, which meant that it was not possible to send reminders to non-responders. Another reason is that participation required considerable effort to keep a contact diary for a full week, which made it difficult for invitees with cognitive issues or difficulties with reading and writing to participate. This may have led to a selection bias for participants without frailty. However, the frailty distribution by age and gender of the study Fig. 2 Weekly number of community contacts per participant by age in survey periods 1 and 2. Plots show the data and model results by frailty and gender population agrees with results reported for the Tilburg Frailty Indicator of community-dwelling older adults [38], which leads us to believe that the study population is representative with respect to frailty of the community-dwelling older population. While participants who were not born in the Netherlands were underrepresented in the study population as compared to the Dutch 70 + population, the distribution of frailty status and number of contacts for participants who were not born in the Netherlands did not differ from the other participants. The participants in long-term care facilities, who were excluded from the analysis, were more frail and had more contacts than Fig. 3 Contact matrices showing age mixing patterns for participants with and without frailty in survey periods 1 and 2. The contact rate can be interpreted as the average number of community contacts per participant per week, if the population were uniformly distributed over the age groups the community-dwelling study population. We believe it unlikely that the observed differences in contacts between survey periods and between participant groups have been affected by unintended deviations from the representativeness of the 70 + community-dwelling study population. The number of reported contacts per day by 70 + participants is within the order of magnitude of what would have been expected based on other surveys that address the general population. A possible exception is that in the second survey period, the 3.9 contacts per day is much lower than the 7.2 and 7.0 contacts per day found for 70 + participants in two previous studies in 2006-2007 [21] and 2016-2017 [22], respectively. Aside from the fact that the study populations and questionnaires were Fig. 4 Fraction of community contacts per participant for four risk behaviour factors: protection, distance, duration and physicality of the contact. Distinction is made between frailty of participants and survey period . The whiskers of the boxplots extend to the minimum and maximum values. Significance levels are denoted by *** , ** , * and ns , according to the Mann-Whitney U test different, several other factors could explain this discrepancy. Most likely, this is because participants had not reverted back to pre-pandemic behaviour in October 2021. However, we cannot rule out other possible explanations: having to fill out the diary every day causes a fatigue effect , and correcting for this effect would increase the average number of contacts by 13%; due to COVID-19 information campaigns, participants better understand what constitutes a contact that could possibly lead to transmission, and they may have filled out the diary more conservatively compared to pre-pandemic contact surveys. During both study periods, two other contact surveys were conducted among the general Dutch population. The PiCo survey was held three times per year among a representative sample of the population in the Netherlands [3,39]. The CoMix survey was held every two weeks on a selected internet panel [40,41]. The 70 + participants in the PiCo and CoMix surveys reported 2.3 and 2.5 contacts per day around survey period 1 and 5.1 and 4.8 contacts per day around survey period 2. While these studies agree, they find a larger difference between the survey periods than this study: 2.8 and 3.9 contacts per day in survey periods 1 and 2, respectively. This could be an effect of differences in study setup, questionnaire, study population, or fatigue. For instance, the 70 + participants in both the PiCo and CoMix surveys are on average younger than in this study, where older age groups are oversampled. Summarising, where we find larger differences between the number of contacts as reported in the current study and the number reported in earlier studies, these can be explained by the different age composition of the surveys and by different contact behaviour in October 2021 as compared to pre-pandemic levels. None of these differences affect the interpretation of our findings. In previous studies [26,27], it was observed that the number of contacts of older adults decreases with age. The results of this study show that in survey period 2, which does not fully but most closely resemble a normal situation, the number of community contacts only slightly decreased with age . Because persons with frailty have on average fewer contacts than persons without frailty, the decrease by age as observed in previous studies without frailty distinction could also be explained by persons transitioning from non-frail to frail status as they grow older. Especially for females, frailty is a better predictor of the number of contacts than age. In survey period 1, with more stringent COVID-19 measures, the number of contacts of persons without frailty did decrease with age. These older persons without frailty interact mainly with persons of similar age who are increasingly frail . As a consequence, their behaviour change indirectly protected their peers with frailty. Younger persons without frailty did not decrease their contacts as much, possibly because they did not interact with persons with frailty as much, and persons with frailty themselves all decreased their contacts, probably to protect themselves. The results of this study reveal how social distancing measures affected the contact behaviour of persons with and without frailty aged 70 years and older during the COVID-19 pandemic in the Netherlands, with the largest decrease in number of contacts for the oldest participants without frailty. These results can be useful in different ways. Frailty questions could be included in contact surveys for the general population, as frailty is an additional indicator for the number of contacts. Stratifying infectious disease models by frailty could increase knowledge on how people with and without frailty are affected under different control scenarios, although it would first be necessary to know how people with and without frailty mix. These results can also be instrumental in public health policy, for instance, by shaping information campaigns on social distancing measures. As the population in many countries ages rapidly, it is becoming ever more important to take the frailty differences of older adults into account to be prepared for future pandemics. --- --- --- --- --- Competing interests The authors declare no competing interests. ---

Background During the COVID-19 pandemic, social distancing measures were imposed to protect the population from exposure, especially older adults and people with frailty, who have the highest risk for severe outcomes. These restrictions greatly reduced contacts in the general population, but little was known about behaviour changes among older adults and people with frailty themselves. Our aim was to quantify how COVID-19 measures affected the contact behaviour of older adults and how this differed between older adults with and without frailty.In 2021, a contact survey was carried out among people aged 70 years and older in the Netherlands. A random sample of persons per age group (70-74, 75-79, 80-84, 85-89, and 90 +) and gender was invited to participate, either during a period with stringent (April 2021) or moderate (October 2021) measures. Participants provided general information on themselves, including their frailty, and they reported characteristics of all persons with whom they had face-to-face contact on a given day over the course of a full week.In total, 720 community-dwelling older adults were included (overall response rate of 15%), who reported 16,505 contacts. During the survey period with moderate measures, participants without frailty had significantly more contacts outside their household than participants with frailty. Especially for females, frailty was a more informative predictor of the number of contacts than age. During the survey period with stringent measures, participants with and without frailty had significantly lower numbers of contacts compared to the survey period with moderate measures. The reduction of the number of contacts was largest for the eldest participants without frailty. As they interact mostly with adults of a similar high age who are likely frail, this reduction of the number of contacts indirectly protects older adults with frailty from SARS-CoV-2 exposure.The results of this study reveal that social distancing measures during the COVID-19 pandemic differentially affected the contact patterns of older adults with and without frailty. The reduction of contacts may have led to the direct protection of older adults in general but also to the indirect protection of older adults with frailty.

Introduction 爷爷: 爷爷好想你们,爷爷想见到你们。爷爷觉得瑞 瑞的中文说的比原来有很大进步。 妈妈: 是吗? 爷爷: 是, 比原来好, 而且爷爷觉得是个飞跃! Grand: Grandpa misses you very much , wants to see you. Grandpa thinks that Rui's Chinese has improved so much. Mom: Really? Grand: Yes really, it's much better than before. Grandpa thinks there is a leap! Mrs K 今天说, 我和Sherry不可以在学校讲中文, 因 为别人听不懂我们说什么。But it was after class, 她 真的很rude。 Mrs K said to us today that I can't speak Chinese with Sherry, because others don't understand what we say. But it was after class, she is really rude. 因为太忙了, 真的是有心无力呀, 真的是, 有心无力, 因为两个孩子, 学校留了好多作业, 所以放在中文上 的时间真的是很少了. Because of our busy life, we have the aspiration but not the energy. Seriously, have the heart and will but not the strength. We have two children, there are so much homework from their schools, so there is really little time left for Chinese. These short excerpts capture the experiences of pride, frustrations, and struggles that many migrant and transnational families encounter when dealing with everyday life involving minority and heritage language in diasporic contexts. In the Zhang family, we see the sense of pride strongly expressed by the grandparent and mother. In Rachel's remarks, we hear her anger over Mrs K's 'othering' demand and her school's endorsement of monolingual practices through language policing ). And in Mrs Chen's account, we feel her sense of powerlessness, her frustrations and struggle with the unequal access to minority language education in the public educational system. These emotional comments and remarks remind us that raising multilingual children and developing minority/HLs are matters that reach far beyond the family domain, because families are always "nested in a wide range of socio-historical, political, cultural and linguistic environments" . As a result, the desired family language practices are conditioned by the sociopolitical environment, filled with linguistic prejudice and hierarchy, that may compel minority language shift. Although family language policy research has grown rapidly in recent years, we continue to grapple with questions about why a large number of minority children become monolinguals despite growing up in a multilingual environment, and what normative concepts about migrant languages and families reflect and reenact institutional power and thus allow marginalisation and educational inequality to continue to prevail. In this paper, we explore how transnational families in the UK find pride in their children's ability to use multiple languages, and how their experiences of raising multilingual children are set back by implicit 'institutional policy' and societal discourses against minority languages, as well as hidden hurdles in public educational systems that create language hierarchies and prevent multilingual development. By focusing on the ideological constructs of 'pride', 'prejudice' and 'pragmatism' , we explore how different language practices are established and what language management activities are implemented in transnational families. --- Pride, prejudice, pragmatism, and family language policy The field of family language policy has gained much recognition in recent years. Much of current FLP research draws on Spolsky's triadic model of language policy, which consists of three interrelated components: language ideology, language practice, and language management. While language practice and language management can be overlapping, the former emphasises language use through socialisation and the latter focuses on deliberate efforts. When examining language ideologies, researchers often find that parents and children attach great symbolic importance to home/heritage languages in relation to the development of an ethnolinguistic identity, emotional engagement, and cultural practices . The feeling of 'pride' is often associated with symbolic values such as "individuals' identity, loyalty or feeling of belonging to an ethnic community" . Duchêne and Heller coined the term "language as pride" to describe how language is legitimised and used as a means for identity work. They argued that language use and practices in the process of legitimisation can invoke 'a sense of belonging', thus mobilising 'feelings of pride'. Research into FLP within transnational and migration contexts has repeatedly reported that HL maintenance is conflated with cultural and ethnic identity . Cultural inheritance and a sense of belonging are frequently framed as a pride in the socialisation discourse of diasporic communities and immigrant families. Curdt-Christiansen's study of Chinese immigrant families in Quebec showed that parents equalled the HL competence with a pride in 'Chineseness'. This perception of pride as reflected in HL competence, a recurring theme in FLP studies, is echoed by many parents of first-generation migrants . Regarding language ideology in FLP, parents see it as their responsibility to socialise their children into the membership of a HL culture and community, thus fostering connections and communications with families and their imagined 'homeland'. Such a pride-based cultural and language socialisation often become a driving force for parents to keep up with "cultural loyalty and linguistic continuity" . A pride-based ideology, however, is not problem-free because 'prejudice' tends to occur when differences are exhibited in language, race, culture and values. Prejudice is often rooted in raciolinguistic ideologies or intolerance of migration and minority languages . Curdt-Christiansen and Huang , in their study of Chinese communities in the UK, found that experiences of racist violence and xenophobic rant often accompanied migration. Earlier Chinese migrants in small businesses were often the target of 'othering' and racism because of their limited and non-standard English. Similar results have been found in studies of other racialised populations in the US, Canada and many other immigration countries . As a result, prejudice is produced in the process of 'othering' and framed by language use and practices that lead to social stratification and inequality. Prejudice and othering can "create a deep sense of marginalisation and exclusion, and take away the sense of 'pride'" from migrant families . As illustrated in Rachel's conversation with the researchers in the beginning of the paper, xenophobia is conveyed through 'official language policing' to control and regulate language practices in schools. Unpacking educational policies against nonstandard English language in the UK, Cushing found that "language policing is a pervasive practice that policy arbiters engage in across all levels", that is, micro , meso , and macro . While Cushing's study focuses on the varieties of English in UK schools, the implications of such language policing extend to languages other than English, as shown in the case of Rachel. Thus, the roots of persistent racial inequality are reproduced through raciolinguistic ideologies in education . When confronted by raciolinguistic ideologies and prejudicial discourses as well as language policing in schools, parents and children may choose a 'pragmatic' solution by giving in to public educational demands and giving up their HL despite their pride in and emotional and cultural attachment to the language. Pragmatism concerns the ways in which families deal with the challenge of HL development and try to make it fit the practical circumstances. Such pragmatical attitude has resulted in contradiction and inconsistency between a 'pro-minority' ideology, expressed by the parents, and an 'anti-minority' language implementation/practice in reality, as shown in many FLP studies . Bohnacker , for example, explored FLP in 105 Turkish families with children 4-7 years old in Sweden. Using a questionnaire that included language practices and home language literacy activities, the study found that while parents held strong pro-Turkish maintenance ideologies, they tended to make limited maintenance efforts. Irrespective of the parents' educational background and country of birth, the study reported that "hardly any families had arranged for alternative private tuition in Turkish" . Curdt-Christiansen and LaMorgia's study of Chinese, Italian and Urdu-speaking communities in the UK showed similar results regarding language management efforts in the families. While parents from the Chinese and Urdu-speaking communities showed positive attitudes towards their respective HLs and believed that HL development plays an important role in cultural practices for ethnic identity and the emotional relationship between generations, they provided insufficient linguistic measures and literacy practices for developing their children's HLs. Similarly, Dekeyser and Stevens examined 300 Moroccan children's language proficiency level in HL and Dutch from an FLP perspective. Their study showed that children's proficiency level in HL was related to their parents' management efforts but not to their attitudes towards HL, despite the parents' pro-HL ideology. These above-mentioned studies and many other relevant studies all point out that the inconsistency between pro-HL ideology and insufficient HL management efforts is related to lim-ited HL support and resources found in public educational systems and community facilities. The limited support and resources may not directly suggest that there is a relationship between prejudice against minority languages and HL provision, but leaving HL education to the parents and ethnic communities as their responsibility is an indication of social inequality and a denial of inclusive education. A consequence of the limited public support of HL education has led many parents to make pragmatic decisions and let go of their HLs . A recent study by Huang and Liao reports that the many challenges met by Chinese interlingual families in Australia lead to inconsistent parental attitudes towards supporting Chinese HL maintenance. The status of the Chinese language in Australia and the children's afterschool activities in English have made the parents reluctant to provide consistent support of HL learning. Curdt-Christiansen and LaMorgia's study in the UK highlighted similar challenges. The parents' frustration demonstrates not only the competing forces between public educational demands and HL development, but also the struggle in the language status between English and HL . 'Having homework to do every day' has pushed HL to the periphery and helped build a linguistic hierarchy that has created power relations by making many migrant parents and children accept 'othering' and prejudice as normalised practices. The normalised acceptance of 'othering' by minority families is an acute illustration that families do not live in a vacuum but constantly interact with broader sociopolitical contexts in which linguistic hierarchy prevails. This reminds us that FLP is inevitably shaped by political and societal ideologies . Cushing's ) studies of UK's educational policies and school language practices point out that racist ideologies about language are pervasive, not only in the lives of minority families but also in schools and society at large. He argues that hostility towards and prejudice against people based on their languages is a form of language oppression that "can intimidate people into modifying their use of language or abandoning it completely" . Such language-based 'othering' requires us to move beyond what is happening within the family domain to look at what is happening in schools and the wider society. In what follows, we present how notions of pride, prejudice and pragmatism as ideological constructs are intertwined to form the mechanisms, the Family Language Policies, through which social inequalities are reproduced. --- The study This study is part of a large research project, funded by the Economic and Social Research Council , that explores how mobility and on-going changes in sociocultural contexts impact family language policy in the UK . As the second large-scale, nationwide project, undertaken 30 years after Alladina and Edwards' project on multilingualism in the British Isles, our project focuses on a multi-level investigation of family language policies in the UK. In this paper, we report data collected from the survey at the national level. In order to enrich our discussion of the ideological constructs of pride, prejudice and pragmatism, we substantiate it with interview data from three Chinese and two Polish families. --- Questionnaire The team developed the questionnaire by drawing on existing research and theories related to FLP. To ensure content validity of the questionnaire, two workshops were conducted, both with experienced colleagues and researchers who were involved in raising multilingual children. Using Spolsky's triadic model, language ideology, practices, and management, as a foundation, an initial draft was constructed based on comments and suggestions received, and then piloted with more than 30 families to ascertain that the questions were clear and would be interpreted as intended. Further revisions were made after the first pilot. The revised survey was piloted again with another small group of families, and the questions were further refined. The finalised multilingual survey consists of five sections: demographic information ; language ideology ; language use for different activities ; language practices after first child ; language management . In the language ideology section, statements were given on a 5-rating scale . The other sections were all multiple-choice questions where answers were selected from a dropdown menu. The questionnaires were then distributed in printed and online forms. The printed versions were distributed in HL schools, at cultural events, in ethnic community supermarkets, and in community centres in different cities. The online survey was uploaded to SurveyMonkey and 问卷星 and sent to our social networks and schools. We received a total of 512 responses. After data cleaning, 470 were suitable for the study. See Table 1 for participants' educational level and length of residence in the UK. --- Semi-structured interviews Semi-structured interviews were conducted in parallel with the survey. We selected families from the Chinese and Polish communities because of their distinctive sociopolitical contexts of migration, demographic characters, and linguistic/cultural practices. Five families were involved, three from the Chinese community and two from the Polish, representing different family types in the two communities with regard to family size, educational level of grandparents and parents, and the different language experiences of the family members. The involvement of these families was not intended to reach generalisability, but to enrich and compliment the survey data. As members of the involved communities, we see ourselves not as 'researchers' to collect facts and information from the families; rather we view them as co-researchers with whom we built a mutual trust and make sense of their accounts through shared knowledge as migrant parents with shared migration experiences . Data collection, thus, is socially constructed through negotiations in Putonghua and Polish. We present the families' views to instantiate how FLPs are established, negotiated and changed based on the families' perceptions of the different languages in their life. Table 2 illustrates the profiles of the families. Interviews were conducted three times as part of the ethnographic research design of the project, either face-to-face in the home or through a social media platform when requested by the families. The overall interview structure was designed as a follow-up to mirror the questionnaire. But as we engaged with the families through socially constructed negotiations, they provided much elaborate knowledge and engaged with us in in-depth discussion about their language practices and experiences. Each interview lasted between 20 and 45 min. The interviews were transcribed in Chinese or Polish first and then translated into English by the researchers and the research assistants . --- Data analysis The questionnaire data were analysed descriptively, based on the three components of language policy -language practice, language ideology, and language management. When coding the data, we used the constructs of 'pride', 'prejudice' and 'pragmatism' to code their different ideologies, and underline patterns of practices, and management measures. For example, the statement, "It is disadvantageous for a child if only the non-English home language are used at home", is coded as prejudice. The interview data were then analysed through the lens of PPP when understanding participants' perceptions of their language and migration experiences, and FLPs. For reliability, two researchers, independent of each other, coded the data and then compared and discussed their results to reach consensus and attain accuracy of the findings . --- Findings The findings of the study are organised according to the three components of the language policy concept -language ideology, language practice, and language management. Throughout the discussion of the findings, we present the way in which PPP are reflected in language ideologies and interact with practices and management by triangulating the survey and interview data. --- Pride, prejudice and pragmatism in self-reported pro-multilingual ideology Within the study of FLP, language ideology has been recognised as the driving force for the establishment of different types of FLPs . As beliefs about prestige, privilege, and discrimination inform language ideologies , we illustrate how these different beliefs about language prestige and language privilege are implicitly instantiated through a variety of statements made in both the survey and interview data. As shown in Table 3, most families reported a strong pro-multilingual ideology and agreed that speaking more than one language had a positive impact on their children's cognition, academic performance, and further development. Pride is illustrated in statements 7-10, 12 and 14, prejudice is reflected in statements 1-6, and pragmatism is shown in statements 11 and 13. In general, there is a strong belief against prejudice as indicated in the high percentage of agreement with statements 1-6. For example, more than 80% of participants disagree or strongly disagree with statement 2 that "Speaking the non-English home language may play a negative role in a child's ability to master English". Statement 1 are used at home) received a relatively low percentage of disagreement , but statements 3 and 4 both received more than 76% disagreements. While these statements indicate that transnational migrant families are aware of the benefits of bilingualism/multilingualism, they reflect largely the cognitive benefits and debunk language confusion and academic disadvantages. Statements 5-6 elicit similar responses from the participants as they disagree or strongly disagree that home language learning is a private matter and that speaking other languages is unacceptable in the UK. With regard to their beliefs about the relationship between home language maintenance and identity in terms of pride , we see also positive responses with 21.8% strongly agreeing and 34.1% agreeing. Comparing to the prejudice statements 2 and 3, and the pride statements 12 and 14 on the benefits of bilingualism, the percentage of positive responses is lower. This may indicate that ethnic and cultural identity in the form of pride is not as essential as other aspects of raising bilingual children. The pragmatic concerns, such as children's academic development and future career, overtake the sentiment of pride in maintaining the home language. We paid particular attention to prejudice statement 1 are used at home), pride statements 7 and 14 to explore whether prejudice is implicitly reflected in their responses to language pride. It was interesting to note that responses showed some inconsistency in distribution patterns: while 35.8% disagree and 33.6% strongly disagree with statement 1, indicating that these participants do not believe such prejudice that using home-language can be disadvantageous for their children, only 20.3% of respondents strongly agree with the pride statement that they feel comfortable when using the home-language in public. The same pattern is repeated between statement 7 and 14 in which over 80% of respondents are happy that many people speak other languages in Britain but only 20.3% feel comfortable using their home-language in public. This phenomenon has been discussed extensively in our interviews with participating families. In the Polish community, speaking Polish in public seems to be a matter that is consciously monitored by the participants. For example, in family PoLA, Penny and Krisztina reflected on their experiences: K: I have Polish friends and when we meet up, we speak Polish. In the beginning, we didn't speak loud or anything but there were those stares from people. And later 1 3 then automatically we speak quieter but there was no situation in which somebody actually scolded us, like "go back to your country", told us this directly. It's more of a feeling… there's such a discomfort. P: I also have such a feeling. It's not that I'm embarrassed that I'm from Poland. When we are at the bus stop or in a shop, it's obvious we will speak Polish, it's easier in Polish so we lower our voice as much as possible because I don't want these stares or somebody telling me 'Speak English'. I prefer to lower my voice or not to speak and wait until we are on our own to be confronted with such stares. The discussion between the Krisztina and Penny underscores the invisible sense of 'fear' from the public 'othering'. They both emphasised the "stares from people" and the "feeling of discomfort" when speaking Polish in public. Although they had never experienced overt racist violence as described in other studies , their social and linguistic reactions -"to lower my voice or not to speak and wait until we are on our own" -revealed a deep-sense of prejudice from the UK society where speaking languages other than English may be seen as deficient and unacceptable in public and educational domains. Speaking LotE in a largely monolingual society like England may give away one's migrant status, social position and economic status. This echoes with a plethora of sociolinguistic studies demonstrating that language attitudes towards and prejudice against certain language speakers and races can create differences and generate marginalisation that leads to social practices of denigration and exclusion . Penny also pointed out another critical issue in the dichotomy of pride and prejudice. She clearly felt the pride of being Polish as indicated by "I'm embarrassed that I'm from Poland". This covert pride, however, tends to be challenged in the public space. Her mother, Mrs A, highlighted the unspoken fear of speaking Polish: Mrs A: But sometimes I speak with P. I ask something. Somebody is coming and she whispers 'silence' and I have to stop talking to her. R: But that's because if anyone comes in, mostly English, I don't want them to stare at us because we speak Polish. So I say 'silence' to wait it out and then we can talk later. And the second thing is for this person not to feel uncomfortable. Again, the fear of 'stares' resurfaced. Acting upon this 'fear', Penny didn't lower her voice with her mother, instead she 'silenced' her mother in order to avoid 'overt' prejudice from the "mostly English". Family PoLA's experience was not a single case, Ms PoLB had a similar linguistic reaction when speaking to her son, Stan, in public. Mrs B: I almost always speak Polish to Stan. After all, it is my emotional language. But I can't help speaking English to him in public. I think it is an automatic reaction. R: Why? Ms B: When I was a student, I used to work at McDonald's. People can be quite rude when they hear that I have a Polish accent…I think I automatically react to 1 3 situations where I don't want people to judge me. If I speak Polish, people know I am from Poland, and that I am an immigrant. But if I speak English, not everyone can identify my accent. If Penny and Krisztina's sense of fear came from 'unfriendly' stares, Ms B's came directly from her working life experience as a migrant person. Her experiences shaped her ways of dealing with prejudice, namely to change/modify her language use in public . While this shift in language use may seem inconsequential, it sends an implicit signal to children indicating that not all languages have a "recognised legitimacy" . In this regard, prejudice through 'othering' has planted a seed for language shift. Similar comments are found in dialogues with the Chinese families as illustrated in the opening excerpt from Rachel. Due to space limitations, we have restricted our analysis of Chinese families to data concerning language practices and management. When examining the descriptive results, we found that there is an inconsistency between prejudice statement 5 is a private matter that should be done only at home) and pride statement 8 . Responses to the former showed 75.8% disagreement . But to the latter, less than 50% positive responses were given with 18.9% strongly agree and 30.2% agree. This inconsistency may suggest that maintaining the home language may be less attainable for minority groups from a pragmatic perspective as it needs public funds to develop different minority language programmes. Cunningham reported that even teachers with a minority background in UK public schools believed that the development of HL is a private matter and that the responsibility remains with families. The inconsistency could also suggest that although most parents believed that learning the home language was not a private matter, they nevertheless believed that some minority languages should not be supported at the cost of public funds. This again shows that not all languages and language-speakers are perceived as equals, not even by the minority-language speakers themselves , and that language prejudice is found in every corner of our society. The tensions and ambiguities arising from the coexistence of different languages, practices and ideologies suggest that linguistic hierarchies invoke power relations in multilingual societies. --- Language ideology and self-reported language practices Similar to many sociolinguistic studies , our findings also indicate perceptible patterns of language change from parent to child generation, as indicated in Table 4. As Table 4 shows, language shift from home-language to English not only took place among siblings, but also between parent and child. While parents used only 11.5% of mainly English when speaking to children, they use 54.5% of English and 23.5% of mixed language during homework sessions. The other noticeable change was in writing activities in which 38.8% reported using mainly English. While these changes may indicate a pragmatic act, namely parents' unconscious language decision in their attempts to align their language practices with that of the school, they also indicated the institutional constraints on providing support and facilities to develop and maintain home languages . In connection with prejudice statement 5 is a private matter that should be done only at home) and the pride statement 8 , the linguistic practices revealed a non-spoken but taken-for-granted language 'prejudice' through an institutionalised practice where minority languages are given little legitimate recognition. When further exploring parents' language practices after their child's birth, we found revealing results demonstrating their unconscious accommodation to societal language practices resulting from a coerced monolingual ideology and invisible prejudice. Table 5 presents self-reported data about the parents' language practices after their first child was born. --- Language behavior NA No Yes My own language practices changed 5.7 48.6 45.6 My mixed use of language increased 3.9 45.9 50.2 I started to use different languages with different people at home 5.7 63.4 30.8 I started to use more English 6.3 57.1 36.6 I started to use more my non-English home language 5.8 67.7 26.5 After my child started nursery or primary school, my own use of English increased 6.1 54.7 39.3 After my child started nursery or primary school, my own use of non-English home language increased 5.4 72.5 22.1 The descriptive data indicated that about half of the parents changed their language practices after their first child was born. A higher proportion of them started to use English more than their non-English home language . The changes were consistent with the language use between parents and children in Table 4. These descriptive analyses indicate that the increasing number of parents using more English with their children might be due to concerns about their children's English language development and also due to pressure from the public educational system, in which a primarily monolingual-rooted ideology in education prevails ). These concerns not only changed their FLPs, affecting next generation's language attitudes, but also 'forced' them to give in to the invisible prejudice reflected in the lack of educational support for minority languages, and choose a pragmatic solution, as illustrated by the case of ChiA: We became very worried about Meimei's communication skills at the nursery. Especially when the teacher said that Meimei hardly spoke at all. But she was very bubbly at home, always talking non-stop. We were so concerned. this was clearly a language issue. That's why we decided to change our home language into English. Then later on, English was used more and more, then Meimei's Chinese became less developed. Now she speaks Chinese like a foreigner. Mrs A recalled and reflected on the critical moment of decision-making when changing her FLP. Before Meimei was sent to a nursery, the family language was predominantly Chinese. But Meimei's uncommunicative behaviour in the nursery had worried the ChiA's parents as they believed that their speaking Chinese was the problem. Such worries appear to be widespread, as existing literature suggests that many parents worry about their children's mainstream language development in the host country . While many immigrant parents, like Mrs A, have to deal with the social realities of public educational demands , the desire to maintain family language practices becomes an ad hoc pragmatic choice. The on-going linguistic competition between minority and dominant languages has involuntarily forced many parents to make a pragmatic decision to discontinue home language learning, as was the case in the Chen's at the beginning of the paper. The sentiment is also shared by other families, such as the ChiB's. R: In addition to speaking to children in Chinese, are there any other activities that you would use to engage them in learning, like watching Chinese entertaining programmes, reading books, singing or doing Chinese related activities? Mr B: We did some in the beginning, but then less and less because we are too busy. Really, we are too busy, we have the heart, but not the energy. As you know the school also give them loads of homework, so the time left for Chinese becomes less and less. Once school is done, everyone is, the kids are, tired. We really don't want to do much. So the only way we keep up their Chinese is to speak. It is hard to spend time reading a book, practise writing, too difficult. Mrs B: just don't have any energy for it, too tired to do much, really, that's how it is. Mr and Mrs B shared their frustration about time constraints for developing their sons' Chinese. While keeping up with schoolwork was prioritised because of the educational demands, such prioritised decisions may overtly and covertly 'coerce' parents to promote English and pragmatically let go of Chinese. Subsequently, the negotiation between educational reality and linguistic continuity has resulted in their compromise for their children's Chinese language development. --- Language ideologies, and language management Taking into consideration the different economic resources and language facilities that families may have, we provided a variety of actions that parents may take as measures for managing and developing their home language. Table 6 shows the preferred choices of actions taken by parents. It is evident from the table that parents were aware of how important it is to provide children with opportunities/exposure to the non-English home language. All the top choices, such as organizing frequent visits to home country and maintaining virtual contact with family members are, however, temporary and periodic activities. These infrequent experiences can be further restricted by travel bans and may become a financial burden on the family. Although maintaining virtual contact with family members is an important way to expose children to their heritage language, these contacts have limited socialisation effects and tend to provide inconsistent language experiences, as expressed by participating families: Real conversation is impossible. The kids can say things like good day, goodbye, then nothing serious. Sometimes, they can say 'have you eaten?' Both the survey data and the interview data provided evidence to illustrate that pragmatic measures, taken when the school language is given priority, can lead to the loss of the home language. More importantly, the loss of the home language becomes the plight of both parents and grandparents, who may be in the process of losing close contact and emotional connection with children who speak less and less their home language . Our interview data showed the emotional reactions to this plight from some parents: In fact, my root is there , just like earlier when I mentioned the communications with families. I feel so guilty that my children can't communicate with my mother, no frequent and deeper communication. But deep in my heart, I want it to happen so much, so much. When the older one was little, before she turned two, I spoke Chinese all the time. But then later it seemed that I had no choice anymore…it's because there were so many problems and issues that you have to deal with in life…when everything falls down upon you, you just don't have any choice. You don't have a choice to speak English then Chinese, and back and forth. What I want to say is that we have our roots there, although we don't speak Chinese . But in our heart and culturally, we are Chinese. As a mother, I want to reconnect to the root, but how should I do it? We have little means to … This emotional confession captured the exact dilemma of the ambiguities inflicted by the invisible 'non-inclusive' language education curriculum. The deep pain and guilt felt by Mrs C underline the processes whereby prejudice against minority languages, promoted by public education, has been internalised as natural and normal, and that failing to provide heritage language education becomes a purely parental fault. The repeated use of "our roots", "in our heart" and "no choice" by Mrs C illustrated her battle against prejudice resulting in the loss to pragmatism. Despite her desire to reconnect to her roots or "our roots", her attempts to keep up with the pride by speak-1 3 ing Chinese to her children and get her children to communicate with their family in China were fruitless. --- Discussion In this study, we explored the FLPs of transnational families in the UK through the ideological constructs of pride, prejudice and pragmatism . By challenging the educational policies and institutional structures that undermine the minority/home language development, our study illustrates how these policies and structures naturalise and normalise the practice of linguistic prejudice in families as reflected in their language ideologies, practices and management. Our study provides illustrative evidence through sociolinguistic survey and interviews to demonstrate how lived experiences of 'othering' and pride are contradicting each other, forcing family language policy to give in to pragmatic measures that perpetuate linguistic racism through implicit, covert and overt raciolinguistic ideologies . In what follows, we present a summary of our findings and discuss three critical aspects in the studies of FLP. --- Linguistic prejudice ideology and FLP Firstly, the findings in our survey indicate that the majority of families did not overtly subscribe to a prejudice ideology, as evidenced by their self-reported pro-multilingual beliefs. Although a large number of parents felt strongly about maintaining and developing their home languages, indicating pride, they tended to view such pride as not essential. Our interview data show strong evidence that pride and prejudice are two ideological constructs that contradict each other. Pride, for example, in the form of ethnic identity and the use of the home language in public can often be challenged by 'racist stares' and xenophobic school policing, as experienced by Polish families and Chinese individuals. The hostile comment from the teacher and the selfcontrolled 'silence' and 'lower voice' from migrants are complex linguistic behaviours mediated by raciolinguistic ideologies. For Mrs K, Rachel's conversation with her friend in Chinese was viewed as a "violation of normative and homogenous linguistic codes" in UK schools and therefor had to be stopped. For the PoLA's, 'silence' and 'lower voice' were the 'self-inflicted' regimentation in support of a prejudice ideology that sanctions ''language oppression" in the UK society . Both acts are products of raciolinguistic policing which demonstrates how perceptions of minority languages can become central to the construction of social problems, thus contributing in a profound way to the shaping of FLP. --- Tensions between pride, prejudice, pragmatism and language practices Our findings of language practices indicate a clear pattern of language shift in the families . This language shift is a direct result of covert and overt prejudices which these families have encountered either in schools, in public domains or elsewhere. As such, some families have formed conflicting practices and management strategies, where what they view as a pride and what they do in reality are contradictory. Such contradictions are not new to the field of language policy as reported in multiple studies . The current study sheds new light on the reality of the lived experiences of minoritised individuals, families and communities, in which linguistic prejudice is foregrounded in relation to FLP. Where there is linguistic prejudice, there is inequality; and inequality will always place "new demands on racialised populations to modify their behaviours" linguistically and culturally , as illustrated by the language practice shift in the Chinese families. The findings indicate that FLP can serve as a critical site for understanding and contributing to the reproduction of social inequality and raciolinguistic practices, which leads us to the discussion of normalised perceptions of home language development as a private matter in FLP. --- Normalised language ideology, home language and pragmatic FLP The findings about language management indicate that the majority of families chose to organise frequent visits to their home country and maintain virtual contact with family members . Only 36% of parents reported sending their children to HL schools. In relation to self-reported beliefs that they consider it beneficial for children to attend HL schools, the results are contradictory. This contradiction requires us to refocus on FLP to unravel the normalised perception of home language learning as a private matter. On the one hand, parents believed that learning HL was beneficial and viewed attending a HL school as a pride. On the other hand, they had little time and few resources to engage their children in HL learning. Although close to 50% of the participants believed that government support was needed for HL development, the interviewed participants largely believed that HL learning was not part of the education agenda. In the public educational system in the UK, HL remains largely invisible, either because of raciolinguistic policing or as a consequence of an institutionally sanctioned curriculum which has led to a normalised ideology that HL education is the responsibility of the parents and the ethnic communities . Such normalised beliefs have invoked emotional distress and a sense of guilt in parents when they conceded to pragmatism and opted for giving up HL. For many parents, this is a logic solution when they face the choice between meeting the public educational demands or keeping up with language loyalty. The institutional pressure for doing well in British schools has forced families to either provide more space for English or give up HL totally. It is not surprising to see that institutional pressure can turn FLP into a mechanism that "supports the raciolinguistic status quo" . Lanza argues that families can provide a space in which "we can choose the language or languages we want to speak, express the ideologies or attitudes we have concerning different languages, and construct the identity we wish to con-1 3 struct through our own language choices." Families are, however, not a safe haven. We have witnessed language changes, language death, and language loss in families and communities, but we have yet to unveil the roots of persistent racial inequality in our society and school system. Although our study is situated in the UK, FLP issues similar to those found in the UK can be found in many other parts of the world. Our study thus indicates that FLPs alone cannot save the minority languages. We call for more research to explore how institutionally sanctioned language practices and raciolinguistic ideologies are shaping FLP in super-diverse societies like the UK. We also call for more studies on home-school collaboration to understand how legitimising the home language in school contexts can provide children with opportunities to develop their multiliteracy knowledge, and to build their confidence in themselves and their pride as multilinguals. --- Conclusion --- --- Xiao Lan --- Authors and Affiliations Xiao Lan Curdt-Christiansen 1 • Li Wei 2 • Zhu Hua 2 Xiao Lan Curdt-Christiansen [email protected] Li Wei [email protected] Zhu Hua [email protected]

In this study, we examine how mobility and on-going changes in sociocultural contexts impact family language policy (FLP) in the UK. Using a questionnaire and involving 470 transnational families across the UK, our study provides a descriptive analysis of different family language practices in England and establishes how attitudes influence the different types of FLP in these families. Complementing the descriptive analysis, we use interview data to understand the driving forces behind the different types of language practices and language management activities, and explore how ideological constructs of 'pride', 'prejudice' and 'pragmatism' are directly related to negative or positive attitudes towards the development of children's heritage language. The findings indicate that migration trajectories, social values, raciolinguistic policing in schools, and linguistic loyalty have shaped family decisions about what languages to keep and what languages to let go. Our paper responds to the linguistic and demographic changes in British society, and makes an important contribution to our knowledge about multilingual development of children in transnational families. Critically, this study shows that FLPs alone cannot save the minority languages; institutionally sanctioned language practices and ideologies have to make a move from limiting the use of these languages in educational contexts to legitimising them as what they are: linguistic resources and languages of pride.

Introduction Quality of life constitutes an ambiguous concept which is described very differently depending on the profession and specialty of the person who defines it [1]. Most researchers dealing with this issue state that there is no unambiguous, precisely defined definition of this notion. According to Delkey and Rurke, quality of life is the sensation of prosperity, satisfaction or dissatisfaction with life, i.e., happiness or unhappiness [2]. In medicine, the concept of quality of life did not appear until the 1970s. Earlier, it had been specified by a definition of health formulated by the World Health Organization: "Quality of life is defined as individuals' perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns. It is a broad ranging concept affected in a complex way by the person's physical health, psychological state, level of independence, social relationships, personal beliefs and their relationship to salient features of their environment". [3] Human health behaviors are already being shaped in early childhood, in the process of socialization. The factors influencing formation of a particular behaviour include, among others, personal patterns at home, kindergarten, school, local community, mass media, advertising, etc. Childhood and youth constitute the decisive periods for health behaviors and lifestyles in adulthood. Health behaviors are the manners of conduct of an individual that have an impact on health, its maintenance and strengthening. It should be noted that health behaviors are not permanent patterns. They are developed throughout life under the influence of various factors, among which we can distinguish at least three groups: predisposing, enabling and reinforcing change in health behaviors [4]. Nowadays, examination of health behaviour is regarded as an important method for measuring the health condition of the population, also constituting the foundation of planning and evaluating for health education, developing preventive programs and health promotion projects. According to the World Health Organization , health does not only mean the absence of illness or disability, but a state of physical, mental and social well-being [3]. Therefore, assessment of health takes into account the subjective sense of health based on personal experiences and reflection, on what we feel and on what kind of people we are. The sources of emotions are located in us and in our environment, and 'a subjective attitude towards own life in the form of its assessment is a consequence of the situations that arose and, simultaneously, constitutes a psychological aspect of the quality of life' [5]. As a result, it is worth analyzing and examining whether Canadian Polonia is satisfied with living in this country and how health factors are evaluated, both in terms of subjective assessment and objective assessment . Statistical data confirms that, despite relatively restrictive regulations and plans for their additional strengthening, Canada still remains an attractive country for newcomers and those striving to become newcomers. It receives around a quarter of a million new immigrants every year. Besides Australia, this country has the highest immigration rate per capita in the world. This positive immigration ratio means that Canada has the highest population growth among G8 countries and one of the highest among highly developed countries [6]. Canada is a country with an immigrant origin, which, due to the standard of livingoffered, attracts crowds of immigrants from all over the world. Toronto is a modern multicultural metropolis that encompasses the whole spectrum of social and cultural diversity [7]. This is a space with a visible ethnic and racial diversity of inhabitants, which respects the symbols assigned to them. Multiculturalism, defined by Fish [8] as strong, manifests itself, above all, in the ethnic diversity of the inhabitants of such modern cities. The metropolis best illustrating this state of affairs is undoubtedly the largest city of Canada, Toronto, founded in 1793, which can be called a multicultural capital of the world. Today, it has representatives of over 200 ethnic groups, which makes the city more ethnically diverse than Miami, Los Angeles or New York. According to the latest national census [9], there are 2.5 million inhabitants in Toronto and 5.5 million in the GTA . In the years from 2001 to 2006, 267,855 immigrants settled within the territory of Toronto [10]. Data from 2011 confirms that 47% of the Toronto population is included in its visible minorities. Reczy ńska presents the most numerous national minorities in the census of 2011 . Considering the multiculturalism of the GTA and the interpenetration of various cultures, nationalities and ancestral descent, it is important to examine how the community of Polonia assesses its life in this metropolis. Therefore, the main purpose of this research was to determine the standard and quality of life of the Polish community living in the Greater Toronto Area, which includes the city of Toronto and four regions: Halton, Peel, York and Durham. After conducting extensive field studies within this scope, this article will present assessment of both objective factors and subjective ones . The respondents' assessment concerning, among other issues, access to the healthcare system and access to sports facilities in the city will be discussed. Evaluation provided by respondents regarding satisfaction with their general health, mental well-being and the possibility of practicing sports, which is associated with maintaining a healthy lifestyle, will be shown. Source: Author's own compilation based on [11]. The origin of the concept of the quality of life is not a simple, clear and obvious one. The main reason for this state of affairs is its ambiguity, the evolution it has been subject to and the fact that, within various scientific disciplines, numerous synonyms were and are used interchangeably with it. The issue of quality of life has developed differently in psychology, where the of interest was limited to issues related to this area of science, and in medicine. However, this does not mean that they each discipling took advantage of its own achievements, rather they enriched each other by developing new research methods [12]. The issue of living conditions and the quality of life of a population is extremely diverse and requires various socio-statistical analyses. Numerous studies in this area have been undertaken as part of the Integrated System of Research on Households that has been implemented by the Central Statistical Office of Poland since 1982. This research was and is still being conducted by various scientific and research institutions ). Until recently, this issue has also been addressed as part of so-called central and inter-ministerial research concerned mainly with social indicators, the concept of families' standard and quality of life, minimum subsistence, the role of social benefits, social poverty, participation of the population in culture, and transformations in the broadly understood level and structure of consumption. Indicators of quality of life include the ability to play current life roles, adaptability, psychological well-being and functioning within social groups. The term 'quality of life' appeared in the second half of the 20th century. Research on quality of life was initiated by Allardt, a Finnish scientist, who suggested that research on the standard of living should be extended by two new elements that make up quality of life: an analysis of emotional states and the sense of existence, of being someone . The beginning of interest in quality of life was caused by disappointment with economic growth and the conviction that increase in material goods was not sufficient to make human life better. Attention was drawn to the fact that, apart from positive effects, economic growth also brings negative ones such as degradation of the natural environment, disintegration of interpersonal relationships and social pathologies. It has been noticed that the satisfaction that a person derives from life does not only depend on the material goods he or she possesses, but also on the ability to satisfy higher needs, such as state of mind, a sense of security, self-fulfillment, and participation in the environment in which we live as residents [13,14]. The scope of the concept of quality of life, [15] suggests that we should distinguish a wide and a narrow approach. Taking into account the nature of the indicators and the level of analysis, the author postulates recognition of quality of life in an objective and subjective sense, and of its relationship with the axiological system. She also puts forward two positions related to different interpretations of needs and value systems. The first, formed on the basis of Maslow's theory of needs, focuses on the analysis of issues related to the category of scarcity. The second, pursuant to prognostic assumptions based on specific concepts of man and of human personality, emphasizes the development needs of both the individual and society. Quality of life constitutes an object of interest for economists, statisticians, sociologists and politicians. Human life is of the highest value and its quality determines satisfaction and possibilities of development for each person and for the whole society. The quality of life of the population is associated with socio-economic development and economic growth. It influences one of the factors of economic growth, i.e., human capital. It determines physical fitness, shaping performance, pace of work and forming the psycho-intellectual condition of people through the conditions it creates for development, education and levels of vocational preparation, which develops their creativity and innovation. Furthermore, subjective feelings in society regarding quality of life and prospects for its improvement shape the economic climate, in particular the investment climate, which translates into the economic situation. There are two approaches to quality of life in the literature on the subject: an objective approach, according to which the objective living conditions of people, measured by means of objective variables, such as GDP dynamics, environmental pollution, infant mortality rate, health, and the material and social living conditions of people constitute quality of life; -a subjective approach that identifies quality of life with the satisfaction of people in various spheres of their lives. According to this approach, quality of life constitutes the individual's sensation of well-being, his or her satisfaction or dissatisfaction with life. Similarly, the definition of the World Health Organization includes an objective and subjective aspect since it regards the quality of life as 'the manner individuals perceive their life position in the cultural context and system of values in which they live, and in relation to their tasks, expectations and standards set by environmental conditions. Therefore, the quality of life is the same as a comprehensive way of assessing the physical health, emotional state, independence in life and the degree of independence from the environment as well as personal beliefs and beliefs by an individual'. [16] Skrzypek [17,18] defines quality of life as a combination of the objective and subjective elements. The fundamental determinants of quality of life include: objective conditions, such as economic conditions, leisure time, social security, housing conditions, the natural environment, health, social environment, etc.; -subjective conditions, which are perceived in a unique way by each human being, and are manifested in his or her well-being, where assessment of living conditions is expressed in terms of satisfaction, happiness, fears, hope and loneliness. M. Rapley [19], one of the researchers on quality of life, stated that quality of life is a synonym of happiness, satisfaction with life, freedom from coercion, a state of complete sense of well-being and not only the absence of illness, objectively good functioning, mental balance, prosperity, self-fulfillment, mental comfort, low unemployment, high income, a good life, joy, good life experiences, a life full of meaningful existence. As Jankiewicz-Siwek and Bartosi ńska [20] note, there is currently no one universal measure that can be used to assess quality of life. It is described by the application of numerous variables that are expressed in various units of measure. These are qualitative features whose measurement is subjective. According to T. Słaby [21], quality of life includes all the elements of human life that are associated with the fact of human existence, of being someone and of feeling various emotional states, resulting from, e.g., having a family, colleagues or friends. From the point of view of social policy, quality of life appears as a synthetic assessment of standard of living and an assessment of the principles of the organization of social life and the nature of social bonds [22]; a measure of subjective satisfaction with life, a measure of well-being, good or bad individual and collective well-being. The presented definitions of quality of life are closely related to Maslow's theory of needs, based on the assumption of the hierarchical nature of human needs. The lowest level in the hierarchy is created by physiological needs. The needs for security and safety constitute the next level. Maslow describes physiological needs along with the needs of security and safety as material needs. Social needs constitute the next level in this hierarchy. Among them, the need to belong is fundamental. In the group of social needs, the most important is the need for a sense of human value. Maslow defined material and social needs as deficiency needs. The group of needs related to self-actualization occupies the highest position in the hierarchy. These include the needs for love, truth, justice, perfection, beauty and meaning in life. The term 'quality of life' covers all these presented groups of needs. In addition to material needs, it also takes into account social and spiritual needs. Therefore, quality of life is satisfaction with meeting all human needs. Słaby [23] believes that 'standard of living' should be determined on the basis of a group of objective measures , while quality of life should be assessed mainly with the application of subjective measures . The former record the existing actual state of affairs, e.g., the level and structure of income, expenditure, consumption, whereas the latter inform about the manner in which this state is perceived and assessed by particular persons or social groups. These are the indicators that are suitable for assessing the quality of life, understood as the level of human satisfaction with the degree of meeting various human needs. Generally speaking, subjective measures can be divided into those that express value judgments and those that relate to the respondent's intentions. Value judgments are, in particular, characteristics of the degree of satisfaction and self-esteem [24,25]. An important and often ignored procedure is the weighing of these spheres of life by including level of satisfaction, e.g., from family life, or from the use of a professional weight index in the assessment of these areas of life and activities. Filipp and Ferring [26] confirmed that such a procedure significantly changes the results of the research on the sense of quality of life. In research concerning the quality of life in situations of critical events and illness, the sense of intensification of change as a result of confrontation with the event, the valency of the 'in plus' or 'in minus' change, as well as the scope, i.e., the number of areas of life in which changes have occurred, are taken into account, along with whether the extent this change is desirable or not. As Wnuk et al. [27] state, in research on quality of life, a value-based understanding of quality is applied more frequently than a descriptive interpretation . Two basic trends in the research and discussion can be distinguished in relation to the possible ways of conceptualizing and operationalizing the category of quality of life. The first one differentiates quality of life in the subjective and objective sense, as well as phenomenalistic and normative understanding. The second results from differences in value systems, ethical attitudes and implemented development concepts [28]. The concept of 'quality of life conditioned by health' appeared in medicine thanks to Schipper. In the Polish literature on the subject, it functions in three variants as: 1. health-dependent quality of life, 2. health-related quality of life, and 3. quality of life conditioned by health status. Each manner of conceptualizing quality of life conditioned by health status should relate to three areas: feelings, functioning and the future. The area of feelings concerns the subjective sense of well-being of the individual in all spheres of life, the area of functioning is related to the physical, cognitive and interpersonal activity of the subject, whereas the area of the future is associated with forecasting changes that may occur in the other two areas [29]. Research on quality of life with reference to health condition has been initiated in the field of medicine in order to assess functioning of the patient depending on ailments and limitations resulting from disease [30]. As emphasized by Sęk [31], quality of life, recognized objectively, constitutes a set of human living conditions, objective attributes of the world of nature, objects and culture, as well as objectively assessed human attributes related to standard of living and social position. These objective indicators of quality of life also include the structure of the human body and the level of functioning of the human biological systems, i.e., health. Subjectively, quality of life is defined as a result of the internal processes of evaluation of various spheres of life and of life as a whole. This is the cognitive-experiential category, and the subject is the most important, if not the only, source of the data pursuant to which a level of quality of life can be inferred. As Sęk et al. [32] state, contradiction is expressed in the fact that individuals asked about the source of their quality of life put health first; they also think that a sense of good health, apart from the sense of security and freedom, can be attributed to a happy person and is the most characteristic for the comparison between quality of life and health. However, the real strength of the relationship between satisfaction with life, i.e., the sense of the quality of life, and satisfaction with life is considerably weaker. The contribution of the quality of a healthy life to a global sense of quality of life is significantly smaller than satisfaction with family life, financial and housing conditions, friends, etc. In the subject literature within the scope of mental health, we find approaches that identify mental health with mental well-being, which has also been reflected in the definition of the World Federation for Mental Health. Within this approach, health is identified with quality of life as one of the dimensions of happiness. However, this is not an attitude that plays a significant role in contemporary health models. With the development of behavioral medicine and health psychology, it has been stated that health can be presented within other theoretical perspectives, biomedical as well as holistic-functional [32]. Quality of life in the city is influenced by a variety of factors of differentiated importance; these include, among others: health, safety, access to services , the possibility of rest and recreation, the quality and friendliness of space , cost of living, transport efficiency, the presence of pollution , and functioning within the community [33,34]. Measurement of quality of life requires designation of the domains of the quality of life to be observed. The report of the Task Force on Multidimensional Measurement of Quality of Life suggests that measurement of the quality of life within the European Statistical System should cover nine domains . Eight dimensions relate to areas connected with living conditions and the ninth concerns mental well-being: 1. Material living conditions 2. Economic activity and working conditions 3. Health 4. Education 5. Leisure time and social relations 6. Economic and physical security 7. State and fundamental rights 8. Quality of the environment in the place of residence 9. Subjective well-being [35]. The report of the Expert Group on Quality of Life of the European Commission [36] presents a proposal for a system of indicators to measure the quality of life. In the process of its creation, it was assumed that this measurement should be based on both indicators of an objective type and subjective indicators . Furthermore, a group of 17 basic indicators has been distinguished within the system of indicators, including health. Despite the extensive literature on the quality and standard of living, there has not been a scientific publication regarding the analysis of the assessment of the level and quality of life of the Polish community living in the Greater Toronto Area before now. This is an important issue from the sociological, geographical, migration and cultural point of view. Health aspects, including an evaluation of satisfaction with the standard and quality of health in the opinion of the Canadian Polonia community, has been undertaken in this research. The capital of Canadian Polonia is Toronto. The city has more than 83,000 people reporting only Polish origin and the same number of people providing some Polish origin. Since the 1940s, Toronto has attracted and continues to attract the largest number of Polish immigrants. Since the 1980s, the Polish community has been mainly concentrated in Mississauga and Brampton, and that is why the research has not only been conducted in the city of Toronto; instead, its coverage included the entire Greater Toronto Area in which the population of Polish origin lives. --- Materials and Methods There are familiar publications by Canadianists who raise in their studies aspects of history, multiculturalism, migration, Polish pastoral activity, causes of emigration and Canada's policy towards people who have e obtained immigrant status However, research taking into account the quality and standard of life of the Polish community in Canada in the context of health would fill a gap in research on the Polish diaspora in Canada. This research complements the historical threads concerning emigration from Poland, in the context of the quality of life of the current Polish community living in the Greater Toronto Area. A large wave of Poles emigrating to Canada took place in the 1960s and 1970s. Today, the people who arrived at that time are 60 years old or older. They are often people currently retired, often requiring medical care, and health has a significant impact on their assessment of the quality and standard of living. In 2011, people aged 60 and over constituted 11% of the world's population; however, by 2050, already one in five people will be in this group. 31.8% of the Canadian populations will be older than the age of 60 [38]. The possibility of active and effective mobility in older people is an important element influencing their quality of life, particularly as it determines their accessibility to health care, cultural institutions, trade and services [39]. According to Reczynska, [40], in 1991 Poland was in second place in the summary of ten countries constituting the most important sources of immigration to Canada. In 1996, Poland shifted to sixth place ; however, it was still ahead of other European countries. The last population census including persons declaring Polish origin is available from 2016. At that time, 259,715 people declared Polish origin in the area covered by the survey , which is a 4% share of the Polish community in the total of all GTA residents . Such a large number of the Polish diaspora in the GTA, clearly demonstrating its distinctiveness in multicultural Canada, constitutes an important topic for exploring and gaining information for sociological research concerning this population in Canada. Taking into account the fact that such a large group lives not only in the city of Toronto but within the entire GTA area, the research was also conducted outside Toronto itself. Contemporary distribution of Poles in Canada has been conditioned, on the one hand, by the processes of development of the young state and, on the other hand, by social changes in the Polish community itself. Until the 1930s, immigrants from Central Europe were directed to agricultural areas in the provinces of Manitoba, Saskatchewan and Alberta by the Canadian authorities. It was only after the Second World War that Poles began to settle mainly in cities. Urbanisation processes resulted in the fact that, as early as the 1970s, 80% of Poles lived in cities, mainly in the large agglomerations of the provinces of Ontario and Quebec. Nowadays, cities, particularly Toronto and subsequently Montreal, Ottawa, Edmonton, Hamilton, Vancouver, Winnipeg, and Calgary, constitute the main centres of the Polish community. Representatives of the most recent waves of Polish migrants from the second half of the 20th century settled mainly in Toronto, hence the strong numerical dominance of the Polish community living there. The most active organizations operate in this city and most of the Polish press titles are published here [41]. 35,151,728 people lived in Canada in 2016, and 1,106,585 inhabitants were of Polish origin, 3.9% of the total population. Ontario, where the Greater Toronto Area is located, is the province with the greatest number of inhabitants of Polonia. 523,490 such people lived in Ontario in the year 2016, while the largest number of people of Polish descent in relation to all residents is in Manitoba, where the percentage of Poles is as high as 6.9%. Most Polish people live in Mississauga, in the county of Peel. This population equals 43,350 people, which is 6% of the total population in Mississauga. A significant number of people of Polish origin live in the county of York, in Vaughan. This number includes 18,265 people, which is 5.96% of the total population of the city. The smallest number of Polish people lives in two towns in the county of Durham, the city of Brock, with 335 people of Polish descent, and the city of Scugog, with 590 people. In 2016, Statistics Canada published census data which confirms that 34 million Canadians use over 200 languages on a daily basis. In the table of Statistics Canada , Polish is the eleventh most widely spoken language in Canada. The Greater Toronto Area includes the regions of Peel, Halton, York, Durham and the city of Toronto itself. The time period of the conducted research covered the years 2017, 2018 and 2019. In 2017, pilot studies were conducted, whereas the actual research in Canada was carried out in two stages. The first stage of field studies took place during the period from 16 August 2018 to 18 September 2018, and the second stage from 10 September 2019 to 19 September 2019. One of the quantitative studies uses a diagnostic survey method in the form of a questionnaire. This method has been applied in this research. The results of the research conducted among the Polish community living in the Greater Toronto Area constituted the fundamental empirical base. The empirical part of the research took the form of a diagnostic survey, conducted by the application of a representative method among the population of Polonia of the Greater Toronto Area. The appropriate research covered a total of 612 respondents, including 583 questionnaires used for the analysis. The research was carried out among individuals who were at least 15 years old. In order to make the research as representative as possible, seven age groups have been differentiated, with the use of a pre-conceived division. These were the following age groups: 15-19, 20-24, 25-29, 30-39, 40-49, 50-59, and over 60 . Conducting research among people aged 15 and over allowed certainty that respondents understood the questions and that the answers provided were reliable. In this article the research results are presented for all age groups together. The questions applied in the questionnaire were formulated in such a way as to refer to the experiences of the respondents in 2017 and 2019, as well as from the period of the last several years. This approach has enabled the capture of an average assessment of the quality and standard of living of the surveyed group of respondents. During the analysis of the quality and standard of living, factors such as gender, age, education, religion, profession, achieved material level, size of household, and regulated residence status were used. This has allowed for a more precise description of the examined group and identification of the relationships between interesting aspects. The survey questionnaire was prepared in two languages: Polish and English, and consisted of 17 questions plus 15 questions in the demographic section. In the quantitative research, answers are most often 'provided' and respondents indicate one of them. The survey questionnaire included questions concerning the following issues: Assessment of specific elements of the standard of living in Toronto Assessment of specific elements of standard of living associated with Polish origin in Toronto Assessment of the possibility of meeting the person's needs in the city, within the scope of particular elements Assessment of satisfaction from particular areas of life Social, demographic and economic features of the respondents In the conducted study, one of the questions concerned the sense of Polish national identity. Respondents were asked if they feel Polish. 97.4% of the respondents answered 'yes'. Only 0.6% of the respondents answered that they did not, and 2% did not have an opinion on the subject. --- Model of Objective Factors and Subjective Factors One of the main aims of the research was to construct an individual model of objective factors and subjective factors and to examine whether the suggested indicators/measures interact with each other. The model has been created on the basis of the prepared questions in the survey questionnaire, taking into account the specific factors that have and may have an impact on the overall assessment of the standard and quality of life. The described model including health factors is presented in Table 2. In the demonstrated research, individual relationships will not be described, and only the factors examined during the research in the Greater Toronto Area will be listed. Terms such as 'living conditions', 'standard of living' or 'living standards' are used interchangeably with the term 'quality of life'. Some use these terms as synonyms. It should be noted that 'standard of living' is determined by objective living conditions, and 'quality of life' is also influenced by subjective factors, such as aspirations and the degree of their satisfaction or perception. The model of objective factors and subjective factors put forward has included demographic factors, social factors, cultural factors, economic factors, political and legal factors, educational factors, geographical factors, economic and health factors that attention has been devoted to in this study. This is a novel research tool to assess the quality and standard of life of the population. This model was described in the article entitled 'Evaluation of the selected components of standard of living and quality of life within a Polish diaspora: the example of the Greater Toronto Area' [42] and particular factors will gradually be described in later scientific studies. This article pertains to the health factors influencing the level and quality of life of the Polish community in the GTA. Assessment of personal well-being, mental health and health constitutes a significant aspect when describing the quality and standard of life. Therefore, the research has included health factors to which due attention has been paid. Table 2 presents the model of objective health factors and subjective health factors that have been considered during the empirical research. In the suggested model of the quality and standard of living, each indicator has been matched with particular indicators/measures, in the factors related to the assessment of the both the level and the quality of life. Mutual correlation between the selected subjective and objective factors constitutes the main assumption of the research. Respondents were asked about access to the healthcare system, leisure time, conditions for rest and relaxation, access to tourism, possibility of practicing sports and about their assessment of satisfaction with health, mental well-being and the possibility of practicing sports . --- Results In the survey questionnaire, subjects from Canadian Polonia were asked about assessment of subjective and objective aspects related to health. Answers to the individual questions can be found in Table 3 and Table 4 . The respondents were asked to rate each of the listed elements, marking their opinion on a six-point Likert scale-from I am very satisfied to I am very dissatisfied; I have no opinion ; and to assess the quality of life where they rated their assessment of health factors from very good to very bad and I have no opinion. The six-point Likert scale has also been used in the question about problems related to health elements, and which, according to the respondents, should be addressed first. The respondents have rated the scale of the problem from a very important problem to an unimportant problem at all, I have no opinion. Table 3 presents the answers of the respondents to the questions concerning assessment of standard of living in the context of the health factors. Respondents were asked to assess accessibility to the healthcare system, possibilities related to leisure time, accessibility to rest, relaxation and tourism, as well as access to practicing sports . In this question, from all the provided answers, the majority of the respondents answered that they are satisfied with possibilities for leisure time . The largest number of 'I am very satisfied' answers referred to access to the healthcare system: 123 answers along with 260 'I am satisfied' answers. With such a high satisfactory response concerning access to the healthcare system, at the same time the largest number of respondents replied that they were dissatisfied with this aspect. There were 73 such answers. 94 of all respondents have replied that they have no opinion regarding accessibility to rest, relaxation and tourism. One of the research objectives was to check how people of Polonia in the Greater Toronto Area assess the quality of life of the selected health factors. Table 4 presents the answers to this question. As in the question regarding standard of living, the aspect related to satisfaction with one's own health was very highly rated by respondents. As many as 120 people evaluated their health satisfaction as very good and 263 as good. Only 16 out of 583 respondents answered that their health has been assessed as very bad and 32 as bad. From the provided answers, it can be concluded that the surveyed people from Polonia in the Greater Toronto Area are also satisfied with their mental well-being, as 114 people rated their mental state as very good and 239 as good. Almost a quarter of the respondents answered that they are quite satisfied with the possibility of practicing sports . The majority of the given answers also indicate satisfaction with the manner of spending their leisure time, assessing it as good, with 361 respondents providing such an answer. In the prepared questionnaire, subjects were also asked which issues should be addressed in the first place to improve their overall level and quality of life. Two aspects related to health have been included in these suggested aspects. These were improvement of access to the healthcare system and increase in sports infrastructure . The answers to this question are in Table 5. According to the respondents, improvement in access to the healthcare system constitutes a very important problem and should be addressed first. There were as many as 231 and 152 replies from the respondents' opinions that this is an important problem. Only 5 respondents believe that improving access to the healthcare system is not a significant problem and 27 subjects think that it is a minor issue. After receiving such answers, it can be safely concluded that improvement of the quality of life related to access to the healthcare system constitutes a very important and substantial issue for people of Polish origin living in the Greater Toronto Area. The respondents assessed issues related to the development of sports infrastructure as less important. 94 people have no opinion at all and only 115 think that this is a very important problem that should be addressed immediately. Another question concerning issues related to health in the survey questionnaire was: 'In the last twelve months did it happen to you or anybody in your household that there was no money for expenses related to treatment, or purchase of medicines?' The respondents had a choice of three answers: 'No, there was not such a situation'; 'Yes, there was such a situation'; 'I do not know, I do not remember ' and 498 people have replied that there was not such a situation. i.e., that they would run out of money for expenses related to treatment or purchase of medicines they needed during the last year. 65 people refused to provide this answer, and 20 respondents admitted that there was a situation where they did not have enough money to cover the costs of treatment last year. --- Discussion Value judgement obviously depends on the structure of needs and the individual system of values, and, in particular, on the individual's concept of the meaning of life. Even a fragmentary review of the literature on quality of life, its nature, objective and subjective conditions [43][44][45], methods of research on the level of quality of life, specific aspects of happiness, especially work, as predictors of quality of life, and the relationship between quality of life and health [46] confirms a considerable diversity of approaches and much contradiction in the results of the empirical research. Numerous studies, both on large population samples, conducted mainly in the USA [47] as well as on groups of healthy and sick people after severe accidents demonstrate that objective living standards are not clear predictors of the sense of the quality of life. Due to the fact that specific people have a much higher financial level , and others have objectively poor health it cannot be predicted that the first group will obtain much higher results in measurements of the sense of satisfaction with life. This applies particularly to measurements of the quality of life projected in the past and into the future. The research conducted by Brickman and others [48] clearly indicates that, while determining quality of life in people who were participants in critical life events during the period of the research, measurements of quality of life related to the time before these events, the time associated with these events and future time display significant differences. Therefore, it should be taken into account that the results will be completely different and assessments of the quality of present, past and future life will perform different functions. Considering important methodological problems, it should be noted that, even when we assume that in the study of the quality of life, we are guided by the individual, subjective assessment of the subjects, care for minimizing 'measurement errors' is of great importance. The research [49] confirms that respondents give higher values for assessment of satisfaction with life in a situation of direct conversation with the researcher than in the case of completing a questionnaire, and that global assessments of quality of life in general are much more influenced by the current situation and the present emotional state of the respondents than when the general measurement of quality of life constitutes the result of the sums of component assessments concerning various distinct areas of life and activity. While comparing specific measures of quality of life, satisfaction and various forms of optimism with health, the most interesting and important aspects regarding health psychology arise. The effects of satisfaction with life, optimism and various forms of trust in life on the development of health, health protection [50][51][52][53] and effectiveness of coping with disease [54] have been widely examined. The sense of quality of life dependent on health is a very good indicator of health condition [55], and also influences the subjective perception of the person's own life situation. The sense of quality of life is shaped by numerous factors, not only related to the characteristics of the person, but also to an understanding of physical health [56], type of disease and the environment in which the patient lives [57]. Family relations, relationships of youth with parents and home life have a significant impact on the sense of the quality of life [58]. Generally speaking, it can be stated that the sense of quality of life dependent on health constitutes a psychological construct describing the physical, mental, social, psychological and functional aspects of well-being and functioning from the point of view of each person [55]. Within the scope of psychology, [59] presents an innovative approach to defining and perceiving quality of life. Here, quality of life can be understood in two ways: either as the sensation of the person's own life as a result of getting to know it, or as the sensation of the person's own life by experiencing it. Quality of life can be considered in the dimensions of reflection on the course and the present state of one's own life, one's own health, and the experience of various mental states during life. Both described solutions refer to the subjective sphere of quality of life, as two aspects of the individual assessment of quality of life. In the event of feeling the person's own life by getting to know it, reflective awareness constitutes the main tool for assessment of its quality. It is worth discussing the results of the research on perception of quality of life by Poles in the 21st century in Poland. According to the research carried out by Social Diagnosis, since 2000 an improvement in the assessment of the current lives of Poles can be observed. An increasing percentage of people evaluate their current lives as at least successful. Furthermore, the sense of happiness in society is gradually increasing and the percentage of unhappy people is decreasing [60,61]. The psychological well-being of Poles has been increasing considerably since the beginning of 2000. Assessment of life satisfaction, according to the assumptions of Social Diagnosis, is largely influenced by so-called partial satisfaction, i.e., satisfaction with particular areas of life. In the years 2000-2013, on the basis of the survey conducted by Włodarczyk [62], the greatest increase in satisfaction in Poles concerned: the sense of safety in the place of residence, the manner of spending free time and the person's own education. The smallest included: satisfaction with sexual life, relations with colleagues and loved ones in the family and satisfaction with marriage. According to the analyses conducted at Social Diagnosis, age has been the most important factor shaping the well-being of Poles for several years. The older the person was, the worse the mental condition of this person was. The second factor in terms of importance for overall mental well-being was marriage, the third was household income and the fourth was related to the number of friends. Analyzing the data concerning the issues related to satisfaction with the health condition of the individual, in 2000 Poles assessed it at 3.24 on a 6-point-scale, where 1 meant very satisfied and 6 meant very dissatisfied. In 2005, the assessment was 3.09, in 2007 it was 3.13, in 2009 it amounted to 3.0, in 2011 the result was 2.93, and in 2013 it was 2.89, which means that Poles assessed their health condition to be better with each year of the survey as compared to previous years. Surveys regarding the quality of life and satisfaction of Poles have also been conducted by CBOS, the Public Opinion Research Centre, periodically since the 1990s. The survey conducted in 2019 confirms that almost two fifths of Poles regard themselves as people who are lucky. The number of those who feel that they have bad luck in life is three times smaller . Every second respondent answered 'difficult to determine ' [63]. Similar to the research conducted by Social Diagnosis, CBOS also carries out surveys on partial satisfaction influencing general satisfaction in the life of Poles. The respondents assessed their health condition to be good in the years 2000-2013. This was the opinion of half of the respondents, while on average about a quarter were dissatisfied. Statistical analysis conducted on the basis of the results of the survey carried out by CBOS confirms that future prospects, place of residence, health condition and marriage are most strongly connected with the general sensation of satisfaction. Among the values respected by Poles in 2005 , 2010 and 2013 , health turned out to be an important value and was ranked second after family happiness as the most important value in life. The results of CBOS studies on these respected values are also confirmed by studies conducted in 2011. The survey was carried out among 1000 adult Poles over 15 years of age . The sample was selected according to statistics of the Central Statistical Office by gender and place of residence. The survey was conducted by Pentor Research International in May-June 2011. Among the values that people subscribe to, and the most important ones for the respondents in order of significance , health was second , after children and family [64]. Research conducted by Panek and Zwierzchowski [35] 2019) concerned assessment of the quality of life of Poles from a national perspective. The survey adopted the structure suggested by the European Statistical System and concerned nine different aspects. Evaluation of quality of life in terms of health satisfaction constituted one of these. The conducted research confirms that the quality of life is considerably and positively determined by being a pensioner rather than being unemployed, and by being a disability pensioner rather than a professionally inactive person, in the area of health. A significant increase in quality of life in the area of health is fostered by an increase in the level of education, living in smaller urban centers, but not in the rural areas, and being a member of more numerous households. On the other hand, quality of life in the area of health significantly decreases with age and especially for women. The increase in quality of life in the area of health is accompanied by a fall in restrictions on access to medical and dental care due to financial reasons and a fall in restrictions on performing activities for health reasons [65]. There are studies concerning older people in the scientific literature on perceptions of health and quality of life. Moore et al. [66] reported that 11 out of 17 studies obtained a strong positive result between good health and satisfaction with quality of life. Raphael et al. [67] also noted a positive link between quality of life and the health of older people in their studies, indicating that healthy people, without chronic diseases, assess their standard of living to be good. According to Farquhar [68], living in a desired environment significantly influences the quality of life. Similar findings were made by Stuifbergen et al. [69] as well as Bowling et al. [70]. Sparks et al. [71] concluded that the only significant positive factor that has an impact on life satisfaction is good health and a well-maintained social position. In contrast, McCamish-Svensson et al. [72] investigated the relationship between support and life satisfaction. They found that satisfaction from contact with siblings was associated with life satisfaction among people over 80, which also significantly contributed to their mental and emotional well-being, which influenced overall good health. Since 1994, the Quality of Life Research Unit at the University of Toronto has been studying the quality of life of adults with physical or psychiatric disabilities and children with developmental disabilities as well as the elderly, young people and adults in the general population. The studies were conducted among Canadian citizens, in 40 small groups in different locations across the country, to discuss the factors which are important in life and what affects quality of life [73]. Respondents provided answers in various thematic areas: democratic rights, health care, education, environment, social programmes and conditions, community, material conditions, economy and employment, and government policy. The studies confirmed that the state plays a fundamental and integral role in providing the components of a good life and influences the overall assessment of quality of life by citizens, especially in the aspects of health and education. Paskulin from the Federal University of São Paulo in Brazil and Molzahn [74] from the University of Victoria in Canada have published data on the quality of life of older people. The studies were carried out on a random 202 older people from Canada and 288 respondents from Brazil. The authors wanted to examine whether there are differences in the quality of life of people in selected regions of Brazil and Canada and whether health satisfaction, aspects concerning the meaning of life, leisure opportunities, satisfaction with housing environment, satisfaction with personal relationships, satisfaction with the ability to perform life activities and possession of adequate financial resources contribute to the perception of the quality of life of older people in the selected regions. As the results of the studies confirmed, evaluation of overall quality of life as well as physical, psychological and environmental areas were higher in the Canadian sample. The authors stressed that health satisfaction, sufficient amounts of money and leisure opportunities contributed to the assessment of a high quality of life in both countries. Health satisfaction was the strongest contributor to a high quality of life, both in Canada and Brazil. Chappell [75] compared the quality of life of older people living in mainland China and those living in Canada. Despite considerably larger levels of poverty, worse education and a health status that was perceived as lower in China, she found that, despite this, an important factor in the quality of life in both cultures is social support and one's own health. Other studies on the quality of life in the Chinese diaspora were conducted by Da and Garcia [76] within the scope of socio-cultural adaptation and change in quality of life in the settlement of recent older Chinese immigrants in Canada. The research was carried out among 31 older women and men who recently emigrated from China to Canada, with the main aim of examining the perception of quality of life in the new place. Their overall perception of quality of life after immigration was characterised by losses and gains as well as influenced by multifaceted factors, such as language, intergenerational relations, economic status and those supporting them. In the literature, there are also studies on the quality of life of Canadian indigenous peoples, known as Aborigines, that include three groups of Aboriginal peoples in accordance with the Canadian constitution: Indians, Inuit and Métis. Familiar studies include that by Salee entitled 'Quality of life of Aboriginal people in Canada. An Analysis of Current Research' from 2006 and studies by Kant et al. [77] who published data on the quality of life of Canadian indigenous people in 2013. The research team conducted these surveys in the province of Ontario and in British Columbia . During the survey, 316 questionnaires were collected. The questionnaire included questions on satisfaction with general welfare, wellbeing, education, employment, health, housing, income, and land use. In order to assess the level of satisfaction, the respondents used a 7-point Likert scale . The studies confirmed that the Aboriginal population is not satisfied with their health condition; there are often mental and psychological problems among this population, in both women and men. The research shows that development of national politics based on drawing attention to mental health in Aborigines and assistance from the government in this regard would significantly improve the quality of life of these peoples in Canada, around 1.7 million people, representing 3.8% of the Canadian population. Studies regarding quality of life in different cultures were also conducted. The LEIPAD project [78] confirmed that good health has a significant impact on the quality of life of older people in Finland, Italy and the Netherlands. The World Health Organization Quality of Life Instrument-Older Adults Module aimed to develop a programme for intercultural measurement of the quality of life of the population. The instrument was tested in 22 countries [79]. While conducting the studies it was found in eight countries that the impact of chronic diseases had a visible effect on the negative assessment of quality of life regardless of cultural differences between countries [80]. Tesch-Romer et al. [81] conducted studies in Norway, Germany, Spain and Israel, hypothesising that good health and quality of life depend largely on good relations within the family, but also on the type of welfare state. In three out of four countries, the results confirmed the assumption that the relationship between family support and quality of life is high where there are low transfers to the welfare state. Keith et al. [82] conducted studies in Botswana, Ireland, the United States and Hong Kong. They did not directly assess quality of life; however, they described the available health services, morbidity, and mortality as well as physical and functional status in each culture. After the conducted studies, they concluded that in order to understand the influence of health on the quality of life and functioning of people, the analysis must be tailored to the surrounding living conditions that affect overall life satisfaction. They noticed in Botswana and Hong Kong that poor quality of life is often due to a lack of basic resources, while in Ireland and the United States, they did not notice this kind of dependence. Minicuts et al. [83] developed a mutual database to identify the common factors of quality of life in Finland, Italy, the Netherlands, Spain, Sweden and Israel. The database included sociodemographic variables, health habits, health condition, physical functioning, social relationships and support as well as health and social services. In each country, respondents indicated that health factors and good health are very important. The Finns and the Swedes rated their quality of life the highest whereas Spaniards regarded their quality of life as the lowest. --- Conclusions Considering the large number of people of Polish origin living in the described area, an important topic related to the assessment of the level and quality of life as well as health factors by Polonia living in Toronto has been discussed. Such extensive research in this field has not been conducted before. The presented material constitutes only part of the research and the issues that have been raised and examined during field studies among the Polish community in the Greater Toronto Area in the years 2017-2019. Based on the provided data concerning health issues, it can be summarized that: Canadian Polonia is most satisfied with possibilities for leisure time and accessibility to rest, relaxation and tourism, Canadian Polonia highly evaluates satisfaction with general health and general mental well-being Canadian Polonia believes that improvement in access to the healthcare system constitutes a very important aspect that should be addressed by city authorities A considerable proportion of the respondents declare that over 12 months they have not run out of money for expenses related to treatment or purchase of essential medicines. While characterizing the level and quality of life, a general model of quality of life, even assuming an approximate one defining the framework for further analysis and empirical research, would be acceptable. Such a model would organize the elements comprising assessment of quality of life to a level corresponding to individual variables or their groups, as well as determine the manner of their inclusion. However, there is no model that is generally accepted in the literature. As a result, the author of the text has put forward her own model, taking into account both elements belonging to objective conditions and subjective ones . The objective of the research conducted among the Polish community in the GTA was to construct a model on the basis of which it was possible to examine the Polish community with regard to numerous areas concerning the level and quality of life . The model of indicators regarding the level and quality of life in relation to health aspects presented in Table 2 may provide a ready-made scheme to explore similar relationships in another research group in a different part of the world. --- Data Availability Statement: Anonymous questionnaires completed by respondents are stored at the author's home. ---

Quality of life constitutes an indicator of well-being, satisfaction or happiness resulting from one's existence. It is often referred to as a standard of living. In general, it is contentment with the fulfilment of one's needs. The main objective of the article is to describe the selected components of the living standards and quality of life within the Polish community of the Greater Toronto Area which includes four regions: Halton, Peel, York and Durham. The model of mutually affecting objective factors (standard of living) and subjective factors (quality of life) will be presented. The specific factors (demographic, social, cultural, economic, legal, educational, geographical and healthrelated) included in field studies among the respondents and based on assigned indices influencing the quality of life in the Polish community of the Greater Toronto Area, will be demonstrated. The major goal of the paper is to present an assessment of aspects related to health factors, both in terms of objective factors (standard of living) and subjective ones (quality of life) by Polonia living in the Greater Toronto Area. Results will be shown on the basis of the survey questionnaire completed by 583 respondents. The questions focused on, among other issues, access to the healthcare system, competence of medical staff and access to sports facilities. Respondents also evaluated their satisfaction with their general health, both physical and mental, as well as the possibility of practicing sports associated with healthy lifestyle.

Introduction Nearly 21 000 individuals in the US were killed by gun violence in 2021, a rate of 6.3 deaths per 100 000 population. 1 A rate of this magnitude has not been seen since 1994, near the late-20thcentury peak in firearm violence. 2 After that, an unexpectedly steady and sizable multidecade decline in violence took place, with the national gun homicide rate decreasing to 3.5 deaths per 100 000 population by 2014. Taking experts by surprise, homicides then began increasing before spiking in 2016 and again in 2020. 3 Firearm deaths have now surpassed motor vehicle crashes as the leading cause of death for US children, with Black children, boys, and children from disadvantaged neighborhoods in cities like Chicago, Illinois, disproportionately impacted. [4][5][6][7] The extreme fluctuation in lethal violence in the US over the past 3 decades means that successive cohorts of children growing up and reaching adulthood in these times experienced much different social worlds, what has been called the "birth lottery of history." 8 Even if they are from the same socioeconomic status, some adults today were raised during an era of relative calm, whereas others, sometimes just a handful of years older or younger, reached their formative years of adolescence and early adulthood during epidemic levels of firearm violence. Despite increasing attention to the burden of firearms on health and mortality, exposure to firearm violence over the life course is not well understood, particularly exposure to nonfatal firearm violence. 9 Much of the focus has been on national estimates of homicides and suicides. Other studies have sought to examine the causes of aggregate changes in violence over recent decades. 10,11 While important, such studies do not reveal when in the life course exposure occurs and whether dynamic patterns differ by race, sex, and birth cohort. Long-term longitudinal studies of exposure to firearm violence are relatively rare, especially those that document how both witnessing violence and personally experiencing violence unfold over the life course and differ by race and sex. 12,13 Rarer still are studies that examine different cohorts who reached late adolescence and transitioned to adulthood during times of historic highs and lows in violence over the last quartercentury. Without studies of multiple cohorts, it is challenging to pinpoint whether and how broad societal changes, such as the 1990s and 2000s crime decline, the steady weakening of gun laws, or the COVID-19 pandemic, might have influenced exposure to firearm violence by age. In this study, we leverage a unique longitudinal design to examine exposure to firearm violence from childhood through midadulthood by comparing the likelihood of being shot as well as seeing someone shot for multiple cohorts of children from Chicago, separated in age by approximately 15 years. Although followed over the same period, 1995 to 2021, the children from these cohorts reached key developmental periods in the life course during much different societal contexts. We highlight cohort inequalities in exposure to firearm violence, as well as disparities by race and sex. Finally, we examine race, sex, and birth cohort differences in spatial proximity to firearm violence in the present decade. --- Methods This cohort study was approved by the Harvard University institutional review board. Adult respondents provided written or verbal informed consent to participate in the study, and parents consented for child respondents. This study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline for cohort studies. --- JAMA Network Open | Public Health Inequalities in Exposure to Firearm Violence by Race, Sex, and Birth Cohort From Childhood to Age 40 Years --- Sample and Procedures The Project in Human Development in Chicago Neighborhoods began in the mid-1990s as a representative sample of 6207 age-eligible children drawn from a screening of more than 35 000 households in a stratified representative sample of 80 of Chicago's 343 neighborhoods. Children falling within 7 age cohorts, including infancy and ages 3, 6, 9, 12, 15, and 18 years, were sampled from randomly selected households and studied over 3 waves of data collection, to the early 2000s. 14 Modal birth years of the original age cohorts were 1978, 1981, 1984, 1987, 1990, 1993, and 1996. In 2011 to 2012, the PHDCN was extended through a random sample of wave 3 participants. Resource constraints prohibited following-up with the entire sample. Therefore, a 60% random sample of 4 of the original 7 PHDCN cohorts with high retention rates at wave 3 and a wide age range was selected, yielding 1057 participants at wave 4 . 15 In 2021, wave 4 respondents were followed-up for a fifth survey wave no matter where they lived in the US, yielding a sample of 682 respondents . We examine 2418 Black, Hispanic, and White respondents from the 4 sampled cohorts, including 1770 original members who were not part of the subsample drawn at wave 4. The difference between the original sample and the analytic sample size reflects subsampling at wave 4 and attrition, which we address through weights, and the fact that our youngest cohort was not asked questions about exposure to firearm violence until they were adults in wave 5. Additional information on the study design has been previously published. 15 The current analysis also draws on data from the Gun Violence Archive , a not-for-profit corporation that aggregates data on incidents of gun violence from more than 7500 sources, including police departments, media, and government. We obtained all incidents throughout the US that occurred within 250 m of the respondents' addresses in the year prior to respondents' wave 5 interview dates in 2021. To illustrate the historical context of the data collection, we plot the ages of respondents and homicide rates in Chicago in Figure 1. The wave 1 survey in the mid-1990s coincided with the beginning of a long decline in lethal violence in Chicago, as in the rest of the country. 6,10 Our oldest age cohort of respondents, born in 1981, reached their teenage years in the early 1990s when lethal firearm violence was at its peak. In contrast, the 1987 cohort, born only 6 years later, reached the formative years of adolescence in a much safer context. The 1984 cohort came of age between these periods, when crime was still high but declining. The youngest cohort, born in 1996, reached their teenage years at the lowest levels of lethal violence in both the US and Chicago in more than 40 years. However, by the time the 1996 cohort reached late adolescence and early adulthood, beginning in 2015 to 2016, Chicago was experiencing a surge in violence. By their 20th birthdays, these individuals were confronted by a dramatically more violent context than members of the other cohorts at age 20 years, despite being born into a period of rapidly declining violence. These patterns motivate our substantive focus on estimating cohort differences, in addition to race and sex inequalities, in exposure to firearm violence over the life course during these changing times. --- Measures and Variables Age when first saw someone shot was calculated from questions in waves 1, 2, 3, and 5. Respondents were asked if they had ever seen or been present when someone was shot, when this last occurred, if they had seen someone shot in the past year, and at what age they had first seen this. Respondents were not asked about the age of exposure in wave 5. The combination of these responses results in either a year of age or interval of ages in which first exposure occurred . Age when first shot was calculated from questions in waves 2, 3, and 5 asking respondents if they had ever been shot, the age this first occurred, and if they had been shot in the past year. Nearby shootings is the count of shootings resulting in a death or injury recorded by GVA that occurred in the past year within 250 m of a respondent's residential location as reported at wave 5. This radius was chosen to capture events happening within the approximate distance of a standard Chicago city block . --- JAMA Network Open | Public Health Race of respondents was reported by their primary caregiver, which was categorized first by ethnicity and then by racial group for those who were non-Hispanic. We excluded 102 individuals of other races because the sample was too small to conduct longitudinal analyses. Sex was recorded by interviewers. Cohort was defined by age at enrollment in the study. Respondents were sampled by eligible ages during the first wave of data collection, which occurred over approximately 3 years. Because of this design, there is variation in birth year within age cohorts. To ease interpretation of results, we refer to cohorts by their modal birth year . --- Statistical Analysis We estimated survival curves of the cumulative percentage of respondents exposed to firearm violence by race, sex, and cohort by a given age using the nonparametric Turnbull maximumlikelihood estimator , a generalization of the Kaplan-Meier estimator applicable to intervalcensored data. 16,17 Accounting for interval censoring is necessary for the seen shot outcome because age at time of exposure is uncertain for respondents exposed in the years between wave 3 and 1 year prior to wave 5. This is less consequential for the been shot outcome, since wave 5 respondents were asked for the age when first shot, resulting in smaller censoring intervals. We estimated associations between sociodemographic characteristics and exposure to firearm violence using multivariable semiparametric Turnbull proportional hazards models. 18,19 The SPT model is analogous to the Cox proportional hazards model but estimates the baseline hazard using the NPMLE to account for interval censoring. A Cox model with interval midpoints as exposure We used a negative binomial regression model to estimate the association between our covariates and counts of past-year shootings occurring within 250 m of a respondent's residence. These models include combined survey design and attrition weights to permit inferences to the population of Chicago children. We performed statistical analyses in R software version 4.2.2 using the interval, icenReg, and MASS packages , with a 2-tailed significance level of P < .05. 17,19,20 Because our analyses are descriptive and not directed at formal hypothesis testing, the purpose of calculating CIs is to draw inferences from our sample to the population. Data analyses were conducted from May 2022 to March 2023. --- Results The 2B). Sex differences in being shot were also significant, with the cumulative exposure of male respondents 5 times higher than that of female respondents by age 40 years . Differences again were age-patterned. Most reports of being shot for female respondents occurred during their midteenaged years, whereas males reported being shot throughout their teenaged years and their 20s and 30s. Sex differences in seeing others shot by age 40 years were more modest: 42.81% of female respondents and 57.64% of male respondents had seen someone shot . Direct experience of being shot was typically highest at all ages for the 1981 cohort, whose adolescence coincided with the 1990s peak in violence . All cohorts displayed rapid increases in incidence of experiencing gun violence in late adolescence, except for the 1996 cohort, whose increase began earlier and was more gradual. The cumulative incidence of experiencing gun violence by the end of follow-up for the 1996 cohort fell between the highexposure cohorts and low-exposure cohort . Similar to direct experience of being shot, the cumulative risk of having seen someone shot was higher for the 1981 and 1984 cohorts . Figure 3 presents adjusted hazard ratios and adjusted incidence rate ratios with 95% CIs from multivariate SPT models of been shot and seen shot, as well as the negative binomial model of nearby shootings. Independent measures of nearby shootings provide a complement to self-reported indicators of being shot and seeing someone shot. We assess proximity to shootings by race, sex, and cohort during a period of elevated violence in 2020 to 2021. Statistical power is low for estimates of been shot due to the rarity of the event . Hence, 95% CIs are wide. The hazard of being shot was more than 3 times as high for Black individuals compared with White individuals. The difference between Survival curves were estimated using the Turnbull nonparametric maximum likelihood estimator for interval-censored events. Interval censoring results in plateaus in curves between survey waves 3 and 5 and regions where curves are not uniquely identified, indicated by diagonal lines. The observed data are consistent with any increasing line connecting the ends of the diagonal segments. At-risk respondent counts are model estimates as the exact number of respondents at-risk at any time is unknown due to interval censoring. Initial at-risk respondents are different between seen shot and been shot outcomes because been shot questions were not asked in the first survey wave. Triangles at base of panels E and F indicate the approximate age of each cohort at the time of the 1994 violence peak in Chicago . compared with the 1981 cohort, which exhibited the highest estimated exposure, the difference between the cohorts was not statistically significant . The 1996 and 1984 cohorts exhibited hazards similar to the 1981 cohort but these differences were also not significant . --- JAMA Network Open | Public Health Analyses of the hazard of witnessing violence found that, compared with White individuals, the hazard of seeing someone shot was more than 4.5 times as high for Black individuals and more than 2.5 times as high for Hispanic individuals . Compared with female respondents, male respondents were significantly more likely to have seen someone shot . Finally, consistent with results presented in Figure 2F, SPT results indicate that the 3 older cohorts had approximately similar hazards of seeing someone shot, but the 1996 cohort exhibited a hazard approximately half that of the 1981 cohort . Figure 3C presents estimates of the correlates of the count of shootings within 250 m of a respondent's residence during the 12 months preceding their wave 5 interview. While there were no significant differences by sex or cohort in proximity to nearby shootings, we observed significant racial disparities. Black individuals and Hispanic individuals experienced higher rates of shootings in close proximity to their residences than White individuals. Notably, we detected no race or sex by cohort interactions for this measure or for having been shot or having seen someone shot. Hazard ratios and incidence rate ratios greater than 1 indicate higher estimated exposure to gun violence. HRs were estimated with multivariable semiparametric Turnbull proportional hazards models. IRRs were estimated with a multivariable negative binomial regression. Reference categories are White for race, female for sex, and 1981 for cohort. Models were weighted to permit generalizing estimates of exposure to the population of children growing up in Chicago in the 1990s. The case weights applied during estimation are the product of the survey design weights and attrition weights at each wave, constructed using estimated probabilities from binary logistic regression models. 15 Observations received the combined weight at their final follow-up period . Results are insensitive to trimming weights to the center 90% of the distribution. Results are also similar with alternative specifications of survival models and without weights . Sample sizes are 2417 respondents for seen shot, 2135 respondents for been shot, and 649 respondents for nearby shootings. --- JAMA Network Open | Public Health --- Discussion In this cohort study, the accelerated longitudinal design of the extended PHDCN allowed us to examine multiple forms of firearm violence exposure from childhood through midadulthood in a population sample. We found significant differences in exposure to firearm violence by race and sex, and our findings on cohort differences point to changing societal conditions as key factors associated --- Implications for Public Health A broader focus beyond fatalities to include nonfatal gunshot injuries and witnessing of incidents is critical for understanding the full health outcomes associated with firearm violence. In addition, the sustained stress resulting from routine exposure to firearm violence can take a cumulative physiological toll on the body, and is associated with damage to the body's regulatory system and the acceleration of aging and susceptibility to disease. [21][22][23] To the extent that major and potentially chronic stressors are disproportionately endured by some sociodemographic groups, as we have shown here, variation in these stressors may contribute to dramatic sociodemographic differences in health. 24 The stress from chronic exposure to firearm violence may also contribute to subsequent violence, through its impact on aggression, or even through the normalization of violence. 25,26 --- Future Research Consistent with the life-course framework of the current study, future work should focus not simply on the consequences associated with having ever seen someone shot or having been shot themselves, but also the timing of the exposure and accumulation of exposures. Furthermore, a focus on the consequences of exposure to firearm violence should examine both near-term and longterm outcomes. Additionally, further research is necessary to explain the variations in firearm exposure by race, sex, and cohort that were identified in this study, including studies that examine the role of residential segregation, as well as residential mobility. --- Limitations This study has some limitations. Although this long-term study uses a representative sample of children who reached the formative years of adolescence at differing times over the last quartercentury, it is based on children originally from Chicago. This raises generalizability questions, but violence rates and trends in Chicago parallel those in other major cities in the US, such as Philadelphia, Pennsylvania. Furthermore, the extended PHDCN data collection followed-up with respondents wherever they moved. Although retention rates compare well with contemporary urban samples and the analyses were weighted to reflect the sampling design and attrition, selection out of the sample on unobserved covariates may bias the results. Another potential limitation, common to self-report surveys, is recall error, such as difficulty in recalling the precise timing of events. A 2020 study 27 using GVA data found undercounting of nonfatal shootings due to selective reporting of firearm violence by the media. However, race or ethnicity of individuals who were shot was not associated with the likelihood of reporting, suggesting the racial and ethnic differences in nearby shootings reported in this study are unlikely to be biased by selective reporting. Moreover, we found similar racial and ethnic differences when restricting our analyses to fatal shootings. --- Conclusions This cohort study reports on previously undocumented inequalities in exposure to firearm violence over the life course by race, sex, and cohort. Life-course exposure to firearm violence was significantly and persistently higher for Black and Hispanic individuals. Sex disparities in witnessing and proximity to firearm violence were small, but men were at significantly higher risk of being shot than women. Cohort differences were most pronounced for having seen someone shot; in particular, the youngest cohort in our study had a lower likelihood of having seen someone shot than --- JAMA Network Open | Public Health Inequalities in Exposure to Firearm Violence by Race, Sex, and Birth Cohort From Childhood to Age 40 Years --- Drs Lanfear and Sampson had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

The past quarter-century has seen both sharp declines and increases in firearm violence in the United States. Yet, little is known about the age of first exposure to firearm violence and how it may differ by race, sex, and cohort. OBJECTIVE To examine race, sex, and cohort differences in exposure to firearm violence in a representative longitudinal study of children who grew up in periods with varying rates of firearm violence in the United States and to examine spatial proximity to firearm violence in adulthood. DESIGN, SETTING, AND PARTICIPANTS This population-based representative cohort study included multiple cohorts of children followed-up from 1995 through 2021 in the Project on Human Development in Chicago Neighborhoods (PHDCN). Participants included Black, Hispanic, and White respondents from 4 age cohorts of Chicago, Illinois, residents, with modal birth years of

Introduction This paper explores changing relationships between parents and early childhood professionals in light of an increased focus on early learning in Danish early childhood education and care institutions. Danish ECEC institutions have a long tradition for collaboration with parents. This tradition is embedded in a Danish and Nordic social pedagogy tradition, characterised by a child-centred approach with a strong focus on children's play and social relations . However, since the 1990s, this tradition has been reshaped by an increased political interest in early learning. Various policy initiatives and legislative frameworks have K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 10 promoted a greater focus on educational content within Danish ECEC . Accordingly, ECEC institutions are required to support learning environments and create learning opportunities for all children . The rise of a learning agenda in Denmark corresponds with a shift from a focus on children's care and socialisation to learning in other Nordic countries, as well as international discourses on early learning and parental involvement . For instance, in its Starting Strong publication series, the OECD highlights parents as important 'partners' for ECEC services and states that ECEC services can help families develop home learning environments . In Denmark, recent changes in the legal framework governing the ECEC sector have stressed the significance of parents for children's development and learning, as well as the importance of a strong collaboration between ECEC and parents . Today, the objectives of strengthening learning communities and increasing parental involvement in ECEC seem to have merged, with ECEC institutions assigned the task of improving children's home learning environments while strengthening their collaboration with parents . ECEC professionals are now expected to guide parents with regard to their children's learning and development, helping them create appropriate learning environments at home. In this paper, I explore how collaboration with parents on home learning intertwines with existing collaborative practices in Danish ECEC institutions. What are the implications of this collaboration on home learning for the relationship between ECEC professionals and parents? How does this new focus on early learning alter their respective roles and responsibilities? Based on ethnographic studies of specific learning programmes and parental involvement, I argue that the growing dominance of a learning agenda within Danish ECEC institutions extends to family settings, appointing parents as learning facilitators for their children, and leads to an institutionalisation of parenthood. --- Parents and home-learning Existing research on parental involvement in early learning within ECEC has emphasised that different ECEC curricula affect how the role of parents is perceived. Countries where a social pedagogy tradition dominates, such as the Nordic countries, tend to promote a form of parental involvement where parents are guided more broadly and have democratic rights. Meanwhile, ECEC institutions in Anglosphere countries, where the curriculum is often primarily focused on children's learning outcomes, usually involve parents to support children's achievement of early learning objectives and ensure their readiness for school . Most research on ECEC and home learning builds on studies conducted in Anglosphere countries. This research is dominated by quantitative studies measuring the outcomes of interventions or programmes promoting home learning . Such interventions and programmes are often designed to enhance children's school readiness by focusing on literacy and numeracy, providing parents with tasks to do at home with their children and instructions on how to engage in home learning activities . However, with their focus on learning outcomes, these studies provide scant knowledge as to how the collaboration between parents and ECEC professionals on interventions or programmes takes place, how the collaboration is perceived by those involved, or how home learning tasks are embedded in everyday life in ECEC and families and entangled with existing relations between ECEC professionals and parents . To better understand how collaboration on early learning and home learning takes place in everyday life as part of existing collaboration, I draw inspiration from studies within a Danish context focusing on collaboration in a broader sense. These studies have shown that, until recently, ECEC professionals' collaboration with parents has first and foremost focused on children's well-being and social skills . Relationships between ECEC professionals and parents are established and negotiated through daily encounters and formal collaboration. Building trustful relations with parents while balancing professional assessments of children are central issues for ECEC professionals . With a stronger focus on the role of parents, it becomes K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 paramount to provide guidance on parenting issues and child development . --- Theoretical framework To address changing relationships between parents and ECEC institutions, I draw on anthropological and sociological perspectives on institutions and parenting. ECEC centres are often described as formative institutions with certain institutional norms guiding the interactions between children and childcare professionals . In this sense, we can understand institutions as particular forms of routinised social practice, taken for granted by those involved and guided by norms and rules . I argue that, if we are to understand collaborative practices with parents, we must address how the collaboration is embedded in the institutional practices and logics of ECEC. Thus, I explore how institutionalised and taken for granted ways of working with learning influence collaborative practices among parents and ECEC professionals in ways that institutionalise parenthood and target parents as objects for learning initiatives. To analyse the expectations towards parents and the respective roles of ECEC professionals and parents, I also draw inspiration from family studies focusing on parenthood and parenting. Such studies have argued that parenthood and parenting norms and strategies are embedded in social, cultural and historical conditions . Parenthood and parenting are influenced by dominant childhood ideals emphasising the crucial role of parents for their children's upbringing . In contemporary welfare state societies, such as Denmark, with an extensive ECEC sector, families and welfare institutions are entangled through shared work and a common concern for children's development and future possibilities . Parenthood and parenting are influenced by welfare state institutions and experts, and parents' social background and living conditions shape their engagement in childcare arrangements and education . To better understand the efforts of parents and ECEC professionals in terms of early learning, I draw on the sociologist Anette Lareau's work on concerted cultivation, as well as parenting culture studies. Lareau describes concerted cultivation as parents' --- K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 careful and comprehensive efforts to stimulate children's development and maximise their potential . Children are to be intensely cultivated by adults. To understand the Danish case, cultivation can be understood as a process involving not only parents, but also ECEC professionals. Parents and professionals engage in a shared effort to cultivate children's development . Whereas Lareau describes concerted cultivation as a middle-class childrearing strategy, other scholars have stressed the normalisation of concerted cultivation, emphasising how middle-class ideals regarding childrearing have spread to other segments of society . Concerted cultivation can be seen in light of contemporary parenting ideals that stress parenting as a labour-intensive emotional, educational and practical endeavour. This kind of child-centred parenthood is increasingly associated with particular parental abilities and skills, and with a deterministic view of parenting that pinpoints the crucial role parents play in their children's development . Parental determinism gives rise to an understanding of parents as in need of expert guidance to become better parents and minimise the risk of issues arising in children's development . Even though parenting culture studies have primarily examined Anglo-American neoliberal contexts, the discourses they have identified on intensive parenting, expert guidance and risk assessment of children and parents are also part of the Danish welfare state . With a comprehensive ECEC sector, the Danish welfare state has appointed ECEC institutions with the task of guiding, supporting and monitoring parents in an effort to prevent some children falling behind. Drawing inspiration from parenting culture studies allows me to analyse the broader implications of collaboration on early learning, including how ECEC professionals offer guidance on early learning to parents, how notions of what constitutes proper parenting are embedded in such guidance, and the extent to which parents comply with this guidance and the tasks they are assigned to support early learning. --- Ethnographic studies of learning programmes and collaboration The paper draws on empirical data from two ethnographic studies of the collaboration between Danish ECEC institutions and parents on learning initiatives for children, conducted 2018-2022 1 . Both studies followed two specific learning programmes integrated in everyday life at participating ECEC institutions to promote children's learning. Both programmes included collaboration with parents. One of the learning programmes, the learning group programme, was developed by the local municipal authority with the aim of enhancing young children's learning, strengthening collaboration with parents and developing a reflexive and systematic pedagogical practice. The learning group programme was organised as learning activities focused on a specific topic for a small group of children for a period of 4-8 weeks. Examples of topics include literacy and language, maths, music and movement, fantasy, senses and social relations. As part of their work, the ECEC professionals regularly informed parents about the learning process, occasionally assigning them small tasks, and organised a meeting with parents in each learning group to assess the process at the end of the programme . The second learning programme, the READ programme, was offered by the municipality with the aim of improving all children's language skills and early literacy before starting school. The programme was introduced in nurseries with children aged 0-3 and included a selection of books for use at the ECEC institutions and at home, as well as instructions for parents. These instructions guided parents in what they should do before, during and after reading a bookin particular, how to talk to their children about pictures and words. The ECEC professionals decided which book each family should read at home during the following week. The books were also read at the ECEC institution. To examine how these learning programmes became part of daily life at ECEC institutions and explore the collaboration with parents, both studies used an ethnographic approach. Three ECEC institutions participated. In each case, this involved participant observations focusing on the learning programme's implementation at the ECEC institution, as well as the professionals' informal and formal meetings with parents 1 Both projects were conducted in collaboration with Associate Professor Pernille Juhl and Associate Professor Allan Westerling, both Roskilde University. --- K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 regarding the specific learning initiatives. Across the two studies, interviews were conducted with 8 ECEC professionals and 16 parents, as well as 2 ECEC managers. Parents whose children participated in the learning programmes were included in the studies. The majority of participating parents can be classified as middle class based on their educational and occupational status. Consideration was given to ethical issues throughout the research process, complying with GDPR and ethical standards concerning, for instance, informed consent and anonymity. All names mentioned in this paper are pseudonyms. --- Collaborating with parents on early learning With their focus on early learning, the programmes expanded collaboration with parents by including prescribed and standardised ways for ECEC professionals to work with parents that supplemented existing practices . Firstly, the focus of collaboration changed, with less focus on the children's social relations and competences, which otherwise often dominate ECEC professionals' informal and formal collaboration with parents . Instead, the work with learning programmes encouraged a focus on the learning content and the children's progress. In their collaboration with parents, the professionals would outline learning aims, inform about learning activities and children's learning outcomes. Even though the ECEC professionals also reflected upon the participating children's social relations and social competences, this information was usually not passed on to the parents. The prescribed ways of involving parents institutionalised new ways of distributing certain kinds of knowledge about learning, promoting learning as something well defined in the form of specific learning goals. This rather narrow way of thinking about learning presented to parents in the learning programmes differed from ECEC professionals' general approach to learning, where learning is embedded in everyday situations and concerns children's development of social and academic competences . Secondly, the collaboration with parents was expanded with new information, tasks and meetings. For instance, in the learning group programme, the professionals sent parents short messages about ongoing learning activities, encouraging them to talk to their children about the activities and to try some of the activities themselves , pp. 9-30 ©2023 story or sing a song at home that had been part of a learning group activity). In the READ programme, parents received books and instructions on how to use them at home. In this sense, the collaboration with parents on early learning extended the logics, objectives and practices of ECEC to family settings through the organisation of learning programmes that include home-based activities and require the active involvement of parents. As such, the work with learning programmes and parental involvement cannot be separated. --- Helping parents to help their child Overall, the ECEC professionals were preoccupied with how they could get parents to Sharing knowledge is a way to integrate people in institutions . Thus, by sharing knowledge about their work, the ECEC professionals tried to bring parents closer to the ECEC institutions. The above examples suggest that sharing knowledge with parents has a dual purpose. Firstly, by presenting their work, ECEC professionals make parents aware of the pedagogical reflections in ECEC. Thus, it is a way for the professionals to emphasise their professional role as experts on pedagogy and learning. Secondly, motivating or inspiring parents to engage in learning activities at home becomes a tool to integrate parents in an already institutionalised way of working with learning. As the manager in the first example explained, it would help parents to be supportive of the institution. The ECEC manager and professionals' statements seem to build on the idea that parents could and would develop their parenting practices and mould their ways of being a parent according to the professionals' distribution of knowledge and taken for granted institutional ways of working . As experts, ECEC professionals contribute to an institutionalisation of parenthood in which parents are expected to develop their parenting practices by mirroring the professionals' learning aims and strategies. In this sense, the ECEC professionals contribute to the promotion of an ECEC-sensitive parenting style, where parents are expected to support the ECEC institution's learning agenda and willingly comply with the priorities identified by ECEC professionals as important for children's wellbeing and development. Overall, this illustrates that parents are targeted as objects for ECEC professionals' pedagogical work . The sociologist Ellie Lee has argued that parents are targets of advice in distinct ways . Targeting parents to ensure proper parenting was part of the birth of the welfare state. Since then, shifting ideas of expert-guided parenthood, often influenced by psychology, have influenced western societies . However, as the examples above suggest, the field of advice has been expanded to include the home learning environment. Consequently, parents are targeted as learning facilitators for their children at home, expected to adopt the learning ideals, principles and practices that characterise Danish ECEC and integrate them within family life. Targeting parents illustrates how parents are not entrusted with the upbringing of their children on their own. On the one hand, parents are identified as important contributors to their children's learning at home. They are seen as a resource for promoting children's learning, with the time and ability to engage in learning activities with their children at home. On the other hand, parents are seen as a risk factor with a potentially negative impact on their children's upbringing and development . They are considered unable to properly support their children's learning and development on their ownand thus in need of expert advice. With help and advice from ECEC professionals, parents are expected to develop so-called home learning environments, hereby contributing to the overall goal of enhancing children's learning within ECEC. --- K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 --- READing books at home: family time vs language training The aim of the READ programme is to help parents support their children's development of language skills so they will become future readers and to prevent learning inequality . In this sense, sending books to family homes is a way of enhancing parents' awareness of the importance of reading books for their children's language development and of reaching out to those parents who do not already read to their children. The participating parents were aware of key municipality and ECEC policies on children's literacy and the strong focus on language and reading. For instance, one father told us: I highly agree that it is a really good idea to get all parents to read to their children, because it is really important if they are to become good readers later on. We have been told that very clearly, which I think is good. However, when asked about their reading practices at the start of the READ programme, most parents stated that they already regularly read books with their children. As a mother described, reading was "a good way of spending time together with my children", adding that she did not give much thought to her children's learning. This corresponds with other parents, who referred to reading situations as moments of intergenerational togetherness, and as a 'cosy' escape from the hustle and bustle of everyday family life. In Denmark, cosiness, hygge, is a dominant cultural trope and something that we Danes strive for in our social life, including family life . When the parents talked about reading and cosiness in relation to family life, they were often referring to existing family routines and everyday practices. In these families, the READ project became part of a steady rhythm of family life. As described by one mother: My child was very preoccupied with it [the book from the READ project] and it was convenient for us to just read it as a bedtime story instead of one of the other ones she normally has . It meant that we read the same book every evening instead of her being allowed to For those families for whom it was easy to integrate the READ project in their daily life, as in the example above, the READ books usually became part of existing bedtime reading practices. The same mother stated that participating in the READ project did not require major changes to their existing practice, continuing: Like, it didn't interfere with our rhythm. I mean, I could imagine that if they said 'You have to watch this film for an hour every day' or 'You have to eat cucumber every day for a week', or, you know, if it was something that we don't normally do, then it would have an impact on our everyday life. But this hasn't done that. This quote emphasises that the READ project can be integrated within everyday family life because it does not interfere with existing routines and practices. Those parents who explicitly stated that READ became part of a daily routine admitted that they had not read the accompanying instructions. However, they felt that they already read to their children in a similar manner. As other studies have shown, middle-class parents and ECEC institutions share many of the same values and engage in what Lareau defines as concerted cultivation . It seems as though the READ programme can easily be integrated into existing family rhythms because of this shared focus on reading as important for young children, but also because it fits well with these parents' ways of organising and prioritising family life. Meanwhile, several parents stated that it would have an impact on family life if their children constantly brought books home from ECEC or were given some other kind of task on a regular basis, as stated in the quote above. As another mother put it, it would feel like "being given homework as a family". However, as long as the READ programme could fit into existing family practices and did not become an extra chore or burden, the parents found it easy to be supportive. --- K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 This is not the case in all families. Across families with diverse backgrounds, family conditions and work lives, such as having a tight schedule after returning home from ECEC, reading specific books selected by the ECEC professionals and following instructions could be an overwhelming task and even a source of conflict . These parents described how they had more children to attend to, many other family activities during the week, and that they thought there were already "so many messages and so much information from the ECEC institution they had to deal with", as one mother phrased it. When learning projects within ECEC cross over into family settings with no consideration for their existing everyday lives, they can become a burden. Overall, an institutionalisation of parenthood that promotes parenting practices mirroring those of ECEC professionals ignores diversity among families and parents, giving an advantage to those families who share many of the values promoted by ECEC in terms of childhood and learning and who can organise their everyday lives to accommodate the demand for home learning. --- Adapting to learning agendas or supporting ECEC professionals? Parents involved in the learning group programme and the READ programme generally expressed a positive view of the learning initiatives during conversations with the ECEC professionals. However, as I will show, the parents' support reflects a more complex attitude to their children's learning and to collaboration with ECEC professionals. This was the case in the learning group programme, which ended with an assessment meeting attended by the parents of participating children, the responsible ECEC professional and the local manager. The aim was to inform parents about the learning process and to gain an insight into the parents' experiences at home. During these meetings, the ECEC professionals explained the learning goals, outlined the learning activities and described observed learning outcomes. During the assessment meetings, parents most often showed an interest in what they were told, nodding eagerly and smiling. While some parents recognised a few of the learning activities from their children's accounts, others explained apologetically that their children had hardly told them anything about the activities. Nonetheless, they expressed a positive view on the learning group, as seen in the following example, where four parents attended an assessment meeting: At first glance, these parents appear eager to express a positive attitude by telling of how they recognise or have implemented learning group activities at home -like Mai's mother. They seemingly comply with the learning agenda by adapting to ideas about children's learning defined by the ECEC professionalsnot by the parents themselves. By complying with the learning group programme, parents implicitly adapt to an institutionalised way of working with learning and learning goals. At the same time, the adoption of institutionalised learning agendas illustrates performative aspects of parenthood, where parents perform parenthood in ways that can be recognised as proper parenthood . Other studies have demonstrated that parents are well aware of ECEC professionals' expectations of them as proper parents . In the case above, the parents perform a role as positive and supportive parents in front of other parents and professionals at the assessment meeting to gain recognition as good parents, engaging with and adopting the ECEC institution's learning agenda within the family. However, even though they were supportive of the learning agenda, most of the parents were less interested in the actual learning outcomes. Instead, in conversations with the professionals and in interviews, they highlighted the ECEC professionals' engagement with the children as the most important aspect of the learning groups, K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 praising their work with the children and emphasising the importance of working with small groups of children . As one of the parents said: "I am so happy that the ECEC professionals see my child [during learning group activities]". Perhaps this explains why the parents insisted that the learning groups made a difference, even when their children had not told them anything about the activities . In this sense, parents' engagement in the learning programme and participation in the assessment meeting is primarily in support of the professionals' work with their children, rather than in support of learning outcomes and home learning. Being supportive of the learning agenda and engaged in the learning group activities is less about the learning agenda itself and more a matter of acknowledging the importance of the pedagogical and social space the learning group provides for their children in the ECEC institution. Thus, embracing a learning agenda promoted by the ECEC professionals is a way of maintaining a role as an engaged and supportive parent and emphasising one's support of the pedagogical work and confidence in the professionals without disturbing a dominant and institutionalised way of involving parents or explicitly expressing disagreement with the learning agenda. --- Conclusion In this paper, I have examined the implications of ECEC institutions being assigned the task of improving children's home learning while strengthening their collaboration with parents. Based on empirical studies of specific learning programmes that promote socalled home learning environments through collaboration with parents, the paper illustrates how such learning initiatives reshape collaboration, institutional expectations towards parents' engagement in their children's lives and the ECEC institutions, and parents' strategies for performing support and good parenthood without embracing all aspects of a learning agenda imposed on them by the ECEC institutions. Until now, the concerted cultivation of children by ECEC professionals and parents has been influenced by an ideal of the 'social child' , with their collaboration primarily focused on children's social competences . While the 'social child' remains a strong ideal within Danish ECEC that both parents and ECEC professionals strive to achieve, collaboration on early learning extends K.I. Dannesboe: Parents as learning facilitators Qualitative Studies 8, pp. 9-30 ©2023 to children's learning in a more academic sense. As shown, the collaboration has been expanded to include new focus on learning and tasks to do at home. I suggest that this kind of extended collaboration is part of an institutionalisation of parenthood whereby the dominant learning agenda within the ECEC sector crosses over into family settings, establishing a collaboration where, in principle, there are no borders between ECEC and family life . The expansion of an early learning agenda influences the roles of ECEC professionals and expectations concerning parental involvement. ECEC professionals are expected to provide guidance to parents with regard to their children's learning and development, helping them develop appropriate learning environments at home. Guidance from various experts is an aspect of contemporary parenthood . However, within Danish ECEC institutions, the nature of such guidance changes, with ECEC professionals instructing parents to perform specific learning tasks at home. Compared to existing guidance structures, in which parents ask for and receive advice on children's well-being, this new guidance appoints parents with tasks that are usually conducted by professionals with pedagogical qualifications. Consequently, parents are instructed to mould their parenting practices in accordance with ECEC standards and professional knowledge. In this sense, the learning programmes represent an ideal of intensive parenting that builds on ECEC institutions' ideas of what constitutes good parenting while ignoring parents' own values and wishes. The roles and responsibilities of parents change when they are expected to embrace a learning agenda promoted by the ECEC institutions and based on political objectives concerning young children's learning. On the one hand, parents are seen as key collaborative partners, on the other hand, as targets for pedagogical intervention. The analysis illustrates that providing parents with guidance to support their children's early learning implicitly promotes the idea that parents should strive to improve their parenting skills while being responsible for the creation of a home learning environment that supports and nurtures their child's development and potential. In this regard, governmental and institutional logics have infiltrated the private realm of the family home, targeting parents as learning facilitators, supporting their children's academic progress. The strong focus on instructing and guiding parents suggests that parents are seen as representing a learning resource for their children and playing a crucial role in their academic development. This understanding of parents as a resource mirrors a dominant ideal of intensive parenting, characterised as a child-centred and labour-intensive parenting style . In a Danish context, the efforts of welfare state institutions to strengthen parental involvement underline that parents' parenting practices are interpreted as crucial for their children's developmentas well as for ensuring well-functioning ECEC institutions . Nevertheless, parents are also seen as a risk, potentially impeding their children's learning and development. Implicitly, the involvement of parents in children's early learning is based on a belief that, if parenting is not in accordance with the predefined learning objectives, their children will not develop Overall, the spread of early learning agendas in the Danish welfare state has contributed to the promotion of an ECEC sensitive parenthood, advancing an agenda of intensive parenting through both explicit expectations and more implicit assumptions concerning parents' ability to mould their parenting practices in accordance with learning activities at ECEC institutions. While policies promoting early learning seek to enhance all children's learning possibilities, their entanglement with the goal of strengthening collaboration with parents accelerates the development of a far-reaching and comprehensive institutionalisation of parenthood that fails to take into account existing practices of ECEC institutions, the diverse conditions and priorities of family life as well as children's everyday lives across these settings. --- The Danish Evaluation Institute . Samarbejde mellem foraeldre og daginstitutioner. Rapport:_Samarbejde mellem foraeldre og daginstitutioner_www .pdf Vincent, C. . 'The children have only got one education and you have to make sure it's a good one': parenting and parent-school relations in a neoliberal age. Gender andEducation, 29:5, 541-555 --- About the author: Karen Ida Dannesboe is Associate Professor, PhD, at the department of Education, Aarhus University. Her main research area is childhood institutions and families as cultural, social and educational contexts of childhood and children's everyday lives. In recent years her research has been concerned with the collaboration between daycare and families in the light of changing early childhood and education policies.

This paper examines the implications of various learning-centred initiatives for the relationship between early childhood education and care (ECEC) institutions and families in Denmark. Since the 1990s, promoting early learning has been a key objective for Danish ECEC institutions, reshaping the Danish social pedagogy tradition. Recently, early learning initiatives have become part of the collaboration with parents on so-called home learning. Based on ethnographic studies of such collaboration, I argue that the expansion of dominant early learning agendas from ECEC to families results in an institutionalisation of parenthood. The analysis shows that parents are expected to embrace a learning agenda promoted by ECEC professionals. They are appointed as learning facilitators who must strive to support early learning at home and improve their parenting skills. Furthermore, parents' engagement in early learning is intertwined with the practical organisation of family life and with ideals of a good family and a good childhood.

Introduction Breast cancer is the leading type of cancer for females across the Arab region. Adib et al. [1] noted that 30.1% of female cancer cases are breast cancer, while El-Attar [2] found that breast cancer comprises between 16.2% and 38.4% of all types of cancer. In second place on the prevalence scale is non-Hodgkin lymphoma , followed by rectal and thyroid cancers. The least prevalent are buccal, gallbladder, and bladder cancers . Meanwhile, earlier studies found that colorectal cancer is the second most prevalent after breast cancer among women in some Arab countries [3,4]. While many studies in the MENA region focused on prevalence, epidemiology, and etiology of female breast cancer, several studies highlighted the psychosocial experiences of the women who had been diagnosed. This study aims to review these studies and identify the main socio-cultural and demographic factors affecting Arab women's psychosocial experiences through their journey with breast cancer. In their study of the epidemiology and management of breast cancer in Arab countries, El Saghir et al. [5] reported that people in the region are reluctant to speak openly about cancer and that it is highly stigmatized, to the extent that many refrain from mentioning it by name. Hence there is a need to better understand the cultural and religious landscape in which breast cancer patients undergo their treatment, the specificities of which may influence their subjective experiences. Recent studies have shown a positive impact of early examination, which leads to early diagnosis and breast-conserving surgery [5]. Bener et al. [6] found that health workers rarely recommend examinations, resulting in women having inadequate knowledge about breast cancer screening [7,8]. Despite breast cancer being one of the most common cancers among women, the uptake of breast cancer screening and breast cancer self-examination is still relatively low. Donnelly et al.'s [9] multicenter cross-sectional quantitative survey conducted amongst Arab women in Qatar reported that in addition to the low levels of awareness and participation in BCS, only one-quarter of the participants reported that their doctors discussed and recommended BCS. Women who did not undergo BCS cited as reasons the lack of physicians' recommendation, fear, and embarrassment [7,9]. Donnelly et al. [9] attribute late diagnosis of the disease to the low awareness of the importance of BCS. They link this low awareness to a variety of factors, mainly age, education, and the absence of a doctor's recommendation. A diagnosis of breast cancer can be disruptive to an individual's physical, psychological, interpersonal, and financial states [10]. Breast cancer and its treatment are associated with a broad range of symptoms that impact physical, social [11], mental , and cognitive functioning. This upheaval can have a profound impact on women's well-being and quality of life. Within clinical domains, quality of life has become an important health outcome measure, used to indicate which facets of an individual's life are most affected by a disease and its treatment [12]. Wisloff et al. [13] demonstrate a link between compromised quality of life and clinical outcomes, including survival. Sociodemographic factors emerge as an important theme in the quality of life literature [14]. Past studies found that factors such as a patient's educational level, age, amount of spousal and familial support, employment and financial status, and stage of disease progression can predict the quality of life of patients with breast cancer [10,15,16]. In this paper, we present the main psychosocial aspects of being a woman with breast cancer in the MENA region. --- Materials and Methods This study is a narrative literature review, in which the researchers reviewed existing studies on the psychosocial aspects of female breast cancer in the MENA region published between 2007 and 2019. We chose a literature review because we aimed to document the main findings by thematically analyzing and integrating the knowledge gained over a whole decade by several researchers in all Arab countries. In this type of review, researchers can summarize shared themes across existing studies, bringing together their conclusions into a holistic interpretation contributed by the reviewers In doing so, we can also summarize similarities and differences, meaning that while we consider the socio-cultural similarities shared by the Arab countries, we can also explore the uniqueness of a specific country that results from its local circumstances. Literature review studies' results are narrative and qualitative rather than quantitative, and they enable the researcher to acknowledge, reflect, and assist the reviewed data. Using a literature review allows us to analyze the psychological and socio-cultural barriers, needs, and challenges facing Arab women with breast cancer. We began by searching academic databases including PubMed, Scopus, Ebsco, and JSTOR. We used the keywords "society", "culture", "breast cancer", "women", "screening", "quality of life", "attitudes", "stigma", "women", "Arab", "Muslim", "family", "coping", "mental health", and "religion". The study included only English language articles and covered both qualitative and quantitative studies. Chapters in books, research reports, Arabic articles, and studies and information published as abstract only were excluded. Both authors, with the assistance of two students, reviewed the titles and abstracts gathered. We found 293 candidate articles, which were further reviewed and reduced to 74 articles . We included only research papers where the main topic was psychosocial experiences and psychosocial aspects of female breast cancer in the MENA. In order to analyze and categorize the content of the articles included, we used conventional content analysis whereby identifying the main concepts and categories is derived directly from the text data [17]. In this type of analysis, researchers determine the presence of certain themes or concepts within the given qualitative data , that repeat themselves across the data. By using content analysis, researchers can identify the presence of certain themes and concepts, as well as identify meanings and relationships between the identified themes and concepts. For credibility assurance, the authors debriefed and discussed the methods and results with experts in the topic of this review from different health sciences fields. The team of experts included two public health researchers, a psychologist, a psychiatrist, a social sociologist, and a medical anthropologist. Furthermore, the researchers evaluated the quality of their review by using the SANRA scale. SANRA Scale is a validated tool developed to evaluate the quality of narrative and literature reviews [18]. The scale includes six items: Justification of the article's importance, Statement of concrete aims of the review, Description of the literature research, Referencing, Scientific reasoning, and Appropriate presentation of data. Each item includes three levels, zero, one, and two, where two indicates that the review meets the criteria for quality. We evaluated the manuscript individually and then met to discuss gaps and ways to improve the manuscript in order to meet SANRA's criteria. --- Results In reviewing the 74 papers on our final list, we identified five main themes that highlight the psychosocial aspects of female breast cancer in the MENA region. These themes are 3.1. Who cares? 3.2. Sociodemographic and cultural determinants associated with breast cancer, 3.3. Awareness and sources of knowledge, 3.4. Quality of life of women with breast cancer, and 3.5. Family support and relationships of women who have breast cancer. 3.1. Who Cares? States, Themes, Types, and Subjects of the Studies Included BSC/BSE was the most frequent topic, discussed in 27% of studies. This was followed by the topic of quality of life of women with breast cancer . The least discussed topic was sexual functioning . Other prevalent themes included attitudes and beliefs regarding breast cancer, its examination and its treatment ; health care providers' roles ; and family support and mental health issues . Some countries conducted a higher number of studies than other countries. Egypt, Lebanon, Palestine, Qatar, and Saudi Arabia were amongst those who published the highest number of studies on the addressed issues. Libya, Syria, Yemen, and Sudan were amongst those who published the lowest numbers of studies. North Africa's countries such as Tunisia and Morocco were in the middle. North African countries are considered francophone countries, and researchers in these countries publish in French more than in English. It is therefore likely that these countries have additional studies published in French, but these were not included in our review due to the exclusion criteria. Among the Middle Eastern countries, Syria, Libya, and Iraq had the lowest number of studies. This can be explained by the political crises and conflicts these countries have been experiencing since 2011, which have made conducting research more difficult. The most frequent research participants were diagnosed women, included in 45% of studies, healthy women , health care workers , the general population , and family members . Most of the studies were quantitative , 20% were qualitative, and 10% were review studies. We argue that the number of the studies, the subjects, and the topics indicate who cares about which aspect of psychosocial experiences of female breast cancer. --- Sociodemographic and Cultural Determinants Associated with Breast Cancer The studies reviewed in this paper indicate that an alarming number of women received their breast cancer diagnoses later in the course of their disease . Socio-cultural and political factors, family values, and religious beliefs were consistently mentioned by the studies as the main factors contributing to women being late in seeking help and thus receiving a late diagnosis. Economic status and political situation were reported as the most influential factors with regard to access to health care. The reviewed studies indicated that a lack of health insurance, low economic status/income, and distance from health care facilities seemed to be the most prominent causes of late diagnosis [1,[19][20][21]. Conflicts and political crises were an additional reason for lack of access to healthcare systems. For example, Palestinian women living in the West Bank faced hardships passing military checkpoints. As a result, they considered the journey to a hospital for screening worthless and postponed seeking diagnosis until the late stages of the disease [22,23]. Religious beliefs were also found to have an impact on women's experiences. When women were questioned about their assumptions with respect to the causes of breast cancer, many linked it to religious causes, such as a test or punishment from God for previously committed sins [24][25][26]. In some contexts, women mentioned unique beliefs regarding the causes of breast cancer, including the increased use of hormonal birth control pills and not breastfeeding. Women believed that Islam promotes and encourages women to breastfeed and utilize natural birth controls for several reasons. One is to protect against breast cancer [25,26]. Furthermore, cultural norms and religious beliefs constituted barriers to early diagnosis and treatment [27,28] Religious practices, such as becoming more devout and praying much more often, as a means of coping with their illness were reported by women, as well as references to tawakkul [29][30][31][32]. In addition, screening and early diagnosis were perceived to be worthless and futile since God is the only protector and healer, even though, opposingly, treatment was claimed to be necessary since "our bodies are lent to us by God" and must be taken care of [25,26,33]. A Moroccan study that interviewed nurses and doctors showed that 60% of nurses believed that breast cancer can be cured by adhering to prayer without any kind of therapy [34]. Moreover, screening was considered to be a breach of a woman's Islamic modesty [35,36]. Some cultural values in the MENA region necessitated having a female doctor for clinical breast examinations and mammograms, which resulted in a delay in diagnosis of approximately 8 months among Libyan, Palestinian, and Egyptian women [25,37,38]. Furthermore, culturally, the term "cancer" has been shown to be associated with either death or baldness due to chemotherapy [29,31]. This seemed to be a more common conception among Arab women, leading them to avoid mentioning it to their families. For example, Palestinian women living inside Israel reported that they felt more comfortable communicating with Jewish women during their chemotherapy sessions than with fellow Palestinians who also suffered from breast cancer [39]. Their comfort speaking about their diagnosis with strangers resulted from the social stigma and the way that cancer is perceived and framed by their close community. Cancer was also linked to being attacked by someone's "Evil Eye", and was believed to be able to be resolved on its own or through the use of home remedies, such as rubbing the lump with olive oil [24,29,32,[40][41][42][43]. --- Awareness and Sources of Knowledge Knowledge about breast cancer risk factors and clinical features is a vital indicator of conducting regular BSE, seeking CBE, and pursuing treatment. Women, mostly working mothers, who had knowledge about BSE practice lacked general information about the frequency and best time of their menstrual cycle to perform BSE [44][45][46][47]. Knowledge of some indicators of breast cancer risk factors, such as the absence of non-lump breast signs, nipple retraction or changes in breast size and/or shape, other health symptoms like weight loss or fatigue, and denial of and fear of finding a lump, made women less likely to perform BSE or undergo a CBE or mammogram [20,27,34,37,42,[48][49][50]. Furthermore, women who successfully underwent a BSE, CBE, or mammogram once did not repeat the screening tests regularly since they believed that the initial negative results accompanied by no breast changes made a second screening unnecessary [36,51,52]. For example, it was found that in the Palestinian Authority, more than 60% of women above the age of 50 had never undergone a mammography and did not know about the need for regular screening tests, while 72% had never had a CBE [22]. Women who underwent CBE also performed BSE regularly, which was also a common trend among Qatari women [26]. It was also noted that women who performed screenings were motivated to do so because they knew someone who had suffered from breast cancer [22,26,32,36,45]. The gap between initial occurrence of symptoms and diagnosis among women of different Arab nationalities living in the UAE ranged from 3 months to 3 years, and this was largely explained by the inappropriate information that women had about the presentation of breast cancer [42]. A huge gap in the knowledge of the risk factors of breast cancer was also noticed among women in northern Saudi Arabia, which may have resulted in delayed diagnosis, as in the UAE [53]. On the other hand, 15.5% of Libyan women reported that they had been falsely reassured by their physicians that their breast lump was benign [37]. The impact of level of education on breast cancer awareness is very debatable. Despite the expectation or hypothesis that women who attended university or who had higher levels of education would have had more knowledge about breast cancer diagnosis, screening, and risk factors, the actual correlation varied between countries. For example, female students attending the University of Assuit and Ain Shams in Egypt and Taibah University in Saudi Arabia were found to have poor knowledge about breast cancer risk factors and clinical features [48,54,55]. Even among those students who had a good level of knowledge about breast cancer and were aware of the existence of BSE, the majority did not recognize the need to perform screenings regularly because of their young age or lack of knowledge on how to perform it. This lack related to a low interest in learning about the topic. Similar results were found in a study of Sudanese medical students who were trainees in the OB/GYN department at Omdurman Maternity Hospital [56]. Likewise, among Omani women who had completed postgraduate studies, almost half had poor knowledge about breast cancer. Similar results were also observed among Jordanian female students of different majors and departments [57]. In addition to the deficiency in information, women reported that their primary source of information was not from health care providers but rather from the media, including the internet, social media, and television [9,26,32,44,45,47,[58][59][60][61][62]. More than half of the subjects in the papers who underwent BSE, CBE, and/or mammograms were advised by the doctors whom they trusted the most. In Iraq, although both students and teaching staff at the Technical Institute of Shatra showed a high percentage of awareness, 73% and 88%, respectively, only 25.4% and 24.4%, respectively, performed BSE [63]. The participants of this study also reported that their principal sources of information were the internet and television , health care providers , and family [63]. In Palestine, only 15% of doctors recommended breast exams to their patients; 48% of them remarked that they did not know if radiotherapy was available to women in Gaza. Similarly, only 19% of female practitioners in Saudi Arabia ordered a mammogram for women over 40 years of age despite their high level of knowledge about breast cancer risk factors [64]. Therefore, the lack of physicians' recommendations for screening was highlighted in most of the countries. Researchers attributed this lack to the embarrassment and shame that derive from gender and body values within the discussed societies [26,37]. Researchers found that the patient's level of education influenced the level of knowledge of risk factors as well as the way women coped with the diagnosis. Women who had a university degree or higher, especially those who were working, tended to experience less severe episodes of depression and anxiety and had better physical functioning [1,[65][66][67][68][69]. In other words, knowledge and education were found to be helpful in improving women's quality of life . Hence, we can conclude that increasing awareness and knowledge around breast cancer can be an alternative to physician recommendation. --- Quality of Life of Women with Breast Cancer According to the WHO, quality of life is defined as an individual's perception of their position in life within the context of the culture and value systems in which they live and in relation to their goals, expectations, standards, and concerns [70]. When Arab women were asked about their definition of quality of life, they generally described their ability to fulfill their roles as mothers and wives [24,28,29,33,71]. Lower income and higher financial stress were associated with a poor quality of life in addition to higher pain severity [71]. In Yemen, younger women had worse scores on the quality of life scales [72]. This could be due to the socio-political crisis in which they live, which is described by many human rights organizations as the worst crisis of the century. Adverse effects related to treatment interfered with general activity and social functioning and hindered women from performing their day-to-day chores as mothers and wives [24,33,71,73,74]. Women described physical symptoms, including fever, nausea, fatigue, vomiting, dyspnea, poor appetite, and arm pain, which greatly impacted them psychologically, leading to depression, constant anxiety, sleep disturbances, and social difficulties. The reverse effect was also evident [67,[75][76][77][78]. Studies revealed that the extent of pain was influenced by the time of diagnosis as well as the mode of treatment. Women who underwent radiation and/or immunotherapy scored best on a spiritual well-being scale [79,80]. Studies in Egypt and Jordan showed that physical and psychosocial symptoms were statistically significantly better in those who underwent breast-conserving surgery rather than modified radical mastectomy [66,81]. In Jordan, lumpectomy surgeries had better outcomes compared with mastectomies [66]. The studies indicated that the progression of women's psychiatric symptoms was significantly associated with advanced stages of the disease, such as metastasis, breast cancer relapse, and multiple tumors, and with continuous post-treatment pain [67,71,77,80,82]. Hence, the stage of the disease was a direct factor in shaping their perception of quality of life. Some of the issues commonly reported by women as consequences of therapy, that their physicians failed to mention, were loss of what represented their femininity, such as breast resection, hair loss, breast pain, lack of desire to engage in sexual activities, and vaginal dryness [24,29,66,74,83,84]. In Morocco, for example, 84% of women with breast cancer who continued sexual activity described it as being increasingly uncomfortable/unenjoyable, painful, and undesirable [77,78,83]. This was substantiated by research done in Bahrain in which women were interviewed about their spouses' reactions to their illness. The women expressed that they felt rejected and weak [85]. Another study in Bahrain showed that husbands sometimes even interfered with the type of treatment their wives were to undergo by insisting on less aggressive forms of treatment or none at all [24]. This indicates that women's quality of life is partially shaped by gender hierarchies within their society. Some women tried to tackle the decline in their quality of life caused by breast cancer by resorting to alternative medicine in parallel with, and sometimes instead of, their treatment plans due to a fear of the toxicity caused by chemotherapy. A high percentage of those women did so without consulting their doctors [43,86]. --- Family Support and Social Relationships of Women Who Have Breast Cancer The studies that were reviewed showed the importance of the role of the family in both diagnosis and treatment stages. For example, divorced and widowed Saudi women scored lower on social well-being than their married counterparts [79]. Patients were affected by the presence or absence of familial support as well as their families' reactions to the illness. An example of the importance of familial support was illustrated in a study performed by Adib et al. [1], which showed that Iraqi migrants living in Lebanon suffered from depression and anxiety more than Lebanese women. This was mainly due to the lack of nuclear family, extended family, and community support, in addition to other factors such as economic status. Furthermore, family structures often change when women are unable to perform their roles as wives and/or mothers [31,40]. While the majority of women shared the news of their diagnosis with their closest family members, other women felt obliged to hide the news from their family members for various reasons. For example, some hid it from their husbands due to the fear of divorce or being forced to accept their husbands marrying a second wife [33,84]. This behavior resulted from the impact of the breast cancer diagnosis and its implications for women's femininity, marriage, intimate relationships, and body image [33,40]. Women hid their diagnosis from their friends due to their fear of becoming a burden and of the social stigma associated with "that" disease. This made women hesitant about mentioning their hospital visits to avoid questions from people who knew them [33,39]. Hiding their diagnosis from family and friends led women not to receive the appropriate treatment in time because they were unable to hide the side effects of ongoing treatment [39]. Some studies mentioned that women hide their diagnosis from their children to protect them from worry and grief. For example, mothers in Saudi Arabia who revealed their diagnosis to their children, despite experiencing a stronger mother-child bond, admitted that the news negatively affected their children's academic performance, which in turn increased the burden on the women [87]. Hiding a diagnosis from children is a protective act resulting from mothers' fear of negatively impacting their children's emotional and academic performance. This is a mothering value, protecting the child despite being in need of support oneself. --- Discussion This review found that several socio-cultural and political factors affect Arab women's quality of life when they are diagnosed with and treated for breast cancer. One of the interesting findings was that few studies addressed sexual functioning, family support, and mental health. These are essential needs that can be reflected directly in women's QOL in the process of healing and facing cancer. We argue that researchers and health care workers should give these factors as much attention as they do the physical symptoms. We assume that the absence of studies on sexual functioning indicates that the sexuality, sexual health, and sexual functioning of women and their ability to express their attitudes and feelings toward it are still considered inappropriate or taboo [88]. Similarly, the social stigma surrounding mental health constitutes a barrier to seeking help for the patient, and a barrier to researchers in asking patients about the issue [89]. In Arab societies, family members tend to hide illness and disease from both their close community and strangers because the cultural value is to keep family issues in the home [89]. This can be explained by the fact that illness and disease are considered weaknesses that society must not see, because if they are seen, the family will be perceived as weak and vulnerable by society. The political events since 2011, beginning with the "Arab Spring", led to continuous crises in several countries, including Syria, Libya, Yemen, and Iraq. We assume that these events have affected women's life in general and have influenced the research agenda within these countries, which reflected by the scarcity of studies on women and breast cancer. In general, in times of conflict women become more vulnerable and women's issues are pushed to the bottom of national priorities [90]. Sidel and Levy [91] argue that armed conflicts may be associated with poor health and poor access to quality medical care, especially for women, children, and the elderly, that is, the vulnerable and those most affected by wars and conflicts. Political conflicts have a direct impact on women's QOL, healing processes, and the general ability to cope with breast cancer. Hence, these women are essentially facing three battles: political, medical, and social. Wars reproduce patriarchy within and between genders and thus require a focus on those institutions that are crucially responsible for the production of masculine identity [90]. Almost half of the studies we reviewed focused their research on women who had been diagnosed with breast cancer, while one-third studied the experiences and attitudes of healthy women. This reflected the patient-centered approach in the studies, where women's voices were the main voices. Despite this, most of the studies were quantitative. While quantitative methodologies tend to analyze phenomena in terms of prevalence and frequencies, qualitative methodologies aim to determine the meaning of a phenomenon and develop concepts that help in our understanding of the phenomenon through those who are involved in it, that is, the human participants. We attest that there is a need for more qualitative or mixed-methods studies in order to better understand women's experiences and needs. This will help to better implement the patient-centered approach in healthcare systems, which highlights patients' perspectives as a basic principle in health care [92]. Socio-cultural factors, such as religion and cultural values, were found to play a major role in BSE and BCS behaviors. The term "cancer" has a negative connotation because it is associated with death and end of life. In her work Illness as Metaphor , Susan Sontag clarifies how fatalistic social perceptions and the framing of cancer, reflected in the language and metaphors used to refer to the disease, negatively influence women's perceptions, experiences, and healing processes [93]. The Health Belief Model [94] considers cultural values and religion as significant variables affecting perceptions of illness and health, with a major impact on managing health and sickness in certain societies. In Arab-Muslim societies, that is all MENA states, religion plays a role on three levels: the holy text , the belief system that derives from the text, and the social-religious practices that derive from both [95]. In several contexts, there may be gaps between the holy text, the belief system, and the religious practice. In our review, the gap was between the perception of illness as a punishment or test that only God can heal, and the belief that our bodies are a gift from God and we have to take responsibility and care for them. This contradiction is reflected in women's screening behavior whereby they say that God gave us our bodies and it is our duty to care for them, yet they also engage in late screening behaviors resulting from fear, embarrassment, and rejection of the idea of being sick. Another example of a gap between the Quranic text and social beliefs is that envy is mentioned in the Quran, the holy book for Muslims, as something people should protect themselves from, but it is not referred to or based on the concept of the "Evil Eye." Arab-Muslim societies use the Evil Eye to refer to envy of others and every bad event they cannot explain. Abu-Rabia argues that the belief in the Evil Eye is embedded in the folklore of fallahin societies throughout the Middle East [96]. We argue that this belief in the Evil Eye influences perceptions of disease and illness, and affects the way women with breast cancer, and indeed other ill Arab-Muslim people, manage their diseases. Religious and social leaders, as well as awareness campaigns in all platforms and forms, play a crucial role in re-structuring belief systems and encouraging women to seek early diagnosis and care for their health. The primary factor to consider in preparing these awareness campaigns is the access of women. Access, in this context, refers to language and location. Studies that considered the role of breast cancer awareness campaigns and investigated their impact on women reflect that these campaigns were ineffective in countries where they occurred abundantly. Women in our review, such as women living in Qatar and Saudi Arabia, recommended using the Arabic language to distribute information. Another common recommendation was to make the campaigns more culturally sensitive by locating informational booths in female-restricted areas instead of public malls, so that women would feel less exposed while visiting the booths [21,38,97]. In addition, in Lebanon it was found that public information campaigns were occurring at a much lower rate in rural areas compared with in the cities [1]. Women in rural areas were found to use more alternative and herbal medicine due to their distance from main health care centers. Few studies mentioned alternative medicine, especially herbal remedies, as a tool women use during their trials to promote healing in parallel to the clinical treatment. Azaizeh et al. [98] found that the eastern region of the Mediterranean has been distinguished from other regions through a rich inventory of complementary alternative medicine, in particular herbal medicine. Our review found that being a working and educated mother does not necessarily indicate better self-care or better awareness of BSE; in fact, we found that knowing someone, such as a family member or friend, who was diagnosed with breast cancer had a stronger impact on self-health management, resulting from fear that prompted women to perform self-or clinical screening. Similar results were found in non-Arab countries, such as in Serbia [99]. We assume that many Eastern and Western societies may express fear as an emotional reaction associated with the diagnosis of a close friend or family member. The difference between societies is in the way their culture shapes an individual's reaction to it. One of the cultural factors found to create delay in seeking diagnosis was a woman's preference in having a female doctor perform the clinical breast examination. The best way to counter this factor is to seek out female doctors. The healthcare system should also construct a culturally tailored system by providing more female doctors. In addition to a lack of female doctors available, our review found that the lack of a physician's recommendation for breast examination was significant. Indeed, this is one of the major reasons women were unaware of the importance of BSE and BCE. Some researchers attributed this lack of physician recommendation to the gender dynamics and cultural values of Arab societies, which can result in embarrassment and avoidance by doctors who would be recommending BSE or BCE to their patients. Lack of doctors' support and women's preference not to discuss health concerns with family members led them to rely on the media, social media, and the internet as their main resources of knowledge. Although some cultural values created barriers to seeking early diagnosis, other cultural values promoted the healing processes. For example, the Arab cultural structure is characterized by collectivism and patriarchy, which was found to play a positive role in the way women were able to deal with treatment. Women reported relying on support from their extended families that lived nearby when they were in need. Extended family members filled the woman's roles in mothering, housekeeping, and managing the daily lives of her nuclear family. Studies from other regions in the world found the same results in terms of extended family support [100,101]. Furthermore, we found that migrant women fleeing war, such as Iraqis, Syrians, and Yemenis who lived in other Arab countries far from their extended families, lacked the necessary support and this was reflected in their poorer quality of life. --- Conclusions Socioeconomic and socio-cultural factors are important in shaping the quality of life of Arab women with breast cancer. These factors are combined with clinical factors, such as the stage of disease and treatment modality. While all women with breast cancer generally share these experiences and impacts, the features of the society that Arab women inhabit make the development of these experiences unique to them. Gender hierarchies and patriarchy, family values and tribal mentality, cultural practices that contradict the holy religious texts, and cultural perceptions of cancer and women's bodies may make the experience of being diagnosed with and treated for breast cancer harder for women living in the MENA. A limitation of this study is that it may not have included all relevant studies because the search was limited to four databases. Indeed, literature reviews cannot include every study on the topic, and this is one limitation of this type of review. We also think that including the studies published in Arabic and French, especially on North Africa's countries, could enrich this review. Further qualitative studies are recommended to explore women's actual lived experiences, especially during the current COVID-19 outbreak. Such studies may shed light on how women's psychosocial experiences are affected during an infectious disease outbreak, a research gap that should be explored during this painful but historical era of COVID-19. --- Supplementary Materials: The following are available online at http://www.mdpi.com/1660-4601/17/18/6802/s1. Table S1: The included studies. ---

Breast cancer, the most common cancer among women in the Middle East and North Africa (MENA) region, is associated with social and psychological implications deriving from women's socio-cultural contexts. Examining 74 articles published between 2007 and 2019, this literature/narrative review explores the psychosocial aspects of female breast cancer in the MENA region. It highlights socio-cultural barriers to seeking help and socio-political factors influencing women's experience with the disease. In 17 of 22 Arab countries, common findings emerge which derive from shared cultural values. Findings indicate that women lack knowledge of breast cancer screening (BCS) and breast cancer self-examination (BSE) benefits/techniques due to a lack of physicians' recommendations, fear, embarrassment, cultural beliefs, and a lack of formal and informal support systems. Women in rural areas or with low socioeconomic status further lack access to health services. Women with breast cancer, report low self-esteem due to gender dynamics and a tendency towards fatalism. Collaboration between mass media, health and education systems, and leading social-religious figures plays a major role in overcoming psychological and cultural barriers, including beliefs surrounding pain, fear, embarrassment, and modesty, particularly for women of lower socioeconomic status and women living in crises and conflict zones.

mediate between these epistemological disputes, offers an alternative ontology that -according to its advocates -'cuts across' or is 'transversal to' the realist/constructionist dualism . This is achieved by displacing the human researcher/observer from her/his central position in the interaction between the world of events and the processes of research. In this new materialist ontology, the capacity of research to engage with events no longer revolves around the disputed capacity of a human observer either to uncover the real mechanisms that produce the social or natural world , or to offer contextual understandings of it . Instead, as will be seen, both events and research processes are considered as material, relational and interacting networks comprising human and non-human components. 'New materialist' approaches have emerged from a variety of philosophical and social theoretical positions, including feminism, biophilosophy, actor-network theory, quantum mechanical theory and posthumanism . Like constructionism, new materialist social theory has concerned itself with the material workings of power, but with a clear emphasis upon social production rather than social construction, and upon matter rather than textuality . Like realists, new materialist theorists concern themselves with the contribution of science to social justice and liberation , but in materiality they see multiplicity, mutability and becoming rather than singularity, stability and being . Among the radical claims of new materialists are that materiality is relational and emergent, plural and complex, uneven and contingent ; that nature and culture are not to be regarded as distinct realms -as both materially affect an ever-changing world ; and that the capacity for agency extends beyond human actors to the non-human and inanimate . Taken together, these distinctive features suggest that new materialism may offer novel opportunities to address the concerns of those involved in analysing social research data and applying it either to explain or to change the world: namely, the relationship between research data and the object of inquiry. We will explore what a new materialist social ontology reveals about the processes and micropolitics of social inquiry, about the relationships between events, research and knowledge, and the kinds of knowledge that different research designs and methods can produce. Our point of entry into an exploration of social inquiry from within new materialist ontology is by considering research as assemblage, a concept that we will examine shortly. From a new materialist perspective, a research-assemblage will comprise the bodies, things and abstractions that get caught up in social inquiry, including the events that are studied, and the researchers. But if research can be assembled it can also be dis-assembled, and this paper will delve inside the research-assemblage, opening up the research process to micro-political investigation, revealing the material interactions between researcher and researched. We will use this mode of analysis to explore a range of quantitative and qualitative research methods and methodologies, to assess the processes that go on within their different research-assemblages, and the micropolitics they produce. From this we will reflect upon how a materialist ontology can inform the design of social inquiry methodologies, raising the possibility of engineering research methods and methodologies by designing into them specific micropolitical objectives. Before this, for the benefit of readers unfamiliar with a new materialist perspective, we summarise the framework upon which this ontology is built. --- 2.3 2.4 2.5 2.6 2.7 micro/macro distinctions, and recognising that materiality and the world and human history are produced by a range of material forces that include physical, biological, psychological, social and cultural . By drawing nature and culture, mind and matter into a single arena, new materialisms radically extend the scope of materialist analysis beyond traditional concerns with structural and 'macro' level social phenomena , to address issues often regarded as 'micro-sociological', because of their association with how thoughts, desires, feelings and abstract concepts contribute to social production . These elisions also mean that new materialism cuts across other social theory dualisms including structure/agency, reason/emotion, human/non-human, animate/inanimate . Consequently, new materialism 'foregrounds an appreciation of just what it means to exist as a material individual with biological needs yet inhabiting a world of natural and artificial objects, well-honed micro-powers of governmentality, but no less compelling effects of international economic structures' . Beyond these commonalities, new materialist scholars have diverged in how they have conceptualised materialist ontology. Thus, for example, Barad retains a concept of agency but extends it beyond human actors to the non-human and non-animate; actor-network theory scholars speak of human and nonhuman 'actants' that have in common a capacity for agency within a heterogeneous network comprising both natural and cultural elements ; while Deleuze, Guattari and their followers prefer to talk of affectivity . Barad describes the 'intra-actions' between matter and knowledge that produce phenomena, while Deleuze andGuattari , DeLanda and Latour consider how the physical and cultural assemble together to produce bodies, social formations and events. To achieve the objective of this paper to draw out features of new materialism that impact upon the ontology of social inquiry, we will use the well-developed and widely-applied conceptual framework deriving from Gilles Deleuze's reading of Spinoza, as developed and applied in the work of Deleuze andGuattari , by theorists such as Braidotti , DeLanda , Grosz and Thrift , and by social researchers such as Fox and Alldred , Renold and Ringrose and Youdell and Armstrong , though we cross reference to other new materialist scholars where appropriate. The DeleuzoGuattarian approach is predicated upon three propositions. Firstly, bodies and other material, social and abstract entities should be regarded not as ontologically-prior essences occupying distinct and delimited spaces, but as relational, gaining ontological status and integrity only through their relationship to other similarly contingent and ephemeral bodies, things and ideas . Assemblages of these relations develop in unpredictable ways around actions and events, 'in a kind of chaotic network of habitual and non-habitual connections, always in flux, always reassembling in different ways' , and work like 'machines' that do something, produce something. An important opportunity afforded by this understanding is that whatever is assembled can be disassembled to see how it works, a characteristic we will exploit fully in this paper. Second, all matter has an 'agential' capacity to affect, rather than being inert clay moulded by human agency, consciousness and imagination . Consequently, the ontology replaces a conventional conception of agency with the Spinozist notion of affect , meaning simply the capacity to affect or be affected: in an assemblage, there is no 'subject' and no 'object' . Rather, an affect is a 'becoming' that represents a change of state or capacities of an entity -this change may be physical, psychological, emotional or social. Affects produce further affective capacities within assemblages , and because one affect can produce more than one capacity, social production is 'rhizomic' rather than linear: a branching, reversing, coalescing and rupturing flow. Affective flows render assemblages constantly in flux, with territorialising flows stabilising an assemblage, while others de-stabilise or de-territorialise it , fragmenting an assemblage. Affects may also aggregate relations in assemblages, while other affects are non-aggregative or 'singular', affecting a single relation within an assemblage in a unique way. So, for example, naming a new pet kitten 'Daisy' is a singular affect, while categorising it as tabby or tortoiseshell is aggregative. These fluxes within and between assemblages create an 'economy' of affects and are the process by which lives, societies and history unfold, 'in a world which is constantly becoming' . The final proposition marks the radical divergence from the exclusive focus in earlier materialist sociologies upon macro-structures, social institutions and economic relations noted earlier. Because thoughts, ideas, feelings, desires, and collective abstractions and 'constructions' can all materially affect and be affected by other relations in an assemblage, they can be treated in exactly the same way as other relations . With this focus upon the materiality of affects and of the actions, interactions, subjectivities and thoughts they produce, the net is cast far more widely than in earlier materialisms , to address materiality across 'micro/macro' and culture/nature divides. --- The research-assemblage: affects, flows and micropolitics These concepts -of assemblages, affective flows and economies, territorialisations and aggregations, provide the means for us to launch our new materialist analysis of social research. Conventionally, social inquiry has been anthropocentric, regarding the researcher as the prime mover in the research enterprise, whose reason, logic, theory and scientific methodologies gradually impose order upon 'data' to supply an understanding, however imperfect, of the world . By contrast, a materialist ontology of assemblages and affects treats the researcher and the researched event, plus the many other relations involved in social inquiry such as the tools, technologies and theories of scientific research, as elements in a research-assemblage productive of a variety of material capabilities in its human and non-human relations. We take this concept of the research-assemblage as our focus for what follows. Deleuze and Guattari described assemblages as 'machines' that link affects together to produce or do something. With this in mind, a 'research-assemblage' can be defined in terms of the multiplicity of affective relations in the research process, including the 'events' to be researched ; research tools such as questionnaires, interview schedules or other apparatus; recording and analysis technologies, computer software and hardware; theoretical frameworks and hypotheses; research literatures and findings from earlier studies; the 'data' generated by these methodologies, methods and techniques; and of course, researchers. To this list may be added contextual relations such as the physical spaces and establishments where research takes place; the frameworks, philosophies, cultures and traditions that surround scientific research; ethical principles and ethics committees; research assessment exercises; and all the paraphernalia of academic research outputs: libraries, journals, editors, peer reviewers and readers. While it is thus possible to disassemble a research-assemblage to disclose its constituent relations, it is far more productive for our analysis to seek out the affects that bind the assemblage together, and we will use Deleuze and Guattari's machine metaphor as the basis for our analysis. To this end, we will treat the research process as if it were a series of interconnected machines that do specified tasks such as data collection, data analysis and so forth . Like physical machines that have been constructed to work in a specified way and produce certain outputs, we will regard the relations within a research-assemblage as engineered to achieve their objectives as a consequence of the particular affective flows between event, instruments and researchers that a methodology or method requires. Thus, a 'data collection machine' would take aspects of an event as its raw materials, and by the means specific to its design, generates 'data'. An analysis machine processes data according to specific rules of logic, deduction or inference, and frequently interprets it within a specific theoretical or conceptual framework to produce 'findings' in the form of generalities or summaries. A reporting machine takes these outputs of data analysis and creates knowledge products for dissemination: theory, policy and practice implications and so forth. Thinking about the affects in these research machines reveals a further interesting aspect of their constitution. Unlike 'spontaneous' assemblages in daily life comprising bodies and body-parts, social institutions, places, clothing and other body adornments, values, scripts and norms), research machines comprise few relations and affects, making them readily amenable to reverse engineering in order to understand how they work. To offer a simple example: a sampling frame can be unpicked to reveal a machine that works by means of a single affect targeted at the various events available to social inquiry. This affect sorts events to be included in a study from those to be excluded . In other research machines, the affects may be more complex: a summary statistic such as chi-squared comprises a series of arithmetic operations that transform 'data' into a single indicator of statistical significance. This means that by unpicking research machines, it will be possible to assess how a change of data collection or analysis method, or of design alters the affective flow in the research-assemblage, and hence what kind of 'knowledge' it produces . However the materialist ontology of social inquiry we are pursuing here allows us to push this analysis of the research-assemblage and its constituent machines a step further. Identifying the affects in the researchassemblage opens to scrutiny the micropolitics inherent in the research machines that do data collection, data analysis and so on. These in turn allow us to explore the micropolitics of the research process, of what happens when events are transformed into 'data', and who gains and who loses in the process . To give an example, in a randomised trial, controlling the experimental conditions and use of statistical techniques together limit the affective capacities of 'confounding' relations found in 'real-world' settings, empowering a researcher to model the 'uncontaminated' affect of an 'independent' upon a 'dependent' variable. By contrast, in qualitative studies a 'naturalistic' approach limits the researcher affective capacities, while enhancing the affectivity of respondents' accounts. The micropolitics of these research-assemblages differ because of the affects that hold them together, and what they do to the relations in the assemblage. Later in the paper, we will reverse-engineer methodologies and methods in just this way, to understand how they work and what their micropolitics tell us about different research approaches. In order to achieve that objective, it will be helpful to set out a more general appraisal of how a research-assemblage works, based on the materialist analysis of relations, affects and micropolitical movements in research-assemblages discussed earlier. --- A model of research micropolitics To develop from within the materialist perspective a model of what happens when an event -for instance, the school trip described by Youdell and Armstrong -is researched sociologically, let us consider the event as an assemblage 'E' comprising a set of relations 'ABC' , linked by affects that make this event do whatever it does. The aim of a research study will be to apply designs and methods that can somehow identify the ABC relations within the E assemblage, explore the affects between these relations, and from this offer an explanation of what E does within its particular social context. However, as we have noted, a research study must also be considered both as an event in its own right, and as an assemblage R. R will have its own set of relations 'XYZ', which are all the paraphernalia of academic inquiry such as the researcher, methodologies, research instruments, theories and so on. As was discussed earlier, these XYZ relations have been purposively assembled in order to engineer specific affective flows within the research-assemblage, with the objective of taking the event-assemblage E or other similar events, and producing a textual or other output that will form the research 'knowledge' of E. Most critically for this model of research, if R is to document, analyse and eventually turn E into knowledge, this requires that the researchassemblage must also be capable of being affected by the affects between relations ABC in the event being studied. in a qualitative study that enables a researcher to make sense of an event being studied.) For the sake of precision, it should be considered that when E becomes the subject of the research assemblage R, the consequent interaction between affects in E and R produces a third, hybrid assemblage, which we will designate R/E, with its own affect economy that links relations A, B, C, X, Y and Z. This economy will be distinct from those in either E or R, though it is the affects in R that will actually produce the research outputs, 'knowledge' of the E assemblage, or the altered sensibilities in the researcher and the research's audience that a constructionist would describe as 'social constructions' of E. It may also produce effects on E itself, such as changes in behaviour due to a 'Hawthorne' effect from being observed, increased individual or collective reflexivity, or impacts on the event due to attention from outsiders. This way of understanding the possible interactions between the affect economies of E and R within a hybrid R/E assemblage enables a sophisticated understanding of the micropolitics of social research. This can be apprehended by considering two opposing 'hazards' often discussed in social research. The first of these occurs when the 'research' relations XYZ within R/E dominate the flow, asserting a powerful effect over the relations ABC of the event-assemblage E. This may happen for various reasons: for instance when a sampling strategy intentionally excludes key aspects of E; by controlling out naturalistic contexts; by imposing a theoretical framework on data; by use of statistics to summarise or generalise; or by textual representation of E. In such cases, R affects radically re-territorialise the affective flow between ABC relations, to the extent that the --- 5.3 'knowledge' produced by R/E no longer reflects the flow within E, distorting its representation in research outputs. This is the situation posited by radical social constructionists, who have argued that modernist research has constructed rather than described its objects and Kitzinger's studies of sexuality). The opposing hazard occurs when the XYZ relations in the research-assemblage have so little affective capacity that the ABC relations are dominant within the R/E assemblage. Now the research process becomes a machine whose outputs are trivial or anodyne rather than analytical; descriptive or journalistic rather than critical. This may occur when affects in the research-assemblage are weak, for instance if the research design lacks a powerful analytical machine or is theoretically uninformed; the research instruments do not possess the capacity to differentiate the relations or affects in the event; or the reporting is literal rather than critical. Occasionally, of course, this affective weakness is seen as an opportunity, for example in case studies that set out to describe specific events; or in 'Delphi' methodologies that aim to gain consensus among experts and thereby offer a definitive statement on current knowledge . However, between these extremes there will be many research situations when neither ABC nor XYZ affects overwhelmingly control the affective flow in the R/E assemblage. Yet in each and every situation, there will be a dynamic tension between the affective flow of E and that of R, a tension that has consequences for the knowledge and representations of the social world that research produces, and potentially for the social world itself. Applying this materialist analysis supplies a means to explore the micropolitics of the research process, understood in terms of affective interactions between event, research methodologies/methods and researcher. --- Dis-assembling the research-assemblage In the remainder of the paper, we will use this materialist understanding of the R/E assemblage to explore in greater detail a range of research machines, their affect economies and the micropolitics of social inquiry these machines produce. The utility and validity of many of these research-assemblages and machines are contested within contemporary social inquiry, but rather than debating the traditional critiques, we have applied our model of research-as-assemblage to make sense of what these designs and methods actually do, and how they work. First we asked what a machine is designed to do, and what outputs it produces. From this, we applied a more critical assessment, imposing the materialist ontology of assemblages and affects to reverse engineer the machine and disclose the affective flows that make it work. We examined how the design of the machine enhances certain affective flows and mitigates or closes down others. This is the affect economy of the machine, which enables it to work in a specific way, and which produces the micropolitics of its interactions with the affect economy of the event studied, with consequences for the knowledge produced, the researchers and the events themselves. To clarify this analysis, we will apply this approach of reverse engineering or disassembling research machines to three methods or techniques and then two research methodologies often used in social inquiry: the survey and the qualitative interview. This analysis will supply the basis for a more critical subsequent discussion of how social inquiry is viewed from a materialist perspective. Analysis of many of the other techniques, methods and methodologies used in social research are then presented in Tables 1 and2. --- Techniques and Methods --- Sampling A sampling machine performs a relatively simple task within a research-assemblage, selecting events for inclusion in or exclusion from a study, based upon two sequential affects. The first affect sets the choice of sampling approach for the researcher to implement; the second affect includes or excludes specific events, applying the appropriate means of selection , as set by the first affect. The affect economy in this machine acts upon study events, systematically including some in the sample and excluding others. Micropolitically, this economy empowers a researcher to achieve the impossible , by restricting the affects from the event population entering the research assemblage according to specific, though arbitrary principles. --- 5.10 The Questionnaire The questionnaire is actually two separate machines: one for questionnaire construction, the other for its administration. The affect in the first transforms variables to be measured into a question that will serve as an indicator , sometimes with preselected permissible responses. The affects in the second act on researcher and respondents, requiring a question to be asked, an answer to be supplied, this answer to be recorded and possibly allocated to a precoded category, and the instrument to be applied consecutively and independently to each respondent in turn, generating completed questionnaires ready to be fed into an analysis machine. The affective economy of this machine extracts information systematically from respondents to produce a dataset. Micropolitically, the machine acts as a filter on the affect economies of events being studied, extracting only certain data, and categorising them according to the affect economy of the instrument rather than that of the event itself. --- Thematic Qualitative Analysis Thematic analysis of qualitative data is a machine that organises and reduces non-numerical data, making it more manageable and amenable to systematic reporting. It entails two affects: a pre-analysis code-generation affect that allocates a code to a range of similar textual occurrences ; and an aggregative affect that takes each piece of data in turn and codes it according to this scheme. The affect economy here acts on raw data from a study to aggregate it, reduce variability, and open it to reporting. The micropolitics of the machine is to reduce complexity in data and aggregate together disparate aspects of an event in ways defined by the analyst . --- Research Methodologies We now consider two common research designs, looking at each as a whole but also as constituted from a number of machines, including those described in the previous thumb-nail sketches and those listed in Table 1. --- The Survey The survey is a social research design assemblage that typically produces a quantitative summary of specific aspects of a study population, as defined by a research question. This is done by a series of researchmachines that a) draw a representative or stratified sample from a population; b) use methods such as questionnaires to gather data on specified measures or indicators ; c) submit these to descriptive or inferential statistical analysis; and d) report a summary of the sample's features, often with an assessment of confidence to generalise these to a population. The affect economy that makes the survey design work derives from the economies of its constituent machines. Affects in the sampling machine supply a means to allocate events to the sample; those in the questionnaire machine select and categorise those features of events to be studied and record them in a form amenable to quantitative analysis; affects in the statistical analysis machine aggregate and manipulate the data mathematically, reducing it to summary values and statistical assessments of probability that efface the complexities and divergences in the events; the affects of the result-writing machine use these aggregated and de-contextualised findings and present them in an effort to answer the study's research questions. Micropolitically, all these machines are highly aggregative: as sampling and questionnaire machines systematise research selection and data collection, they restrict which affects from the event can become part of the R/E assemblage; the analysis machine aggregates the affective capacities of the event into numerical metrics, simplifying and thereby reducing the granularity of the event-affects represented in the research outputs; the writing machine imposes the constraints of a narrow research question on the affects in the event. In these ways, a survey powerfully privileges a researcher perspective over the events it studies. --- The Qualitative Interview Here we examine qualitative interviewing as a research methodology designed to provide 'rich descriptions' of a social event or events by interrogating accounts elicited from social actors. Within this researchassemblage are a number of machines that: a) set and refine the research question; b) select a sample -usually according to a purposive strategy to overcome the limitations of a relatively small sample size ; c) use individual or group interviews to gain in-depth data on interviewees' affective backgrouud, engagement with the issues being studied, and reflections on these issues; d) undertake some kind of qualitative analysis, typically one that categorises data [6] 5.11 5.12 5.13 into themes either 'grounded' in the data or deriving from a pre-defined theoretical or conceptual framework, or imposes some other structure upon the data; and e) report these data in a textual format that often includes extracts from the interviews conducted. The affect economy of this methodology is again provided by its constituent machines. The purposive sampling machine selects subjects to interview based on expectations of their affective economies, and often aims to maximise diversity rather than to ensure data is representative of a wider population. An interview schedule is a simple affect that determines which elements of the subjects' affective engagements with the topic can be reported, while the question/answer format underpinning the interview method also governs the material gathered. The qualitative analysis machine organises, aggregates and reduces the textual material, enabling themes to be developed or explanatory constructs to be developed; writing produces a second-order account of the events being studied, as interpreted first by interviewees and then by the researcher, and illustrates the account with 'representative' or 'allusive' quotations from the interviewees. The micropolitics of qualitative interviewing reflect the interactions between researcher, interviewees and the events they describe. The researcher's questioning role and the answering role of subjects produce inequality, while the choice of research question, sampling and interview schedule machines all impose a framework on the affects admitted into the R/E assemblage. However, the interview format does enable respondents to control the accounts they offer, and researchers may actively try to hand some power back to interviewees, to 'give them a voice'. As noted earlier, the systematising and aggregating affects in the thematic analysis machine privilege the analyst's account over those of respondents; this is reflected in how the data are reported -typically within an imposed structure that establishes the researcher's unitary account of the event, with interviewees' accounts used selectively to justify the researcher's answer to the research question. These detailed descriptions of methods and designs demonstrate how we have moved from what a machine does, to how its affects make it work, and finally what this means for the micropolitics of the hybrid R/E assemblage. Tables 1 and2 summarise other methods, techniques and designs that may be reverse engineered in the same way. The first table is divided into research techniques, data collection and data analysis methods, and aspects of data presentation. Table 2 analyses the principal research designs from across the spectrum of social inquiry. Of course, social research is continually innovating new designs and methods, often highly antagonistic to positivistic science, and these too are amenable to analysis, to reveal their unique affect economies and micropolitics. --- Discussion: re-assembling social inquiry Assessing this range of research methods, techniques and designs from a materialist perspective suggests three features of research-assemblages and their micropolitics that bear on the concerns of this paper with social research theory and practice. First, most research-assemblages possess just the few relations and affects needed sequentially to perform simple tasks of interconnected machines. By breaking researchassemblages down into their component parts, a materialist analysis can exploit this sparseness, to reverse engineer research machines to disclose their affects and what they do, both de-mystifying the research process, and allowing research designs and methods to be manipulated to serve particular ends. Second, generally speaking, the affective flows in most of the research assemblages and machines reviewed have in common that they produce simplicity where there was complexity, definition in place of indeterminacy, and evenness where there was variability, with consequences for the knowledge social inquiry produces. Earlier we noted that affects come in two forms: an affect that produces a change in one relations in an assemblage are 'singular' in their effects, while 'aggregative' affects clump relations together in some way. Scrutinising Tables 1 and2, it is striking how many research methods, techniques and designs have affective flows that aggregate event affects in one way or another. From selecting which events to study, to which features of events to subject to attention, through analytic methods that categorise or reduce data and frame it theoretically or conceptually, to reporting findings and assessing their significance, affective flows in the research-assemblage act upon the affects of the events being studied to produce research outputs that are aggregations, generalisations or summaries. From this perspective, survey and interview methodologies reveal less divergence than might have been predicted from the rival claims of their advocates. Aggregation influences how events are presented in research reports, tending toward a loss of detail or granularity in what is been described. Whether it is through a descriptive statistic such as a mean or a confidence interval, or a thematic presentation of qualitative data, aggregations have the effect of excluding the outliers and aberrations that in social life may be extremely significant; the social world thus presented seems more bland and less exceptional than it might to those immersed within it. On occasions, the aggregative character of research is subverted, enabling singular flows within event-assemblages. Examples include case study designs 6.4 6.5 6.6 6.7 --- 6.8 that offer detailed descriptions of events and their contexts; analyses that seek out aberrant events or utterances rather than the 'representative' illustrations often favoured by researchers when reporting data; and studies that set out to change the world rather than simply understanding it, including emancipatory or transgressive action research that involve participants in a research setting in undertaking research themselves in order to improve some aspect of their world , feminist/queer research that sets out to challenge and disrupt patriarchal and heteronormative orthodoxies , and 'schizoanalytic' and 'diffractive' research designs inspired by DeleuzoGuattarian and Baradian theory that aim to trouble received wisdom and produce new possibilities within the settings researched. Third, these kinds of aggregating affect economies within research machines and assemblages territorialise affects in the E assemblage, shifting the micropolitics of the research-assemblage and the hybridised R/E assemblage firmly toward the researcher's agenda. This is unsurprising; after all research machines have been designed to enable researchers to do research, though this assessment rejects any argument that research is in some way a 'neutral' event. A materialist analysis of the interactions between event, research process and researcher supplies a nuanced view of the micropolitics of research-assemblages: by analysing the affects in the hybrid R/E assemblage, it is possible to assess in what ways, and to what extent, event affects have been territorialised by the affect economy of the research process, and which machines and designs have done this. One outcome of such a micropolitical analysis of research is a refinement of a strict constructionist perspective on research, which has tended to regard knowledge of the world produced by social inquiry as highly contingent upon the conditions of its production by researchers , or constitutive of the objects it describes. So, for instance, in the progressive territorialisations by psychological sciences of nonnormative sexualities , Rose argued that 'evidence, results, arguments, laboratories, status and much else' contributed to the construction of differing 'truths' about LGBT sexualities as pathology, then deviance, then diversity. Up to a point, a materialist analysis would assent to this assessment: as has been seen, an aggregating and territorialising research-assemblage will dominate the affective flow in a hybrid R/E assemblage, reshaping the ABC affects of the event in the research's image. However, the materialist analysis differs in two key ways. Firstly, as was noted earlier, the ABC affects of an event cannot be entirely erased from an R/E assemblage: except in intentionally fraudulent or fictitious research reports, or where the machines in a research-assemblage are wildly inappropriate , something of the event's affective flow will inevitably remain in the research outputs. Pulling apart a research-assemblage to interrogate its machines and their affects can certainly disclose these flawed processes, but more generally, it can specify and evaluate precisely what aggregations and territorialisations a research-assemblage has wrought upon its subject-matter, and hence the extent to which research outputs provide useful knowledge of events. Secondly, if disseminated into the social arena, these outputs may indeed feed into the event-assemblage upon which they are based, potentially altering the latter's affective flows and what it can do -for instance, by producing guilt in people involved in a 'taboo' sexual practice. However, from an assemblage perspective, the aggregated and territorialised output of a research-assemblage is only one among the many relations that contribute to an eventassemblage, and only if the research-assemblage's affective capacity is exceptionally powerful might an event's affect economy be fully overwhelmed. Furthermore, analysis of the affect economies within a research-assemblage can enable precise assessment of the extent to which it has re-constituted an event in its own image. Though powerful cultural forces always have the potential to influence the affect economy of a research-assemblage, social inquiry is not doomed to inevitably and uncritically recapitulate these social and cultural forces. If this nuanced understanding of the relational micropolitics within event/research hybrid assemblages offers a different assessment of research knowledge from constructionist suggestions that social inquiry inevitably produces contingent, ideological knowledge of the events it investigates, there is also clear water between a materialist analysis of the social world and a realist perspective. We have shown how affects in the research assemblage inevitably must interact with those in the eventassemblage if research is to be other than merely descriptive, denying the possibility of unmediated translation of an observed event into research knowledge of it for a comprehensive critique of correspondence theories of truth). At the same time, the materialist understanding of events as relational -6.9 6.10 6.11 6.12 6.13 comprising affects that interact ceaselessly to produce new capacities -suggests a social ontology of flux and becoming rather than of being and stability. While the opportunity afforded by materialist ontology to peer inside the machines and assemblages of social inquiry has permitted us to assess the research process micropolitically by unpacking its affect economy, perhaps of greater significance is the potential also to manipulate these micropolitics. We can, if we so wish, design and re-engineer research-assemblages and machines to include or exclude specified aggregative and territorialising effects, thereby innovating creative research-assemblages that produce specific capacities in researchers, data and the events studied. Though this could be simply a charter for removing as many aggregating affects as possible from social inquiry , we would suggest there is potential for a more sophisticated response. Alongside such efforts to develop less aggregating and territorialising research-assemblages, the capacity afforded by a materialist perspective upon research to critically assess the affects that make research machines work suggests a potential rehabilitation for a range of methodologies sometimes seen as highly territorialising: surveys, experiments and quasi-experiments. At the same time, proponents of such designs must temper their enthusiasm with an explicit commitment to review reflexively and account for the aggregations and territorialisations they produce in R/E hybrid assemblages and hence in their findings . These can be acknowledged and their impacts predicted, probabilistically if not absolutely. This may also supply a new justification for applying a mix of methods within studies, some highly aggregative but analytically powerful, others less analytical but intentionally non-or even dis-aggregative. For instance, a study might combine a descriptive case study that produces a rich picture of the concerns and values of research participants in a setting with an intervention that attempts to alter aspects of the setting to address these concerns and values. A subsequent evaluation might combine aggregative quantitative measures with opportunities for participants to offer their own unmediated assessments of any improvements, and use the research outputs to challenge policy or improve their living environment. Mixing methods and methodologies in this way does not mean that the aggregations of particular methods are somehow 'cancelled out'. But because researchers can estimate precisely what aggregations their methods entail, the consequences for knowledge-production can be accurately predicted and acknowledged when reporting findings and drawing conclusions. In this paper we have taken the precepts of new materialism and run with them, to explore what they mean for social inquiry. What then of the suggestion that new materialism resolves the contradictory perspectives on knowledge production between realism and constructionism ? It would be a neat coda to the paper to argue that our analysis has revealed a new resolution of this dualism. But we would suggest that this is not an appropriate conclusion; instead, what a new materialist perspective on social inquiry has done is to change the terms of reference for the discussion. Rather than trying to answer an anthropocentric question of belief or scepticism about the possibility of gaining knowledge of the world independent of human culture, what our analysis has done is to look at what research actually does. We have used the 'turn to matter' and a DeleuzoGuattarian materialism of assemblages and affects to disclose the micropolitical workings of the research-assemblage as it hybridises with an event-assemblage. We have thereby offered a means to assess the effects of every element of the research process on both events and on the knowledge that research produces. Materialist analysis of research-assemblage micropolitics, we would simply conclude, supplies an innovative take on the ontology of social research that opens up possibilities for how research is designed and undertaken, what capacities the research assemblage produces in researchers, research tools, audiences and events, and how a range of methodologies and methods may be critically applied and combined in the pursuit of useful understanding of the world and human culture. --- Notes Ontology concerns propositions about the fundamental nature of things and the kinds of things that exist, while epistemology addresses how these things can be known by an observer. For some readers, the recognition that agency extends beyond humans will be reminiscent of the emphasis in Actor-Network Theory upon non-human 'actants' that contribute to social production alongside human agents. This convergence between ANT and Deleuze and Guattari's materialism is also evident in Latour's understanding of the social as relational and assembled, and Law's discussion of a heterogeneous network of forces deriving from both the natural and social worlds. Our terms 'singular' and 'aggregative' replace the somewhat obscure DeleuzoGuattarian terms 'molecular' and 'molar' that derive from physical chemistry. The machinic analysis offers a materialist understanding both of research's design and its failings. Research machines do not work perfectly -they break down, misfire, go into reverse or self-destruct. This can occur because of legitimate but faulty/inappropriate affects in the machine , or due to extraneous sources such as reticent or disruptive research-participants, unpredictable factors in the physical or cultural environment of a study, or unprofessional or fraudulent behaviour by researchers. For this exercise, we exclude the subjective biases, prejudices or commitments of researchers and other 'extraneous' affects from our analysis of the research-assemblage. In practice, these will also contribute to the R/E assemblage and its micropolitics. A second level of coding may be undertaken to further reduce the variability in the data, but this again comprises just two affects: one that develops the aggregation/coding scheme, and a second that aggregates the data into an even more limited number of categories or themes. Sometimes code-generation will overlap aggregation, with codes generated 'on the hoof' as new categories emerge from the data. Our analysis of the interactions between event-and research-assemblages is cognate with Barad's ontology of events and observers, but without recourse to quantum mechanical theory.

This paper explores social inquiry in terms of the 'research-assemblages' that produce knowledge from events. We use the precepts of new materialism (and specifically DeleuzoGuattarian assemblage ontology) to develop understanding of what happens when social events are researched. From this perspective, research is not at root an enterprise undertaken by human actors, but a machine-like assemblage of things, people, ideas, social collectivities and institutions. During social inquiry, the affect economies of an eventassemblage and a research-assemblage hybridise, generating a third assemblage with its own affective flow. This model of the research-assemblage reveals a micropolitics of social research that suggests a means to interrogate and effectively reverse-engineer different social research methodologies and methods, to analyse what they do, how they work and their micropolitical effects. It also suggests a means to forward-engineer research methods and designs to manipulate the kinds of knowledge produced when events are researched.

Introduction This paper illustrates the ADA Project, an action-research dealing with design adaptations in homes of severely disabled people. The focus of the project on accessibility in interior spaces and on people with extremely specific needs defines a challenging field of action for the application of Universal Design principles [1]. We believe that the lesson learnt in the implementation of the project can contribute to the debate on Universal Design and its limits, and to its effort to serve collective goals [2]. --- Home adaptation Home is the primary space that contains human bodies, where primary needs and many desires are expressed. Its environment both shapes us and is shaped by our attempts to adapt it to our changing needs and requirements [3]. Disabled people's functional limitations, however, compromise this bidirectional process and their capacity to respond to the 'environmental pressure' of their domestic spaces [4]. In addition, the more severe their limitations are, the more they are confined to home, this being their main -if not the exclusive -spatial domain of their existence. This is what makes the home a crucial area of intervention for disability policies. Home adaptation is a practice to address this unbalanced condition by adapting home's physical environment to improve accessibility [5]. Several studies have investigated home adaptation from different perspectives [6]. Some studies have focused on home adaptation outcomes, such as increasing disabled people's autonomy in their daily activities and well-being, facilitating caregivers' work, reducing injuries and the need for institutionalization, therefore limiting social and healthcare costs [7]. But also on other aspects, such as: the decision-making processes that lead to the implementation of adaptations, the relationship between clients and professionals who design, build or install them; or even on how adaptations can transform the meanings disabled people see in their homes, or on how they can affect family relationships [8]. The assessment aimed at defining home adaptations to be implemented in a home is complex, and several protocols have been developed for this purpose. Two of the most acknowledged examples are the Swedish ENABLER and the American I-HOPE [6]. Some protocols assess domestic spaces by checking to what extent they comply with local or national accessibility regulations. Nonetheless, as these regulations and the standards they are based on often fail to consider the real and diverse needs of disabled people [9], such protocols tend to result in a narrow focus on standardized accessibility and an increase in unnecessary adaptations, compromising their potential. As a result, other needs highly valued by disabled people and that pertain to meaningful domains of human experience are overlooked. Another limit of some home adaptation approaches is to neglect the role of the social, cultural, and economic dimensions that also structure the person-environment relationship. In contexts where home adaptation policies have been organically implemented -Northern Europe, the United States and Australia -the negative effect of these limits is often mitigated by a favorable context which increases the availability of compensating options, both public and private. In other contexts, such as Italy, these positive systemic factors are often lacking, and disability policies are fragmented and poorly funded. Moreover, the market of accessibility products and services is still inadequately structured and disabled people have restricted access to supplementary resources. These limits can profoundly diminish the positive impact of home adaptation projects. --- The ADA Project ADA is an action-research project on home adaptations funded by the Tuscany Regional Government. This project is aimed at people with severe disabilities, with the goal of increasing their autonomy and well-being, and that of their caregivers [5]. Despite having a regional scope, the project falls within the intricate Italian national context of disability policies, affected by "multiple definitions of disability across sectors and regions, leading to disparity in access to support and services. [...] Disability continues to be defined through a medical perspective, and the revised concept of disability, as proposed by the National Observatory on the Status of Persons with Disabilities, is not aligned to the Convention [on the Rights of Persons with Disabilities] and lacks binding legislation at both the national and regional levels" [10]. This also affects the reliability of statistical data on disability and the quality and funding of accessibility policies, limiting the capacity of such policies to effectively address disabled people's accessibility issues, especially in their domestic environment. In this context, ADA was one of the first projects in Italy to propose home adaptation in a unified regional framework, introducing a series of methodological and operational innovations, some of which are discussed in this article. ADA pursues its goals through two actions: 1) providing each participant with a set of Accessibility Recommendations , which are suggested home adaptations to overcome the accessibility issues that hinder the disabled person from carrying out the activities they perform or would like to perform at home; 2) providing access to funding to carry out the home adaptations that the participant decides to implement. Adaptations can include modifications to spaces, furnishings and equipment and the adoption of assistive and home automation technologies. For this purpose, an interdisciplinary workgroup assesses the conditions of participants, carrying out an approximately 90-minute site survey at the home of each disabled person. The assessment is made using the Ada Assessment Model , a multidimensional tool developed by the research team of the University of Florence that coordinated the ADA Project. The AdAM follows two data gathering strategies. First, during the site survey, the AdAM is used in a relational setting, with an in-depth interview that involves the disabled person, their caregiver, or both, depending on the functional profile of the disabled person. The goal is to outline an in-depth profile of the disabled person which, beyond covering some essential socio-economic aspects, focuses on three areas: 1) the functional autonomy of the disabled person; 2) their social network; 3) their relationship with the domestic physical environment. This last aspect is investigated by visiting and discussing the domestic spaces with the disabled person, or their main caregivers. Second, after the site survey, the AdAM is used to summarise this qualitative analysis in a series of quantitative assessments: these, in turn, are used to compare cases with different conditions and to structure the distribution of public funds. --- Personalization Acting in a domain defined by the homes of people with very specific profiles, the main strategy of ADA to achieve the Universal Design goals is to maximise the personalisation of its interventions [2]. The tactics implemented to achieve this strategy -as detailed in the two following paragraphs -aim to: a) reliably define participants' profiles and the relationship with their domestic environment; b) suggest Recommendations consistent with these profiles; c) promote equality in access to public funding. --- Interdisciplinarity A first tactic towards personalisation is interdisciplinarity, an approach that shaped the project from the outset. The university research team that developed the methodological and operative framework of the project was made up of architects with expertise in accessibility and sociologists with expertise in the human-environment relationship. The team consulted with physicians, physiatrists, physiotherapists, social workers, experts in assistive and home automation technologies, administrative personnel, political sponsors at the Regional Government, and representatives of disabled people's associations. Interdisciplinarity is also a key element of the workgroups involved in the site surveys. Each ADA workgroup, in fact, is made up of three figures: a physician, a social worker, and an architect with expertise in accessibility. The workgroup could include other specialists depending on the specificity of the assessed case. The physicians and social workers of the workgroups were already members of the Multidimensional Assessment Units , the formal bodies that in all Tuscany health districts assess the conditions of disabled people. With ADA the multidimensional units were expanded, including a design member with specific expertise in accessibility and the human-environment relationship. This enriched the analytical, evaluative and design potential of the multidimensional assessment. The interdisciplinarity of ADA is even more significant in the Italian context, which stands out for the lack of established occupational therapy. Occupational therapists have extensively contributed to the theoretical development of home adaptations. In their practice they connect various disciplines and promote an activity-focused approach similar to that advocated by ADA . In Italy, however, there are less than 4 occupational therapists for every 100,000 citizens, whereas there are 145 in Denmark, 80 in Norway, 70 in Germany and 56 in England [10]. --- Relational Setting and Centrality of Activities ADA suggests home adaptations through an assessment made in a relational setting that directly involves the disabled person and, when needed, their caregivers. What is assessed is how domestic activities are hindered by the spatial environment where the disabled person lives. The assessment also considers the personal, functional, and social conditions of the disabled person, their perspective and needs, and those of their caregivers, also addressing how this multidimensional profile can change in the future. The person-environment relationship, therefore, is not primarily viewed from a standardised perspective but, rather, considering the domestic activities that the person actually performs or would like to perform if the spaces allowed it. This assessment posits that activities do not have the same value for all disabled people. On the one hand, only some activities are essential for all disabled people . Other activities, instead, are assessed only if they are both compatible with the functional profile of the disabled person, and significant to them. In other words, the ADA framework considers that the assessment cannot be limited to the intrinsic value of activities, which can be extrapolated from the comparison with standards; rather it also investigates their extrinsic value, which instead accounts for the actual needs and desires of each person in their specific context. A relational setting directly involving participants is crucial for such a granular assessment of disabled people and their homes. Equally necessary are interdisciplinarity and competence among the workgroup's members, who only have a 90-minute site survey to gather all the information needed for the assessment. These factors form the basis for highly personalised Recommendations, tailored on the specific needs and expectations of the disabled person in her specific context. On the contrary, a general check of regulatory compliance with predefined standards can hardly allow such a high level of personalisation. --- The impact of ADA Two experimental phases of ADA have been completed: the pilot phase, covering only two health districts of Tuscany ; and a second phase, in all its 34 health districts . Both phases followed the same methodological and operational principles, albeit with some alterations to address the leap in scale. In total, all 362 participants of the two experimental phases received the Recommendations, for free. Considering that the cost of such a service on the private market has been estimated in a range that spans from 1,000 to 2,000 euros, the overall economic value of all the Recommendations ranges between 362,000 to 724,000 euros. However, a comparable service in terms of quality and interdisciplinary validation is hardly available on the private market. Around 75% of all participants received public funding to implement some of the suggested adaptations, for a total funding of 900,000 euros. Public funding totally covered the costs of adaptations for participants with low income, while for the others the funding partially covered the adaptations' costs. --- The impact on people with disabilities A qualitative ex-post evaluation of the effects of the pilot phase of the ADA Project was conducted by a sociologist with expertise in the human-environment relationship, one year after the introduction of the adaptations [10]. The in-depth interviews of 36 participant that received the Recommendations and the visits of their adapted homes illustrated the wide range of effects of the Project. First, adaptations produced new opportunities of autonomous appropriation of domestic spaces by the disabled persons . That happened both allowing the use of spaces that were not accessible before, or that were only usable with the support of others. The home devices that were introduced also improved both social relations and safety at home. Adaptations improved the accessibility of the home and autonomy of the disabled persons by allowing not just essential activities but also other activities that participants described as particularly relevant for their personal fulfillment. These adaptations also allowed a better use of caregivers' psychophysical, time and economic resources. The analysis of the symbolic changes produced by the adaptations revealed that the suggestion in the Recommendations of 'universal' solutions that could not be directly attributed to the world of devices for disabilities was strongly appreciated by participants, as it reduced the perception of stigma associated with disability. This increased the implementation rate of the suggested solutions, given that participants were free to choose among the adaptations proposed in the Recommendation. This perception promoted the disabled persons' appropriation and territorialization of the adapted spaces, but also -when their cognitive conditions allowed that -the recognition, strengthening and expression of their identity as persons, and not just as disabled persons. The participation to the project, in fact, triggered in participants a new capacity to observe and conceive their domestic spaces, pushing them to fine tune the adaptations after the project or find other minor issues, sometimes coming up with new solutions to overcome them. This revived agency promoted in participants the perception of themselves as active and intentional producers of their domestic space, a condition usually diminished by the fragilities directly or indirectly connected to disability. This is highly significant for people who must spend most, if not all, of their daily life confined in such space. In the interviews the ADA project was often described as a turning point in participants' relationship with public institutions dealing with disability. Participants appreciated the on-site visits at their homes and the workgroup listening attitude, both interpreted as signs of attention and care by the institutions to the participants specific conditions. This is particularly relevant considering the Italian context of general mistrusts towards public institutions and disability policies. As a result, some participants even offered availability to share their positive experience to future participants, especially to help them in the implementation stage. The implementation of the adaptations, in fact, is often a difficult step to undertake, especially for people affected by the typical social and cultural fragilities that are associated with disability. And, unfortunately, ADA did not have enough resources to follow participants in such stage. --- The impact on the culture of accessibility The social, cultural, and economic impact of the Project goes beyond the effects of the adaptations on the disabled persons and their proximate social network . ADA, in fact, promoted the diffusion of the multidimensional approach to accessibility in the wide community of actors directly or indirectly involved in the project. Opening the interdisciplinary assessment to architects with expertise in accessibility allowed a cultural and professional growth in all the workgroups members. This affected around 100 professionals, considering both the group of physicians, social workers and architects directly involved in the workgroups, but also the neuropsychologists, rehabilitators and experts in assistive and automation technologies that were consulted to address the most complex cases. This improved the spectrum and the accuracy of the assessment of the functional, social, and environmental profile of the participants. But it also affected the process of definition of the personalized adaptations suggested in the Recommendations. On the one hand, the impact of the adaptations was higher than expected by the workgroups social and health members, improving their awareness about the effects of the physical environment on the wellbeing of disabled persons. On the other hand, the architects had the chance to improve their understanding of the functional, health, or social conditions of participants, and to consider the effects of their design solutions on such conditions, often in unexpected ways. Discussing design solution in the interdisciplinary workgroups, in fact, allowed to avoid some unintended consequences of the adaptations. For example, suggesting motion-activated wall switches in the home of a person with a limited capacity to control their fingers could end up reducing the chances to train that residual capacity, thus accelerating its further reduction. Such unintended consequences are more common than expected and avoiding them requires a case-by-case multidimensional assessment of the supposed efficacy of the solutions for the specific person they are designed for. It was also estimated that the participants contacted about 1,000 professionals to implement the adaptations. This group includes a wide spectrum of professional figures, such as designers, artisans, builders, installers, and retailers of building material, furniture, and devices. This group also was exposed to the contents, the interdisciplinary approach, and the language of the Recommendations. For the relatively unstructured sector of services for accessibility, the project meant a chance for both a cultural and an economic growth, given that this sector was the final receiver of the money spent to implement the adaptations. Considering both the public and the private funds spent by the participants who had to co-finance the adaptation, the total funds that this sector received are estimated around 2,000,000 euros. The positive outcomes of the two experimental phases of the Project led to the recent introduction of ADA as a structural action of the Tuscany Regional Government's disability policies. This decision built on the explicit recognition of the innovative approach of the ADA Project and, specifically, of its effort to implement the principles of the UN Convention on the Rights of Persons with Disabilities and of the International Classification of Functioning , promoting a framework of evaluation of disability in terms of integration between the people and the physical environment they live in. Therefore, accessibility experts are now taking part to the works of the Multidimensional Assessments Units, and the AdAM is the official assessing protocol. In 2017 the ADA Project was awarded with the title of "Good practice" by the International Design for All Foundation. Then, in 2018, the project won the International Design for All Foundation Award. --- Conclusions: ADA and the Universal Design approach The ADA Project is an attempt to introduce an innovative strategy to improve home adaptations protocols. Albeit some limits of the Project related to the implementation phase, we believe that this strategy has a positive impact on the effectiveness of adaptations; specifically, on how they improve the relationship that severe disabled people have with their domestic spaces, increasing their autonomy and wellbeing, and that of their caregivers. The project strategy and tactics also aim at overcoming some of the limits of Universal Design and of other design approaches promoting inclusion and social sustainability, as illustrated by Arenghi et al. [2]. When such approaches are applied both to private domestic spaces and to address very specific needs, their success lays on their capacity to maximize personalization of their actions. As seen, in the ADA Project such personalization strategy implies an articulated multidimensional assessment method, that is applied prior to any design choice. The assessment does not just focus on the relationship between the disabled person, with their functional limitations, and the space they live in. This relationship, in fact, is analyzed considering an articulated network of other factors, such as social, cultural, and economic ones. Therefore, the design solutions that are proposed consider a wide variety of conditions of their final users, not just addressing their basic needs, but also their subjective desire and aspirations. Defining and applying such strategy required a great interdisciplinary effort, a principle often much more stated than adopted. Such effort was based on the competence of the workgroup members involved both in the assessment and in the definition of the Recommendations, a result of the training stages activated during the project. These stages prepared the workgroup members to take a much greater responsibility than the one needed to check the compliance of physical spaces with regulations or standards, an approach that, although useful, in the end is often only apparently objective. We believe that the availability of the workgroup members to get involved and not hide behind the shield of a formal assessment is another example of the cultural outcomes of the project. These results complement those produced by the relational setting adopted in the assessment of the functional, social, and environmental profiles of the disabled persons. This approach promoted the direct involvement of the disabled persons in the assessment. As seen, this produced effects that go beyond the primary expected outcomes of the adaptations, namely improving physical accessibility: it fostered empowering processes, with a re-activation of participants' agency, and an improvement of their trust in the public institutions. Such effects are strongly connected to the realization of social sustainability, one of the main goals of Universal Design. In conclusion, we believe that the ADA Project is an example to clarify how, when acting in the private and domestic sphere, the 'universal' attribute stated in the Universal Design label should not be intended as a quality of the field of application of the specific proposed design solutions, as it happens when operating in the public or collective space. Rather, in domestic spaces the tension towards universality is played in the effort to create the conditions to design solutions that answer to extremely specific conditions, often made of the unique mix of personal, functional, social, cultural, economic, and spatial conditions each person lives in. Adopting this posture requires an articulated strategy aiming at personalization. But, in turn, this only allows effective solutions if it is informed by inclusive and universal principles and goals.

Home adaptation is a practice that addresses accessibility issues in the domestic environment of disabled people, introducing modifications to the spatial environment or devices which improve their autonomy and wellbeing, and that of their caregivers. Protocols developed to define the right adaptations for each home mainly rely on checklists to verify the normative compliance of the physical environment to predefined accessibility standards. However, these protocols fail to address the complexity of the social, cultural, and economic dimensions that structure the person-environment relationship, thus compromising the efficacy of the adaptations. The excessive rigidity of such approach relates to the current debate on the limits of Universal Design when applied to the domestic environment, and especially when directed to people with specific needs. As an example of a more productive approach, this essay illustrates ADA, a public funded action-research project that proposes home adaptations for severely disabled people. The paper discusses the innovative strategy of ADA, based on a high level of personalization, and its main tactics: interdisciplinarity, relational setting, and centrality of activities, both in the assessment of the users' profiles and spaces, and in the design of the adaptations. This essay also evaluates the impact of the project, showing how in specific domains, such as those of ADA, personalization is the key to achieve the inclusive and sustainable goals of Universal Design.

INTRODUCTION Adolescence is a critical and challenging time when children become independent individuals, form new relationships, develop social skills, and learn behaviors that endure throughout life. 1 42% of the world population are young , of which 1.2 billion are adolescents . 2 Nearly one-third of India's population are young. 3 In this transition from childhood to adulthood, adolescents are exposed to injurious entities like tobacco, alcohol, violence, accidents, mental health issues as well as sexual health issues such as Sexually Transmitted Infections , sexual exploitation, teenage pregnancy etc. 2 3000 adolescents die every day mainly due to preventable causes like accidents, complications of pregnancy, HIV/AIDS etc. 2 Significant proportions of young people experience risky sexual activity, do not receive appropriate care, and experience adverse reproductive health outcomes. Over 35 % of all reported HIV infections in India occur among young people. 4 Every year, approximately one out of every ten adolescents, even in developed countries, acquire STIs each year. More than one million teenagers become pregnant each year, out of which 60% are unwanted pregnancies. 5 Globally, about 1 in 10 girls under 20 years of age have been subjected to forced sexual acts at some point in their lives. 90 % of adolescent girls who reported exploitation say that their first abuser was someone known to them. Globally one in three adolescent girls aged 15-19 years had been the victims of emotional, physical, or sexual violence at some point. 6 In most instances, children do not report these to their parents due to inadequate communication skills or lack of healthy relationships. Today, there are more than one billion 10-19-year-olds, 70% of whom live in developing countries. They grow up in completely different circumstances than their parents, with better access to formal education, a growing need for technological skills such as internet skills, exposure to new ideas through media, telecommunications, etc. The Covid19 pandemic led to schools closing and use of online education further paved way for adolescents to become more freely familiar with mobile phones, internet, laptops, social media etc. Rates of sexual initiation during young adulthood are rising in many developing countries. As per the National Family Health Survey 5 report of India, women aged 20-24 years married before age 18 years was 23.3 % and those aged 15-19 years who were already mothers or pregnant at the time of the survey was 6.8 %. The adolescent fertility rate for women aged 15-19 years was 43. 7 According to the National Family Health Survey 5 report of Kerala, 2019-20, the total women aged 20-24 years married before age 18 years is 4.1 % in urban and 8.2% in rural areas. Women aged 15-19 years who were already mothers or pregnant at the time of the survey was 1.8 % in urban and 3.0% in rural areas. 8 The World Health Organizations' mission on adolescent SRH is to contribute to a world in which its importance is understood, accepted, and supported. 9 It is a common practice in many schools to arrange sessions on menstrual hygiene, reproductive health etc., but such sessions remain open only to girl students. Kerala, though one of the most literate states in India, was not open towards the idea of sex education to students, apprehensive that it will promote immorality among the youth. The state is yet to implement sex education in syllabus. Parents who believe in the age-old fact that 'their children will gradually learn when the time comes', worsen this. The discomfort many parents feel about talking to their children about sexuality further impedes their ability to provide guidance. 9,10 Many adolescents today have patchy knowledge on SRH issues which most often comes from information shared by their same sex peers who may or may not be well informed. This can lead to misinformation, myths, making them vulnerable to unprotected sex, unwanted pregnancy, STDs, unsafe abortions etc. In the state of Kerala, dominating in both healthcare and literacy, it would be interesting to identify the awareness on SRH issues. This study aimed to assess the awareness of adolescents on SRH issues that may help policy makers, program planners, and implementers to design appropriate interventions to address the same. --- Aim: To study Sexual and Reproductive Health awareness among high school students in Kerala. --- Objective: 1. To list the sources of information on Sexual and Reproductive Health issues among adolescent high school students in Thrissur district, Kerala. 2. To study the awareness on Sexual and Reproductive Health issues among adolescent high school students in Thrissur district, Kerala. Sampling method: Multistage random sampling was used for this study. Due to the COVID 19 pandemic and successive lockdowns, it was possible only to collect data from 2 schools. A sample of 338 was achieved. The sample of adolescents thus obtained was stratified based on gender into male sand females; and based on grade of studying into 8 th , 9 th, and 10 th . --- MATERIALS & METHOD Data collection: A Self-administered questionnaire in English language was used. Due to the COVID 19 lockdown and subsequent closure of schools, few responses were recorded through Online-Google forms sent via email to the students. --- Study tool: A pre-tested and validated structured questionnaire prepared based on the WHO Questionnaire for the young was used to collect data. 12 Few questions were modified to account for the cultural and regional differences expected in this study. The questionnaire consisted of 2 parts-A, B. A. Socio-demographic details of the student B. Sexual and reproductive health issues: The sexual and reproductive health issues were divided into 10 domains- private parts types of touch pubertal changes menstruation masturbation sexual contact pregnancy contraception Sexually transmitted diseases sexual abuse. Each of the domain had open-ended question on whether they have heard about the domain, what they understood by the domain, and their source of information regarding the same. The reliability of the questionnaire was assessed with a Cronbach's alpha value of 0.8-Good reliability. A pilot study was conducted to validate the questionnaire, prior to the study period. --- Data analysis: The data obtained was coded and entered to MS Excel worksheet and analyzed using SPSS Version-23. Results of baseline characteristics of the study subjects were explained in terms of frequency, percentage, mean and standard deviation. Ethical consideration: Institutional Research Committee and Ethical Committee clearance were obtained prior to starting the study. The following tables 2 to 14 shows the awareness of adolescents regarding questions asked on sexual and reproductive health issues. Only the responses of those adolescents who answered the open ended question were considered. --- RESULTS --- Sociodemographic --- DISCUSSION In the present study, the major sources of information for adolescents regarding SRH matters like private parts, type of touch, response to sexual assault and menstruation was their mother. When it comes to puberty changes, both mother and school were the major sources. Media was the most common source of information regarding Masturbation, sexual contact, pregnancy, contraception, STDs, and sexual abuse. Aunts, uncles, and grandparents were the least family members mentioned as a source of information on SRH. In a study in West Ethiopia, large proportion quoted people other than family as a source of information about SRH, particularly their friends . 13 Among mentioned source of information for SRH issues, television, and school accounted 48.3% and 42.6% respectively in Kasiye et al's study; the preferred source of information on SRH was school for 76%. 14 But in a study by Yesus et al, the most constantly mentioned source of information for SRH were school followed by friends . 15 In Visani et al study, the main sources of SRH information for adolescents were health facilities, followed by internet as it was easy access. 16 A cross sectional survey done in Hamadan showed that adolescents of both genders seek information about SRH as well as relationship advise from their friends further than from any other sources. 17 A quarter of the study sample of Rada et al reported mass media as a source of information. 18 It's intriguing to note that sources such as schools, doctors, and health staff, were very inadequately represented. In the study of Mohammadi et al, adolescent boys considered their friends to be the primary means of information. They preferred to talk to their friends about sexual issues and relationships. 19 The results of a study by Mazloomi et al on-university scholars also showed that their main resources for sexual health were friends followed by special books. 20 Survey results of Soltani et al reported that after friends, mothers were the source for girls, while movies and internet for the boys. 17 Similar results were found in Iran. 21 In 2014, Mousavi et al did qualitative exploration which revealed that mothers were the primary source for knowledge regarding puberty, but friends were preferred for sexual subjects. Also, girls admitted the fact that their friends' information wasn't always accurate and that they would have liked to acquire this knowledge from their mothers rather than friends. 22 In Ethiopia, teenagers preferred to gain their knowledge from school, radio, TV, and other media, while parents were only the alternate preference. 14 It's interesting to note that, due to disappearance of the taboos and once constraints, adolescents in Chile considered the knowledge acquired from parents, teachers, and health professionals as the most dependable. 23 It seems that, cultural and religious influences and sense of shyness and not being comfortable with the parents is what encourages teenagers to further use of the internet, media, magazines, and friends as sources of information. 24 In a study in India by Shankar et al among adolescent girls in the slum areas of Pune, majority felt comfortable in confiding regarding reproductive health issues with friends rather than parents, teachers, or doctors. 25 Analogous results were observed in a study by Sandhya et al in Kerala. 26 The major source of information regarding HIV was from mass media , followed by friends , teachers , and least was from parents and siblings in a study conducted in 2016 in Mysore. 27 In discrepancy are two studies conducted in America by the National Campaign to help Teen Pregnancy which surveyed nationally representative samples of youth 12 to 19 years of age. Unlike the former studies which asked about most important sources of information, these concentrated on who or what most influenced their actual sexual decision-making. 28 In the current study, less than half of adolescents knew rightly what menstruation was. When it comes to knowledge on masturbation,70.3% adolescents answered rightly .99.03% didn't know exactly what sexual contact was . HIV/ AIDS was the most common STD known to adolescents, followed by Genital Herpes. Many adolescents thought that STDs could be acquired via using public toilets;10.4% of adolescents answered that STDs can be transmitted by talking to an HIV/ AIDS case or sharing food with an STD case. While, in a study in Gumuz region of Ethiopia, most of the respondents knew rightly regarding period;84.2% respondents knew about STIs. HIV/ AIDS was the most known STI , followed by Gonorrhea. 16 While in another study according to Kasiye et al,59.9% only knew about menarche .57.3% of the repliers were knowledgeable about SRH matters;97.8% had heard about STIs; majority of the repliers heard about HIV/ AIDS followed by syphilis . But only 34.9% of the repliers had heard of LGV.95.3% of the repliers had heard about HIV/ AIDS prevention methods [being faithful to partner, abstinence, avoiding of sharing sharp materials which accounted ]. Majority had heard about contraception and Norplant was the generally heard system of contraception followed by injectables which accounted 73.7% and72.2% respectively. 14 While in the present study, Condoms and Birth control pills were the most common contraceptives known to adolescents. further than half of the adolescents knew rightly about sexual abuse; females knew more than males. Condoms were the most extensively known contraception among adolescents while emergency contraceptives were least known . Also, 78% of subjects knew about physical signs of adolescence and 88% were aware of the need for adolescent health according to a study by Agarwal et al in India. 29 Less than 30% of the girls were apprehensive of condoms and oral contraceptive pills. Regarding harmful effects of indulging in adulterous physical relationships, further than half of the participants were concerned about bringing a bad name to themselves and their family, while RTIs, HIV/ AIDS, failure to pursue further education and unwanted pregnancy were cited as other ill effects according to Shankar et al in Pune. 25 The difference in the findings suggest the difference in cultural backgrounds, taboo associated with SRH communication among parents as well as teachers and lack of adolescent friendly health facilities in India. --- CONCLUSIONS The major sources of information for adolescents regarding basic SRH issues like private parts, type of touch and menstruation was their parents. When it comes to puberty changes, parents and schools were the major sources. Media was the most common source of information regarding more sensitive issues like Masturbation, sexual contact, pregnancy, contraception, STDs, and sexual abuse. There is poor awareness of adolescents regarding most of the SRH issues. India has the largest adolescent population, and it is a country known to have strict traditional values and age-old cultures and customs. Adequate information through parents may improve the knowledge on these issues. --- LIMITATIONS Due to the COVID 19 pandemic, the calculated sample size could not be achieved to get more accurate results. Few of the responses were obtained through online google forms. It was difficult for students to fully express their views on the open-ended questions. --- RECOMMENDATIONS Age-appropriate comprehensive sexuality education must be incorporated into their curriculum effectively to improve their SRH. Further research to understand the adolescents and parents' perspectives on SRH communication is recommended.

Background: Adolescence is a critical and challenging period of life. Many adolescents today have patchy knowledge on SRH issues, which may lead to risky decision making concerning their health. Kerala, though one of the most literate states in India, is yet to implement comprehensive sex education. Objectives: To list the sources of information and to study the awareness of adolescents on Sexual and Reproductive Health (SRH) issues among high school students in Kerala.A cross sectional study was conducted among 8 th , 9 th , 10 th class students. Multistage random sampling was conducted and a sample size of 338 was achieved. Data was collected via a self-administered questionnaire and analyzed using SPSS Version-23. Results were explained in terms of frequency, percentage, mean and standard deviation. Results: Mean age of the study population was 14.2 ±0.88 years; most of them were females (58%) and followed Christian religion (49%); belonged to urban areas. The major sources of information for adolescents regarding SRH matters like private parts, type of touch, response to sexual assault and menstruation was their mother. When it comes to puberty changes, both mother and teachers were the major sources. Media was the most common source of information regarding Masturbation, sexual contact, pregnancy, contraception, STDs, and sexual abuse. There is wide disparity in the awareness of adolescents on major SRH issues. Conclusions: India has the largest adolescent population, and it is a country known to have strict traditional values and age-old cultures and customs. Adequate information through parents may improve the knowledge on these issues. Age-appropriate comprehensive sexuality education must be incorporated into their curriculum effectively to improve their SRH.

INTRODUCTION Smoking is the leading cause of preventable illness, disability, and death in the United States [1], where it is responsible for 1 in 5 deaths [2]. This rate is disproportionately higher among veterans [3]. Data collected from a national stratified sample of veterans using Department of Veterans Affairs healthcare suggest that 50% of veterans who served during military operations in Afghanistan or Iraq have a lifetime history of smoking and 24% remain current smokers [4]. Unfortunately, effective smoking cessation treatments do not reach most smokers and the use of smoking cessation aids is limited [5]. There is a clear need for new approaches that increase the reach of cost-effective smoking cessation interventions. Studies of the general population indicate that African American smokers are less likely than Caucasian smokers to quit successfully [6][7][8][9][10]. This disparity is accounted for in part by the fact that African-Americans are less likely to be screened for tobacco use, receive cessation advice, and prescribed nicotine replacement therapy than White smokers [9,11,12]. Among veterans using VA healthcare, however, there are not racial differences in screening and utilization of current smoking cessation services [13]. Specialty VA clinic-based tobacco cessation programs appear to effectively reduce smoking for all veterans. Although VA clinic-based care is highly efficacious, these programs are infrequently attended, which limits their impact on prevalence, disease impact, and economic costs of smoking [14]. There is potential to leverage the internet to improve reach of smoking cessation interventions by avoiding barriers that limit participation in specialty care. Smoking cessation treatment may be particularly well suited to delivery via the internet through online chat groups, bulletin boards, email contacts with experts, individually tailored information and feedback regarding behavioral skills, problem solving training, and advice regarding medication usage. Treatment can be provided in "real time" and at the convenience of the user for as long as he/she needs it which may help prevent relapse [15]. Recognizing the potential of the internet as a source of medical information, the VA implemented a web-based portal in 2003 called My HealtheVet that provides evidencedbased health information, access to medical records, patient-directed reminders and secure messaging [16]. While the internet could be used to help decrease smoking among veterans, one limiting factor may be veterans' access to the internet. There are relatively few studies on the use of the internet by veterans [17][18][19][20]. A study of over 7,000 veterans collected in 2010 suggested that 66% of veterans using VA healthcare use the internet [17]. Consistent with prior surveys of access to the internet among veterans [20], there was evidence of a racial disparity in internet access where African American veterans were significantly less likely than other racial groups to access the internet [17]. A recent survey of Afghanistan/Iraq era veterans using VA healthcare, however, suggested internet access among younger OEF/OIF veterans was nearly universal [21]. While the majority of veterans report using the internet [17,21], relatively little is known about access to the internet at home. Access to the internet in the home, may be particularly important for the success of web-based self-management interventions. A 2010 study of VA users with diabetes found that 59% of respondents reported having access to the internet at home [16]. Home access to the internet was associated with several socio-demographic variables including younger age, higher income, and more education [16]. No previous study has examined internet access in the home among VA healthcare users that smoke. Access to the internet at home may be an important barrier to electronic health and mobile health smoking cessation interventions. The current study explored possible racial and other socio-demographic disparities in access to the internet at home among veteran smokers. The specific goals of the current investigation were to describe the socio-demographic characteristics of veteran smokers with and without home internet access, assess factors associated with internet utilization and ascertain demographic correlates of utilization of a guideline-concordant, state-of-the-art web-based smoking cessation intervention. --- METHODS --- Participants and Procedure Data were collected as part of a randomized smoking cessation effectiveness trial that compared a web-based smoking cessation intervention to a facilitated referral to VA specialty clinic-based smoking cessation treatment [22]. The methods and outcomes for the parent trial are fully described elsewhere [22]. Veterans with tobacco use were identified from electronic medical records based on the presence of ICD-9 code for nicotine dependence or identification of current smoking status based on results of required annual screening. VA EMR smoking data has been well validated [23]. Veterans were sent an introductory letter that described the study and included a toll-free opt-out number. Those who did not decline were proactively recruited by telephone between 2010 and 2014 and asked if they were currently smoking tobacco and willing to quit smoking in the next 30 days. Eligible patients included current smokers who: a) were enrolled at the Durham VA Medical Center for primary care and b) were willing to make a quit attempt in the next 30 days. Patients were not excluded if they did not have access to the internet. In total, 2756 veterans were mailed an invitation letter. Eighteen veterans called to opt-out of the study and were not recontacted; 1,516 were called by telephone to determine eligibility and willingness to participate; of these 1,103 were excluded ; and 413 were enrolled. Participating veterans completed a baseline survey that included demographic data, smoking characteristics, mental health screens, and access to the internet. A dichotomous measure of home-based internet access was used: "Do you have access to the Internet at your home?". Additional items queried frequency of use. Participants randomized to the internet smoking cessation intervention were provided a free, lifetime membership to the full, enhanced version of QuitNet® . They were provided with information on how to register on-line via a QuitNet® home page branded for the study. Participants were encouraged but not required to register and use the site. Study staff discussed options for accessing the Internet with those who reported a lack of internet access. For those randomized to receive the internet intervention , objective data on utilization of the website were provided by QuitNet®. Any utilization of the site was monitored for a period of 12-months. Perceived utility of the site was assessed with a single item on a 7-point Likert scale, "On a scale from 1-7, where one means "not at all useful" and 7 means "very useful", how useful was the QuitNet® program in helping you to try to quit smoking?" Similarly satisfaction was assessed by "On a scale from 1 to 7, where one means "not at all satisfied" and 7 means "very satisfied", how satisfied are you with the QuitNet® information and services?" An additional item measured how difficult the site was to understand; "On a scale from 1 to 7" where one means "not easy to understand" and 7 means "very easy to understand" how easy was the information/advice from the QuitNet® program to understand?" Participants were also asked if they would recommend QuitNet® to a friend. Nicotine dependence was assessed with the 6-item Fagerström Test for Nicotine Dependence [24]. Depressive symptoms were measured using the 10-item version of the Center for Epidemiological Studies Depression Scale [CES-D 10; [25]], which has been used extensively in epidemiologic studies [26]. A score of 10 or higher indicated probable depression [25,27]. The Primary Care PTSD screen [ [28]; PC-PTSD] was used to screen for clinically significant PTSD symptoms. The PC-PTSD has demonstrated good reliability and validity [28]. A score of 3 or more was used to indicate probable PTSD [29]. Alcohol misuse was assessed with the well validated AUDIT-C [30,31]. Scores of > 5 for men and > 4 for women were used to indicate probable alcohol abuse [31]. --- Analyses Descriptive statistics were calculated to characterize demographic attributes of smokers with and without a home internet access. Logistic regression analyses were used to test bivariate associations between demographic and clinical factors and access to a home internet connection. Multivariable logistic analyses were conducted to examine the unique association between home internet access and each predictor while adjusting for all other demographic and clinical variables. A similar approach was used to examine frequency of internet use and use of the web-based smoking cessation intervention. These latter models were also adjusted for home internet access . The following variables were dichotomized to avoid problems with small cell sizes: marital status , employment status , education , and era of service . In models examining use of the internet smoking cessation intervention , race was additionally dichotomized . All statistical analyses were performed using SAS, Version 9.4. --- RESULTS The mean age of the sample was 43 years , 51% were White, 39% African-American, 16% were female, and 48% served during the Afghanistan/Iraq era. Many participants screened positive for probable PTSD , depression , and alcohol abuse . The majority of participants reported having home internet access . Seventy percent of smokers used the internet daily, 80% used the internet at least on weekly basis and 13% percent of participants reported they never used the internet. Table 2 shows both unadjusted and adjusted effect sizes of demographic and clinical variables associated with having home internet access. In this sample of veteran smokers, older age and African-American race were significantly associated with reduced odds of having internet access at home. Higher levels of education, being married, and being currently employed were all associated with increased likelihood of home internet access. Higher levels of nicotine dependence, screening positive for PTSD, and screening positive for alcohol misuse were also positively associated with access to the internet at home in bivariate results. Adjusted results from the multivariable logistic regression model indicated that smokers who were younger, married, and White were more likely to have home access to the internet. Only seventy-two percent of African-American veteran smokers reported having a home internet connection compared to 88% of White veterans. Adjusted results examining using the internet on a daily basis indicated that older Age, African-American race, and lower education were associated with lower odds of using the internet on a daily basis. Longer travel times to a VA medical center were also associated with reduces odds of daily internet use as were higher levels of nicotine dependence. Not surprisingly, having access to the internet at home was strongly related to using the web on a daily basis. Factors associated with never using the Internet were largely the same, but in the opposite direction. Older age, less education, and a lack of home internet access were all significantly associated with reports of never accessing the internet . Among those randomized to the web-based smoking cessation intervention, 84% had access to the internet at home and the majority used the web on daily basis. Only half of those who were given free access to the full version of QuitNet®, actually registered for and used the website. As shown in Table 4, several socio-demographic factors were associated with using the site in bivariate results. Older age and African-American race, were associated with significantly lower odds of using the site while being employed, being married, and increased education were all associated with increased odds of registering and using the website during a quit attempt. Those who had access to the internet at home were significantly more likely to go online and register on QuitNet®, although 19% of those who did not have internet access at home , still registered and used the site. In fully adjusted analyses, only internet access at home was uniquely associated with using the site. None of the other socio-demographic variables were significantly related to use of the site after controlling for internet access at home. Among those who registered on the site, there were no differences in the level of utilization as a function of age, education, employment, marital status, or race. An analysis of items used to evaluate the perceived utility, ease of understanding, and satisfaction revealed few socio-demographic associations. Overall, participants reported the web-based intervention was moderately useful , easy to understand , and were satisfied with the intervention . Eight-six percent of respondents said they would recommend the intervention to a friend. Although older age was associated with less use of the site in bivariate results, older participants tended to rate QuitNet as more useful . There were no other statistically significant socio-demographic associations with perceptions of the site. --- DISCUSSION Smoking is disproportionally higher among veterans. Among veterans the prevalence of smoking is even higher among those that use VA healthcare and those with lower socioeconomic status [17,19,16]. To our knowledge, this is the first study to examine internet use among a sample of veteran smokers who use VA healthcare. The majority of smokers reported having access to the internet at home. Although results of the current study suggest increased penetration of internet use compared to previous studies [17,19,16], we found evidence of significant socio-demographic disparities in access and use of the internet among VA users who smoke. Veterans who were African-American, older, less educated, had longer travel times to their nearest VA facility, and who were more nicotine dependent were significantly less likely to use the internet on a daily basis. Compared to 88% of White veterans, only 72% of Black veterans reported having access to the Internet at home. Even after adjusting for other socio-demographic and clinical variables, race remained significantly associated with home internet access. Socio-demographic disparities in internet access and use may exacerbate health-related disparities associated with smoking as eHealth interventions are widely disseminated. Among those participants randomized to receive free access to a guideline concordant, stateof-the-art smoking cessation website, there were significant bivariate associations between use of the QuitNet® intervention and age, race, marital status, education, employment, and service era. Older age and African-American race were associated with lower odds of using the site while being married, more educated, currently employed, and serving during OEF/OIF increased the odds of using the website during a quit attempt. In adjusted analyses, however, only access to the internet at home was uniquely associated with use of the webbased smoking cessation intervention. These results indicate that access to the internet at home is an important barrier to the uptake of web-based smoking cessation interventions. Indeed, none of the socio-demographic variables were significantly associated with the intensity of the use of the site among those who registered. Similarly, race and education were not significantly associated with perceived utility, ease of understanding, or satisfaction with the site. Consistent with national data showing increasing penetration of internet use over the last fifteen years [32], the rate of internet use among the current sample of smokers using VA healthcare is higher than previous reports of internet access among veterans [17,19,16]. Rates of home access to the internet were also higher than a sample of VA users with diabetes [16] recruited earlier from the same medical center as the current study. Despite increasing access levels, significant disparities in access and use of the internet continue to exist. Nationally, smoking rates are significantly higher among those with less education and income. The burden of tobacco-related morbidity and mortality is disproportionately higher among low income, less educated and underserved racial/ethnic minority populations [33,34]. Disparities in access to web-based and mHealth interventions may exacerbate current disparities in health care outcomes. Several of these variables could be readily addressable. For example, efforts to increase access to technology and the internet among low-income and ethnic-minority populations could increase utilization of eHealth resources. For example, based on promising pilot data [35,36], we are currently examining the effectiveness and cost-effectiveness of a mHealth smoking cessation intervention that involves giving smokers internet-capable smart phone devices during the intervention period. In the VA, a program could be developed that would provide surplus computers to veterans paired with computer training and tutorials on how to use the Internet to access health information [16]. Consistent with previous work [17], veterans screening positive for mental health conditions in the current study were as likely to have home internet access as other veterans. Internetbased and eHealth interventions could be a promising modality to increase the reach of mental health treatment for veterans. There is increasing evidence to support eHealth and mHealth interventions aimed at treating anxiety, depression and substance misuse [37][38][39]. There are a number of limitations associated with the study. First, while the trial from which these data were collected [22] used proactive recruitment methods to maximize generalizability of results, participants were limited to veteran smokers who used VA healthcare and who reported they were willing to make a quit attempt in the next 30 days. Thus, the sample may not be representative of all veteran smokers. We provided participants with free access to the paid, full version of QuitNet®. If smokers had been asked to bear the costs associated with a commercial intervention, one might expect even greater socioeconomic disparities in utilization. While we encouraged veterans to register on the specially branded internet site, we were unable to send them an email or text link to the site due to VA privacy rules. It is possible that socioeconomic disparities may have been smaller if the process to register and use the site was made as simple and user-friendly as possible. We did not collect qualitative data on barriers or reasons participants chose not to register and use the internet smoking cessation site. More research is needed to examine other potential barriers to the use of web-based smoking cessation interventions. In conclusion, although the majority of veteran smokers enrolled in a large smoking cessation trial reported having access to the internet at home, sociodemographic variables including race, age, and education were associated with the frequency of internet use. The VA and other large healthcare systems must remain aware that there are disparities in use and access to the internet, and that these gaps in access must be addressed to minimize racial and socio-demographic health disparities.

Introduction-Access to the internet at home may be an important barrier to electronic health (eHealth) smoking cessation interventions. The current study explored possible socio-demographic disparities in access to the internet at home among veteran smokers. Methods-Data from participants proactively recruited and enrolled in a randomized smoking cessation effectiveness trial (N = 408) that compared a web-based smoking cessation intervention to VA usual care were used to examine the demographic attributes of smokers with and without internet access at home. Multivariable logistic regression was used to examine associations between demographic factors and home internet access. Data from patients randomized to the internet arm of the study (N = 205) were used to ascertain correlates of utilization of the intervention website. Results-While the majority of the sample (82%) endorsed access to the internet at home, veterans who were African American, older, and not married were significantly less likely to have home internet access. Veterans who were African American, older, less educated , had longer travel times to the nearest VA facility, and increased nicotine dependence were less likely to access the internet on a daily basis. While several socio-demographic variables (e.g., age, race, education,

introduction Nursing accounts for approximately 59% of the healthcare workforce 1 and is an essential profession for the operation of public, private, and philanthropic healthcare systems. The essentiality of nursing stems from the indispensability of nurses, nursing technicians, and nursing assistants, working daily in the healthcare network, a labor locus in which the minimum conditions for the development of decent work in human healthcare must be conceived in order for the paradigm of essentiality to prevail, based on human dignity, on the valuation of the worker, and contractual work relations 2 . Nursing is a predominantly female profession, although it has been undergoing a process of masculinization over the past three decades. Data from the Federal Nursing Council 3 , show that there are currently 450,770 nursing assistants; 1,611,639 nursing technicians; and 670,581 professional nurses registered in Brazil. Among these workers, 85.1% declare themselves to be female, while 14.1 declare themselves to be male. Approximately 50% of this contingent has worked in the frontline during the public health crisis resulting from the COVID-19 pandemic. In Brazil, the nursing workforce is concentrated in large urban centers, with 56.8% of workers employed in the 27 state capitals and their metropolitan areas, and 40.9% employed in the over five thousand cities around the country 3 . Given its demographics, nursing, better than any other profession, represents how women have been treated in both the labor market and society throughout history 4 . These are people exposed to miserable wages, exhausting hours, unjust periods of rest, prejudice, moral and sexual harassment, precarious transportation, insecurity, violence, and an overload with home and family responsibilities 5 . As in teaching, gender inequalities are obvious. Double employment is a reality imposed by low wages on professionals in this sector. Working in two or more health units makes earning a decent salary possible. The study, Profile of Nursing in Brazil 6 , shows that at least 25% of the category work two jobs. Pressured by low wages, another significant portion of the categories that make up the nursing class is pushed towards informality, having to do "odd jobs" and "extra activities" to survive, such as nanny services, caregiver for older people, beauty salon work, housekeeping, civil construction services, private security guards, delivery services, among others. The work overload of nursing professionals is a brutal reality and is present in 75,000 healthcare facilities in the public sector and 60,000 in the private sector 7 . In general terms, the life of a nursing professional in this space is divided into four stages: the beginning of professional life -up to 25 years of age; the stage of professional training -between 26 and 35 years of age; professional maturity -between 36 and 50 years of age; and the phase defined as "professional slowdown"between 51 and 60 years of age 3 . According to official metrics, 59.3% of all nursing teams work in the public sector, 31.8% in the private sector, 14.6% in the philanthropic sector, and 8.2% in teaching activities 8 . Due to low wages and exhausting working hours, only 40% of the people who make up the nursing teams could take on some form of professional training after graduation. This scenario can favor the stagnation of professionals, who tend to remain attached to their careers for fear of losing rights and benefits, with no prospect of professional growth and development 9 . Conversely, in a clear demonstration of resilience, approximately 30% of the nursing assistants and technicians attended graduation to become a nurse. In the same sense, the absolute majority want to continue their studies to reach a higher level, demonstrating great commitment to their career 6,10 . Nevertheless, the pursuit, mastery, and monopoly of knowledge bring nursing closer and closer to autonomy and social prestige 7 . Despite the barriers, the profession broke with the history of philanthropic and religious assistance provided to the sick, at Santas Casas de Misericórdia , and became professional. The charitable and altruistic image associated with nurses has become a thing of the past. Today, the activity combines the sociological elements that project its future: self-knowledge, a vast job market, and social recognition 4 . During the pandemic, Brazilian nurses cared for 34 million patients infected with COVID-19, applied 519 million vaccines against the disease, and comforted hundreds of thousands of COVID-19 patients, of which 680,000 died 11 . This work retrieved and enhanced the narrative of heroism 12 , to the point that people went to their windows to applaud the work of brave professionals who risk their lives in the frontline, without protection, yet with exceptional courage 13 . As never before, nursing became the object of editorials and starred in reports in the media. Category representatives were invited to give their opinion on public health problems 14 . Thus, nursing professionals' history and demands were at the center of the national political debate 15 . However, despite the narrative, the pandemic revealed to society that, behind the hero's esthetic, there is a category that is tired, exploited, and exposed to the risks of an inhuman healthcare system, with no respect for safety and health rules or the human aspects of work processes. According to official data, in Brazil, COVID-19 infected 64,629 nursing professionals and killed 872 of them 16 . One out of every three nursing professionals who died of COVID-19 died in Brazil 17 . The country accounts for 3% of the world's population and 11% of total deaths caused by the new coronavirus 18 . This was an unprecedented tragedy, which reveals the level of precarious health to which nursing professionals are subjected. During the COVID-19 pandemic, 60% of Brazilian nursing professionals experienced a lack of institutional support to cope with the disease; 21% felt devalued by their bosses; 40% experienced some type of violence in the workplace; 33.7% suffered discrimination in their own neighborhoods; and 27.6% were victims of discrimination on the way between home and work, as people believed that the worker carried the virus and, therefore, represented a walking risk. As a consequence, 15.8% of the professionals declared that they suffered from sleep disturbances, 13.6% reported irritability and frequent crying, 11.7% said they were unable to relax and felt stressed, 9.2% had difficulty concentrating or slow thinking, 9.1% reported losing pleasure in their careers or lives, 8.3% reported a negative feeling about the future and negative or suicidal thoughts, and 8.1% faced appetite or weight changes 19 . During the pandemic, Cofen received 6,200 complaints about the precarious working conditions of nursing professionals in the care of COVID-19 cases and, through the Regional Nursing Councils , supervised 8,674 healthcare facilities 20 . Cofen dealt with a sanitary storm driven by experimental guidelines drawn from rudimentary knowledge. The situation changed into a scenario that exacerbated inequalities and showed that we were not all in the same boat. This metaphor makes it clear that the most fragile vessels are those that take healthcare professionals in the frontline 21 . This is evidenced by the fact that, in nursing teams, 65.9% of people claim they suffer from burnout, mainly due to an overload of responsibilities. Living with pain, suffering, and illness, as well as with poor working conditions, has led to a bleak outlook. Only 34% of nursing professionals exercise throughout the week; that is, 66% are declared sedentary. One-fourth of the category has comorbidities, such as hypertension, obesity, lung diseases, depression, and diabetes, and more than 70% show signs of extreme exhaustion, in addition to physical and psychological sequelae resulting from the pandemic 4 . Strictly speaking, they are individuals who are prevented from leading a healthy life. In this scenario, the studies: "Healthcare professionals' work conditions in the Covid-19 scenario in Brazil" 22 and "Invisible healthcare workers: work conditions and mental health in the Covid-19 scenario in Brazil" 23 , whose unpublished results subsidize this work and will be demonstrated below, present revealing data on the situation of the profession in Brazil and make a relevant contribution to the advancement of knowledge on the conditions of the largest healthcare workforce in the country, Nurses. In view of the above, this study aimed to describe the sociodemographic and health profile of nursing professionals in the context of the COVID-19 pandemic. --- Method This is a descriptive study, with a cross-sectional outline, built from data obtained in studies "Healthcare professionals' work conditions in the Covid-19 scenario in Brazil" 22 and "Invisible healthcare workers: work conditions and mental health in the Covid-19 scenario in Brazil" 23 , coordinated by the Center for Strategic Studies of the Oswaldo Cruz Foundation's National School of Public Health . These two studies originated from a nationwide matrix project and counted on the participation of 8,897 professional nurses and 11,469 nursing assistants and technicians, respectively. The nursing professionals in the study sample worked in the frontline of the fight against the COVID-19 pandemic in public and private health institutions in 2,200 cities across the country. The studies were made public on social media and institutional contacts by the national and regional entities that bring together healthcare workers, including the Federal Nursing Council and the Regional Nursing Councils. The research database consisted of an online questionnaire containing closed-ended questions, applied between June and December 2020, using the Research Electronic Data Capture The studies used non-probabilistic sampling, the snowball model, starting from the social media of the actors involved to access the group that constituted the target audience of each study so that the choice of participants did not follow a random template. The questionnaire was applied for self-completion and free sharing, and data analysis was performed using descriptive statistical methods. The matrix project, coordinated by Fiocruz, was approved by the Research Ethics Committee under No. 4.081.914 . --- results This study included a sample of 20,393 nursing professionals, of which 8,897 were professional nurses who had participated in the study "Healthcare professionals' work conditions in the Covid-19 scenario in Brazil" 22 , and 11,496 were nursing technicians and assistants that had participated in the study "Invisible healthcare workers: work conditions and mental health in the Covid-19 scenario in Brazil" 23 . As shown in Table 1, these professionals are distributed throughout the country, with both professional nurses and nursing assistants and technicians being predominantly based in the Southeast region. There was a lower concentration of nurses in the North region and of nursing assistants and technicians in the Midwest region . A concentration of professionals working in the capitals of the Brazilian states stands out: professional nurses and nursing assistants and technicians , while in cities in interior regions: professional nurses and nursing assistants and technicians . According to the demographic profile , most of the professionals who participated in the study were female , with a concentration of professionals in the professional maturity phase , followed by those in the professional training phase , with 48.2% and 35.1% of professional nurses, respectively, and 48.5% and 29.3% of nursing assistants and technicians. Regarding color/race, most professional nurses were white . Among nursing assistants and technicians, there was a predominance of people of mixed race . As shown in Table 3, the work profile showed that, in the COVID-19 pandemic scenario, most nursing professionals worked from 21 to 40 hours a week . Notably, however, 30.9% of professional nurses and 29.1% of nursing assistants and technicians worked between 41 and 60 hours a week. These results show more than one employment relationship among nursing professionals participating in the studies. The hospital system constituted the main group of healthcare facilities in which nursing professionals worked during the COVID-19 pandemic: 56.2% of professional nurses and 65.2% of nursing assistants and technicians worked in hospitals. Despite the role of nursing in coping with the COVID-19 pandemic, professionals who worked in the frontline were at risk of contamination. It was observed that 24.4% of professional nurses claimed to have received a positive diagnosis for COVID-19. Among nursing assistants and technicians, this percentage was 38.5% . Many nursing professionals who worked on the frontline in the fight against COVID-19 had pre-existing diseases, which may constitute a risk factor for the increase in the severity of the disease and death. The study shows that approximately one-fourth of professionals claimed to have pre-existing conditions, including high blood pressure , obesity , and pulmonary diseases . --- Discussion The results of this study express the profile of Brazilian nursing in the context of the COVID-19 pandemic and denote that these professionals were essential for the health care of people affected by the disease. They also point out that the distribution of nursing professionals in the different regions follows the spatial distribution of the population, with a greater concentration in the Southeast region, the most developed region with the greatest availability of healthcare resources and services in the country 24 . This disparity in the distribution of nursing professionals may have intensified in the COVID-19 pandemic scenario, due to the availability of resources, with a more substantial impact on the economically more vulnerable population 25 . It is not objectively about decentralizing the healthcare system nor about the division of federal, state, and municipal responsibilities in achieving low, medium, and high-complexity services to citizens 26 . If that were the case, the system would work equally and universally, regardless of regional disparities, as provided for in the constitution. In a literal expression of reality, we are dealing with inequality stipulated mainly by the economic bias of a country marked by structural inequalities that date back to the colonial period, with a deep stratification of classes 27 . Indeed, these asymmetries in the distribution of the nursing workforce reflect and represent the social and economic inequalities of Brazilian society 28 . More specifically, they essentially represent the difficulties and prejudices faced by women in the pursuit of social ascension, financial emancipation and professional independence 29,30 . Stifled by low wages, a significant number of professionals are forced to keep more than one job, in addition to still being responsible for domestic and maternal chores. It is no exaggeration to say that these workers face triple shifts, with no due recognition and appreciation 31 , and even in the face of these difficulties, they engage in their work and demonstrate energy, resilience, and enthusiasm, strengthening their commitment to their careers 9,32,33 . In the meantime, the hypothesis is corroborated that nursing is discriminated against, subjugated, and deprived of fundamental rights, since it consists mostly of women. The sexism impregnated in society is reproduced in labor relations in healthcare services and causes the sedimentation of a culture based on patriarchy, which subordinates women and promotes the unfair division of labor 30,34 . At the same time, men who exercise the profession are not in a comfortable position either. The study results indicate that, all things considered, they are also affected by the prejudice that outlines Brazilian society. Thus, they also become ill because of social injustices and excessive responsibility, adopting a sedentary life due to a lack of time and resources for leisure and physical activities 35 . Inequalities affect all aspects of gender issues; the nursing profession is no exception. Most white people are graduates , while most mid-level workers are declared black or brown. Centuries of slavery has led to racism and the establishment of unequal levels of access to education and a career in nursing, demonstrating how racial and social relations have been reproduced up to the present 36 . In addition, working 41 to 60 hours a week in healthcare facilities -as approximately one-third of Brazilian nursing professionals doand remaining fully healthy is impossible 37 . Long working hours associated with high levels of occupational stress generate psychological illness in nursing professionals [38][39][40] . In addition, dealing with others' pains, with the chance of becoming infected, with the risk of becoming ill, and with the overload of responsibilities led these individuals to face a damaged life with no margin and room for self-care and self-knowledge 41 . Despite the visibility and recognition achieved by nursing after the COVID-19 pandemic, these factors have yet to make the reality of the category better, at least for now 42 . Except for the advancement of wage floor regulation, sought for at least three decades, there is no promising discussion in progress about the necessary reduction of the workload to 30 hours a week 43 , nor has there been a reduction in the risks involved in the exercise of professional activity. This is reflected in the increase in stress and the prevalence of pre-existing diseases, such as hypertension, obesity, and lung diseases, manifesting as risk factors for death in different contexts and scales, mainly due to infections and work-related accidents 26,38,44 . In general, the situation of nursing directly impacts the quality of care offered to users of healthcare services. Demanding quality and not providing the required investment is not reasonable. Any self-respecting solution should not be restricted to inputs and technologies but must also reach the human sphere of work processes, as it is impossible to achieve better rates in the providing of care and not tend to the urgent needs of these workers, who still seem to have a long road ahead. The situation presented by this study is not recent, but it accumulates aggravating factors over time. The ills that affect the category and the tra-jectory of the nursing struggle for rights, recognition, and better working conditions have been a hundred-year reality in Brazil. It dates back to 1926, when graduates of the Nursing School of the National Department of Public Health , dissatisfied with their work conditions, began to organize themselves politically to expose the hardships of the profession 45 . Since then, this process has intensified to the point of becoming a latent issue in the public debate. Nursing professionals are the backbone of the Brazilian healthcare system. Evidently, all workers in the multidisciplinary team are essential for the functioning of healthcare facilities. However, nursing is the category closest to patients. It is defined as essential in pain care, the administration of medication, therapy management, the search for human comfort, immunization of the population, and the management of life and death, which are core activities of those who work in the segment. --- conclusion The results of this study allow one to outline some aspects related to the profile of nursing, in the COVID-19 pandemic scenario, such as the predominance of female professionals, with an age corresponding to the stages of training and professional maturity; heterogeneous distribution of professionals in the national territory, with a higher number in the Southeast region; concentration of professionals working in state capitals and hospital units; the presence of more than one employment relationship; and many professionals with pre-existing diseases, especially systemic arterial hypertension, obesity, and lung disease. By profiling the category, this study contributes to the definition of public policies and adequate approaches to the need for these profes-sionals in the country, primarily concerning the most vulnerable segments of the population. The World Health Organization advocates that being healthy does not mean being free of disease, but having living conditions that guarantee physical, mental, and social well-being, including access to healthcare services, the guarantee of which depends on the appreciation of nursing professionals. Healthcare conditions in Brazil were unfavorable, and the COVID-19 pandemic aggravated this. The lack of protective equipment, the shortage of professionals, low wages, exhausting hours, and the high rate of illness and death among our nursing colleagues have shaken foundations that were already fragile. From this perspective, nursing has been plagued by practically insurmountable difficulties over the past three years; at the same time, it gained public respect in society, given the number of editorials and tributes rendered to the category during this period. However, the tributes did not translate into better working and living conditions for nurses, nursing assistants, and technicians, at least in the short run, as evidenced by the data in this study. Thus, overcoming this reality and building a better future for nursing and healthcare involves meeting the demands expressed in the aforementioned research coefficients, which reveal essential information for the scientific and political understanding of the subject. collaborations BMP Santos, AMF Gomes, LG Lourenção and NP Freire contributed to the concept and development of the study, data collection, acquisition, and interpretation, writing and final approval of the article. ICKO Cunha, AJCA Cavalcanti, MCN Silva and D Lopes Neto contributed to the writing and final approval of the article.

an emphasis on the presentation of potential perspectives arising from this historical process and recent events.

Introduction In Sub-Saharan Africa, chronic childhood malnutrition is a major cause of child morbidity and mortality [1]. Though under-nutrition is still a major treat, the three simultaneous dimensions of malnutrition-undernourishment, micronutrient deficiencies, and over-nutrition-co-exist in developing countries [2]. Nutritional problems emanate from different underlying causes which include inadequate household food access and caregiver feeding practices [3,4]. Mothers, who are the primary care takers of children in this region, are responsible for purchasing, preparing, and household food allocation that affects the nutritional status of children. Food-related decision making is a complex matter, and mothers need to consider the food supply, which is related to agricultural-food systems that affect the demand for and use of food [5]. Women's decisions on what to feed their children are also influenced by price and environmental issues such as availability, type of outlets, food types, and location of shops, as well as social factors including knowledge of nutrition, social status, preference, peer pressure, time and effort required to prepare meals, prior bad or good food-related experiences, and cultural expectations [5][6][7][8][9]. These factors have a cumulative effect on the food purchasing practices of the mother, thereby determining what food is provided to the children. Available evidence highlights the importance of parents and caregivers as the primary influencers and gatekeepers of children's eating behaviors [10]. Therefore, understanding what influences mothers' choices of food for their children is key to designing effective strategy to improve children's diet and nutritional status. However, there is limited research on this issue in low-income countries. As urban settlements are increasing in Sub-Saharan Africa, food provision at a household level is becoming even more complex. Issues related to food systems in African cities have certain similarities [11] though much of them are not fully understood [12]. Due to urbanization, there has been a large shift in low-income countries from the traditional diet, rich in cereals and fiber towards more processed, calorie-dense foods that are rich in sugar and fat [13,14], in line with the global trends [15]. These rapidly growing food habit changes in low-income countries [16] are exacerbating the existing nutritional challenges [17]. In many urban areas, processed and ultra-processed food items are regarded as symbols of high economic status and are widely consumed [18]. A recent study in Ethiopia also showed that the consumption of saturated fats, sugary beverages, processed meats, cholesterol, and sodium is increasing in urban areas [19]. Considering that half of Africa's population will be living in cities by 2050 [14], it is very important to understand the shifts in food habits and decisions that enforce the changes with further details to better inform policy and programs. In cities that are rapidly growing, like our study area, people have limited access to cultivation land and are therefore highly reliant on purchased food [20], making the urban environment quite a different living space than the rural areas. The growing urbanization is disrupting the traditional eating/feeding practices, forcing households, particularly mothers, to navigate through the local market and make choices depending on their knowledge and economic status. Unfortunately, the food industry exploits the inherent human preference for energy-dense, smooth , salty, fatty, and sweet foods by making them widely available and affordable [21]-appealing for both mothers and their children. In financially constrained African cities, energy-dense but cheap nutrient-poor foods have become more accessible compared to healthy foods [6,22]. A combination of high rates of urban poverty, dependency of urban households on the market, and fluctuating food prices intensifies the food insecurity concerns among urban residents, leaving parents under the tremendous stress of making tough decisions on food choices. Caregivers, mostly mothers, who are responsible for providing to their children have to consider the options carefully before choosing what to buy. Having a more nuanced understanding of what influences mothers is key to designing effective interventions. Therefore, this paper explores urban mother's food decision making through investigating her experiences within her food environment. --- Conceptual Framework This paper draws on the food environment conceptual framework developed by the Agriculture, Nutrition and Health Food Environment Working Group [23]. This framework draws its foundations from theoretical socio-ecological concepts and expands upon it to conceptualize the interaction with the wider food environment. The food environment according this framework has two domains that share an inter-related set of physical, economic, and socio-cultural dimensions. The external food environment domain consists of food availability, prices, vendor and product properties, marketing, and regulations, while the personal food environment domain encompasses accessibility, affordability, convenience and desirability at the individual level. In this study we focus on the personal domains and have used the definitions of Herforth and Ahmed [9] of the domains, which are more succinct and in line with our work. The availability and affordability domains focus on food availability in the market and the households' ability to purchase food, respectively, while convenience considers how time and cost interrelate. When time is a scarce resource, ease of access and preparation may be even more important than the actual financial cost. Finally, desirability is seen as made up of internal factors and external factors . Though this framework served as a guide for organizing our results, this exploratory qualitative work concerning women's decision of what to feed their children revealed more nuanced motivations which lie behind these very concrete factors. Turner et al. [23] in their review also point out the research gap in understanding the personal factors as well as the underutilization of qualitative studies. Hence, this study is sought to address this disconnect by focusing on individuals and understanding how they navigate within their wider food environment. --- Study Population and Methods This study was conducted in Addis Ababa, Ethiopia, from March through November 2016. Addis Ababa is home to approximately 4 million people and is one of the most populous cities in Sub-Saharan Africa. Over the past two decades, the city has experienced a phenomenal economic, social, and spatial transformation which is expected to continue and is primarily attributed to the country's double digit economic growth, rapid urbanization, and an expanding population [24]. Increased economic prosperity has translated into relatively more disposable income that can be used for purchasing foods, including those foods outside the traditional Ethiopian diet. In line with this, the retail sector in Addis Ababa is currently very diverse, ranging from the traditional open air market to modern supermarkets [24]. Despite the growing economy in the country, Addis Ababa is faced with poverty-aggravating factors such as massive rural-urban migration, poor housing, and unemployment, factors which counteract the progress of the city. Majority of the employed households , have a monthly income of about US $68 and dependency ratio of 28% [25]. For this qualitative study, participants were sampled purposively with the help of the health extension workers who are employed by the government to provide primary health care at the neighborhood level. A maximum variation sampling technique was used to capture a wide range of perspectives related to child feeding by including mothers with children of a diverse age range to account for the different developmental stages of the child relative to their dietary needs [26]. Additionally, efforts were made to include mothers from various socio-economic levels in order to capture potential differences in their involvement in community work, educational status, work situation , and number of children. Members of the research team approached each woman identified for the study and explained the study objectives, confidentiality issues, data use and handling, and their right to withdraw from the study at any time without requiring a reason. After obtaining informed verbal consent from each participant, interviews were carried out in the privacy of their home or in other private locations where mothers felt comfortable to be interviewed. The dialogical approach as per naturalistic inquiry guided the interviews. This involves a more fluid conversation style of inquiry with open-ended questions versus a question-answer interview [27]. Initial questions were developed in consultation with the research team and were based on relevant literature. The questions included: "Can you tell us a little about your daily activities?", "What types of food do you usually feed your child?", "Are there any reasons why you give those food items to your child?", "What are the popular foods given to children in your area?", "What street foods are available in your area?", and "Which street foods are commonly given to children?". Some probes were also used to further facilitate dialogue, for example if a mother referred to her own different childhood foods, the questions "What is different from the way you were raised?" or "How do others raise their children with regards to child feeding?" were asked. The first author observed several of the interviews and reviewed all field notes at the end of each day, discussing with the team how to fine-tune the interview process. Throughout the study, additional questions were added as new issues were raised, as is typical of emergent design in qualitative research. All interviews were carried out by experienced research assistants who had received training on interview techniques and the objectives of the study. A total of 36 interviews were completed; they ranged from 20 to 109 min in length. All interviews were tape-recorded, transcribed verbatim, and translated into English. Analysis was an ongoing process and formally commenced in the field, including fieldwork observations and daily recording of field notes. Thematic analysis was employed when all interviews were completed. To conceptualize what influenced mothers' decisions in providing food for their children we used the lens of the personal food environment, which considers four dimensions: affordability, availability, convenience, and desirability [9,23]. Two of the authors then read and re-read the transcripts and identified themes independently. After the initial series of mapping, defining, and refining as per Braun and Clark [28], the dominant themes were discussed until consensus was attained among all the research team members. The themes were also presented in two seminars where feedback from other researchers promoted further clarity. --- Ethical Considerations Ethical approval for this study was obtained from the Institutional Review Board of Addis Continental Institute of Public Health. Before approaching individuals, the necessary permission was sought from the Lideta sub-city and all district health offices. All interviews were carried out in the privacy of the mothers' homes; necessary precautions were taken to avoid intruders during the interview or avoid any situation that could make the mother uncomfortable. Data were then kept securely and were only accessible to the research team members. --- Results A systematic analysis of the qualitative data revealed four major themes: Mothers give-in to a child-driven diet; quick-fix solutions versus the privilege of planning; keen awareness of food safety, nutrition, and diet diversity; social, familial, and cultural influences. Each of the themes is elaborated below. --- Mothers Give-in to a Child Driven Diet For this theme, mothers expressed how they gave in to the child's requests in many instances: they talked about giving food that the child preferred rather than what they thought was good for them, due to fear that the child may refuse to eat at all. "I don't know why he chose these foods . I cook very delicious foods but he refuses to eat them, especially lunch time. I don't know why he only wants to eat instant noodles . . . So I often prepare that for him." Other mothers used food as a reward or an appeasement mechanism when the child was continuously crying, pleading, and nagging. In such instances, the food given appears to be less thought out and more impulsive. As this mother's story conveys, the most accessible food at that moment, be it street food and/or sweets, was given to soothe the child: "She mostly eats lollipops, whenever I go to work she starts crying . . . So I buy her a lollipop or biscuit to calm her down before I leave." Mothers also reported that what they pack in their child's lunch box is highly influenced by fear that her child will feel inferior to other children and is driven by the child's peer pressure. Mothers shared that they feel guilty when not being able to provide variety: "I know kids love to have a variety of foods which are also good for their growth... But one may not afford to do that always . . . when they say 'shiro again?' . . . you feel something inside and try to do something." Children hear and see different things at school and the neighborhood, and based on that they demand their mother to do something similar. In this example the child dictates what needs to be packed: "Now that he sees different foods at school he says cook this and that . . . When I ask what to pack for school. He says: rice; macaroni, silsi with bread, egg sandwich." Some food retailors specifically target children, positioning their stores or vending stands alongside schools and thus tempting the children. When the children see those products, they start crying and pleading. "There is 'jelati' . . . chips, samosa, pasti ... there are many things sold especially at the doors of every school." This mother highlights, with the increasing street vendors around school, how difficult it is to adhere to the healthier home diet. Also, when the children see their friends having the street food, they want to act exactly like their friends. "When he sees children eating chips bought from the street, he also wants us to buy it for him. I tell him it will make him sick . . . and I will make it for him at home but he doesn't agree, he wants the one from the streets. The thing is you can't escape . . . they are everywhere and continuously tempting him." --- Quick-Fix Versus the Privilege of Planning Participants gave detailed accounts of the wide range of factors that influenced and constrained their decisions and choices in purchasing, cooking for and feeding their children. Mothers who were financially better off expressed a more planned style; they bought food items they needed in bulk, they opted for quality rather than quantity, and since they had space both for storage and cooking, they prepared most of their food at home. "We buy teff for a month . . . other ingredients like shiro and berbere for a year. Oil and the like we buy every three or four months." In contrast, other mothers said they were acutely aware that they were often unable to afford food and had to manage on a limited budget, devising different means to deal with this constraint. An approach that was mentioned is using sub-standard quality food to make sure the kids have something in their stomach: "I sometimes even consider collecting spoiled fruits and vegetables from people's house, even if I have to 'pour sweats down my ass' . . . I will do anything to bring home food for my kids." Another strategy followed by some mothers was to buy food from street venders that sell in small portions or to buy cheaper items as they encountered them rather than buying food in bulk and from their regular stores. Mothers also admitted to borrowing food items from their neighbors or taking store credits to cope with shortages as depicted in the example below: "The most common thing we do is, if I am baking injera my neighbor may come and take 1-2 injera, and I will do the same when I have shortage." Mothers in urban settings often express time constraint due to their engagement in economic activities or their social obligations. Thus, they sometimes tend to opt for quick-fix meals due to their convenience and ease of preparation and buying. "Packed food like cerifam . . . If I become very busy, I will give him by mixing it with boiled water, but I do this seldom." --- Keen Awareness on Food Safety, Nutrition, and Diet Diversity Almost all mothers who participated were well aware of the local nutritional recommendations; being particularly well informed about the importance of exclusive breast feeding, the correct timing to initiate complementary feeding, and the importance of balanced diet. They also agreed that nutrition knowledge influences their food choices: "She can't have other food, she is too small for anything . . . she is just 45 days old. She will not start anything till she is 6 months. Even with my older child, he was entirely on breast milk until he was 6 months. It is after 6 months that I started food for him." Despite the fact that mothers were aware of the dietary recommendation, for the most part there appeared to be a 'gap' between nutritional awareness in theory and putting this knowledge into practice. "I don't think that chocolate is good; first it damages teeth; and for future as others said, it causes diabetic mellitus . . . But to avoid his disturbance, I buy it for him... regardless of the consequences." In some instances, despite the mother's intention to provide diverse food for her family, due to her lack of in-depth knowledge on what counts as diverse, her efforts fell short of achieving what is best for the child. This mother indicated her idea of diversifying the food was to prepare stew with the same ingredient but changing the color and taste. "If we prepare potatoes and carrots as a yellow sauce, tomorrow we will prepare potatoes with red sauce." Mothers demonstrated an awareness of food safety. Most of the mothers indicated that safety concerns are one of the things they consider in making decisions on the food products to give their children. They indicated that they are not happy when their children eat street food or outside the home as they fear contamination due to unhygienic food handling practices. Environmental contamination, especially of the street food, is another concern they voiced. Despite these foods being cheaper to buy, mothers revealed they preferred to prepare them at home instead. "It's cleanliness . . . it might not be that clean they might be careless about their utensils... too much dust blows on it . . . the cleanliness of the environment is worrisome." The other main concern was related to adulteration and fear of expired products. Mother mentioned they took extra care to select untampered products: "I don't buy packed juice . . . I don't even like when adults drink it let alone children . . . because the expiry date may have passed . . . I heard traders import products passed their expiry date and sell it here by falsifying the expiry date . . . So I don't trust what is written on product." Mothers stated that they received dietary/food-related information from both informal and formal sources including schools, advertisements, and health workers . The informal sources included: word of mouth, peer, rumors, and media. "What I hear from media, from my friends, from my family about food and child feeding are important, I put in to practice what I hear from the different sources." The selection of food items for children was declared to also be influenced by the perceived health benefits or restrictions for different family members. If a family member has certain health condition that restricts him/her from eating certain food items, the family's food habit will also change since most families eat together, and thus the children's food consumption pattern could also be limited. " . . . my husband suffers from 'rhee' on his foot, therefore since it is not recommended for him to consume meat . . . the family also limits meat consumption to once a week." --- Social, Familial, and Cultural Influences Mothers expressed that despite the cost of food, special holiday food was prioritized. Mothers indicated they would do everything in their power to make sure their kids got a good holiday meal. Animal source foods were identified by the mothers as hard to incorporate in a routine diet but formed part of holiday meals. As this expression from a mother captures: " . . . animal source foods are not easy to find . . . we can't afford eating eggs, meat or chicken regularly . . . those are special holiday meals for our family." The religious background of the family also influenced the type of food the children ate. As there are religious restrictions among both Muslim and Christian followers, there are certain foods that are not allowed to be consumed at all times and certain foods during fasting times. "Eating ham is not allowed in our religion." "For breakfast we all eat together the same food, usually eggs or firfir except for Wednesdays and Fridays on which we don't eat any dairy or any animal source food." The selection of food was also shown to be dictated by the perceived social norm. Mothers' stated their desire to fit in and make sure their children identified with a given group and gained social approval. "I usually give my children the same breakfast as the other kids in the neighborhood . . . think that is customary with families around. Bread with tea/milk is said to be good for children . . . especially milk since it has lots of nutrients." This mother's expression also captures that social norms motivate the consumption of trendy foods which are more processed and not part of the staple diet: " . . . There is much difference. Currently you prepare and give mashed meat like sandwich, macaroni, previously you won't even get macaroni, let alone feed." Mothers also indicated that their familial roots had an influence on their food selection; in particular, some types of food were adapted from their ancestral roots. The intergenerational transmission of traditional foods is reflected by this mother's expression: "For example I am from Gurage ethnic group so sometimes I prepare kotcho with cottage cheese when it is not a fasting day... that is what I was used to in my upbringing." For most of the mothers, child-care and food preparation for the whole family was their responsibility, whether she was involved in additional income-generating activity or not. They considered all child-care responsibilities, from morning to evening, as the mothers' responsibility, and fathers were only responsible for income-generating activities. "I am responsible for all of the food purchasing and cooking . . . It is not convenient for him . . . He has to leave very early in the morning and comes back home quite late . . . his role is providing money." Though the gendered role of a mother as the preparer and feeder while fathers are the providers was typical for most women in our study; there were a few exceptions. For example: "On his days off, for example on Sundays he helps out with everything starting from babysitting . . . we help each other." --- Discussion This qualitative study found that mothers' decisions on what to feed their children is influenced by a variety of factors including the children's preference, their ability to plan for food, the information they have on nutrition,diet, and food safety and cultural/religious concerns, as well as changing societal and family values. These findings highlight the complexity involved in maternal decision making with regards to the food they gave their children. The data in the study supports the importance/compatible use of the conceptual framework of the food environment as a basis for understanding and explaining how mothers make food choices for their young children. Within each theme, different dimensions of the external and personal food environment are reflected; in the first theme the influence of child preferences in maternal decisions was evident. This is in line with other studies whereby children's taste, preference, and their hunger level [7] were identified to have an influence. Mothers in this study succumb to their children's preferences not just because of the persistent pressure from the children but also due to added pressure from advertisements targeting children and mothers. Previous studies have shown that adverts have an effect on young children's subsequent attempts to influence shopping toward advertised foods which in turn drive their eating and/or snacking habits [29,30]. However, there are some child-oriented innovative initiatives in developed countries that encourage kids to make more nutritious food choices and become more physically active [31]. In this regard, social marketing campaigns in developing countries like Ethiopia are lagging behind commercial marketing in creating appeal for a more healthier diet alternatives. With what appears to be the child in the driver's seat when making daily food choices, contextually suited innovative strategies are important to counteract the effects of advertisement and create a more discriminating audience that rejects processed foods and picks a healthy diet. A good example would be recreating/remarketing traditional foods into more peer-driven/modern versions, and school programs would be a natural choice for implementation. It could also be argued that changes in family size and structure could be possible explanations why mothers give in, first of all since the birth rates in the study town are below replacement level [32]. Children are considered precious and privileged, and thus everything is done for their delight. Secondly, since mothers often work full-time they neither have the time nor the energy to cook, especially something their kids/family do not like to eat. This dimension of convenience has been supported by studies from Singapore and India: mothers were overwhelmed due to limited time availability to prepare, customize, and feed their children appropriately and adequately [33,34]. Considering the emphasis put on mothers as the ultimate purveyors of nutrition to their children, it is essential to support her in her endeavours and that includes supporting her health and wellbeing. Affordability or financial/resource availability is the third dimension which was identified as one of the drivers for mothers' decision in choosing food for their children, those who were better off had the opportunity to plan for a healthier diet while those with scarcity had to strategize to avoid hunger. Studies from high-income countries support that the better-off have a better opportunity for a healthy diet, since healthy items are more costly than refined, sugary, and fatty foods [35,36]. The participants in our study were very creative in strategizing the procurement and preparation of food. The various strategies used by families to cope with short-and long-term changes in household income and food insufficiency in this study have also been documented in other studies; measures taken include compromising quality and quantity of food consumption [37,38]. Additionally, financial constraints also force mothers to opt for ready-to eat food like the instant noodles and to go for "modern" dishes like pasta, bread rather the traditional foods which require more resources to prepare. A review by Ruel and colleagues also highlighted that the food and fuel price increase during the 2007-2008 crisis forced households, especially for the urban poor, to increase consumption of street foods which can be a cheap source of energy and are also time savers [39]. The fear of incurring greater cooking fuel costs for the traditional preparation as opposed to the instant foods [40] contributes to the nutrition transition. Our participants were often knowledgeable about the child nutritional needs; however, we observed gaps between knowledge and practice: there was confusion about what constituted food diversity and they opted for giving sweets even though they know it is not good for the children. In a review, Worsley [41] argues that "nutrition knowledge" is a necessary but not sufficient factor for changes in consumers' food behaviors.The behavior itself is influenced by a number of environmental and intra-individual factors, including motivations. In our previous work findings highlighted that mothers were functioning in an overwhelmed and thus compromised emotional state [42]. Though one would expect that that mothers' emotional state would have a role in the interplay between motivational factors and information processing, studies weighing in on emotional aspects as a driver are lacking. Given that food choice and eating are already identified as highly complex and context-driven social practices [43] we need to consider additional dimensions like emotional status of the individual making these complex decisions in future studies. Food safety concern was one of the drivers of food choice. Food safety is a rising concern where poor agricultural, manufacturing, and hygiene practices contribute to a high rate of contamination with pesticide residues and heavy metal as well as microbial contamination [44,45]. Participants were almost unanimous in their declarations that all outside food is unhealthy. This was substantiated in a study from Ghana where respondents perceived food being offered in fast-food restaurants as unsafe and were thus concerned about several food safety issues [46]. Though food prepared in the house was perceived as healthy, studies are not widely available to determine whether traditional cooking techniques are indeed healthy. Herforth et al. [9] reflected that the term "desirability" as part of the food environment does not simply refer to preference, but rather to the external influence on preference like quality and sensory properties of food, knowledge, and norms of the consumers, in alignment with the findings of this study. Mothers' food decisions are reinforced, driven by, or stem from traditions originating from religious, rituals, cultural taboos, or what they have learned from elders and their families. Furthermore, in this study fathers' role was depicted as that of salaried work outside the house and as the main financial provider. There were some exceptions whereby fathers engaged and shared the responsibilities of child care and feeding, which may indicate traditional gender roles in transition. This finding is consistent with other findings from Ethiopia [47]. This study however focused on mothers, thereby curtailing deeper exploration into the role of fathers. Considering this as a limitation and the scarcity of literature on fathers and feeding, we would urge researchers to explore the role of fathers in child feeding. The findings of this study indicate the importance of understanding the broad food environment which is crucial to the development of policy interventions aimed at improving the food supply as well as any nutirition-related goals. This is especially true for cities in low-income countries similar to our study setting which have experienced fundamental changes in the food systems [48] and are encountering emerging nutritional concerns of obesity and non-communicable diseases. However, there is lack of research focusing on food environment in low-income countries. This paper contributes to bridge that gap in information. By focusing on a rapidly changing urban area, this study highlights the complexity of the setting as well as the unique challenges that urban mothers face as they try to meet their responsibilities in feeding their children in a changing and complex food environment. This could serve as an vital input in the design of effective context-specific interventions. This qualitative work also fills the methodological gaps and serves as foundational knowledge for further qualitative exploration and quantitative studies. Quantifying the level of influence each of the different factors have on the food decision making would be an important adjunct to design effective interventions. --- Reflexivity Though maximum efforts were made to minimize bias by training the data collectors to be sensitive, warm, and spend more time with the mother to create rapport, one of the limitations of this study is the possibility of social desirability bias as mothers would want to be perceived as "good mothers". They may not always opt to tell us what they actually do but what they think researchers want to hear. Despite that, mothers appeared to appreciate a chance to express themselves and verbalized a host of challenges they deal with in feeding their offspring. Throughout the research, issues related to trustworthiness were reflected upon by keeping an audit trail, having an iterative consulation with diverse research groups, and presenting results in different seminars. Mothers were selected from different districts and they represented diverse social environments. Additionally, Addis Ababa is one of the fastest growing cities in Africa, rendering these results as potentially transferable to other fast-growing cities on the continent, but consideration of how contexts differ remains at the discretion of the reader. In any case, it would be a good starting point for replication. --- Conclusions The findings of this study reveal a range of influences that mothers are faced with when making decisions on what to feed their children. Mothers living in rapidly urbanizing cities are pressured to make food choice decisions in contrary to their knowledge due to changes in the social practices, their roles at home, ever-present media advertisements, and rapid changes in the food environment. The extent to which each component influences mothers remains to be explored. Given that food choices and eating are highly complex and context-driven, future research should investigate the extent of these factors and how they interact to influence food choices made for children. For example, the extent to which financial constraints, cooking time and skills, infant acceptability, and food safety concerns contribute to the decision making process, particularly for lower-income parents, should be studied. Furthermore, the child-driven diet opens a whole range of possibilities for interventions that focus on the child as a purveyor of nutrition knowledge to the family versus the top-down approach. ---

Mothers carry the prime responsibility for childcare and feeding in low-income countries. Understanding their experiences in providing food for their children is paramount to informing efforts to improve the nutritional status of children. Such information is lacking in Sub-Saharan Africa. To understand what influences urban mothers' food acquisition and their motivations for selecting food for their children, 36 in-depth interviews were carried out with mothers having children under five years of age. Interviews were conducted in the local language, audio-recorded, transcribed, and translated into English. Data were analyzed using thematic analysis which led to the identification of four major themes: mothers give-in to a child-driven diet; quick-fix versus the privilege of planning; keen awareness on food safety, nutrition, and diet diversity; and social, familial, and cultural influences. The findings indicate that child feeding practices are influenced by interlinked social and environmental factors. Hence, nutrition education campaigns should focus on targeting not only families but also their children. Attention should also be given to food safety regulations, as well as to the much-needed support of mothers who are struggling to ensure their children's survival in low-income countries.

Introduction In recent years there has been much emphasis on strengthening the relationship between health system actors and community members. One such tool is social accountability, defined here as "citizens' efforts at ongoing meaningful collective engagement with public institutions for accountability in the provision of public goods [1]. " Scorecards, social audits, and participatory budgeting, popular social accountability tools, are used to facilitate processes that creates opportunities and spaces for those seeking services and those responsible for providing them to come together and mutually identify barriers and solutions to improve their services. Joy Moncrieffe has argued that social accountability is best understood as a relational process in which people's values, attitudes, and interactions change as a consequence of participating [2]. These changes in respective norms, values, attitudes, and relationships toward each other and themselves as a result of this dialectic interaction has been borne out in the evidence [3,4]. Recognizing the importance of values and attitudes on the delivery and performance of the health system is by no means new. We can see how values and attitudes affect health seeking behaviour -in the way community social norms surround women's decisions about health care [5], or in how providers' bias and beliefs about a service user affect how they counsel and treat them [6]. Moreover, the quality of people's interactions also affects the performance of the healthcare systems as reflected in the respectful care movement [7], and more recently in health system responsive literature [8]. These trends are part of the increasing recognition of "soft skills" or the "software" of the health system, that is the values, norms, attitude, communication skills and collaboration practices at work [9]. Others regard the performance of the health system as the product of interactions between system 'hardware' and the 'software' [10]. Supporting social accountability processes is one of several ways to change values and attitudes of both those using and those providing services. The affective effects of social accountability has been widely documented and captured as governance related outcomes such as increased community participation [3,11,12], and increased confidence among women to claim rights and make demands [13][14][15]. Other relevant changes include the responsiveness of duty-bearers [8,16], and their increased awareness of community needs [15,17,18]. In addition, the type and quality of interactions between duty-bearers and those claiming their rights has been acknowledged in the increased community engagement in decision-making [3,4], more meaningful providerservice user interaction [3,4], and in enhanced mutuality and trust [15,[19][20][21]. Yet these important changes in service users and providers' values and attitudes are often not measured because they pose several methodological challenges [18,22]. One notable exception is CARE's Women's Voice tool, a validated psychometric measure that captures these variables and provides an innovative starting point to capturing values, attitudes, and interactions [23]. As part of a more extensive complex intervention study on social accountability in the context of contraceptive services [24], the authors of this paper first adapted and validated CARE's measures of service users' attitudes and behaviours [25], and then used the measures to understand the processes by which change was affected. In the Community and Provider driven Social Accountability Project's theory of change these values, attitudes and interactions are set as intermediate outcomes along the casual pathway and are directly influenced by the social accountability process under study. In this paper, we share the findings about how service users' values, attitudes, and interactions changed as a result of a social accountability initiative in Ghana and Tanzania. --- Methods --- Study design This study was a quasi-experimental, pre-post evaluation cross sectional study of service user's values, attitudes, and interactions as part of the larger Community and Provider driven Social Accountability Project undertaken in Ghana and Tanzania. [24]. The CaPSAI Project contributes to the evidence on the effects of social accountability and participatory processes in the context of a family planning and contraceptive programmes. The study was designed according to the Medical Research Council guidance on complex interventions and was based on a theory of change using a co-designed intervention to account for the complexity of SA processes [24,[26][27][28]. This study accounted for the multiple components required to track the different levels and interrelated outcomes, including changes in values, attitudes and interactions that this paper focuses on [24]. The constructs assessed in this study are drawn from the CaPSAI Theory of Change in Fig. 1, these are the intermediate outcomes of expanding inclusive and effective negotiated spaces, empowered health workers, and empowered women and community members [24,25]. The theory of change was developed to link the intervention components and the intermediate and distal outcomes . The development and validation of the measures used in this pre-post evaluation study have been outlined elsewhere [25]. This study aimed to gauge change in the validated items at preand post-the intervention, and measure intermediate outcomes. The CaPSAI intervention is a co-designed social accountability process that is responsive to local contexts and practices [26,27]. In the community surrounding a facility, community members and health workers were separately trained by civil society organisations in their rights and entitlements and each group is then taken through a process by the civil society organisations to generate issues and then rank what they consider to be a priority. The civil society organisation leading the process then bring together the two groups, community members and health workers, in a facilitated interface meeting to jointly identify shared issues and solutions to address them. The community members and health workers together then met with local government officials to discuss their concerns and priorities and make action plans to respond to the claims made. These action plans were then followed up and reported on over a 6-month period. The social accountability process was composed of nine steps led by a national civil society partner over a sixmonth period. Each step and its purpose are outlined in Table 1. A full description the intervention has published in the WHO Community and provider-driven social accountability intervention for family planning and contraceptive service provision: experiences from the field [27]. --- Participant selection and enrolment Possible study participants were identified in eight intervention facilities in each country during consultations by health providers who referred them to study staff. Screening was done by the study staff using the eligibility criteria that included age who were attending the FP services of health facilities involved in a participating study facility, were first-time or continuing users contraception and provided consent to participate. Following the consent or assent process, the interviews were conducted in person at the facility, or an appointment was set later. The same eligibility criteria for participants were used in all intervention sites. --- Data collection A sample of over 750 women aged 15 to 49 years accessing contraceptive services was interviewed prior to the start of the intervention and an additional 750 women aged 15 to 49 years after the intervention in each country. Sampling for the service users survey was calculated using a priori sample size calculation with the ratio of ten subjects per item ratio and guidance of more than 500 which equals a very good sample for validation [25,26,29,30]. The calculation was based on 75 items. The preand post-survey were taken eight months apart. A total of 118 questions were asked of respondents upon leaving a facility, and only 44 scale items and 11 domains were included in the following analysis. The other items included questions about demographic status, reproductive and family planning history, relationship status, income, occupation, and religion. The 11 domains measured were driven by the theory of change, which was informed by existing empirical and theoretical work on social accountability. After defining the domains, we identified existing validated measures for each domain. We drew on CARE's Women VOICES tool, a validated measure the aimed to capture similar intermediate outcomes concerning maternal health in Malawi [23]. In addition, we included three domains to represent the CaPSAI theory of change -namely 'knowledge and awareness of rights, self-efficacy and political capabilities. The items were validated through consultations with experts and with World Health Organization Forms Committee. The instruments were psychometrically tested at baseline in Ghana and Tanzania and the findings were published [23]. Table 2 outlines the domains that were included in the final analysis. Post-intervention survey was conducted at the facility in a private location. In Ghana, a total of 15 data collectors were trained over a 3-day training workshop from 4 to 6 April 2018 and a 2-day refresher from July 9-10 2019, respectively. In Tanzania, a total of 15 data collectors were trained over the survey over a period of 5 days from 19 to 23 March in 2018 and a 5-day refresher training was conducted from 29 to 2019 to 2 August 2019. Data collection was conducted using a tablet-based questionnaire to capture real-time data using OpenClinica and was later uploaded onto a secure server. In Tanzania, the first survey data collection started on 26 March and was completed on 25 May 2018. The second survey took place from 2 to 2019 to 3 October 2019. All respondents chose to be interviewed in Kiswahili. In Ghana, data collection for the first survey started on 9 April 2018 and was completed on 4 June 2018, and 46.4% chose to be interviewed in English, while 53.6% chose to be interviewed in Akan. The second survey took place between 10 and 2019 and 10 October 2019, and 54.5% chose to be interviewed in English, while 45.5% chose to be interviewed in Akan. --- Patient and public involvement It was not appropriate or possible to involve patients or the public in the design, or conduct, or reporting, or dissemination plans of our research. --- Ethical consideration Once assessed as eligible, respondents completed the informed consent process. There were no incentives given to women and girls to participate in the study. However, study participants who agreed to participate were reimbursed for their travel cost, where it was permitted by country-specific ethical requirements. In Ghana, the research team supported the travel cost to the facility with five Ghana cedis given after the interview. In Tanzania, no reimbursements were given. For adolescent participants identified to participate in the study, research staff explained what the Table 1 Steps of the social accountability intervention [27] Step --- Title Purpose and description Step one Preparatory work To prepare the necessary institutional and social permissions and approvals required to implement the social accountability process in a community. --- Step two --- Introduction to local authorities and gatekeepers To ensure that local authorities and gatekeepers are both aware and supportive of the process. Step three --- Mobilization and introduction of the project to the community members To spread awareness of the project among the local community and ensure inclusive participation from a wide range of community members. Step four --- Health, rights and civic education with community members To share information about sexual and reproductive health and rights and entitlements with community members. To work together to explore any perceived gaps or shortcomings in the services they receive. To generate discussion on the local issues that faced by the population. Step five --- Health, rights and standards of care sensitization with health actors To share information about srhr and entitlements with health providers. To work together to explore any perceived gaps or shortcomings in the services they receive. To generate discussion on key issues faced by local people. --- Step six Prioritization meeting with the community To ensure that a diverse range of community members identify and rank the most pressing issues related to fp/c information and services in their community. --- Step seven Prioritization with health actors To ensure that health care providers identify and rank the most pressing issues related to fp/c information and services in their community. Step eight --- Interface meeting and joint action planning To share the assessments separately generated by community and health actors and then to jointly identify areas for improvement. To develop an action plan to ensure concrete measures are taken to improve services and/or maintain good practices. Step nine --- Regular ongoing monitoring and follow-up To track if progress has been made in the jointly-agreed action plan, by regularly following up with both the community and health authorities. To present an opportunity to involve high-level duty bearers or third parties in addressing unresolved issues. --- Statistical analysis To detect statistically significant differences in the distribution of sociodemographic characteristics between the groups at pre-intervention and post-intervention in each country, we used a chi-square or a Fisher's exact test for categorical variables; and a t-test for comparison of continuous variables. For comparison of continuous pre-and post-intervention intermediate outcomes a t-test was used. We conducted a two-sided test, with type 1 error at 5% level. Statistical Analysis System Version 9.4 was used for the statistical analyses. --- Results --- Demographic characteristics The demographic characteristics of the study population were similar in the pre-and post-intervention samples, see Table 3 for each country. In Ghana, most women were 21 to 35 years of age, 70.1% in the pre-group and 72.5% in the post-group, respectively. In Tanzania, the 21 to 35 age group also made up the majority of the study population, 75.6% in the pre-group and 75.7% in postgroup, respectively. There were significant differences in the percentage of women who have completed primary education in Ghana, with 25.9% in the pre-group and 14.4% in the post-group, respectively. There was no difference in the percentage of the study population who had completed secondary education, 9.2% in the pregroup and 10.8% in the post-group, respectively. There were no differences in the percentage of women who have completed primary education in Tanzania, with 56.2% in the pre-group and 53.2% in the post-group, respectively. There was no difference in the percentage of the study population who had completed secondary education, 20.9% in the pre-group and 22.6% in the postgroup, respectively. In both Ghana and Tanzania, there were significant differences in marital status between the pre-and post-intervention groups. In Ghana, there was a significant difference in the percentage of injectable users in the pre-and post-group. In Tanzania, there was a significant difference in the percentage of male condom users, standard days method, Lactational Amenorrhea Method, rhythm method and withdrawal in the pre-and post-group. --- Changes in service user's values, attitudes, and interactions in Ghana In Ghana, there were several statistically significant changes found, see Table 4. There was an increase in the women's participation in household decision-making after the intervention and there was a positive change in how people perceived they were treated by the providers. However, the service users' knowledge of health rights declined, as did their perception of quality. Service users also noted a decline in one's ability to attend and to participate in community meetings. There were highly significant declines in awareness of accountability mechanisms among respondents and in the collective efficacy found in the community over the course of the study. --- Changes in service users values, attitudes and interactions in Tanzania In Tanzania, there were several statistically significant changes, see Table 5. Over the course of the study, women's participation in household decision-making was also seen to improve. Service users also perceived that they were treated better by health workers. There were also improvements in the ability to attend and participate in community meetings and in mutual responsibility for and support of services. However, women's knowledge of their health rights, the perception of service quality, their knowledge of accountability mechanisms to make claims and the sense of collective action to bring about change were all seen to decline over the study period. --- Comparing the findings across the two settings Table 6 compares the findings across the two settings. In both settings, there were significant improvements in women's participation and in household decisionmaking perceived treatment by health workers. In both settings there was a decline in women's right knowledge, the perception of service quality, awareness of accountability mechanisms and collective efficacy in community. The results for several of the constructs were different between the two countries, namely the ability to attend and participate in community meetings, mutual --- Discussion The CaPSAI theory of change suggests that changes in the values, attitudes and interactions of the community and health providers are central part of the social accountability process leading to health-related outcomes. When comparing the results within and between countries, there are mixed results. There were only two domains where there were significant positive changes, and in five domains there were significant negative changes in the study period. This is different to the study in Malawi that used similar intermediate outcomes and reported changes in 7 out of the 13 constructs assessed [31]. There were a higher number of positive changes in Malawi compared to the findings of this study. In Malawi, there were 3-4 cycles of the social accountability process completed by the endline, whereas in the present study, only one cycle of the social accountability process was completed over a six-month period and, therefore, there was a lower level of exposure. Bearing in mind the shorter time and limited exposure of the CaPSAI intervention, there are two significant positive trends across both countries, namely increased participation of women in household decision-making and decreased perceived mistreatment by health workers when visiting the contraceptive services, though this was not significant. These findings align with related research in this area, it is well-established that community group engagement can improve women's decision-making power and consequently, women's ability to make and act on decisions is linked to contraceptive use [32,33]. There is also evidence that women's self-help groups have had positive impacts on women's empowerment, social support, health service use and outcomes [34,35]. In addition, in both sites, there was a significant decrease in the mistreatment by health workers perceived by service users, service users reported less disrespect and abuse. This suggests that the intervention is having an effect on some of the attitudes and interactions of both service users and services providers in a relatively short period of time. There were more significant positive changes in the Tanzanian context. In each country, the national partners adapted the eight-step process to respond to the local context and practices [26]. This resulted in social accountability interventions taking different forms but retaining conceptual fidelity [26]. In the Tanzania model, nominated community members were trained in social accountability and remained present and active throughout the intervention period. The trained community members led the monitoring teams alongside the health staff and local health authorities to assess progress against the action plan, reporting back to both the communities who elected them as well as the local authorities who had been engaged. This may have created a more localised and sustained social accountability process that can be seen to positively affect the more interaction-related constructs. We also found negative results on several of the main outcomes, namely knowledge of health rights, quality of services, awareness of accountability mechanisms, and collective efficacy. Several possible explanations can account for this. Interpreting these results is complex and does not necessarily mean an adverse outcome. One explanation of these negative/ insignificant effects on some of the outcomes at post intervention could be attributed to either selection bias as observed through differential distribution of some outcomes between women studied at these two time points or possible intervention effects. We did not account for confounding variables in a formal regression model, and this limits our ability to explain some of these results. Another possible explanation of these results is that there has been a failure of the intervention, for example, the group at the facility that was sampled were not exposed to the intervention or there were conflicting interests between individuals that discourage joint action and collective efficacy. This could be a limitation in our theory of change, these intermediate outcomes may take a longer time to build up and could not be achieved with one cycle of the intervention. This study had several limitations. Firstly, the intervention occurred at the community level to affect change at the facility, and the measures were taken at the facility. With this sampling strategy, it is not possible to capture those participants who attended other facilities not sampled, particularly private sector providers. Research has clearly shown the extensive role of the private sector in providing family planning and 35% of women are using private sector services in the Sub-Saharan Africa [36]. Secondly, some of the constructs examined relate to long-entrenched power relationships that can require a longer timeframe to observe changes [31]. It is worth repeating that this study could have benefitted from at least an additional cycle before the endline. Finally, there are limitations of a pre/post design, especially given that some characteristics were different between pre-and post-groups. With this study design, it is not possible to assess if the differences at endline could be attributable to selection bias or systematic differences between the groups . This possibility cannot be ruled out. --- Conclusion The CaPSAI intervention set out to change the values, attitudes and interactions between the community and those providing contraceptive services. Sugh outcomes related to knowledge , whereas others related to attitudes and others were experiences . There were changes in different directions, and this suggest that we need to further examine how changes take hold and bring about certain outcomes in different timeframes. --- Data Availability The data that support the findings of this study are available from the World Health Organisation but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the World Health Organisation. Please contact Petrus Steyn: [email protected]. --- Authors would also like to acknowledge the support of the Chief Executive Director of Ifakara Health Institute, Honorati Masanja, the CaPSAI data manager in Tanzania, Selemani Mmbaga and the IHI administrative team . The authors would also like to acknowledge CaPSAI Project implementation lead, Victoria Boydell and implementation team members, Mary Awelana Addah and Osei-Bonsu Gyamfi from Ghana Integrity Initiative, Ghana; Alice Monyo, Happy Myovela and Joshua Nkila from Sikika, Tanzania. Implementation support was provided by Heather McMullen and Vernon Mochache . Additional analytical input was provided by Soe Soe Thwin and data management support was provided by Antonio-Lucio Fersurella . Additional conceptual input on the CaPSAI protocol was provided by Ian Askew and Karen Hardee . The CaPSAI Project team would also like to acknowledge the support of Roseline Doe and Edwin Swai from the WHO country offices in Ghana and Tanzania, respectively. --- --- Declarations Ethical approval and consent to participate All methods were carried out in accordance with relevant guidelines and regulations. CaPSAI Project master and country protocols were approved by technical and ethics review committees at the World Health Organization . Additionally, the country protocols were reviewed and approved by the Population Council Institutional Review Board and Ghana Health Service Ethics Review Committee in Ghana. In Tanzania, the protocol has been approved by Ifakara Health Institute Institutional Review Board and the National Institute of Medical Research review board , as well as the NIMR/Mbeya Medical Research and Ethics Review Committee . Informed consent was obtained from all subjects and/ or their legal guardian. --- --- Competing interests The authors have no competing interests to declare. ---

This study evaluated the effects of community engagement through social accountability on service users' values, attitudes and interactions. We conducted a pre-post study of the community and provider driven social accountability intervention (CaPSAI) over a 12-month period among 1,500 service users in 8 health facilites in Ghana and in Tanzania (n = 3,000). In both countries, there were significant improvements in women's participation in household decision-making and in how service users' perceive their treatment by health workers. In both settings, however, there was a decline in women's knowledge of rights, perception of service quality, awareness of accountability mechanisms and collective efficacy in the community. Though CaPSAI intervention set out to change the values, attitudes, and interactions between community members and those providing contraceptive services, there were changes in different directions that require closer examination.

INTRODUCTION P rimary care is a core tenet of health care in developed nations, and is widely regarded as a natural nexus for coordinating clinical care, public health, and community-based services. 1,2 Widening health disparities continue to plague society, however, despite attention to integration of services across the health care spectrum. In recent years, there has been increasing recognition of the nonclinical factors affecting health, namely, the social determinants of health. 1,3 The World Health Organization describes the social determinants of health as the "conditions in which people are born, grow, work, live and age, and the wider set of forces and systems shaping the conditions of daily life." 4 Social determinants can present a major challenge to delivering effective primary care, because they often compound the complexities in addressing patients' health concerns. 5,6 This issue is recognized by primary care professionals, as 85% of respondents in a survey of 1,000 physicians found social determinants of health to play as important a role as medical needs, but many lacked the confidence to address those social needs and believed their patients' health suffered as a result. 7 This impression is supported by literature showing the social determinants of health to be associated with elevated rates of morbidity, mortality, and other adverse health outcomes. [8][9][10][11] Expanding the role of primary care professionals in addressing the social determinants of health is feasible with initiatives that invest upstream and expand the health care team. 12 In the Unites States, health systems increasingly emphasize social determinants through bundled payments, capitation funding models, and, more recently, penalties to hospitals for 30-day readmissions. 13 Efforts to transform primary care delivery, such as placing more emphasis on components of the patient-centered medical home model, 14 have increased the capacity of primary care teams to address patients' needs in a broader interdisciplinary context. These initiatives should address concerns that primary care professionals are not compensated for incorporating the social determinants of health into practice, or that they would be operating "out of scope." Quality primary care is associated with attenuation of the harmful health effects of income inequalities, 1,15 making primary care a highly appropriate environment for measuring and intervening on social determinants of health. High-quality data are critically important for identifying underlying challenges with social determinants of health and for evaluating interventions designed to mitigate them; however, individual-and practice-level data on social determinants of health are rarely available within primary care settings, and information on the best way to collect and use these data has not been widely established. At the Manitoba Centre for Health Policy , we have a unique resource for investigating the association of social determinants and primary care in the Manitoba Population Research Data Repository, which holds more than 30 years of linkable individual-level administrative data from health and social sectors on nearly the entire population of the Canadian province . To better inform strategies addressing social determinants of health in primary care, we defined 11 social complexity factors using population-based administrative data and examined the association of these factors with the quality of care received by primary care patients. We hypothesized that, because of the multilayered nature of addressing socially complex patients' health concerns, these patients would receive poorerquality care than patients living without social complexity factors. We anticipate that our findings will provide insight into the social needs of primary care populations, and support the development of focused interventions to address social complexity in primary care settings. --- METHODS This cohort study examined the association of social complexity with quality of care received by primary care patients in Winnipeg, Manitoba, Canada. The study was approved by the University of Manitoba Health Research Ethics Board and the Health Information Privacy Committee at Manitoba Health. --- Study Setting and Participants Our cohort included all 627,047 Manitoba residents who made at least 3 ambulatory visits to a primary care clinician based in the major urban center of Winnipeg between April 1, 2010 and March 31, 2013. Under Manitoba's universal health care system, patients have free choice of the primary care clinician they visit. We excluded those living in nursing homes and those not living in Manitoba for the entire study period , for a total of 626,264 patients in the study cohort. --- Data Sources The administrative data used to identify primary care patients and define social complexity factors came from 16 health and social data sets in the Manitoba Population Research Data Repository housed at the MCHP . These data are routinely collected during the administration of the universal health care system, social services, and the education and justice systems in the province. The repository includes a health registry of almost all provincial births . Military personnel and their families, members of the Royal Canadian Mounted Police, and Indigenous groups are federally insured, and their inclusion in the databases is variable depending on the source. The deidentified data can be linked at the individual level to conduct research within the secure MCHP environment using scrambled numeric identifiers. Data in the repository have been extensively documented and validated for this type of research. [16][17][18][19] A detailed description of the data quality assessment and linkage procedures at MCHP is available elsewhere. 20,21 --- Social Complexity Factors We defined a set of social complexity factors related to housing, income, mental health, newcomer status, family structure, and involvement with the justice system. Selection criteria were based on literature that shows these social complexity factors are linked to poor physical and mental health outcomes. [22][23][24][25][26][27] We also drew heavily on the work of Schaink et al, 6 who identified social determinants of health in the psychosocial dimension of patient complexity of care. Ultimately, we assessed 11 social complexity factors: high residential mobility ; lowincome quintile ; social housing resident ; income assistance ; major mental health diagnosis ; newcomer ; child of a newcomer ; teen mother ; child of a teen mother ; in care of Child Welfare Services ; and involved with justice system . --- Main Outcome: Quality of Primary Care We used a total of 26 primary care quality indicators to assess the care that patients received. Previous research has identified these indicators as important measures of quality of care in family practice, 28 and demonstrated that they can be assessed using administrative data. [29][30][31] Three of the quality indicators were developed at MCHP for this study based on current literature highlighting their relevance to high-quality primary care. [32][33][34] --- Statistical Analyses We used generalized linear mixed models to assess the association of number of social complexity factors with primary care quality. These models accounted for the clustering of patients by physician and the resulting correlated observations. Using a logit link and binary distribution, the models were multilevel logistic regression analyses that produced robust odds ratios comparing individuals with 1 to 5 or more social complexity factors against those with none. Contrast statements were included in the regression models to assess the linear trend of number of social complexity factors with the odds of the outcome. All models included sex, age, income, and morbidity level as covariates. Morbidity was assessed using resource utilization bands , which are derived from the Johns Hopkins Adjusted Clinical Group Case-Mix System. 35 RUBs are a simplified ranking system of a person's overall morbidity from 0 to 5, taking into account all diagnoses attributed to him/her during medical visits and hospitalizations in the preceding year. Diagnoses are grouped together on the basis of the level of resources necessary for delivering health care for those conditions. These categories include minor illnesses ; more severe but time-limited conditions ; chronic illnesses ; illnesses resulting from structural problems that are generally not curable ; and psychosocial conditions . 35 To address multiple comparisons, we used an α of .01 to assess statistical significance. As many quality outcomes correlated with each other, using strict Bonferroni correction would have been too conservative. All analyses were hypothesis driven and prespecified. Data management, programming, and analyses were conducted using SAS version 9.3 . --- RESULTS Characteristics of our study cohort are presented in Table 1. The final cohort consisted of 626,264 Manitobans . As expected, patients were evenly distributed across income quintiles, and the majority had moderate morbidity levels . Figure 1 shows the prevalence of the 11 social complexity factors in the study cohort and the distribution of patients according to number of factors. Slightly more than one-half of the study cohort were living with at least 1 social complexity factor; 4% were living with 5 or more of them. Associations between the number of social complexity factors and indicators of the quality of primary care are presented in Table 2. We observed several highly significant associations. When patients living without social complexity factors were the comparator group, there was a downward trend in the odds of receiving appropriate prevention and screening care as number of social complexity factors increased. In other words, the more factors a patient was living with, the less likely he or she was to receive, for example, a mammogram or vaccinations. Conversely, there was an upward trend in patients' odds of using health services as their number of social complexity factors increased, relative to patients living without any. Stated another way, the more factors a patient was living with, the more likely he or she was to seek ambulatory or emergency care. Even patients living with only 1 social complexity factor had a sizable reduction in the odds of breast cancer screening and a substantial increase in the odds of ambulatory visits . Not shown in Table 2 is continuity of care, which we assessed as a continuous measure using the Bice-Boxerman continuity of care index. 36 The index measures both the frequency of ambulatory visits and the dispersion of visits among different clinicians. Values range from 0 to 1, where 0 indicates that all visits were made to different clinicians, and 1 indicates that all visits were made to a single clinician. In our study, patients with no social complexity factors had an average index of 0.53 , and the index decreased by an average of about 0.01 with each addi- a Compared with patients having no social complexity factors. b For trend across factor categories. c On a scale of 1 to 5, with 5 being lowest acuity. tional social complexity factor. For example, the continuity of care index was 0.52 with 1 social complexity factor, 0.51 with 2 factors, and 0.50 with 3 factors, and this linear trend was significant . Compared with peers having none, patients having more social complexity factors also had worse outcomes for some of the chronic disease management indicators, such as diabetes management and congestive heart failure management . Indicators of medical care for community-dwelling older adults, including prescribing of benzodiazepine and Beers List drugs , also showed an upward-trending association with number of social complexity factors. Even patients living with only a single social complexity factor had a sizable reduction in the odds of receiving an eye examination as part of their diabetes management and a sizable increase in the odds of benzodiazepine prescriptions among geriatric patients . --- DISCUSSION This study demonstrates how linking secondary data from the health and social sectors can provide insight into the underlying role of social determinants in influencing health care delivery and patient outcomes. Our findings show the pervasive nature of the social challenges faced by the primary care patient population; slightly more than one-half of our study cohort were living with some degree of social complexity, and this disadvantage had a nearly universal negative impact on the quality of care they received. For the majority of the 26 primary care quality indicators we measured, an increasing number of social complexity factors was associated with poorer outcomes. --- Findings in Context Recent health system reform in Canada and the United States has recognized the importance of addressing the needs of complex, high-needs patients. Although the primary care quality indicators we used in our study do not specifically address the 5% of the population who have been shown to account for 60% or more of health care costs, [37][38][39] we have developed administrative data definitions to identify a population of patients who are likely to contribute disproportionately to health care costs and primary care professional workload. By defining the specific social complexities that affect the health of this population, our study provides evidence underscoring the importance of developing and using interventions to address future health care needs. The imperative for addressing these issues goes beyond the financial costs of care, as social justice dictates that society take action on these glaring health disparities. 40,41 Access to high-quality individual-level or practicelevel data on the social determinants of health is critical for research focused on health inequities in primary care. A number of recent studies have demonstrated the use of primary and secondary data in this context. Administering brief surveys in primary care settings to flag patients' key social complexity factors has been shown to be feasible and acceptable to patients and clinicians, 42,43 and has spurred the development of interventions to link patients with resources and services in their communities. 44,45 The necessary integration of these data into electronic medical records is in many cases still too limited to provide contextual information for clinical decision making and the organization of care, however. 1,42 Analyses of secondary socioeconomic status data, such as those available in the Manitoba Population Research Data Repository at MCHP, can mitigate some of the cost of, and time required for, primary data collection. This repository makes large representative patient cohorts available for research purposes, granting us the ability to develop definitions for the social determinants of health and measure new indicators of primary care quality. The individuallevel data and cross-sector linkage capabilities of the repository allow for analyses of unparalleled depth and breadth. Although data collected for administrative purposes are not always research ready, the data in MCHP's repository undergo rigorous quality assessment to minimize biases related to misclassification or missing data. 20 In jurisdictions where administrative data repositories lack the breadth and depth needed for holistic analyses of social determinants, using geospatial technologies to link neighborhood-level socioeconomic data to patient addresses in electronic medical records may prove valuable. 46,47 --- Limitations Despite the many strengths of this study and the administrative data on which it is based, there are some limitations that must be acknowledged. Although income quintiles are a widely accepted measure of socioeconomic status, one consequence of using the first quintile to define people as having low income is that 20% of the study cohort falls into this category by definition. We found, however, that this proportion was comparable to another widely used and validated measure of poverty, the low-income cutoff, a threshold below which families devote a larger share of their income to household necessities. 48 There are a number of behaviors related to poor health, such as substance misuse, that would likely have been important indicators of social complexity in primary care. Although we did not define substance misuse as a social complexity factor , health behaviors are accounted for in the RUB morbidity measure insofar as they are captured by clinical diagnosis codes. Likewise, variables describing access to care, including transportation barriers and hours of operation of primary care clinics, were not available to us in the administrative data. Additionally, this study examines quality of primary care in an urban population; although the findings are widely generalizable to other urban settings, 17 further research is required to examine whether a similar relationship between social complexity and quality of primary care exists in rural and remote communities. --- Practice and Policy Implications Our findings highlight striking differences in social complexity across the primary care patient population and show the higher demands placed on primary care clinicians caring for socially complex patients. We have also identified the specific social needs of primary care patients. The potential for ensuring patients receive care in their home practices now exists, thereby making it possible to identify specific interdisciplinary team members best suited to individual practices. Expanding interdisciplinary team-based models of care and exploring alternative funding models that acknowledge the greater complexity of addressing the social determinants of health in the primary care setting could help to achieve better health equity for vulnerable patient populations. Funding support: This study was funded as part of a 5-year contract between the Manitoba Centre for Health Policy and the Government of Manitoba. --- Disclaimer: The funders had no input into the study design, implementation, or interpretation of the findings. The conclusions are those of the authors alone, and no official endorsement by the funders was intended or should be inferred.

PURPOSE In primary care, there is increasing recognition of the difficulty of treating patients' immediate health concerns when their overall well-being is shaped by underlying social determinants of health. We assessed the association of social complexity factors with the quality of care patients received in primary care settings. METHODS Eleven social complexity factors were defined using administrative data on poverty, mental health, newcomer status, and justice system involvement from the Manitoba Population Research Data Repository. We measured the distribution of these factors among primary care patients who made at least 3 visits during 2010-2013 to clinicians in Manitoba, Canada. Using generalized linear mixed modeling, we measured 26 primary care indicators to compare the quality of care received by patients with 0 to 5 or more social complexity factors. RESULTS Among 626,264 primary care patients, 54% were living with at least 1 social complexity factor, and 4% were living with 5 or more. Social complexity factors were strongly associated with poorer outcomes with respect to primary care indicators for prevention (eg, breast cancer screening; odds ratio [OR] = 0.77; 99% CI, 0.73-0.81), chronic disease management (eg, diabetes management; OR = 0.86; 99% CI, 0.79-0.92), geriatric care (eg, benzodiazepine prescriptions; OR = 1.63; 99% CI, 1.48-1.80), and use of health services (eg, ambulatory visits; OR = 1.09; 99% CI, 1.08-1.09). CONCLUSIONS Linking health and social data demonstrates how social determinants are associated with primary care service provision. Our findings provide insight into the social needs of primary care populations, and may support the development of focused interventions to address social complexity in primary care.

Background The South African burden of disease reveals an extremely violent society with the highest reported intimate femicide rate [1,2]. Most South African social contexts are characterised by oppression of women. At a fundamental level, disrespect for the feminine seems validated by cultural norms and values which prioritise males over females in multiple ways. This normative framework impacts negatively on the quality of people's relationships and their self-expectations. Despite a progressive constitution and legislation, IPV is still regarded as culturally acceptable, and thus, in many contexts, is normalised [3]. The health consequences of intimate partner violence can be categorised as fatal and non-fatal [4]. Fatal outcomes include femicide, suicide, maternal mortality, antepartum haemorrhage, abortion, stillbirth and AIDS. Non-fatal consequences comprise burns, fractures, chronic pain syndromes and mental illness, problems with hearing and sight, arthritis, seizures, headaches, sexually transmitted infections , HIV, and pelvic inflammatory disease. Indirect consequences of IPV include stomach ulcers and other gastrointestinal disturbances, heart disease, hypertension, unwanted pregnancy, low birth weight and premature labor [5]. Indirect health consequences also extend to abusers compromising their partners' health by withholding medication, changing a prescription, cancelling appointments, or keeping partners awake [6]. The impact on a woman of even a single incident of physical violence in an intimate relationship should not be underestimated. Use of any violence in a relationship can dramatically alter the balance of power, destroying respect, openness and trust and resulting in a permanent sense of inequality, threat and loss [7]. Recent findings of the World Health Organisation's multi-country study on women's health and IPV suggest that the mental effects of violence last long after the violent episode. Moreover, cumulative abuse impacts powerfully on health [4]. International research on IPV undertaken within mental health care institutions and settings is scarce. One study set in New Zealand, England and the United States found that over 50% of women traumatised by IPV suffered a psychiatric disorder. Most notable were the elevated rates of mood and eating disorders [8]. Similarly, in an English primary health care context IPV showed a strong association with most mental health conditions, particularly if experienced during the preceding year [9]. Most research has focused on the nature and prevalence of IPV, but relatively little has been published on interventions or models for care. Baldwin-Ragaven testifies to the proliferation of peer-reviewed articles measuring the problem and documenting the consequences of our failure to act, commenting: "For any other disease process as costly in financial and human measures we would demand answers, find cures, and disseminate evidence about interventions. What is it about IPV?" [10]. There is a clear moral argument that health providers should attend to the problematic impact of IPV on health. Women experiencing IPV present to all health care settings, usually without naming the IPV problem. There is evidence that women appreciate inquiry about IPV and can benefit from intervention [11][12][13]. There are no existing guidelines for the management of IPV in primary care that have been operationalised in the South African setting. Currently IPV is largely unrecognised by primary care providers and in the few cases that are diagnosed the standard of care is fragmented, poorly coordinated, lacking in continuity and missing important aspects. The best attempt to address this issue has been seen in the PC101 guideline that was developed for nurses, where an approach to the abused patient is included under mental health [14]. This guideline is currently being evaluated in the Western Cape. Our study aimed to implement and adapt the first published protocol on screening for and management of IPV in the South African primary care setting and to recommend a model of care for IPV. At the time of this study this was the only protocol that was on offer for testing in clinical practice. --- Methods --- Study design Professional action research is ideal for innovating alternative health service delivery systems [15]. It is one of four typologies of action research [16]. Professionals work collectively on a problem identified from their practice with the aim of improving practice and conceptualising their learning. Such professional action research closes the gap between theory and practice, while learning remains highly contextualised. A contextually appropriate application of this is the co-operative inquiry group method, which was adapted by Mash and Meulenberg-Buskens to develop medical education materials, and has been the overarching action research methodology utilised in this project [17]. A co-operative inquiry group implemented a published South African protocol for the screening and management of women experiencing IPV [18]. The underlying assumption was that participants in the inquiry group would create new knowledge from their concrete experience; by observing and reflecting thereupon; by forming abstract concepts and generalisations; and by testing the implications of these concepts in new situations. The co-operative inquiry group, which included the authors, worked with the standard four-step action research cycle: planning, action, observation and reflection [17]. --- Setting Primary care is offered through a network of community health centres and clinics. Community health centres provide services through larger multidisciplinary teams in more urban areas and first contact could be with a nurse or a doctor. Health centres may provide a range of services including mental health, family planning, services for STIs and HIV, maternal and child health, orthopaedics, dental and emergency care. Fixed and mobile clinics are smaller nurse-run services and are particularly common in rural areas. At fixed clinics doctors usually consult on a weekly basis and are employed to support the nurses and see more complicated patients. In order to test the protocol in rural and urban settings, two urban and three rural facilities, serving historically disadvantaged communities free of charge, were purposively selected in the Western Cape. The two urban facilities are health centres, while the three rural facilities comprise one health centre and two smaller clinics. They were selected for being representative of primary care facilities in the region, having mental health and other services as required by the protocol, a sufficient workload to ensure enough participants, and a private room. Site A is situated in a formerly designated 'black group area' and Site B in a former 'coloured group area'. Both are urban health centres serving economically disadvantaged residents who are reliant on these health services. Each facility serves approximately 400 patients a day. All general primary health care services are provided by a team of health care providers , a social worker and a psychiatric nurse. Both sites offer preventative services in the form of family planning, immunisation, voluntary counselling and testing for HIV and tuberculosis clinics. After-hours services comprise the trauma unit at Site A and the adjacent maternity facility at Site B. Site C is situated in a former 'coloured group area' in the main town of the Cape Winelands district adjacent to a provincial hospital. This community health centre serves the town and surrounding farming district. Unusually dedicated practice teams offer continuity of care, familyoriented care and the integration of chronic and acute patients. The practice teams depend on effective collaboration between the clinical nurse practitioners and doctors [19]. Sites D and E are clinics that lie just outside of a rural town in the Witzenberg sub-district. Agriculture, predominantly fruit farming, is the main local industry and consequently many of the study participants were farm workers. Site D is in a former 'coloured group area' and Site E in a former 'black group area'. Both are underserviced islands away from the hub, designed to supply labour for town residents and farmers. --- Formation of the co-operative inquiry group The co-operative inquiry group consisted of five people, although participation varied according to availability. The first author engaged fully with the implementation of the protocol. Facilitation of the group process was primarily the responsibility of the second author, who did not engage with the implementation of the protocol. Four members were engaged with implementing the protocol. Members who implemented the protocol were chosen for being registered professional nurses with relevant language skills and interest in IPV. They are referred to below as the study nurses. --- The protocol for IPV management Comprehensive assessment and management was performed according to the protocol [18]. Originally this had been conceptualised as a specialist service. It involved: universal screening for IPV; a supportive primary care provider relationship; systematic history of abuse and any attempts to enlist assistance from police, legal service providers or courts; forensic documentation; referral and reporting of abuse to the justice system; emotional status; participant's verbal report of previous results of voluntary counselling and testing for HIV; casefinding for pregnancy and STIs; other special investigations as indicated; safety assessment; safety planning; referral to local services and follow-up appointment. All of this was conceptualised in the protocol as the responsibility of the primary care provider. --- Engagement with action by co-operative inquiry group members Initial training of co-operative inquiry group members included an overview of: IPV; the study purpose; action research methods; use of the protocol; how to collect forensic evidence; the Domestic Violence Act of 1998; use of mental health assessment guides; and communication skills. Primary care providers from the facilities were trained on site by the principal researcher to screen for IPV and refer to the study nurse. Providers were equipped with a laminated menu of possible screening questions and were requested to ask one of every adult female over 18 years [18]. Screening for study participants was performed over a period of four to eight weeks at each facility and 168 women were referred to a study nurse for assessment and management as per the protocol. Women were then invited to give feedback on their experience and 74 per cent returned for the follow-up visit one month later when they gave feedback to a different member of the inquiry group. On-site support and mentoring of the study nurses was provided by the first author, who implemented the protocol at one site and conducted follow-up interviews at another. Five focus groups were conducted with the primary care providers to explore their experience with screening patients. These findings are reported elsewhere [20]. One focus group interview was held in the urban area with a male psychiatric nurse, four female clinical nurse practitioners, an emergency/trauma-room nurse and a female doctor. One rural focus group interview was held with doctors and nurses who had referred women to the study and another focus group interview with those who had not made any referrals. Both groups included a spectrum of doctors, from specialist family physicians to medical officers, and psychiatric, maternity and primary care nurses of both genders and all races. A final rural focus group interview included two lay counsellors, a nurse manager and a clinical nurse practitioner, all female. Interviews explored their experience of screening and initial management of identified women. Sixteen key informant interviews were conducted. Participants were purposefully selected on the basis of their expertise. Eight interviews were held with members of the Department of Health who included the facility managers, relevant programme-specific managers and the head of chronic care for the Western Cape Province. Six interviews were held with academics working with gender issues or forensic medicine. Two interviews were held with leaders from non-governmental organisations with a specific focus on IPV [21]. These interviews explored different perspectives on how the emerging model for addressing IPV in primary care could work. Interviews were digitally recorded, transcribed verbatim and analysed using the framework method [22]. --- Reflection by co-operative inquiry group members Each co-researcher kept field notes to record key experiences, thoughts, emotions and reactions. Five co-operative inquiry group meetings were held over a period of 14 months for collective reflection and planning. All discussions were digitally recorded and transcribed. The first author circulated a summary after each meeting. --- Building the final consensus The first author reviewed all transcripts and field notes and conducted a qualitative content analysis. A final meeting six months later provided an opportunity to reach consensus on the proposed model [21]. Thereafter a consensus of the group's learning was circulated for validation. --- Ethical considerations Ethical approval was obtained from the Health Research Ethics Committee of Stellenbosch University and permission to conduct the study from the Department of Health. Where requested, a psychologist provided support to co-researchers experiencing vicarious traumatisation. --- Results --- Role of primary care providers The protocol recommended that all women should be screened by primary care providers. In practice, however, primary care providers selected women for screening. It was clear that primary care providers did not support universal screening, but some were willing to consider the hypothesis of IPV when the patient's presentation indicated it as a possibility: "How are we specifically going to identify that? It doesn't fit in with any . . . we have enough to keep us busy the whole day. We don't really have time to pay proper attention to it and to explore the problems further." '. . . that went way above my head. Ask every patient! I sort of probably decided on the first day I'm only using this for certain patients." Most primary care providers were nurses with a task-orientated and bio-medical approach to assessing patients, who worked in clinics with high workloads. Primary care providers often felt they needed to protect themselves against further demands from patients or managers: "There is not a culture of support for health providers. They are ordered to do more and more tasks with no extra staff. Providers expressed feeling overwhelmed, exhausted, frustrated and unsupported . . . that spills out onto the patients where they're not really that supportive of patients." ". . . many times . . . providers do want to, but they know what is waiting outside the door . . . so we should not label all health workers as disinterested, its actually the system itself that squeezes them till a point that they can't . . ." (CIG 6, p.20, l. [7][8][9][10][11][12][13][14] An approach of selective case finding rather than universal screening, initial clinical care and referral to someone with the capacity to provide more comprehensive assessment and counselling appeared to be the best fit with the current realities of South African primary care. "If health workers know their role is to identify the issue and only provide clinical care before referring client to the intimate partner violence champion it actually fits in with their paradigm "I have identified this and can . . . move it over there!" "The quality counselling we gave isn't feasible to expect from others. I noticed psychiatric nurses don't do it, social workers, nurses, and doctors also don't!" Such feedback resulted in a modified protocol relying on a selective case finding approach based on cues identified from analysis of participants' medical records. The identification of these cues from the medical record is described more fully elsewhere [20]. The original protocol suggested a number of different screening questions. However, a new question, "Are you unhappy in your relationship?" was thought to work best as it was "harder for people to disagree with that statement if it is true for them." . "How are things going in your relationship?" was also considered an appropriate option. --- IPV champions It was recognised that the inquiry group members who performed the assessment instinctively worked with a guiding style [23]. The full assessment took 60-90 minutes, which is impractical for primary care providers who are expected to spend an average of seven minutes per patient [24]. Nevertheless, women benefited from telling their stories to an empathic listener as well as from specific assistance: "It must be someone who wants to do the work; who is interested, who is motivated; who will give that listening heart and support . . ." Reflecting on the attributes needed in a provider who attends to IPV, the inquiry group agreed that such a person could be any member of the multi-disciplinary team with the following characteristics: desire to work with IPV; empathy and good listening skills; respect for client confidentiality and autonomy; efficient case manager; collaborative approach to problem-solving; and effective multi-disciplinary team work. Implementing this model of IPV care therefore would imply that such a person should be identified and set aside for this work in the Department of Health or Social Development, both of whom have some responsibility for the issue: "I think it is an excellent idea to use somebody outside because in the clinics it's really hectic and I think this system will be very much helpful for the client." --- Four key components The inquiry group categorised the different components of the protocol into four broad areas: clinical, social, psychological and legal. The model that emerged was that the clinical component could be implemented by primary care providers, while the others required an IPV champion, as described above. The clinical component consisted of recognising cues suggestive of IPV as an underlying issue , treating injuries, forensic documentation if necessary, attention to unwanted pregnancies , syndromic management of STIs and HIV testing. This work is already an accepted part of the primary care providers' role. The social component comprised a detailed and empathic history of the abuse, an assessment of the woman's immediate risk of injury or death, mobilisation of social support and planning for emergency situations. Survival issues such as procuring maintenance for children and disability grants were included. The psychological component comprised screening for common mental health problems and referral for more formal diagnosis and treatment if needed. The legal component included a history of previous attempts to utilise the police or courts, information about legal rights and assistance to obtain a protection order or lay a charge. Referrals from study nurses significantly improved women's chances of success with accessing legal rights. Victim empowerment units at police stations also played an important part. --- Assessment of abuse After testing the protocol in the first phase of data collection, we reordered it to better fit our experience of the interview's flow. The original protocol was forensically oriented. For example, abuse was framed as assault that would require forensic documentation using body charts. Yet only a third of the urban sample were injured, so we reformulated the forensic component as an appendix for ease of use, and tested it in the rural phase of data collection which followed [21]. The first phase of data collection also suggested the need for a more comprehensive assessment of the patient's social context and so a genogram was inserted as part of the consultation to document the patient's household and family context. See the Additional file 1 for our modified protocol. --- Danger assessment Use of the safety assessment and plan in the original protocol exposed grounds for doubt about its validity in the South African context. It seems based on the assumption that the IPV survivor is badly injured, and cannot return home to an abuser, who may kill her. Our experience was at odds with this. Only two survivors in our sample moved to safe housing, for example, and many were not living with the perpetrator but needed to stop him coming to their home. A quarter of perpetrators were at home reflecting South Africa's high levels of unemployment. Consequently the accuracy of the scoring for danger assessment in this context seemed questionable since "Has he threatened to kill you?" scored the same as "Is he presently at home?" . The scoring also emphasises the presence of a firearm. Ownership of a legal gun, combined with being a security worker, has been found to dramatically increase risk of femicide [25]. Yet paradoxically, while a firearm suggests a higher risk of injury or death, the absence of a firearm does not necessarily imply lesser risk. The inquiry group members debated whether it helped a woman, who has nowhere else to go, to be told she is at severe risk. However, feedback from participants indicated that the safety assessment rating had helped them to reframe their situation from a different perspective. As the assessment actually measures a woman's risk of repeated assault or murder, the inquiry group felt that "danger assessment" would be a more accurate description of this process. The group felt that a woman's risk of suicide, and even matricide, should be part of a danger assessment. In general the inquiry group felt that the validity of the items and their scoring to determine risk of injury or death needs further evaluation. A revised danger assessment is found in the Additional file 1. --- Safety planning The protocol's suggestions for safety planning were based on relatively affluent circumstances. Access to credit or bank card and car keys did not match most participants' socio-economic reality. However, participants reported that it had been helpful to learn about the importance of having one's critical documents in order and copies hidden safely, as well as a bag packed. The safety plan appeared to have more relevance when there was extreme physical violence. Also, given the danger abusers present when their partner leaves them, and for two years afterwards [26], key informants stressed the importance of a careful history and to inform her that if she is planning to leave him, she must tell no one. --- Need for better mental health assessment and care At the outset the authors realised that the original protocol over-emphasised forensic issues and underemphasised mental health care. Women survivors of IPV have a high prevalence of mental disorders [8]. The cooperative inquiry group therefore incorporated a more comprehensive component on mental health problems based on the World Health Organisation's toolkit for the recognition, diagnosis and treatment of mental disorders in primary care. The second author had previously adapted this toolkit for the South African primary care context. This included a checklist to identify the possibility of a mental disorder [27]. Women were then referred to a mental health nurse or doctor to make a formal diagnosis and treatment plan. The decision to expand the section on mental health was validated by the experience of assessing women and reviewing their medical records: "The thing that jumps out at me is that psychiatric medication is number four on the list [of reasons for encounter]. . . That's bizarre doesn't happen in any other primary care grouping." (CIG 6; p.24, l. [18][19][20][21] Although the mental health nurse was the most obvious person to refer patients to, there were concerns with her/his capacity to deal with IPV: "I got the sense that the psychiatric nurse would value other things over something like domestic violence . . . they might put more of a premium on your regular psychiatric disturbances/conditions" (CIG 6, p.5, l. [15][16][17][18] Alcohol abuse was the commonest substance women admitted to using. Given the prevalence of chronic pain in this sub-population, it was decided to add use of analgesia to the substances participants were questioned about [28]. --- Need for ongoing containment and personal empowerment process The inquiry group recognised that IPV is most often dealt with as an acute event in primary care, but is more applicable to a chronic care model. Women may take time to decide on what to do and may require support through multiple attempts at changing their lives. A once off comprehensive assessment and counselling session with referrals is unlikely to be sufficient. The concept of IPV champions was congruent with emerging local chronic care policy which promoted 'champions' who provide continuity of leadership for chronic care and practice. In addition the local chronic care program, which mainly focused on non-communicable diseases, had produced a five-week personal empowerment process focused on issues of self-efficacy, self-care and motivation to change. In the final inquiry group members felt that this could help women with IPV and should be included in future interventions . --- Need for ongoing community-based support groups Following such a personal empowerment process the inquiry group recognised the need for longer term social support. In fact many women spoke about their desire for a support group and even their willingness to establish one and be trained as a facilitator: ". . . something valuable about a group process, especially for isolated, depressed, marginalised women. There's a lot that can really help them to value themselves more, to feel more connected, to be supportive to each other." A system of support groups can be beneficial in scarcely resourced contexts with overburdened health systems. There is also scope to incorporate job skills development as a lack thereof entraps many women. Adolescents were reported to "fall through the gaps" within contemporary health systems where their needs are very poorly met. Further, the onset of IPV difficulties in teenagers was emphasised by key informants [21]. Given the stigma that surrounds IPV, the group could be referred to as a women's health or interest group. Support groups were not implemented during the study period. --- Discussion --- Key findings The model of care for women with IPV that emerged from the co-operative inquiry group process is illustrated in Figure 1. The first tier of the approach relies on the primary care provider, usually a nurse, to recognise cues suggestive of IPV. An analysis of the women's medical records and review of the literature suggested the cues shown in Figure 1 [20]. When these cues are recognised the provider asks a question such as "Are you unhappy in your relationship?" If IPV is disclosed as an underlying issue then the provider will deliver important clinical aspects of care and refer to a local IPV champion. We developed a one page flow chart for assessing and managing such patients to prompt the provider in the hope of moving from the fragmented approach of the past to a more clearly defined package of clinical care. The second tier of the approach relies on an IPV champion. Two IPV champions may provide a service at the primary care facilities on a rotational basis throughout a subdistrict. Here patients are assessed and assisted comprehensively and then invited to participate in a five week empowerment group. The IPV champion could be either someone set aside by the Department of Health, with the recommended attributes and motivation, or a social worker from the Department of Social Development who would work in collaboration with primary care services. Whoever is identified in a specific sub-district would provide comprehensive assessment and assistance by taking a history of abuse and attending to social, psychological and legal issues. Training of the IPV champion should pay equal attention to the therapeutic and motivational interviewing skills as well as expertise in the different areas of IPV management. The IPV champion should also facilitate the five-week personal empowerment group process following the initial assessment. A skilled facilitator in such a program could create a supportive environment for discussion of difficult issues and build awareness around protection orders, human rights, effective parenting, conflict management skills and so forth. It could offer a potentially powerful process of personal transformation, blending educational and therapeutic value to provide a context where women could be stabilised, educated and treated. Broader than the medical model, this group process could offer participants a rare opportunity to attend to the meaning of their own life, reflect on their behaviour, choices and way forward. Note, this program was not implemented by the inquiry group during the study period. The third tier of the approach relies on the establishment of community-based support groups that would support women in the longer term. These groups could be established with the help of the IPV champions and the Department of Social Development as well as local non-profit organisations. Adolescents and young adults should have a group of their own, with women older than 25 years comprising another group. Both groups should initially be facilitated by a social worker, psychiatric nurse, occupational therapist or psychologist. Over time they may become self-sustaining. --- Comparison to literature Four factors have been identified that increase provider self-efficacy for IPV screening: institutional support; effective screening protocols; thorough initial and ongoing training; and immediate access/referral to onsite and/or offsite support services [29]. Our model allows for easy access to onsite support via the IPV champion and recommends a constructive approach to effective screening. Previous reviews have only considered universal screening as a valid approach [29] and studies have been critiqued for focusing too much on whether programmes work and less on how they work [30][31][32][33]. Given the cues presented by women experiencing IPV and our experience of trying to implement universal screening in the South African primary care context, we argue that selective case finding of women with a higher risk of IPV is a more constructive approach in resource poor settings [20]. Clearly future implementation of the model will require institutional, and at best intersectoral, support as well as initial and ongoing training. Successful interventions, which alter provider behaviour, should address both predisposing factors as well as enabling factors [29]. When compared with Colombini's typology of IPV models, the model suggested here is primarily one of "facility-level integration" in that a patient will receive initial screening, assessment, management and a small group empowerment process from different providers at the same facility [34]. Provider-level integration of care, in which a single provider offers a comprehensive service, is more typical of high resource settings and most of the literature refers to this model. For example individually tailored counselling sessions for pregnant women [35]; one-to-one advocacy interventions [36]; individual cognitive-based therapy [37]; and one-to-one advocacy for pregnant women [38]. Systems-level integration suggests a model where the service is offered in multiple sites within a system such as a sub-district. In the model suggested here the community-based support group component extends the service to the system level. The communication style suggested in the model is typical of a guiding stylecollaborative, empathic, evocative, supportive of choice and control and yet directive in its focus and process [23]. The guiding style is seen as appropriate when engaging with brief behavior change counselling. As these women survivors of IPV are facing difficult decisions about change and behaviour it is not surprising that this approach makes sense. One can therefore make a conceptual link between communication skills for consulting the IPV survivor, brief behaviour change counselling and motivational interviewing. This may be useful in the approach to training IPV champions. The structure of the 5As that is currently seen as best practice for brief behaviour change counselling may be useful in the approach to discussing mental health problems, safety planning and use of legal rights and resources [39,40]. --- Limitations of the study design Mash and Buskens suggest quality criteria for action research [17]. In retrospect we were still too contaminated by an empirical-analytical mindset to fully innovate, experiment and implement change as part of the process. This dynamic was compounded by the fact that membership of the co-operative inquiry group was not consistent. Further, the group ownership of the inquiry process had to be held in tension with the first author's requirements for a doctoral study. The development of reflectivity also varied between co-researchers. --- Implications and recommendations The intervention depicted in Figure 1 is currently being piloted in the rural Witzenberg sub-district of the Western Cape as a collaborative project between the Departments of Health and Social Development. The piloting of this IPV model will involve further monitoring and evaluation and will be reported on. If the model is successfully implemented then the Department of Health intends to implement more widely. Further work should also be undertaken to look at the incorporation of this approach to IPV into the HIV and TB programmes, which are often separate vertical programmes in South African primary care. The incorporation of the model into emergency and perinatal units within community health centres should also be considered. --- Conclusion This study demonstrated the feasibility of implementing a model for recognising, assessing and assisting women survivors of IPV in South African primary health care. The original protocol had to be adapted in terms of its expectations of the primary care providers, overly forensic orientation, lack of depth in terms of mental health, validity of the danger assessment and safety planning process, and need for ongoing empowerment and support. The inquiry process led to a model of comprehensive and intersectoral care that is integrated at the facility level and which is now being piloted in the Western Cape, South Africa. --- Additional files Additional file 1: Modified Protocol. IPV assessment tool for use in 1 st stage of 2 nd tier of model. --- Competing interests The authors declare that there are no competing interests. --- --- Authors' information

Background: Despite extensive evidence on the magnitude of intimate partner violence (IPV) as a public health problem worldwide, insubstantial progress has been made in the development and implementation of sufficiently comprehensive health services. This study aimed to implement, evaluate and adapt a published protocol for the screening and management of IPV and to recommend a model of care that could be taken to scale in our underdeveloped South African primary health care system. Methods: Professional action research utilised a co-operative inquiry group that consisted of four nurses, one doctor and a qualitative researcher. The inquiry group implemented the protocol in two urban and three rural primary care facilities. Over a period of 14 months the group reflected on their experience, modified the protocol and developed recommendations on a practical but comprehensive model of care.The original protocol had to be adapted in terms of its expectations of the primary care providers, overly forensic orientation, lack of depth in terms of mental health, validity of the danger assessment and safety planning process, and need for ongoing empowerment and support. A three-tier model resulted: case finding and clinical care provision by primary care providers; psychological, social and legal assistance by 'IPV champions' followed by a group empowerment process; and then ongoing community-based support groups.The inquiry process led to a model of comprehensive and intersectoral care that is integrated at the facility level and which is now being piloted in the Western Cape, South Africa.

Introduction The term synergy in science was used in neuromuscular physiology by Sherrington [1] when he described the integrative action of the nervous system. The concept was further developed as a process involved in self-organization by the theoretical physicist Haken [2] and the biologist Corning [3]. Exploring "Google Books Ngram Viewer" for the word "synergy," ranging the years 1800 to 2000, shows that the use of the word synergy starts after the year 1900. In modern economics, the term is widely used. A Google Scholar search of the terms "synergy" and "economics" revealed 216000 results of "synergy" and "mergers" 37300; of "synergy" and "Adam Smith" 12700; of "synergy" and "invisible hand" 6610; of "synergy," "invisible hand," and "simulation" 1540. This last search showed the paper by Marris and Mueller [4], review of the literature entitled "The Corporation, Competition and the Invisible Hand" where the word synergy appears only in the title of one of the papers in the 5 pages of references. This was the work by Carter [5] entitled "In search of synergy: A structure-performance test." Here, Carter wrote that "the possibility of synergistic effects to diversification is recognized by most if not all observers." So evident is the importance of synergy to economists that few write about it. More recently, research on virtual artificial societies or computer simulations of social dynamics have shown their worth in illuminating how the aggregate of various simple interactions might produce phase transitions and the emergence of novel properties of the system and even novel phenomena. These modern computer simulations, specifically agent based simulations, allow us to explore complex economic phenomena. Examples include the complexity of exchanges [8] and money dynamics and banking catastrophes [9]. However, agent based simulations have not been incorporated in mainstream economics [10], nor have they unveiled until now in detail the working of the invisible hand of the markets. Agent based simulations are a powerful tool in clarifying fundamental aspects of the working of complex economic phenomena. Its potential in visualizing fundamental concept in very simple economies will be explored here so as to avoid the limits "of ascertaining all the data whose utilization in the allocation of resources is the 2 Complexity great merit of the market system" , and which are not easy to determine analytically in more complex economic settings. Here, a deeper analysis of the working of synergy in market economies is presented, using experiments with computer simulations written in different languages that improved the robustness of my previous findings and provided more details of the economic mechanisms that trigger synergy. As complex system science has shown with many examples , the emergence of complex phenomena, such as synergy, is the product of interactions of many elements. Presenting a simple analytical expression of this process creates a false illusion of simplicity. Thus, analytical formalisms that give this false impression are avoided and the focus is placed on the details of the algorithms that allow the synergistic processes to emerge. Smith in his book The Wealth of Nations [11] described the operation of the market as follows: "Every individual necessarily labours to render the annual revenue of the society as great as he can. . .. He, however, does not have the slightest intention of promoting the public interest or is aware that he is promoting it. He intends only his own gain and is led, as in many other cases, by an invisible hand that makes him promote a cause that does not form part of his intentions. This is not a disadvantage for society. By pursuing his own interest, he frequently promotes that of the society more efficiently than if his interest were the latter. I do not know of much good dispensed by those who strive to represent the common good. It is not from the benevolence of the butcher, the brewer, or the baker, that we can aspire to our dinner, but from their attention to their own interests." In another part of the book he writes "The greatest improvement in the productive powers of labor, and the greater part of the skill, dexterity, and judgment with which it is anywhere directed or applied, seem to have been the effects of the division of labor. . .. It is the great multiplication of the productions of all the different arts, in consequence of the division of labor, which occasions, in a well governed society, that universal opulence which extends itself to the lowest ranks of the people." These are beautiful descriptions of phenomena where the interactions at the individual level bring as a consequence dynamics significant only at the social level, without individual activity being conscious of it. It is a fascinating phenomenon but difficult to study using traditional experimental techniques. The discovery of the invisible hand of the market is a major achievement of humankind. It recognizes the absence of centralized social cohesive forces and discovers forces of the market that explain our social dynamics. More detailed studies of the effects of division of labor have been published , but all failed to grasp analytically in ways acceptable to the natural sciences the emergence of synergies in economic markets due to division of labor. This inability to grasp numerically these phenomena has led some economists to conclude about the analytical intractability of all details in complex economies. Prominent among these thinkers is Hayek [13], who coined the term "Economic Calculus" when referring to this fundamental analytical limitation of economic analysis. He said [14] "[economics] has become too ambitious by applying standards of rigorousness. . . to the empirical science of economics where there are definite limits to what we can positively now; that we shall see more clearly what economics can do if we separate that logical groundwork -the economic calculus as I have called it -from its use in the empirical science of economics; and that, though this science is of great help in all-important issues of the choice of an economic order and of the general principles of economic policy, its power of specific prediction is inevitably limited -limited by the practical impossibility of ascertaining all the data -those very data whose utilization in the allocation of resources is the great merit of the market system." In theory, the adaptation to dynamic fitness landscapes by biological populations of individuals with complexly interrelated gene assembles can be studied with an evolving multiplex network analysis where population evolves in a multilayered network with nonlinear complex interactions between the nodes of the different networks. Adaptive networks have been described for simple networks with linear interactions , However, the problem tackled here requires a dynamic multiplex network that produces adaptive evolution of a population of agents with a layer of networks representing the spatial environment where agents interact and which provides the selection forces that act on individuals, a second layer that mirrors the allelic composition of the genes in the genome of each agent and its intraindividual interactions, a third layer reflecting the interindividual interactions of the allele of all genes in the population, and layers that code for the nonlinear interactions of each allele in each gene locus. Only with these layers can we attempt to represent the agents and their reproductive interactions, the genetic modulation of individual fitness, kin selection, and inclusive fitness. Multilayered networks have shown to be unable, up to now, to explain cases as complex as that of the evolution of sex [16]. In contrast, agent based simulations provide an expedite route to study these phenomena. This rational is at the base for the choice of agent based simulations for studying the emergence of synergies in economics. Simulations with the agent based simulation Sociodynamica [17] allow for exploring abstract virtual economies that are far simpler than real ones but already so complex that the experimenter may lose the integral view over the interactions between environment, agent behavior, pricing mechanisms, and environmental heterogeneity in the market. This might happen in simulations of economies where realistic price dynamics were included that showed that division of labor was the strongest predictor of successful economic performance [18]. However, it was not clear in these simulations, if this effect of division of labor was exerted through the price dynamics of the economy or through other means. In order to pinpoint the source of the synergies achieved by division of labor, the model was simplified until the emergent effect of division of labor disappeared. Stripping out effect of pricing on the market dynamics did not eliminate the effect of division of labor. This allowed us to follow in detail the features that make division of labor work, making the system amenable to analytical analysis. Figure 1: Representation of the landscape of two virtual economies. shows a result from a simulation with free prices in an economy collecting food and minerals by omnipotent agents, whereas shows the same but for agents specialized either in mining or farming. --- The Model The agent based computer simulation model Sociodynamica is a freely available agent based simulation model, first written in Visual Basic, and has previously been used to study the effect of altruism and altruistic punishment on aggregate wealth accumulation in artificial societies and to grasp the dynamics of complex markets . These models are completely mechanical in nature, and individual incentives may emerge trough an evolutionary process that makes agents with the right combination of incentives or behaviors survive and those with the wrong combination to become eventually extinct. The features revealed by Sociodynamica are very similar and, in cases identical, to those revealed by Sugarscape, an agent based model developed independently by Axelrod [20] or the model developed by Axtell [8]. In all three cases, Walrasian solutions in which an auctioneer centrally computes prices cannot be made more efficient than the decentralized alternatives based on free and heterogeneous agents making these decisions. These results support the proposition of Adam Smith that markets are ruled by an invisible hand that coordinates the different kind of labor rendering markets efficient. Specifically, simulations in complex economic setting showing omnipotent agents performing all tasks produced less aggregate wealth than simulations where three different agents performed different tasks, such as farming, mining, and trading [18]. This counter-intuitive result was partly due to the fact that optimal prices and conditions for trade were different for each agent, depending on its spatial position in the virtual world. Omnipotent agents had to assume average solutions to balance their different tasks. Therefore, they never traded at optimal prices and optimal quantities according to their spatial position. Here the same model was used to run simplified scenarios, until the effects of prices dissipated, to reveal fundamental economic features that allow the emergence of synergies from division of labor. The model simulates a virtual society of agents who farm and mine for foods and minerals, respectively, analogous to the model "Sugarscape" by Axelrod [21] and also trade their surplus according to different economic settings. The agents inhabit in a continuous flat two-dimensional toroidal world that was supplied with patches of agricultural land and separate nonoverlapping patches of mines . Diverse agents were distributed at random on a fine-grained virtual landscape with resources. Simulations depended on the type of movement of agents, and, thus, to simplify interpretation of results, agents were simulated as immobile entities. Their individual utility function was defined by two resources. Each time step, any agent that happened to be located over one of these resources acquired a unit of the corresponding resource, accumulating its wealth, either as sugar or food and/or as spices or minerals . Agents spend a fixed amount of each resource in order to survive, consuming each of them at a basal constant rate . Both resources were consumed and metabolized similarly, but food was 3 times more abundant than minerals . Each patch remained in the same place during each simulation run and the resources inside them were replenished continuously. Agents perished when they exhausted any of the two resources. Success in gathering and trading resources was defined by variables that produced behaviors that made them unable to compete successfully for resources. These variables included type of movement, spatial positioning, price thresholds for selling each of the resources, price threshold for buying the resources, and type of agent. During the simulation, natural selection weeded out unsuccessful combinations of these variables. The total population of agents was maintained constant by creating the required amount of new agent necessary, each with randomly assigned initial parameters. Initial parameters were the type of agent, the random spatial position, and the initial amount Complexity of money used to start trading resources . The amount of money for each agent varied according to its trade balances. Agents gain money when selling food and/or minerals and lose money when buying them. Agents traded the resources they possessed with other agents. In order to trade, they had to find a partner with the desired resource, and they had to have agreement over prices. The trade could be among any agent in a population of omnipotent agents without "division of labor." When simulating division of labor, agents specialized in collecting food or collecting minerals or collecting neither but engaging only in trade. Here, agents were subdivided into three categories. Farmers which specialized in collecting only resource 1; Miners which extracted only resource 2; Traders specialized in trading minerals for food when encountering a farmer, and food for minerals when encountering a miner. Food collectors traded only with mineral collectors and Traders, mineral collectors traded only with food collectors and Traders, and Traders could interchange resources with all types of agents. Trades were allowed only between agents spaced at a distance not larger than the "contact horizon" of the trading agent. Each time step, all buyers searched for potential sellers of the required good by contacting randomly up to 10 agents in the area defined by this contact horizon. If finding a seller with the wanted goods at or below the price defined by the buyer, a trade was executed using the price of the seller. Trades were limited to the amount of money available to the buyer and the amount of goods possessed by the seller, unless credit was simulated. Variation of this contact horizon allowed simulating different levels of globalization or integration of economic agents. The effect of the degree of globalization on the economy can thus be measured quantitatively, a feature that is not possible with real economies . Prices were initially assigned to each agent for each resource at random from a range of values defined by the experimenter and then varied according to supply and demand as experienced by each individual agent. That is, at the end of every time step, after finishing a tournament of trades in the market, each selling agent attempting to sell parts of its excess of resource that could not find a willing buyer because of the price it asked for reduced its reference price by an unit. And each buyer that could not find a seller willing to sell the desired resource at the desired price increased its reference price for that resource. In this way, each agent maximized its self-interest by selling each resource at the maximum price possible and buying at the lowest. Various processes were simulated. A first process of the simulation was the balance between income of resources and their consumption . For survival, agents were required to conform to Ir > Cr. ( )1 Income can be either by direct gathering or by trade Ir = Gr + Tr. (2 ) Here each agent has to balance two resources in order to survive. A utility function is simulated so that U remained positive for the agent's survival and Ur = -Cr. ( A second process of the simulation was the dynamics of traded resources. These resources can increase or decrease, according to the balance of resources bought and sold Tr = Br -Sr. (4 ) The amount of resources bought and sold depends on the availability of money and the price paid for the resource by the agent Br = M a Pr a , Sr = M a Pr a . The amount of money of agent depends on the amount spend buying and the amount gained selling resources M a = M a 0 + Msr -Mbr, where M a 0 stands for the initial amount of money supplied to each new agent j. Ur was calculated every time step for the population of agents so that the total accumulated of wealth for each resource Wr = Σ a Ur. Agents with 𝑈 1 <= 0 or 𝑈 2 <= 0 were eliminated and substituted by new ones with default properties, as an analogy of broken companies that are replaced by new start-ups. Here the focus is on the age of the agents as the most relevant variable for assessing the benevolence of an economic system. The average age is a measure of the probability of survival of individuals in the population, but other measures are possible [18]. The aim was to pinpoint the features that allowed the emergence of the synergies of the market due to division of labor. Simulations of virtual worlds with one type of agents , two types , and 3 types allowed determining the effects of increasing complexity of labor structure. Simulation with homogeneous and heterogeneous distribution of resources allowed assessing the effect of the environmental complexity, and simulations with different contact radius provided insights to the importance of synchronization between trading agents. Synchronization here means allowing the meeting of buyers and Traders. A longer and more detailed description of the simulations is provided in Jaffe [18] and the detailed program in Visual Basic is available in the help feature of the program. Sociodynamica has been rewritten in C# and JavaScript. Simulations with these versions can be run with parameters chosen at will by downloading Sociodynamica. The Visual --- Results Figure 1 shows two examples of the output of simulations with free prices after 200 time steps. Figure 1 shows the virtual world when simulating omnipotent agents; Figure 1 reflects the outcome when division of labor was included in the simulations. The figures reflect the effects on the economic dynamics of introducing division of labor. In the figure with omnipotent agents, agents over mineral fields are smaller than those over fields with food. In contrast, the figure from simulations including division of labor showed that Miners over field of minerals were very wealthy, and so where many farmers over field with food. The price agents were willing to pay for minerals was higher among farmers ) and lower among Miners ) when division of labor was simulated. In the case of the omnipotent agents, prices agents were willing to pay for minerals or food seemed to be randomly distributed ). This example reveals that omnipotent agents made suboptimal trading decisions. They sold the resource they had accumulated in more abundance to any other agent willing to take it, whereas specialized agents ) traded only the resource they were collecting; that is, Miners only sold minerals to Farmers and Farmers sold only food to Miners. The trading patterns of omnipotent agents produced less wealth in the long term than that of specialized agents, even though the cognitive complexity of the algorithm omnipotent agents used was more complex. Bright green field is covered with "food"; darker green field is covered with "Minerals"; the lightest green is devoid of resources. Each agent is depicted as a colored sphere. The color of the body of the sphere describes the type of agent: Farmers are yellow, Miners are red, and omnipotent agents are black. The width of the bubble is proportional to the amount of food and minerals accumulated by the agent and the height by the amount of money the agent possesses. The thickness of the border of the sphere is proportional to the perceived cost of living calculated as Food Price + Mineral Price and the color of the border ranges from blue to yellow. The more red or even yellow the higher the ratio between the Food price and Mineral price. Agents with red and yellow borders pay more for minerals, whereas those with blue or black borders pay less for minerals compared to what they are willing to pay for food. As multivaried complex data is difficult to visualize, as the attempt in Figure 1 shows, data will be presented also in a more classical format. The difference between the two simulations shown in Figure 1 is presented quantitatively in Table 1. As intended in this simplified model, the effect of centralized or decentralized pricing was stripped out, so that this feature did not affect results. Agents in simulations with 2 types of agents performed much better economically, achieving much longer life-spans in average than agents in simulations with only one type of agents . Increasing the complexity of the division of labor by including Traders, that is, agent that did not collect resources but only bought and sold them, using the same algorithm as that of omnipotent agents, did not further increase the efficiency of the resulting economy. Eliminating heterogeneity in the environment eliminated the advantage of division of labor . Simulations where fields of minerals overlapped in all its extension with field of food produced quantitative results that were identical between virtual economies of omnipotent agents and economies with division of labor with two type of agents . Results show that the division of labor, in order to work properly, required an adequate coordination of actions with synergy between the agents. For example, omnipotent agents sold whatever resource they had, independently of what they collected, whereas Farmers and Miners only sold the product they collected and bought the one they did not collect. Including Traders in simulations introduce noise in the economy unless they had a specific function besides trade, such as providing credit, which benefited society in economies where prices could be adjusted through demand as shown in Table 3. Other features also increased the possibility of improving the economy. In Figure 2 the effect of an improved ability to trade is shown. Here, the maximum contact radius determining the distance at which potential Traders could be spatially separated was varied. As expected, results show that, at greater maximum contact radius, the economy performed better. This trend, however, was not linear. Prices had a very strong nonlinear relationship with the contact radius. This relationship strongly depended on the topology of the resource distribution simulated, as was shown before [18]. --- Conclusions The experiments presented here showed that the same results are obtained with different software written in different computer language. This makes it unlikely that the result are due to coding artifacts rather than the logic of the causal network explored. The results presented here identified features that are indispensable for making the emergence of synergies, due to division of labor, possible. These are as follows: Heterogeneity: spatial or temporal heterogeneous environment and/or agents. Complementarity: complementary activities of agents, with divergent optimization options. Synchrony: markets help synchronize agent's actions. The larger the contact horizon between participants of the market is, the more efficient the market forces act. These features allow the social system to decrease its entropy, increasing the information managed by the system, and at the same time increase the economic work performed. These two features have been proposed to thermodynamically characterize a synergistic process. Thus, economic activity can be meaningfully characterized using thermodynamic concepts providing a way to quantify synergistic processes [23]. The results are relevant to empirical findings in real economies. For example, the first feature is interesting in that it recognizes the fact that the complexity of labor structure mirrors the complexity of the economic environment where it works, and vice versa. This feature is a consequence of economic activity being molded by the environment to which it has to adapt. In human societies economic wealth of a country and economic complexity are linked very strongly [24]. This suggests that economic complexity, of course, can only be achieved with complex division of labor. This requires increased availability of information. This last requirement triggered the search for this relationship in real economies , showing that management of scientific information is more closely related to economic output of a society than the complexity as measured by Hausmann and Hidalgo [24]. The second feature is illuminating as it allows predicting when division of labor might improve economic activity and when not. Division of labor might prevent inefficient trades and/or might make more efficient trades possible. Tasks, which converge in skills, might not require specializations, whereas tasks that require very different types of skills will benefit more of division of labor. It can be argued that more intelligent agents might be capable of performing several tasks and, thus, rather that division of labor, economies benefit from more complex or intelligent agents. This might be true but complexity and intelligence have their costs and might be suboptimal if different simple agents can handle the problems with higher efficiency. This seems to be the case in the social evolution of ants, where a negative correlation between brain development of individuals and social complexity was evidenced [25]. The third feature is possibly the more difficult to manage in real situations. Experience shows that on-time synchronized productive chains allow specialization of tasks to be more productive overall, whereas low division of labor is more tolerant to inefficient supply chains. These features explain many a difference between highly developed economies and ones with incipient industrialization and poor services. Empirical studies showing how heterogeneity in a network can create heterogeneity of behavior [26] exist. This is one of the ways social systems might achieve synergistic synchronization of different activities. More research in this area is urgently needed in order to explain differences in economic development between countries [27]. The identification of these 3 features as fundamental in allowing division of labor to elicit economic synergies might seem trivial. But browsing the literature, a great number of reasons have been postulated to explain the synergies created by economic markets. The problem seems far from solved . Even if found to be trivial a posteriori, the simulations helped to identify the relevant features and discard superfluous ones. Bowles [29], for example, defined the invisible marker mechanism as a Nash equilibrium and its Pareto optimal, where the self-interest of each actor yields an outcome that maximizes the wellbeing of each. Here more details of how this can be achieved are presented. The three features unveiled here seem to be very general, relevant to system dynamics, biology, human society, and real economies. Possible empirically falsifiable predictions based on these tree features are as follows: Division of labor should be more developed in societies that exploit a greater variety of resources [24]. Better communication and economic instruments are provided by more sophisticated financial systems which in turn provide better opportunities for synergies between economic actors in more complex economies. Novel communication technology, such as the Internet, can also broaden the contact radius of economic agents improving synchrony [30]. More division of labor leads to more incompatibilities between skills required to perform them and thus to more diverse specialized education [31]. Larger contact horizons or more globalization improves the synergies unleashed by division of labor as they broaden the scope for more diverse interactions between economic agents Though new research purposefully designed to answer these questions should be designed. Division of labor is related to synergies associated with fundamental aspects of social systems in biology [32]: In human society, economic synergy and division of labor have important relationships. Research showed that more complex economies requiring more division of labor accumulate more wealth and produce higher economic growth [24]. Empirical data showed that an even better predictor than economic complexity for future economic growth in developing countries is the scientific knowledge estimated by the amount of academic activity. Even more striking, the type of division of intellectual activity in a country is a much better predictor than the total complexity or the absolute amount of academic research performed [33]. That is, division of academic labor that prioritizes basic natural sciences over applied sciences and social sciences is much more efficient in producing future economic growth. These results, in the light of the findings of the present simulations, show that much remains to be learned about the quality and quantity of division of labor and its effect on economic activity. More interdisciplinary research is needed to improve our understanding of this very fundamental phenomenon. This exercise shows that computer simulation of simple economic agents can generate nonlinear dynamics that resembles real life features of known economic system. Simulations of very complex systems produce complex results that may become intractable even with sophisticated statistical analysis. Here this limitation is avoided by focusing on very specific and fundamental problems. The simulations presented revealed fundamental features that allow division of labor to create economic synergies. This feat is impossible in complex real situations, but the insights gained in simple systems help in understanding synergies in more complex ones. Simulation models, besides having a potential in experimental economic research, are a fantastic tool to make complex phenomena visible to human understanding and thus should have a potential, if properly adapted for that purpose, in didactic games for teaching economics at all levels of educational and academic specialization. Science learned through smart games based on simulations, such as made available at http://bcv.cee.usb.ve/juegos/intro en.html, might reduce self-serving cognitive biases among lay people, professional practitioners, and decision makers, improving the rationality of our society and thus, hopefully, its economic performance. --- Complexity Use of resource 1 and 2 agent-BRC1; agent -BRC2 Assessment of GDP GDP = GDP + agent --- Show and Plot Plot Shows the agents according to their total resources indicated as the sqr of the diameter. The vertical size of the agent is proportional to the total wealth of the agent's money. The color of the bubble depends of the type of agent as indicated at the left bottom of the screen. The thickness of the border is proportional to the perceived cost of living . The color of the border gives the ratio MinPrice/ FoodPrice. Yellow and then red give the highest ratios. Blue borders indicate that food prices are higher than mineral prices. Black bar at the bottom indicates a length of 100 pixel. --- Internal Parameters General Parameters 𝑋 : Spatial dimension 1 𝑌( --- Degradation of Resource due to Consumption . --- Efficiency of Consumption . It is amount of resource assimilated. When EfC > 10 then productivity is simulated: Agent collects = * price . . Frequency in 𝑡steps distribution changes. --- Frequency of Change in Distribution Consumption of Resource. Rate of exploitation: resource = resource -DNR BRC: Wearing or passive use of resource agent-BRC Resource 1 can be modeled as a renewable resource , whereas resource 2 as a nonrenewable resource , by assigning DNR = 0 and DNR = 1 respectively Food Reserve . It is minimum amount of food needed for agent to engage in transactions of any kind. --- Min Food for Reproduction . It is the amount of food that needs to be accumulated before reproduction can start when Simulation Scenario is 1 or 4. --- Type of Economy (EconoT --- Appendix Sociodynamica creates a virtual society where agents exploit and compete for resources and share resource 1 among them, according to the settings defined by the internal parameters and the external parameters. The agents may acquire renewable and nonrenewable resources through work; they may accumulate those resources and commercialize them. At the same time, agent may acquire resources through commerce. --- Conflicts of Interest The author declares that there are no conflicts of interest regarding the publication of this paper. --- Submit your manuscripts --- Stochastic Analysis International Journal of

Inspired by Adam Smith and Friedrich Hayek, economists promoting free markets postulate the existence of invisible forces that drive economic growth. Simulations with Sociodynamica allowed the emergence of market forces in virtual economies, showing that the synergistic working of division of labor in complex settings favors a stable state where all actors benefit (winwin interaction). By visualizing the detailed dynamics underlying this phenomenon in a simple virtual economy, the elements underpinning the synergistic effect on economic output produced by the division of labor between agents could be dissected. These are heterogeneity or spatial or temporal heterogeneous environment and/or agents; complementary activities of agents, with divergent optimization options; and synchrony. Markets help synchronize agent's actions. The larger the contact horizon between participants of the market is, the more efficient the market forces act. These features allow for social processes that increase the information available and increase simultaneously the capacity of producing useful economic work, that is, synergy. This insight, although trivial if viewed a posteriori, improves our understanding of the source and nature of synergies in real economic markets and might render economic and natural sciences more consilient.

Introduction Approximately 8 million Ukrainians have been displaced and five million children had their education disrupted by the war in Ukraine. By March 2023, over 165k refugees from Ukraine were recorded in the UK . The refugees, mostly women and children as adult men are banned from leaving the country , needed to rapidly adjust to the local customs, speak a new language, and form social networks within the new community they are thrown into. Such stressors, on top of potential exposure to war violence, could lead to mental health difficulties among refugee children . Access to services that could help relieve some of these stresses may be useful to promote well-being in the children. We describe here a community art-based project led by a team of volunteer artists in collaboration with health, education and academic professionals, which was formed in April 2022 as an immediate response to the needs of newly arriving displaced Ukrainian families in London. The project consisted of five community art workshops that were open to Ukrainian refugee children. Each workshop was dedicated to creating or decorating art objects using mixed-media techniques. Similar to other community art projects in the displaced communities , this project was aimed at promoting enjoyment, health, socialisation, and health behaviour . Furthermore, the project was aimed to enable skill learning, creativity, and to build a renewed sense of agency , facilitated by the community art principle that emphasizes participatory, creative "experience" and social processes rather than the final aesthetic of the produced artwork, which provides a non-elitist entry into the arts for all community members . As well as being a direct response to the community needs, the workshops were created as a pilot for a larger community art project for Ukrainian refugee children and potentially other displaced communities in the future. To provide evidence-based support for future projects, we evaluated the feasibility , the acceptability of the workshops and the influence of the workshops on the well-being, health behaviour and socialisation and the children's enjoyment using secondary data collected from the refugee children and their parents. We hoped to provide information about the feasibility of conducting similar future community art-based projects catered for the needs of Ukrainian refugee children, to demonstrate the positive influence, or the absence of negative effect, of the workshop sessions on the well-being of the participating children as reported by themselves and their parents, and to finally reflect on any lessons learned from the available data. --- Methods --- Design This paper outlines secondary analyses of observational data gathered from the art-based community project to examine the feasibility, acceptability, and influence of the workshops on the well-being and healthy behaviour of the children . These workshops were delivered to non-randomised groups of Ukrainian children without control participants. --- Community art workshops The community-art workshops, with a working title My World Workshop, took place from June to July 2022 and were organised by a team of volunteer artists in collaboration with the Children and War UK, founded and curated by the first author MW. These workshops were carried out in a community art studio in West London. The workshops involved a team of independent artists and volunteers comprising school educators from a London-based Ukrainian School, clinicians, and some higher education researchers. The project consisted of five weekly, half-day workshop sessions, where children created or decorated art objects. These workshop sessions received input from different disciplines including art, health science, and pedagogy, and were critically informed by the Displacement Art movement and object theory . The children's experience of displacement and exposure to war were not explicitly disclosed, but there was an anticipation that the art objects and their creation might reflect thoughts or insights that were related to experiences of geographical dislocation and forced migration that might affect the children during the workshop. Therefore, each session was organised with Ukrainian-speaking trauma-specialist counsellors on site. The sessions were curated with the recurring themes of "home", and "hope", selected in recognition of the children's displacement, while both honouring the memories of home and nurturing positive experiences and belonging in their new environment, in accordance with the community-art principle, displacement art, and object theory . As such, the sessions were built around art activities that combined mark-making and object-making that reflect on both themes. During the workshop sessions, children were presented with blank, symbolic tabula rasa objects, paper, clay, or fabric, to creatively mark-make upon, such that the objects brim with meaning. In addition to the individual work, some exercises included the making of communal art pieces to encourage the children to create their community and sense of collective agency. Each session was focused on one art project: Workshop 1: Decorating a Key invited each child to ornately gild a lucky house key, and to draw/write a secret wish on a paper tag tied to the key; Workshop 2: Designing Own T-shirts gave children the opportunity to self-express on a white t-shirt serving as a blank canvas; which was accompanied by a group activity the Wish Catcher, where the children created and tied personal paper messages on a giant, communal mobile hoop, in keeping with Ukrainian folklore traditions of hanging strichka that marks celebratory feasts, symbolising fortune; Workshop 3: Working with Clay taught the children the skill of making Own Bunting enabled the children to create their own decorative style for their rooms at home, using collage, sewing, and fabrics; and finally Workshop 5: Percussion Music taught the children to decorate percussion instruments made from white plastic bottles filled with dried pulses. The finished instruments were used in a free-style jam session, led by a professional musician . --- --- Measures A range of validated and non-validated questionnaires and inventories were used during the project. Some measures were adapted and shortened from validated measures to reduce the burden on the children. The measures assessed the mental health and psychological well-being, socialisation, and enjoyment of the children, as well as the feasibility of the families attending the workshops. The measures were translated when a Ukrainian version was unavailable. Unless otherwise stated, all questionnaires were completed up to one week before and immediately after the completion of the 5-week workshops. --- FIGURE 1 Example Art works produced by the children. Photographs of the art works produced by some children: a t-shirt with a child's drawing of the war in Ukraine, clay "hugging" bowls made of air-drying clay and a clay pinch-bowl technique, a wish catcher with a message to hang onto the communal hoop with a strichka ribbon tagged with a drawing of a Ukrainian flag saying "We have won. We have peace", a percussion instrument made of a found material, in this case a plastic bottle, decorated with bows and stickers, and a child places haberdashery, beading, and lace onto fabric bunting triangles. Photographs: Michaela Wenkert 2022©. The self-rated CRIES-8 was used to measure the impacts of traumatic events in the forms of intrusive re-experiencing of the events and avoidance of reminders/feelings associated with those events . This 8-item scale was designed for children aged 8 years and above who can read independently. It contains 4 items measuring intrusion and 4 items measuring avoidance. It yields a total maximum score of 40, with higher scores indicating higher post-traumatic stress symptoms. --- Emotional symptoms questionnaire The Emotional Symptoms Questionnaire was adapted from the Adolescent Wellbeing Scale . The scale was chosen instead of the self-rated Strength and Difficulties Questionnaire below due to its accessible wordings for children. To reduce burden for the children, we used six items of the scale to explore their emotional symptoms: "I enjoy the things I do as much as I used to", "I like talking to my friends and family", "I feel like crying", "I feel very bored", "I look forward to things as much as I used to", "I am easily cheered up". The questionnaire yielded a maximum total score of 12, where higher scores indicate more negative emotions. --- Strength and difficulties questionnaire The SDQ , is a 25-item screening questionnaire assessing five domains of emotional symptoms, conduct problems, hyperactivity/inattention symptoms, peer relationship problems and prosocial behaviour. Negative emotional symptoms of the children, reported by their parents, were indexed using emotional symptoms domain of the SDQ with a total score ranging from 0 to 10, where higher scores indicate more negative emotions. --- Columbia impairment scale The CIS measured general impairment of different domains of functioning including relationships with family members and peers and academic performance at school. The parent-rated scale contained 12 items, and each was scored from 0 to 4 , leading to a maximum total score of 48, where higher scores indicated higher functional impairment. --- Sleep questionnaire The brief non-validated sleep questionnaire was created based on the prompts question of the Kiddie Schedule for Affective Disorders and Schizophrenia . The parent-rated questionnaire contains four items: "Does your child have problems falling asleep?", "Does your child wake up at night?", "How frequently does your child have nightmares?", and "Is it difficult to wake your child up in the morning?", each rated on a scale from 0 = "not at all", 1 = "1-2 days", 2 = "3-4 days" and 3 = "5 days or more", yielding a total maximum score of 12, where higher scores reflect more sleep problems. --- Socialisation questionnaire The brief 5-item non-validated Socialisation Questionnaire explored the workshop's role in encouraging and facilitating socialisation in the children. The inventory contained 5 items: "I like spending time with others in the workshops", "The workshops help me make new friends", "The workshops make me feel less lonely", "I talk more often/with more people during the week", and "I am confident in new situation", rated 0 = "True", 1 = "Somewhat True", 2 = "Not True". The total maximum score is 10, with higher scores indicating higher socialisation difficulties. --- Mood and motivation state questionnaire The Mood and Motivation State Questionnaire was adapted from Maurizio et al. and assessed the mood and motivational state of the children before and after activities in Sessions 2, 3 and 4. The self-rated questionnaire contained 4 items each consisting of opposite statements, from "Bored" to "Interested", "Bad Mood" to "Good Mood", "Tired" to "Awake", and "Stressed" to "Not Stressed", to be rated from 1 to 5, which will yield a total score up to 20, where higher scores indicate better mood or motivation. --- Workshop planning and procedure Prior planning of the workshops took place between April and June 2022 and consisted of meetings between MW, DO and representatives from the Ukrainian school and Children and War UK Charity, which helped formulate the workshops' objectives and required organisation. All members of the team received training on trauma provided by two qualified clinicians . The workshops were co-designed by artists, psychologists, and educational staff members, and underwent pre-piloting involving some Ukrainian families and the local community. Each workshop started with a "welcome" and some refreshments for the children and their parents. Then the children were led to the art studio while their parents had the opportunity to socialise in an adjoining welcome room. Children were seated typically in small groups of 5-8 around workshop tables. They were introduced to the session and given a demonstration of the workshop activity in Ukrainian. Then they were given art supplies and encouraged to freely engage in creative expressions. Up to ten volunteers took part in each session to support the children's activities. Several were Ukrainian speaking and provided communication translations between the children and the English-speaking volunteers. Most activities were easily demonstrated visually. Each session was concluded with refreshments and children were encouraged to bring their artwork home with them. Data collection was conducted by Ukrainian-speaking volunteers under the supervision of the first author SL. --- Analysis plan The influence of the workshops on trauma, mood, daily functioning, and sleep was explored by comparing these dependent variables before and after the workshop completion using a linear mixed model, with time as fixed effects, and compound symmetry covariance structure. Post-hoc explorations adding age group and time × age group or sex and time × sex to investigate whether age or sex moderate the pre-to post-workshop changes, i.e., whether the main interaction effects of time × age or time × sex as indicated by Fstatistics, were significant. As there were only few pre-school children with substantial missing data, we also carried out a sensitivity analyses including only school age children. Furthermore, the influence of each session on the dependent variable mood and motivational state of the participants was examined by comparing their ratings before and after each session using paired t-tests. All the children's experience of friendship and their and their parents' views about the workshops were assessed by combining the percentage ratings of "somewhat" and "yes" on each item denoting their presence. Free-text answers from the families were translated into English from Ukrainian and grouped according to themes and described narratively. All quantitative analyses were conducted with IBM SPSS Statistics for Windows 27 . --- Results --- Recruitment and data collection Fifty-three pupils were screened using the CRIES-8 by teachers, and 24 pupils with the higher scores were invited to take part in the community art workshops. Of the 24 pupils invited, eight declined to participate. These spaces were allocated to other pupils on a firstcome first-serve basis, resulting in a total of twenty-two participants in the workshops aged from 4 to 13 years . The children and their parents fully completed the questionnaires on site. However, there was no mechanism to gather data from the families outside the sessions. Therefore, datasets were completely missing from the families when they missed a session. --- Acceptability of the workshops 3.2.1. Workshop attendance The families attended on average 3.1 of 5 sessions on average , with 60.8% of the participants attending at least three out of 5 sessions. Sessions 1-5 were attended by 14, 15, 16, 13 and 15 children, respectively. Data from n = 15 parents indicated that the children missed sessions primarily due to sickness and travelling . Anecdotal information was provided by school staff who stated that a few families missed some sessions due to parents' attendance at peace protests against the war. --- Children's and parents' views about the workshops Most children who provided data at the end of the workshops reported that they enjoyed the activities , and that the activities helped them forget about stressful things . They felt that they have learnt new skills . Only one child reported that the workshops were somewhat too long . Most wanted to attend more workshops and would recommend the workshops to others . Free-text feedback was given by ten child participants, mostly related to the workshops' contents and the desire to have more sessions . Lastly, the three most popular workshops among the children were: "Designing Own Tshirts", "Percussion Music", and "Working with Clay". All parents providing data at the end of the workshops found it easy to register their children into the workshop , and to fit in the workshops around other family activities . The workshops did not interfere with the children's other commitments . Parents thought that the activities were exciting and that their children appeared to enjoy and to derive positive influence from the workshops . They would recommend the workshops to other parents . Parents' free-text suggestions revolved around similar themes as their children, with the theme of satisfaction and/or gratitude being the most dominant , followed by the desire to have more sessions . --- The workshops' influence on the children's well-being and socialisation Pre-and post-workshop ratings are presented in Table 2. Children self-reported impact of traumatic events score appeared lower at postrelative to pre-workshop period, although non-significantly, t = 2.04, p = .054. This was driven by lower intrusive re-experiencing of traumatic events, t = 2.51, p = .021 to the children postworkshops, but not by avoidance of reminders/feelings associated with the event, which did not differ significantly between pre-and post-workshops, t = 1.30, p = .21. The children also reported significantly reduced negative emotion, t = 3.50, p = .006. Parents reported the children to have significantly reduced negative emotion, t = 2.19, p = .043, and sleep disturbance, t = 2.96, p = .015. Finally, they also reported a weak but nonsignificant reduction of functional difficulties from pre-to postworkshops, t = 1.999, p = .073 . Post-hoc analyses revealed no significant interaction between age group or sex and the pre-and post-workshop changes of the dependent variables, F age × time ≤ 1.32; ps age × time = .39-.95, and F sex × time ≤ 1.31; ps sex × time = .28-.70. Sensitivity analyses involving only school-age children revealed a pattern of diminishing workshop influence which remained significant only in the context of self-reported negative emotion, t = 3.09, p = .015, and parent-reported sleep disturbance, t = 2.80, p = .021, and non-significant for all other measures, i.e., parent-reported reduction of the children's negative emotion, t = 1.99, p = .068 and functional difficulties from pre-to post-workshops, t = 1.96, p = .079, and selfreported of trauma, t = 1.62, p = .12, and intrusive trauma Children's and Parents' views about My World Workshops. Children's and parents' views on the workshops and how they can be further improved. Darker blue/red indicates "true", lighter blue/pink indicates "somewhat" and grey indicates "false". TABLE 1 Participants views on how to further improve the workshops. --- Themes n Quotes Children's views Workshop content suggestion 3 "To have more stories and ideas", "entertainment for older teenagers", "to have more supplementary materials". More sessions 3 "To learn more", "continue", "the first one, too little". Satisfaction/gratitude 2 "I am happy with everything", "thank you" Not knowing what to improve/ no improvement needed 2 "I do not know", "nothing" Parents' views Workshop content suggestion 1 "Go out to the yard/open space' More sessions 2 "We'd like to attend more-It's not easy to say goodbye", "continue" Satisfaction/gratitude 6 "I was pleased with everything", "all super", "I enjoyed it all-very interesting session", "all super, thanks", "all is good, thank you", "all is well" No improvement needed 1 "Nothing was needed everything was fine". Themes arising from the free text answers from the participants and their families. Child report of negative emotions were measured using the adapted Adolescent Wellbeing Scale, while the Parent report of their children's negative emotions were measured using the score of emotional symptoms of the Strength and Difficulties Questionnaire. Lukito et al. 10.3389/frcha.2023.1260189 re-experiencing, t = 1.99, p = .063, and avoidance, t = 1.01, p = .33. Children reported their mood and motivation scores increasing after compared to before the activities in Session 2 , t = 4.01, p = .001, and Session 4 , t = 2.60, p = .023. Their scores did not increase significantly in Session 3 , t = 1.97, p = .068 although the change of mean scores was in the expected direction. Post-workshops, children reported that they liked spending time with others in the workshops , that the workshops helped them make new friends , feel less lonely , talk more often/with more people during the week , and feel more confident in new situations . --- Discussion This study aimed to provide an evaluation of the feasibility, the acceptability, and the influence of the pilot communityart project My World Workshops on the psychological well-being, health behaviour and socialisation of the Ukrainian refugee children and their families, using secondary data analyses collected before, during and after the workshops. The workshops showed an overall good level of feasibility for recruitment and data collection. There was a sufficiently high level of interest from the families to take part in the workshops, such that nearly all workshop spaces were taken even though some potential participants declined to take part during the initial invitation. Completion of the questionnaires could pose a significant burden to children, but this appeared not to be the case during the data collection for the workshops. Therefore, it was promising that all participants, including the younger children, could complete all the questionnaires fully. The success could be attributed to the selection of shorter measures and the administration of the questionnaires by Ukrainian-speaking volunteers on sites. However, the absence of data collection mechanism outside sessions did lead to some missing data. Gathering data outside sessions is essential and is recommended for future workshops. This is likely to be a joint effort between school staff and Ukrainian-speaking volunteers/researchers. The workshops showed an adequate level of acceptability from the point of view of attendance, with over 60% of participants attending at least three of the five sessions. More importantly, the families reported overwhelmingly positive views about the workshops. Most children derived enjoyment from the workshop Psychological well-being and socialisation of the participants. Reports of child-rated impacts of traumatic events and negative emotions, of parentrated sleep problems, negative emotions, and functional impairments measured pre-and post-workshops, mood and motivational state before and after a session, and socialisation of the children post-workshops. Child-rated measures are coloured blue, parent-rated measures about the children are coloured pink. Darker blue/red indicates "true", lighter blue/pink indicates "somewhat" and grey indicates "false". Lukito et al. 10.3389/frcha.2023.1260189 activities, which helped alleviate stresses and allowed them to learn new skills as was the case in previous qualitative reports in displaced communities . Most of the children expressed wishes to have further workshops and would recommend them to others. Parents indicated that it was not difficult to register their children into the workshops and to fit them in around other family activities. All parents viewed the workshops as exciting, enjoyable, and benefitting their children, and would recommend the community art workshops to others. Our findings regarding the influence of the community artbased workshops on the children's psychological well-being, health behaviour and socialisation are generally positive, in line with past findings of the use of creative arts to increase wellbeing in young people [as reviewed by ]. The art form used in the studies included in these reviews were typically performative, presumably chosen due to their appeal to older adolescents aged 11-18 years for whom the interventions were designed. This resonates with the free-text suggestion to include "entertainment for older teenagers" from one workshop participant. Should these workshops be made available for older adolescents, expanding the art medium of creative expressions to suit the wider age range might be beneficial. The brief community-art workshops differed from the more intensive and/or longer-term art therapy provided for traumatised child refugees of a similar age range . Nevertheless, these workshops appear promising for providing relief from trauma, and improving mood and sleep disturbances among child refugees with elevated trauma scores. Each workshop session, particularly in Sessions 2 and 4, is also associated with improved mood and motivational states after relative to before activities, indicating the rapid influence of the workshops on the children's emotional well-being. Taken together, these findings suggest that the community-art based workshops will find applicability in children who do not otherwise meet the threshold for receiving clinical care, either to influence their well-being in the longer term or to provide short-term emotional relief. The influence of these workshops on the well-being of the children appeared not moderated by sex or age groups , suggesting the versatility of the approach to different child subpopulations. When the analysis was repeated to include only school-age children, who constitute the largest portion of the sample, our findings suggest a downward trend in the influence of these workshops on children's well-being. However, it's important to note that such a trend might be a result of reduced statistical power for detecting effects in smaller subpopulations. Replicating the analyses with a larger sample that includes a more even distribution of preschool and school-age children is necessary to arrive at a more robust conclusion. Finally, most children reported more socialisation and confidence after the completion of the workshops. Overall, the workshop completion thus appears to be associated with increased skill learning, creativity, and a renewed sense of agency, which fulfilled the core aim of the project . There are some limiting factors to the interpretation of these findings. First, the absence of mechanisms for collecting data outside sessions from the family have led to missing data from participants who were absent from some sessions. The problem of missing data was partly mitigated by the linear mixed modelling statistical approach, but missing data could be associated with self-selection, which biases the conclusion of this evaluation in a positive direction. Second, our evaluation was based on observation from a small sample of participants, which limited the power for detecting smaller effects including potentially any moderating influences of subpopulation characteristics on the pre-to post-workshop changes. Third, while the association between the completion of these community-art-based workshops and the children's psychological well-being and health behaviour is promising, such potential benefits might still be attributable to non-specific developmental changes over time due to the absence of a control group. Future studies thus may consider the inclusion of an appropriate control group to establish more rigorous evidence for the specific benefits of these workshops for the children. --- Conclusion In conclusion, based on the available data, community art-based workshops appear to be feasible and acceptable, and they exert a positive influence on the psychological well-being, sleep behaviour and socialisation of Ukrainian refugee children that do not seem to be moderated by sex or age groups of the participants. The artist-led community-art workshops are valuable for Ukrainian refugee children. A provision for a larger number of participants should be considered, with an appropriate experimental control to yield rigorous evidence of the workshops' positive influence on the well-being of the refugee children. --- Author's note MW is an Independent Artist and Researcher in London, United Kingdom. LT is a clinical psychologist and trustee at Children and War UK, London, United Kingdom, and WY is Emeritus Professor of Applied Child Psychology at the Institute of Psychiatry, Psychology and Neuroscience, King's College London, United Kingdom, and trustee at Children and War UK, London, United Kingdom. participation in this study was provided by the participants' legal guardians/next of kin. Written informed consent was obtained from the individual for the publication of any potentially identifiable images or data included in this article. --- Data availability statement The data are available from the first authors upon reasonable request. Requests to access these datasets should be directed to [email protected]. --- Ethics statement The studies involving humans were approved by King --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Objectives: Approximately 8 million Ukrainians have been displaced by the war in Ukraine and five million children had their education disrupted. Here, we report an evaluation of (1) the feasibility (i.e., recruitment), (2) the acceptability (i.e., attendance, participants' views) and (3) the influence of a pilot community artbased project on the well-being, health behaviour and socialisation of Ukrainian refugee children in London, UK. Methods: Twenty-two refugee children aged 4-14 years from St Mary's Ukrainian school in London took part in five weekly art workshop group sessions led by a team of volunteer independent artists based in a community art studio in West London in collaboration with Children and War UK. Analyses were conducted on measures of the children's psychological well-being, health behaviour, and socialisation; collected from participating children and their parents through the workshops. Results: The community art workshops received sufficient interest from parents during recruitment. Child participants and their parents expressed overwhelmingly positive views and high satisfaction towards the workshops and their activities. While the workshops were conducted without a control group, changes in psychological well-being and health behaviour and socialisation were in the expected direction. The workshops were associated with reduced impact of intrusive re-experiencing of traumatic events ( p = .021), negative emotion ( ps = .006-.043; rated by children and by their parents, respectively), and sleep disturbance ( p = .015). Mood and motivational states increase relative to before activities within sessions ( ps = .001-.023). Conclusions: The artist-led workshops are a valuable community project associated with high satisfaction and potentially increased well-being in Ukrainian refugee children. A provision for a larger number of participants should be considered.

Introduction National trends show that African American adolescents, relative to most other demographic groups, are currently more religious, and show relatively fewer declines, despite drastic decreases in religiosity among youth in the United States over the past 25 years . Although important to recognize given the emphasis on religiosity within the African American population , these broad findings may be limiting because they fail to acknowledge diverse patterns of religiosity among African American adolescents. This omission is critical because of the known protective effects religiosity confers on the mental health and psychological and social wellbeing of African American youth . In a recently published paper , three different profiles of intrinsic religiosity were identified in a lowincome African American adolescent sample, attesting to religious heterogeneity. Differences in well-being, specifically goal-directedness, coping, emotion management, and life satisfaction, across the religiosity profiles were reported in that study, however parent and family correlates of these profiles were not examined. Given the importance of family in African American culture , the principal role of families in shaping the religious development of their children , and the fact that there have been few empirical investigations of the transmission of religiosity within African American families , the current study focused on associations of maternal and family factors with these previously identified profiles of religiosity in this low-income African American adolescent sample. Although adolescents are given increasing autonomy as they age , parents still exert a great deal of influence on adolescents, particularly in areas of values. With respect to adolescent religiosity, parents may influence their children through a number of mechanisms including modeling, the religious activities and practices in which they engage, and through the family context, including the general environment of the home and the quality of the parent-adolescent relationship . The present study builds on two lines of work, both of which touch on the mechanisms of modeling and family context: research focused on parent and family influences on African American adolescents' religiosity generally, and research focused on the benefits of religiosity for low-income, urban African American youth. The first line of work cuts across African American families at all socioeconomic levels, and taps into cultural values important for many African Americans . The second line of work specifically tests the potential protective effects of religious involvement for low-income African American youth. As noted by Taylor and Chatters , a majority of African Americans report that religion is critical to their lives. African Americans also emphasize interdependence and connectedness among family members, and care and concern for all members of the family, including extended family, strong work ethics, reciprocity, restraint, and reverence or respect for authority . Although sparse, several researchers have conducted empirical investigations that reflect this cultural focus or have examined linkages between African American parents' religiosity -beliefs, practices, or a combination of the two -and adolescents' religiosity. In a study of 166 early adolescents and their mothers from diverse socioeconomic backgrounds, almost half of whom were African American, Laird et al. assessed maternal and adolescent views of the personal importance of religion and the frequency of their attendance at religious services. They also linked these constructs to adolescent outcomes. Maternal and adolescent importance and attendance were highly correlated. The importance adolescents ascribed to religion and their religious attendance was associated with higher levels of self-control, which in turn was associated with lower levels of antisocial and rule-breaking behavior. In a similar study with 154 African American Christian middle-class adolescent-parent dyads , Butler-Barnes et al. examined associations between parents' religious socializationassessed with a 5-item measure encompassing both religious beliefs and practices -and adolescents' self-reported relationship with God, communication with God, and psychological well-being. Although parents' religious socialization was not associated with adolescents' relationship with God or communication with God in correlational analyses, religious socialization interacted with adolescents' reported relationship with God and communication with God to affect psychological well-being. This was particularly evident for the model with communication with God. Communication with God was positively associated with psychological well-being, but having parents with moderate to high levels of religious socialization enhanced this effect. Finally, Halgunseth et al. interviewed African American adolescents and parents from 130 two-parent households in a study that explicitly examined the roles of both mothers' and fathers' religiosity on African American adolescents' religious beliefs and practices. Maternal religious beliefs and practices were associated with the beliefs and practices of both sons and daughters. However, fathers' beliefs were only associated with sons' beliefs; fathers' practices were linked to the practices of both sons and daughters, but were stronger for daughters than for sons. The above empirical studies illustrate that African American parents from diverse socioeconomic backgrounds communicate their religious beliefs and practices to their children. Further, evidence suggests that this communication affects their offspring, either by shaping the religious beliefs and practices of children in the home, altering the family environment, directly affecting adolescents' behavior, or some combination of these mechanisms. In addition to research that emphasizes the relevance of religion to African Americans across the socioeconomic spectrum, the current study builds on a second line of work focused on the benefits of religiosity for low-income, urban African American youth. This line of research acknowledges the heightened levels of stress to which low-income, urban, African American youth are exposed, especially compared to non-minority youth, and specifically tests the potential protective effects of religious involvement for these youth. A limited number of empirical studies have tested the protective function of religious constructs in samples of low-income, urban, African American adolescents. Among them, Grant et al. tested the protective effects of religious involvement in the relation between life stressors and internalizing symptoms in a sample of 224 low-income African American middle school youth. Religious involvement mitigated the impact of life stressors on symptoms for girls, but not boys. In a second study, Butler-Barnes et al. examined religious coping as one of two protective factors in the relation between exposure to community violence and achievement motivation beliefs in a sample of 380 low-income African American high school students who were at risk for academic failure. Butler-Barnes et al. found that in their sample of low-income youth, religious coping had both main effects and protective effects. That is, youth who reported relying on their faith to cope in times of difficulty had higher overall levels of achievement motivation beliefs. Additionally, for males, but not females, religious coping helped to mitigate the negative impacts of exposure to community violence on achievement motivation. These studies, while limited, illustrate the potential promotive and protective effects of adolescent religiosity for African American youth who are at risk for high levels of exposure to stressors. A limitation of the extant literature, however, is the lack of empirical data on associations between parent religiosity and adolescent religiosity among low-income, urban African American samples. --- The Current Study The present study addressed this gap in the literature, building on work that described three distinctive profiles of intrinsic religiosity during adolescence in a sample of low-income, urban, predominantly African American youth . Operating from a socialization perspective , the present study examined multiple maternal and family influences on adolescent intrinsic religiosity, both for their independent and unique contributions. Intrinsic religiosity, in contrast to extrinsic religiosity, refers to religious beliefs and private religious practices ; these beliefs and practices generally demonstrate a commitment or devotion to the divine , and are not as overt or observable as extrinsic religiosity. It was anticipated that mothers of youth whose intrinsic religiosity remained high would have higher levels of religious attendance, religious coping, and religious commitment than mothers of youth whose religiosity was low and remained low or mothers of youth whose religiosity declined over the study period. It also was anticipated that youth high in religiosity would report higher levels of felt acceptance by mothers than youth in the other two profiles. In terms of emotional climate, it was anticipated that mothers of youth high in religiosity would report a more favorable emotional climate. That is, lower hostility, higher positive emotions at home, and lower negative emotions at home, than mothers of youth whose religiosity declined or whose religiosity was low. --- Method Participants Participants included 326 African American adolescents and their maternal caregivers who participated in a larger four-year longitudinal study on stress, coping, and adjustment, which took place in a mid-sized city in the mid-Atlantic region of the U.S. At baseline, 54% of the adolescents were female . Most of the maternal caregivers were the youth participants' biological mothers, but grandmothers , other female relatives , adopted mothers , stepmothers , and fathers' girlfriends also were represented in the sample. Most of the youth reported they lived with their biological mother most of the time; only 20.1% reported that they lived with their biological father most of the time. Family structure varied: 42.5% of caregivers never married, 31.7% were married or cohabitating, 23.7% were separated or divorced, and 2.1% were widowed. Socioeconomic status also varied, but most of the sample came from low SES backgrounds. Median weekly household income at time 1 was $301-400, with 17.6% reporting household earnings of $200 or less per week. The most common reports on maternal education level were no high school diploma , high school diploma or General Education Diploma or some college but no degree . Thirteen percent of the maternal caregivers had an Associate's or Vocational degree; only 8.3% of maternal caregivers had a bachelor's degree or higher. In the previous study youth had been classified, using Latent Class Growth Analysis , into three profiles based on their pattern of intrinsic religiosity over the four study years. Figure 1 presents these profiles for the African American youth in the current study, which represented 91% of the original sample. These profiles included groups with: low levels of intrinsic religiosity that changed little over the study period , levels of intrinsic religiosity that began moderately high but declined significantly over the study period , and levels of intrinsic religiosity that began high and remained high . Only participants identifying as African American were included in the present study in order to evaluate the contributions of maternal and family factors to African American adolescent intrinsic religiosity. Data were collected between December 27, 2004 andJune 17, 2009. --- Procedure The Institutional Review Board at the first author's University approved all study procedures for the larger investigation. Families were recruited through community events and agencies, by participant referral, and through flyers distributed in low-income neighborhoods. To be eligible, participants had to be the female caregiver of a 5th or 8th grade child during the first time point of data collection, and speak English. Sixty-three percent of the eligible families who were approached enrolled in the study. This figure is better than those of many community-based studies for recruiting participants from disadvantaged neighborhoods . Caregivers provided written consent and adolescents provided written assent for participation in the study. Interviews were conducted face to-face with visual aids in families' homes, unless otherwise requested by the family. Two trained research team members interviewed the caregiver and adolescent separately in different rooms. The research team was diverse and included both white and African American interviewers and male and female interviewers. Comparisons to assess for systematic biases by interviewer race or gender revealed none . Interviews lasted approximately 2.5 hours and families were compensated $50 in Wal-Mart gift cards for each time of participation. --- Measures Demographics-Demographics assessed in the study and used as covariates in the analyses included adolescent gender and grade level, both reported by the adolescent. Maternal education and weekly household income from all sources were completed by the maternal caregiver. Response options for maternal education included: no diploma; high school diploma; GED; some college, no degree; associate's degree; vocational degree ; bachelor's degree; master's degree; and advanced degree . Response options for weekly household income included: less than $100 per week; $101-200 per week; $201-300 per week; $301-400 per week; $401-500 per week; $501-600 per week; $601-700 per week; $701-800 per week; $801-900 per week; and $901 or more per week. Maternal caregiver religious attendance-At each time point, caregivers responded to the question "How often do you attend religious services?" using the following response scale: 0 , 1 , 2 , 3 , 4 , 5 , 6 , or 7 . Maternal caregiver religious commitment-Religious commitment was assessed at each time point with the 10-item Religious Commitment Inventory-10 which is appropriate for individuals in a variety of faiths. A sample item is "I spend time trying to grow in understanding of my faith." Caregivers rated each item on a scale ranging from 1 to 5 . Worthington et al. report good reliability and validity. We did not collect data on the specific faiths with which individuals were affiliated. Cronbach alphas in the present study ranged from .95 to . 96 across time points. Maternal caregiver religious coping-Coping by turning to religion was assessed at each time point with a 4-item subscale from the COPE . A sample item is "I try to find comfort in my religion." Caregivers responded to items using a scale ranging from 1 to 4 . Carver et al. provide extensive reliability and validity information on the COPE. Cronbach alphas in the present study ranged from .84 to .92 across time points. Maternal hostility-Maternal hostility was self-reported at each time point with the fiveitem subscale from the Brief Symptom Inventory . A sample item is "Feeling easily annoyed or irritated." Items were rated on a 5-point scale from 1 to 5 with higher scale scores indicating higher symptom levels. The BSI, a brief form of the Symptom Checklist , is a valid and reliable measure of symptomatology . Cronbach alphas in the present study ranged from .74 to .80 across time points. Felt acceptance from maternal caregiver-At each time point adolescents reported on the extent to which they felt accepted by their maternal caregiver using the 20-item parental acceptance-rejection subscale of the Child Report of Parent Behavior Inventory . A sample item on the CRPBI is "Understands your problems and worries." The CRPBI has good discriminate validity , distinguishing between delinquent and non-delinquent youth. Using a 3-point Likert scale, adolescents rated the extent to which they felt the statements were representative of their maternal caregiver/mother, from 1 to 3 . Items were reverse coded so that higher scores indicated greater felt acceptance. Cronbach alphas in the present study ranged from .86 to .91 across time points. Family emotional climate-The 10-item negative dominant and the 10-item positive dominant subscales of the Family Expressiveness Questionnaire , reported by maternal caregivers at each time point, were used to assess family emotional climate. Items represent a range of negative and positive emotions and behavioral actions typical of many families. A sample item from the positive dominant subscale is "Expressing deep affection or love for someone" and a sample item from the negative dominant subscale is "Quarreling with a family member." Items are rated on a 9-point scale ranging from 1 to 9 . Caregivers were instructed to complete the measure with respect to the family with whom they currently lived, including themselves, their spouses/partners , children, and other individuals living in the household. Halberstadt reports good reliability and validity for the measure. Cronbach alphas in the present study ranged from .81 to .84 for positive emotional climate and .78 to .83 for negative emotional climate across time points. --- Data Analyses An initial series of repeated measures analyses of variance were conducted for each predictor to determine if there were any effects of time, or if the construct could be collapsed across time. These models were run with and without imputing data to determine how imputation might affect the results. Profile membership was the predictor in these models, and adolescent gender, grade level, household income, and maternal education were included as covariates. There were no significant effects of time, no significant interactions of time with profile, nor any substantive differences between the models run with and without imputed data for any of the seven predictors. Based on these analyses, summary scores were computed for each of the seven predictor variables that averaged data across the four study time points using all available data. See supplemental Table 1 for a summary of these analyses. Descriptive analyses were conducted first, including comparisons of means of the predictor variables across each profile, and inter-correlations among predictors. Multivariate analyses were conducted next, in order to evaluate the unique contributions of maternal religious constructs and emotional climate variables to intrinsic religiosity profile membership, after accounting for demographic covariates. Two sets of logistic regression analyses were conducted, with three models in each set. The first model included demographic covariates . The second model included demographic covariates and maternal religious constructs . The third model added family climate variables . This approach enabled a comparison of the relative contribution of maternal religious constructs over and above demographic covariates, and of family context variables over and above demographics and maternal religious constructs. In the first set of analyses, youth in the low religiosity profile were compared with youth in the high religiosity profile . In the second set of analyses, youth in the decliner religiosity profile were compared with youth in the high religiosity profile . --- Results --- Attrition Analyses Seventy percent of the sample was retained across the four study time points. Adolescents who had data at all four time points were compared with adolescents who were missing data at Time 4 on adolescent gender and grade level using Chi square analyses and on maternal education, household income, maternal religious attendance, commitment, and coping, maternal hostility, felt acceptance from parents, and positive and negative emotional climate in the home at Time 1 using independent samples t-tests. Chi square analyses indicated that adolescents in middle school were more likely to remain in the study than adolescents in high school [X 2 = 5.51, p = .02]; there were no differences by adolescent gender [X 2 = 3.13, p = .09]. There were no differences at Time 1 on maternal education, t = -0.93, p = .36, household income, t = 0.74, p = .46, maternal religious attendance, t = -0.80, p = .43, maternal religious commitment, t = -0.29, p = .77, maternal religious coping, t = -0.21, p = .83, maternal hostility, t = .03, p = .98, positive emotional family climate, t = -1.67, p = .10, negative emotional family climate, t = 0, p = 1.00, or felt acceptance from parents, t = 0.69, p = .40. --- Descriptive Information on the Study Constructs Table 1 presents descriptive information on the religious and emotional climate predictor variables by religiosity profile. There was a significant mean group difference on maternal religious attendance, F = 10.63, p < .001. Youth with low levels of religiosity , and youth who declined in their religiosity differed from youth with high levels of religiosity. There also was a significant mean group difference on maternal religious commitment, F = 9.54, p < .001. In this case, youth with low levels of religiosity differed from youth who declined or who had high levels of religiosity. Mean group differences were significant for parental acceptance, F = 11.97, p < .001, and the patterns mirrored those for religious attendance: youth with low or declining levels of religiosity differed from youth with high levels of religiosity. Mean group differences were observed for both a positive, F = 3.69, p = . 026, and negative F = 6.12, p = .002 home emotional climate. For a positive home emotional climate, differences were observed been youth with low and high levels of religiosity ; for a negative home emotional climate, differences were observed between youth with declining levels and youth with high levels of religiosity . There were no significant mean group differences on parental religious coping, F = 2.74, p = .066, or parental hostility, F = 2.81, p = .062. Table 2 presents correlations among the predictor variables in the study. The predictors were uncorrelated with adolescent gender , and only felt acceptance from mothers was correlated with grade level, r = -.15, p = .008, indicating that younger adolescents felt more accepted. Maternal education was significantly positively correlated with religious attendance , religious commitment , religious coping , and a positive emotional climate in the home . Household income was significantly positively correlated with religious attendance and significantly negatively associated with maternal hostility . --- Multivariate Analyses Table 3 presents the results of the logistic regression models predicting membership in low versus high religiosity profiles from demographic, maternal religiosity, and family climate variables. As seen in table 3, in Model 1, both higher maternal education level and lower grade level were associated with membership in the high versus low religiosity profile. In Model 2, grade level remained significant, and none of the maternal religious constructs was significant. In Model 3, felt acceptance from mothers, in addition to grade level were uniquely associated with membership in the high versus low religiosity profile. Table 4 presents the parallel results for the models comparing membership in high versus decliner religiosity profiles from demographic, maternal religiosity, and family climate variables. As seen in the table, in Model 1, higher maternal education, lower grade level, and being female were associated with membership in the high religiosity profile versus the decliner profile. In Model 2, adolescent gender and grade level remained significant, and higher maternal attendance at religious services was associated with membership in the high versus decliner religiosity profile. In Model 3, adolescent gender, grade level, maternal religious attendance, and felt acceptance from mothers were uniquely associated with membership in the high versus decliner religious profile. In a final exploratory analysis we repeated the above analyses, comparing membership in the decliner religiosity and low religiosity profiles. The only significant term in the model that differentiated these two profiles was maternal religious commitment . --- Sensitivity Analyses As noted in the Data Analyses section, the first set of analyses for the manuscript included repeated measures ANCOVAs in order to determine if there were any significant changes in any of the seven predictor variables over the four time points of the study , or if religiosity profile membership interacted with time . These analyses were run both with and without imputed data to see how data imputation affected the results. There were no significant effects of time, no significant interactions of time with profile, nor any substantive differences between the models run with and without imputed data for any of the seven predictors. Therefore, each of the seven predictor variables were collapsed across time using all of the available data. --- Discussion As a group, African American adolescents are more religious than their peers, and have experienced fewer declines in religiosity despite a historic rise in secularism in American culture over the past 25 years. Yet, broad characterizations of this demographic fail to acknowledge the diverse patterns of religiosity among African American youth. Further, there is a paucity of empirical investigations focused on the transmission of religiosity within African American families. A recent study with low-income African American teens identified three distinct profiles of intrinsic religiosity; operating from a socialization perspective, the present investigation addressed gaps in the literature by examining multiple maternal and family influences as contributors to religious profiles identified among adolescents in the previous study. Both the independent and unique contributions of these influences were assessed. Univariate analyses revealed that the modeling variables -specifically maternal religious attendance and religious commitment -as well as family climate variables -specifically felt acceptance from mothers and the emotional climate of the home -differed across the three adolescent religiosity profiles, but in unique ways. At the univariate level, maternal religious attendance and maternal acceptance appeared to be important in differentiating both the low and declining religiosity profiles from the high religiosity profile. In contrast, maternal religious commitment appeared to matter in differentiating the low religiosity profile from the declining and high religiosity profiles. Family climate operated in different ways, depending on whether the climate was positive or negative. A positive emotional home environment distinguished low versus high religiosity profiles, while a negative emotional home environment distinguished decliner versus high religiosity profiles. These findings highlight the importance of taking a multi-faceted approach when studying the transmission of religiosity within African American families . These findings also are consistent with reports showing that maternal religiosity is crucial to the religious socialization of their children . In an intergenerational study of religious socialization, Gutierrez et al. found that family religious socialization predicted adult self-reports of religious importance, commitment and, spirituality beyond the effects of control variables such as age, gender, education, and household income. Furthermore, they found that maternal religious influence contributed the greatest proportion of the total explained variance in the model. Moreover, within the family religious influence variables that were explored , grandmothers' religious influence contributed the second largest proportion of the overall variance. This finding has been attributed to gender orientation roles, which describe women as the bearers of culture in households: responsible for transmitting important cultural traits and values to children . The findings in the current study extend these reports by identifying specific activities that affect the religious profiles of African American adolescents: maternal religious attendance and maternal level of religious commitment. Findings from the multivariate analyses revealed that, after accounting for demographic covariates, only felt acceptance from mothers uniquely differentiated adolescents in the low intrinsic religiosity profile from adolescents in the high religiosity profile, while both maternal religious attendance and felt acceptance from mothers uniquely differentiated adolescents in the decliner intrinsic religiosity profile from adolescents in the high religiosity profile. Findings on the significance of felt acceptance in differentiating adolescents in the three religiosity profiles are consistent with previous work highlighting the importance of perceived acceptance on general adolescent development. Perceived acceptance has been associated with better academic outcomes and psychosocial adjustment among adolescents . Findings in the present study contribute to this body of literature by illustrating how felt acceptance from mothers is associated with the religiosity profiles of African American adolescents. It is interesting to note that adolescents in both the high and decliner profiles started off with fairly high levels of intrinsic religiosity. Youth who declined in their intrinsic religiosity, versus youth who remained high, came from homes rated by mothers as having a more negative emotional climate, reported feeling less accepted by their mothers, and had mothers who reported attending religious services less frequently. Conversely, youth who had high levels of intrinsic religiosity and remained high over the study period, reported feeling more accepted by their mothers and their mothers reported attending religious services more frequently than youth who declined in religiosity. The extent to which adolescents perceived their mothers' behavior as congruent or incongruent with general religious philosophies in the household likely affected the likelihood of adolescents experiencing declines in their own religious beliefs. Perceived congruence may facilitate stable religious development while incongruence may increase the likelihood of losing or forsaking previously ingrained or socialized practices. The linkage of congruence with stability in intrinsic religiosity and incongruence with a decline in intrinsic religiosity during adolescence may be directly connected to newly developed cognitive skills, specifically abstract thought, and the ability to think and reason about the possible. With the development of abstract thought, adolescents are much more critical of inconsistencies they see in the behaviors of others around them . These findings strongly suggest family context variables that are crucial for religious development, an important index of positive development, in African American adolescents. They reveal that beyond prescribed religious beliefs, behaviors also are essential: behaviors that foster a positive family climate at home promote stable religious development. However, exhibiting behaviors that are inconsistent with acceptable religious beliefs and practice, and behaviors that cultivate a negative family climate in the household, negatively affect the religious development of African American adolescents and increase the likelihood of witnessing declining religious beliefs and practices. This has implications for interventions targeting the volatile period of adolescence, which can be tasking for many family units. It also reinforces the importance of mothers' own behaviors, family dynamics, and practices in this stage of child growth and development. The study had a number of strengths. Methodologically, these included use of multiple indicators of maternal religious involvement as correlates of adolescent religiosity; use of multiple indicators of the quality of the family environment; focusing on intrinsic religiosity, which has been studied less frequently than extrinsic religiosity; focusing on different profiles of intrinsic religiosity over the course of adolescence; including both maternal and adolescent perspectives in the models; and focusing on a low-resourced sample that broadens the understanding of linkages between maternal influences on adolescent religiosity within this demographic. Despite these strengths, the exclusive focus on lowincome African American families limits the generalizability of the study. Second, due to budgetary constraints, fathers were not recruited into this study and therefore the influence of paternal religiosity on adolescent religiosity profiles could not be examined. Some studies examining the effects of parent religiosity patterns on adolescent religiosity show that specific paternal beliefs and behaviors may influence child behaviors . However, in comparison to the influence of maternal beliefs and behaviors, paternal beliefs and actions seem to carry less weight. Nevertheless, it is important for future studies to examine possible unique influences of paternal religiosity patterns on the outcomes of African American youth from low-resourced settings. Finally, this study used longitudinal data summed across time, strengthening the inferences that can be drawn from its results. However, the data is still correlational in nature and thus precludes the ability to make causal inferences. --- Conclusion African American adolescents are more likely than their peers in other racial/ethnic groups in the United States to subscribe to religious beliefs and practices . However, a focus on African Americans as a group fails to recognize the religious heterogeneity within this demographic ---an omission with real consequences given the known protective effects of religiosity on adolescent well-being. Further, there is a paucity of literature examining parent and family contributions to African American adolescent religiosity. Building on a previous study, the current investigation examined specific maternal and family correlates of three previously identified profiles of intrinsic religiosity among African American adolescents from low-resourced neighborhoods. The results suggested that both maternal behavior -attendance at religious services and religious commitment that evidenced in the form of concrete behaviors -as well as the emotional climate of the home -the extent to which adolescents felt loved, valued and accepted by their mothers, and the care and concern family members expressed to one anothercontributed to adolescents' intrinsic religiosity profiles. Feeling loved, valued, and accepted by one's mother in concert with having her model frequent attendance at religious services uniquely predicted whether adolescents "kept the faith" or declined in their intrinsic religiosity over the study period. These findings contribute to a more nuanced understanding of the role of family contexts in the socialization of religiosity in African Americans households and may help to explain factors associated with well-being and risk-taking in this demographic. --- Supplementary Material Refer to Web version on PubMed Central for supplementary material. Results from latent class growth analyses for African American adolescents, collapsed across grade level. This graph displays changes in means levels of intrinsic religiosity across the study period for the three identified profiles. --- Biography Wendy Kliewer is a Professor in the Department of Psychology at Virginia Commonwealth University. Dr. Kliewer's research centers on the broad theme of risk and resilience, with specific attention to cumulative stressors, their impacts on a broad array of functioning, and protective factors that mitigate risk. She has long-standing interests in interdisciplinary, cross-cultural research, and is committed to training the next generation of scholars to continue to do research that matters. Dr --- ---

National trends show that African American adolescents, relative to most other demographic groups, are more religious, and show fewer declines in religiosity, despite drastic decreases in religiosity among youth over the past 25 years. These broad findings are limiting because they fail to acknowledge religious heterogeneity among African American teens. Further, there are few empirical investigations of the transmission of religiosity within African American families.

INTRODUCTION Women's political participation is essential for the formation of democratic societies and the promotion of inclusive governance. Nevertheless, despite progress made in recent years, women's representation and participation in political processes remain a significant obstacle in many nations, including Nepal. Understanding the factors that influence women's political participation is crucial for developing strategies that effectively address existing barriers and promote gender equality in political decision-making. The deeply ingrained patriarchal norms and gender biases prevalent in Nepalese society are a significant influence in women's political participation. Often, traditional gender roles, cultural practices, and societal expectations restrict women's access to political spaces, hindering their ability to participate actively and effectively. In addition, socioeconomic factors, such as limited educational opportunities, economic disparities, and lack of financial resources, can impede women's political participation. According to recent data, Nepal has made some progress in increasing women's representation in politics, with women holding 33% of the seats in the federal parliament as of 2021 . However, women's participation in politics remains limited, especially in rural areas, and there are still significant barriers to their full engagement and leadership in the political sphere. Various studies have explored the factors that contribute to these barriers and challenges, including socio-cultural norms and expectations, economic and educational disparities, political bias and discrimination, and legal and institutional barriers . However, there is a need for further research that examines these factors in greater depth, particularly from the perspectives of women themselves. It aims to contribute to the development of evidence-based policies and interventions that promote gender equality, empower women in positions of political leadership, and nurture an inclusive and participatory democratic system. Nepal can make significant progress toward achieving gender parity in political representation, ensuring women's voices are heard, and utilizing the full potential of its diverse population to shape the country's future by addressing the barriers identified by this investigation. Therefore, this study aims to investigate the factors affecting women's political participation in Nepal through a qualitative approach, with a focus on the experiences and perspectives of women who have engaged in political activities at various levels. By exploring the lived experiences of women in politics, this study seeks to contribute to a better understanding of the challenges and opportunities for women's full participation and leadership in Nepal's political landscape. --- LITERATURE REVIEW Women's political participation in Nepal has been the subject of several studies in recent years. These studies have identified a range of factors that contribute to the underrepresentation and marginalization of women in politics. In this section, we review some of the key findings from these studies, organized around the themes of socio-cultural factors, economic and educational disparities, political bias and discrimination, and legal and institutional barriers. Socio-cultural factors, such as gender stereotypes, patriarchal norms, and traditional gender roles, have been identified as significant barriers to women's political participation in Nepal . For example, women are often expected to prioritize their domestic and family responsibilities over their political involvement, and they may face social stigma and discrimination if they deviate from these expectations . Additionally, women's political participation may be perceived as a threat to male authority and power, leading to resistance and backlash from male-dominated political and social spheres . Economic and educational disparities between men and women have also been found to hinder women's political participation in Nepal . Women are often disproportionately affected by poverty and lack of access to resources, which can limit their ability to participate in political activities and campaigns . Moreover, women's low levels of education and literacy can also impede their political participation and leadership, as they may lack the knowledge and skills necessary to engage in political discourse and decision-making . Political bias and discrimination against women have been identified as another significant barrier to their political participation in Nepal . For example, women may face gender-based violence, harassment, and intimidation when running for political office or participating in political activities . Additionally, women may be excluded from political decision-making processes and opportunities due to the male-dominated political culture and networks . Legal and institutional barriers to women's political participation in Nepal have also been identified in previous studies . For example, Nepal's electoral laws and regulations may not be gender-sensitive, and they may not adequately address the specific needs and challenges faced by women candidates . Moreover, women's participation in political parties and their representation in leadership positions may be limited by the lack of affirmative action policies and quotas . Previous studies have identified a range of factors that affect women's political participation in Nepal, including socio-cultural norms and expectations, economic and educational disparities, political bias and discrimination, and legal and institutional barriers. However, there is a need for further research that explores these factors in greater depth and from the perspectives of women themselves. --- METHODS This qualitative study aims to investigate the factors affecting women's political participation in Nepal. The study will be conducted through semi-structured interviews with 20 women who are currently or have previously held political positions in Nepal. Sampling will be done using purposive sampling to select participants who have experience and knowledge related to women's political participation. Inclusion criteria will be women who have held elected or appointed political positions at the local, provincial, or national levels. Participants will be recruited through snowball sampling, where initial participants will be asked to refer other eligible participants, they know. Data collection will be conducted through in-depth semi-structured interviews, which will be conducted in Nepali language. The interviews will be recorded and transcribed for data analysis. The interview questions will be developed based on the literature review and will cover the following topics: personal experiences of political participation, perceptions of barriers to political participation, perceptions of facilitators of political participation, and recommendations for improving women's political participation. The data will be analyzed using thematic analysis. The transcripts will be read multiple times to identify common themes and patterns. The themes will be developed through an iterative process of coding and categorization. The analysis will be conducted using NVivo software to aid in organizing and analyzing the data. Ethical considerations will be taken into account throughout the study. Informed consent will be obtained from all participants before the interviews. Participants will be informed of their rights to confidentiality and anonymity. The study will adhere to the ethical principles of the Belmont Report, which include respect for persons, beneficence, and justice. The study will contribute to the understanding of the factors affecting women's political participation in Nepal and provide insights for policymakers to improve women's participation in politics. --- RESULTS AND DISCUSSION Twenty women who have held elected or appointed political positions at the local, provincial, or national levels were interviewed for this study. The average age of the participants was 42 years old, with a range of 28-58 years. The majority of the participants were married , had completed higher education , and were from urban areas . Thematic analysis of the interview transcripts identified several key themes related to the factors affecting women's political participation in Nepal. The following themes were identified: Personal experiences of political participation: Most participants reported experiencing discrimination and gender bias in the political arena. They also reported facing challenges related to balancing their political responsibilities with their family and personal obligations. Barriers to political participation: The most commonly reported barriers to women's political participation were social and cultural norms that limit women's mobility and restrict their access to resources and opportunities. Other barriers included lack of support from family and community, lack of financial resources, and harassment and threats. Facilitators of political participation: Participants reported that having a supportive family and community, access to financial resources, and having positive role models were important facilitators of their political participation. Recommendations for improving women's political participation: Participants suggested several strategies for improving women's political participation, including increasing women's access to education and resources, implementing quotas for women's representation in political positions, and raising awareness about the importance of women's political participation. Overall, the findings suggest that women in Nepal face significant challenges to their political participation, including discrimination, gender bias, and social and cultural norms. However, there are also opportunities for improving women's participation, such as through targeted interventions to increase access to resources and support. These findings have important implications for policymakers and practitioners seeking to improve women's political participation in Nepal. The results of this study highlight the complex and multifaceted nature of the factors affecting women's political participation in Nepal. The findings are consistent with previous research that has identified social and cultural norms, discrimination, and lack of resources as major barriers to women's political participation . One of the most significant findings of this study is the importance of family and community support in facilitating women's political participation. This is consistent with previous research that has found that family and community support can be a critical factor in overcoming the barriers faced by women in politics . Policymakers and practitioners should consider strategies that engage families and communities in efforts to support women's political participation. The findings also suggest that policies such as quotas for women's representation in political positions can be effective in increasing women's participation. This is consistent with previous research that has found that quotas can help to increase women's representation and provide a platform for their voices to be heard . However, it is important to note that quotas alone are not sufficient to address the underlying social and cultural norms that limit women's political participation. Another important finding is the need to increase women's access to education and resources. This is consistent with previous research that has found that education is positively associated with women's political participation . Policymakers and practitioners should consider interventions that provide women with the skills and knowledge necessary to participate in politics and increase their access to financial and other resources. Finally, the findings highlight the need for continued efforts to raise awareness about the importance of women's political participation. This is consistent with previous research that has found that awareness-raising campaigns can help to change social norms and attitudes towards women in politics . Policymakers and practitioners should consider strategies that engage a range of stakeholders, including media, civil society organizations, and political parties, in efforts to promote women's political participation. In conclusion, the findings of this study provide important insights into the factors affecting women's political participation in Nepal. The study highlights the need for a multifaceted approach that addresses the social, cultural, and economic barriers to women's political participation. Policymakers and practitioners should consider a range of strategies, including family and community engagement, quotas, education and resource provision, and awarenessraising campaigns, in efforts to promote women's political participation and advance gender equality in Nepal. --- CONCLUSION Despite the constitutional and legal provisions and the significant progress made in recent years, women's political participation remains low due to several social, cultural, economic, and political factors. The study identified various challenges that hinder women's political participation, including gender discrimination, lack of education and awareness, patriarchal attitudes and beliefs, cultural norms and traditions, poverty, and political violence. Efforts to enhance women's political participation should focus on addressing these challenges by implementing policies and programs that promote gender equality, women's empowerment, and participation in decision-making processes. These initiatives should target women from marginalized groups, including Dalits, Janajatis, and Madhesis, who face multiple forms of discrimination and exclusion. Moreover, to ensure sustainable progress, these efforts should be supported by the active involvement of various stakeholders, including civil society organizations, political parties, government agencies, and the media. Overall, this study contributes to the existing literature on women's political participation in Nepal and provides insights for policymakers, practitioners, and scholars working in this field. Future research

This qualitative study investigates the factors affecting women's political participation in Nepal. The research employed semi-structured interviews with 15 female politicians and 10 male politicians to collect data. The findings indicate that despite constitutional and legal provisions, women's political participation remains low due to various social, cultural, economic, and political factors. Challenges include gender discrimination, lack of education and awareness, patriarchal attitudes and beliefs, cultural norms and traditions, poverty, and political violence. The study emphasizes the need for policy and program interventions that promote gender equality and women's empowerment to enhance women's political participation. The involvement of various stakeholders, including civil society organizations, political parties, government agencies, and the media, is crucial for sustainable progress.

Introduction Female breast cancer ranks as the first commonly occurring malignancy in China [1,2]. It poses a significant health threat to Chinese women. In 2020, 420,000 newly developed breast cancer cases were diagnosed in China, and 120,000 deaths were registered, accounting for 19.9% and 9.9% of the global female cancer incidence and mortality rate, respectively [1]. Because of the sustained socioeconomic development and rapid urbanization, China's contribution to the global breast cancer rates is increasing every year, especially in younger women [3]. Within the past two decades, breast cancer incidence among Chinese women has increased by 3-5% annually, which is significantly higher than the average annual global increase of 0.5% [4]. The age-standardized rate of breast cancer in China increased from 21 to 29 per 100,000 women between 2003 and 2011 [5], with large differences between urban and rural regions [6,7]. Obviously, breast cancer prevention and control are challenging in China. Early routine screening is considered imperative to reduce the mortality rate of this highly treatable malignant disease. The breast cancer mortality rate in the United States decreased by 40% between 1989 and 2017, mainly due to increased screening utilization [8]. Furthermore, a study by McCormack et al. [9] in 5 sub-Saharan African countries reported that an estimated 28-37% of breast cancer-related deaths can be prevented through early detection and adequate treatment. Therefore, close attention must be paid to the screening behaviors of women for early prevention and treatment of breast cancers. Screening for early detection of breast cancer has been widely promoted elsewhere [10][11][12], but it has not been prioritized in China. Currently, there has been no organized national breast-screening program; the Chinese government does not offer free screening services for women [13]. Additionally, breast cancer screenings are not covered by the Chinese universal healthcare insurance. The China Anti-Cancer Association has encouraged asymptomatic women older than 20 years to carry out BSE monthly [14]. Simultaneously, women older than 40 years, or younger if they have a high risk, are recommended to undergo mammography in combination with CBE or ultrasonography every 1 to 2 years [14]. However, breast cancer screening rates have remained low among Chinese women [15][16][17][18]. A survey conducted in 2010 reported that only 21.7% of 53,513 Chinese women aged 18-107 years had been screened [17]. A 2013 large-scale national study found that only 22.5% of Chinese women aged 35-69 years were screened [18]. Therefore, gaining a better understanding of the determinants of breast cancer screening is important to improve screening rates. Evidence from different countries and territories suggests that low socioeconomic status is a substantial barrier to the utilization of breast cancer screening [19][20][21][22][23]. Women with lower SES have access to a smaller pool of health promotion resources [24], which may limit their participation in health-promoting programs such as breast cancer screening. Furthermore, a study in China reported that patients with lower SES had a later breast cancer stage at the time of diagnosis and a poor overall prognosis [25]. Several socioeconomic factors, including lower income, education and occupation, residence in rural areas, and single marital status, have been determined to be strong predictors of low patronage of breast cancer screening services [19][20][21][22][23]. However, research on the influence of SES on breast cancer screening patronage among women in China is limited. SES is a robust determinant of health promotion programs, referring to the position of an individual in the social stratification system [26]. SES is generally measured using educational attainment, income, occupation, or a composite of these variables [27]. Using single SES variables can only improve the understanding of the relationship between a certain aspect of SES and breast cancer screening. However, SES is inherently complex and multidimensional [27]. Individual SES variables cannot capture the entire concept of SES [28]. The SES index is a composite measure, which can overcome the aforementioned problems [27]. It is generally constructed using a combination of SES variables [27]. Nevertheless, few studies have investigated the association between the composite SES and breast cancer screening patronage among Chinese women. It is also not clear whether the SES index is a stable variable and whether it changes the association of SES indicators with breast cancer screenings. This study aimed to assess the association between the SES index and female breast cancer screening practices in Eastern China, while considering the differences in screening utilization of women with different SES. We hypothesized that the SES index may significantly affect female breast cancer screening patronage, and women with high SES may be more likely to engage in breast cancer screenings. --- Materials and methods --- Design and participants We carried out this cross-sectional survey in Anhui, Eastern China. A snowball sampling strategy was utilized. Data on breast cancer screening practices questions for most countries were limited to women of reproductive age [17,22,23,[29][30][31][32][33]. In addition, given the increasing incidence of breast cancer and considering the onset age for breast cancer at younger age in Chinese women [3][4][5][6][7], we conducted this study targeting Chinese women aged 18 to 70. We distributed an anonymous web-based questionnaire through WeChat/Weixin between January 2020 and May 2021. We encouraged the respondents who had received the link to the survey to forward it to their friends and relatives. Before the survey, all participants read the study objectives and consent information and provided implied consent after clicking "Continue" to complete the online questionnaire. On average, it took 10-15 minutes for participants to complete the survey. Participants who completed the survey were rewarded with electronic cashable coupons. Participants who were registered but did not live in Anhui and had a history of breast cancer and cognitive impairment were excluded. Duplicated responses were blocked by checking the unique ID of each participant. De-identified data were collected and analyzed to ensure participant confidentiality. We computed the sample size using a 95% confidence interval and a statistical formula [34]. Considering a possible invalid response rate , a minimum sample size of 461 was determined. A total of 2,223 eligible participants were voluntarily enrolled in our study. Data from the 407 questionnaires were excluded due to missing values . Ultimately, we obtained 1,816 complete and valid questionnaires for further analyses, yielding an effective response rate of 81.69%. The sample size of this study exceeded the minimum requirement. A sufficient sample size ensures the credibility and repeatability of our results. The Ethics Committee of Bengbu Medical College reviewed and approved the protocol of this study . The study was conducted in accordance with the ethical standards of the institutional and national research committee and the Declaration of Helsinki. --- Measurements The initial questionnaire was developed based on our previous studies [16,35], and an extensive literature review [19][20][21][22][23]. We assessed content validity by sending the first draft of the questionnaire to two public health experts and two breast cancer experts. Each expert was invited to rate each item based on the relevance of the item content. They were also asked to give comments on the general formulations of the initial questionnaire. We conducted a small-scale pilot survey from a convenience sample of 90 women of different ages. Those who gave informed consent voluntarily and met the inclusion criteria were included. They were required to complete the questionnaire, to comment regarding the ease of understanding and acceptability of each item. We improved the questionnaire according to the expert's recommendations and the pre-survey results. The actual study samples did not include the pilot population. We conducted the exploratory factor analysis and reliability test to assess construct validity and internal consistency reliability. The Kaiser-Meyer-Olkin value of the questionnaire was 0.876, and bartlett's test of sphericity value for p was 0.000. EFA with iterated principal factor extraction resulted in 4 factors with initial eigenvalue estimates above 1.0 and 64.229% variance explained. The Cronbach's alpha coefficients was 0.780. A 2-week test-retest examination was conducted with the participation of 50 volunteers from a pilot sample. During this time, participants could not consult the replies they had given when the questionnaire was first administered. We used the Cohen's Kappa to assess test-retest reliability, and the Cohen's kappa value was 0.53-0.79. The final questionnaire included items for the sociodemographic profile, SES, breast cancer knowledge, and breast cancer screening practices. Sociodemographic data, including age, marital status, number of children, location of residence, education level, annual household income, and occupation, were collected. Participants were also asked whether they had received breast health education. SES was assessed using three direct indicators, which were education level, annual household income, and occupation, and one composite variable, which was the SES index. Five education level categories were used: junior high school or below, high school or vocational school, junior college, university, and masters and above [36]. We used three income levels to assess annual household income: <50,000 CNY , 50,000-120,000 CNY, and >120,000 CNY. Occupation was classified into two major groups: non-high-tech and hightech [37,38]. We conducted a factor analysis to construct the composite SES index using the three SES indicators [39]. We utilized a varimax rotation with Kaiser normalization. The principal component analysis was used as the extraction method [39][40][41]. The principal component analysis results showed that the Kaiser-Meyer-Olkin value was 0.661, and Bartlett's sphere test was significant . This indicated a correlation among three SES indicators. Eigenvalues of >1 were applied as the extraction criteria [41]. The results showed that only the first principal component was extracted, which explained 61.34% of the total variance variation. Each indicator score was multiplied by its standardized factor loading. Referring to previous studies [39][40][41], we calculated a composite SES index / 0.613. We applied the median to categorize the SES index scores into low, medium, and high. We used ten questions to assess breast cancer knowledge, covering breast cancer risk factors, possible symptoms, and screening strategies. The responses had to be "true" or "false". A score of 1 was assigned for the correct response, while incorrect responses were scored 0. Higher total knowledge scores represented higher levels of breast cancer knowledge. We used three questions to ask participants about their breast cancer screening experiences. The questions included the following: "Have you performed BSE regularly during the past 12 months?"; "Have you undergone CBE in the past 2 years?"; and "Have you received mammography screening within the past 2 years?". The answers had to be "Yes" or "No". BSE was defined as ''a woman examines her breasts". CBE was defined as "a health professional uses his/her hands to detect breast changes". Mammography was defined as "a radiographic imaging of the breast by a mammography machine". --- Statistical analysis We used IBM1 SPSS1 Statistics for data analysis. We applied χ 2 tests to observe the proportions of breast cancer screening programs by different types of personnel. We used the Kruskal-Wallis test to examine the differences in breast cancer knowledge because the data did not follow a normal distribution. We used a bivariate Spearman's correlation analysis to evaluate the potential correlations between socioeconomic factors and breast cancer screenings. We utilized multivariate logistic regression to estimate the association between the SES index and female breast cancer screening utilization, applying the forward stepwise method. Three multivariate regression analyses were performed to determine whether breast health education, marital status, and location of residence significantly affected breast cancer screening patronage after accounting for the SES index. The same order of entry was utilized for three sets of regression analyses, with the SES index entered in step 1, followed by the breast health education in step 2, and lastly, marital status and location of residence entered in step 3. Before introducing independent variables into the regression analysis, the variables were assessed for multicollinearity and interactions. Conservatively, the two-tailed p-value of <0.05 was considered statistically significant. --- Results --- Demographic information of participants The average age of 1,816 participants was 34.01 years and ranged from 19 to 70 years in this study. Of the total sample, 41.2% were university-educated, 86.3% had an annual household income of 50,000 or higher, 52.9% worked in high-tech jobs, 19.27% reported having practiced BSE, 12.89% indicated ever having undergone CBE, and 3.52% reported having received mammography. The rate of BSE was higher than those of CBE and mammography. In comparison, married urban women who had children received breast health education showed higher rates of breast cancer screening patronage. The results of the χ 2 tests showed that breast cancer screening patronage was significantly correlated with most of the sociodemographic variables. Except for mammography, the rates of BSE and CBE increased with the improvement of the SES index . Fig 1 shows the proportions of BSE, CBE, and mammography among women with different SES indexes. Another sociodemographic profile is reported in Table 1. The proportion of correct responses to all knowledge questions was higher for women with higher SES indexes than those with lower SES indexes . Furthermore, results from the Kruskal-Wallis test showed that the mean total knowledge score of women with high SES indexes was 7.96 , and it was significantly higher than the score of women with medium and low SES indexes. The comparison of breast cancer knowledge among women with different SES indexes is provided in Table 2. --- SES index and breast cancer knowledge --- Bivariate correlation analysis of socioeconomic factors and breast cancer screening Spearman's correlation analysis indicated that education level, income, occupation, and the composite SES index were all positively associated with the three breast cancer screening practices . Table 3 shows the specific Spearman correlation coefficients. --- Associations of the SES index with breast cancer screening practices Tables 4-6 summarizes the associations of the SES index with the three breast cancer screening practices. The results indicated that the SES index was statistically significantly positively associated with female breast cancer screening practices. Model 1 showed that women in the medium to high SES index groups had significantly higher odds of BSE and CBE than those in the low SES index group. However, women in the medium SES index group were more likely to undergo mammography. Specifically, women with medium and high SES indexes had 2.113 and 3.323 times the odds of performing BSE than women with low SES indexes; 1.863 and 3.320 times the odds of undergoing CBE than women with low SES index; and 3.739 and 3.453 times the odds of screening mammography than women with low SES indexes. Model 2 showed that the strength of the associations between the SES index and three breast cancer screening practices slightly decreased after the addition of the following item: "Ever received breast health education". However, the odds trend was consistent with the results of regression model 1. Controlling for other variables, breast health education was positively linked to female screening utilization. Specifically, women who received breast health education had 3.017 , 2.029 , and 1.880 times the odds of engaging in BSE, CBE, and mammogram practice than women who did not receive breast health education. Model 3 showed that the strength of the influence of the SES index on female breast cancer screening practices decreased after the addition of marital status and the location of residence based on model 2. However, the odds trend was still consistent with the results of model 1. Controlling for other variables, married urban women had higher odds of engaging in breast cancer screening than single rural women. Specifically, married women had 2.183 and 1.693 the odds of practicing BSE and CBE than unmarried women, and urban women had 1.484 , 1.725 , and 1.961 the odds of participating in BSE, CBE, and mammogram practice than rural women. --- Discussion To the best of our knowledge, this study provides the first evidence of the relationships between female SES and breast cancer screening practices using the composite SES index. Consistent with findings of previous studies [19][20][21][22][23], our results confirmed that socioeconomic disparities exist in breast cancer screening utilization in Eastern China. SES indicators were positively associated with the utilization of breast cancer screening. In addition, we found that the composite SES index was an important factor related to female breast cancer screening practices. Our findings provide an evidence base for policy formulation to design intervention strategies and improve the screening uptake in China, and ultimately promote more equitable breast cancer outcomes. Our results demonstrated that the rates of breast cancer screening practices were critically low among women living in Eastern China, which needs improvement. Most local women failed to participate in breast cancer screening. This implies that several complex factors may be at work. An important reason is that China still has no nationwide population-based breast cancer screening program because of the large population, insufficient screening equipment, funding obstacles, and inadequate insurance coverage [13]. Consequently, the government should put forth greater efforts to increase the coverage and penetration of breast cancer screening. Specifically, it is urgently needed to establish population-based national breast cancer screening programs and national mass-media campaigns to increase participation rates. https://doi.org/10.1371/journal.pone.0279107.g002 Our results showed that higher education, income, and occupation increased the likelihood of participation in breast cancer screening. Similar results are reported in other countries, and they show the positive impact of socioeconomic factors such as education [21,23,42,43], income [21,23,42], and occupation [42,43] on breast cancer screening. Among these SES components, education is considered a prerequisite for health, reflecting health awareness, and decision-making competence [26]. Occupation reflects prestige and skill-related achievements [26]. Income reflects the ability to pay for more health inputs, which is closely linked to the level of education and occupation [26,27]. This implies that higher education, higher income, and belonging to higher occupational classes were related to healthier lifestyles, access to more health promotion resources, and easier access to preventive healthcare services [23]. Socioeconomic disparities in breast cancer screening participation provide valuable information for public health programs. Reducing breast cancer screening disparities requires policy initiatives to address SES component inequalities. Another noteworthy finding of our study was that the composite SES index was an independent determinant of breast cancer screening practices. We found that women with higher SES indexes had a greater likelihood of breast cancer screening participation. Regression analysis also showed a strong influence of the SES index on breast cancer screening. This highlights the importance of SES in influencing female breast cancer screening in Eastern China. Our results are similar to the findings of other international studies [19][20][21][22][23]44], reporting that lower SES was a major determinant for non-participation in breast cancer screening. It is imperative to develop optimal national breast screening programs to increase promotion of early screening among women with low SES and who are under-screened. Notably, our results indicate that a higher SES index has a positive impact on women's breast cancer knowledge. Similarly, a study by Yavari et al. [45] showed that women with lower education had insufficient knowledge of breast cancer detection, compared with highly educated women. This indicates that women with high SES tend to have more health-related knowledge [24], which may contribute to their participation in health-promoting programs [16]. Educational campaigns about cancer should proactively reach vulnerable women and promote widespread access to knowledge about breast cancer. Our study also found that the influence of the SES index on breast cancer screening practices varies with health education, marital status, and location of residence. This demonstrates an association between the SES index and these characteristics of women. In addition, being married, receiving breast health education, and living in urban areas increased the odds of breast cancer screening practices. Rakowski et al. [46] in the United States and Al Rifai et al. [23] in Jordan highlighted the positive impact of health education on participation in breast cancer screening. Furthermore, results from a systematic review documented that being single and living in a rural area were associated with the low participation in breast cancer screening in Latin America [21]. Therefore, cancer screening interventions need to take these factors into account. It is worth noting that we found that the SES index is a neutral and stable variable, which did not alter the relationship of three SES indicators with breast cancer screening in the study population. Previous studies analyzed separate bivariate associations between selected SES indicators and breast cancer screening [19-21, 23, 42, 43]. However, each dimension of SES can only reflect different actual resources. The SES index is represented as a new comprehensive indicator [27,40]. Compared with a single indicator approach, the SES index can be used to explore the composite contributions of educational attainment, income, and occupation to breast cancer screening uptake. Similarly, scholars in other fields also confirmed that the SES index is a robust measure for SES assessment [27,39,40]. Several potential limitations should be considered while interpreting our results. The main challenge with cross-sectional design is the inability to provide inferential causality. Potential sampling bias should be noted because people willing to participate were enrolled in this study, that is, the sample was not randomly obtained from the source population. Some limitations of self-report questionnaires, including item clarity, social desirability, or response bias, may have also affected the reliability of the study results. This survey was conducted during the COVID-19 pandemic, and thus, breast cancer screening practices during this time may not reflect true rates. Of note, other useful information was not included in this study, such as health insurance [21,22], health status [20,42], and unhealthy behaviors [23,42]. Furthermore, SES is constantly changing throughout one's life, which may have influenced our findings. Also, there is no indication of whether the current SES index is related to screening behaviors which may have occurred many years in the past. Finally, according to the CACA guidelines at the time, BSE and CBE should begin at the age of 20, and mammography screening should begin at 40 years. We included women 19-39 years of age in our analysis for both BSE, CBE, and mammography screening utilization. Not aligning our study design precisely with recommendations provided by the CACA poses as a limitation in our study design. Notwithstanding the limitations described, our study provides useful information for designing breast cancer screening intervention strategies and promoting more equitable breast cancer outcomes. Our findings could be used as a historical reference. In the future, more indepth, well-designed longitudinal and prospective research will be required to address these potential study limitations. Step 1 model fit, -2 log likelihood = 538.866, Cox and Snell R 2 = 0.008, Nagelkerke R 2 = 0.031, Model χ 2 = 15.076 . Step 2 model fit, -2 Log likelihood = 533.210, Cox and Snell R 2 = 0.011; Nagelkerke R 2 = 0.043, Model χ 2 = 20.733 . Step 3 model fit, -2 log likelihood = 525.983, Cox and Snell R 2 = 0.015; Nagelkerke R 2 = 0.057, Model χ 2 = 27.960 . P-Values were calculated using logistic regression analysis. �� p < 0.01 � p < 0.05. https://doi.org/10.1371/journal.pone.0279107.t006 --- Conclusions This study found that female breast cancer screening rates were low among Eastern Chinese women, despite its increasing prevalence. Socioeconomic inequalities in the patronage of female breast cancer screening exist in Eastern China. Furthermore, socioeconomic factors strongly influence breast cancer screening practices and breast cancer-related knowledge. Most importantly, using the composite SES index, this study confirmed the effect of SES on female breast cancer screening practices. The composite SES index was found to be positively and significantly correlated with female breast cancer screening patronage. There was an interaction between the SES index and those characteristics, such as health education, marital status, and location of residence. Our results provide strong evidence further validating that SES is a multidimensional construct, and the SES index is a robust variable for SES assessment. Future breast health promotion programs should develop tailored interventions targeting these women with lower SES to promote the patronage of breast cancer screening, especially lower educational levels, insufficient income, and lack of employment. --- The data from this study are available, and there are no restrictions. We have uploaded the minimal anonymized data set necessary to replicate our study findings as Supporting Information files. Among our Supporting Information files, one separate XLSX file is labelled "journal. pone-D-22-06263-s001", and another separate SAV file is labelled "journal. pone-D-22-06263-s002". --- --- Data curation: Shunhua Zhang. Formal analysis: Shunhua Zhang. --- Funding acquisition: Shunhua Zhang. Investigation: Shunhua Zhang, Hairuo Wang, Binhao Liu, Jianwei Yu, Yizheng Gao. Methodology: Shunhua Zhang. --- Project administration: Shunhua Zhang. Resources: Shunhua Zhang. Software: Shunhua Zhang, Hairuo Wang, Binhao Liu, Jianwei Yu, Yizheng Gao. Supervision: Shunhua Zhang. Validation: Shunhua Zhang. Visualization: Shunhua Zhang. --- Writing -original draft: Shunhua Zhang. Writing -review & editing: Shunhua Zhang.

Breast cancer is the most prevalent malignancy affecting Chinese women, and early routine screening is incredibly important for its prevention and control. This study aimed to investigate the role of socioeconomic status (SES) in female breast cancer screening practices using the composite SES index.This cross-sectional study involved 1816 women in Eastern China. Data were collected using an online self-administered questionnaire from January 2020 to May 2021. We used principal component analysis to construct the composite SES index using educational level, annual household income, and occupation. Logistic regression was used to analyze the association between the SES index and breast cancer screening utilization.Of the participants, 19.27% reported having performed breast self-examination, 12.89% reported having undergone clinical breast examination, and 3.52% reported having received mammography. The SES has a significant influence on the patronage of female breast cancer screening in Eastern China. The composite SES index was found to have a positive and significant association with breast cancer screening practices. An interaction was found between the SES index and patient characteristics, such as health awareness, marital status, and location of residence. In addition, the SES index had a positive effect on breast cancer-related knowledge.Socioeconomic inequalities existed in Eastern China and were related to breast cancer screening patronage. The composite SES index was identified as a significant determinant of breast cancer screening practices. Our results highlighted the negative impacts of

Introduction Social inequality in health and disease is a pervasive phenomenon, both over time and across societies. The mechanisms underlying it, however, are not fully understood . Earlier research has suggested that one important cause of social inequality in health is differences in quality of health care . Quality of health care comprises both the effectiveness of and access to health care . Focusing on a crucial aspect of access to health care, namely the doctor-patient relationship, we examine the effects of different models of doctor-patient interactions on educational inequalities in self-reported health. Education is a crucial indicator of socioeconomic status and low education is an important risk factor for poor health . The lower educated suffer from poorer health compared to their better educated counterparts; for instance, with respect to self-rated health or mortality . While remaining largely stable over the life course, education is a major predictor of future life chances. It is strongly connected to access to resources important for health, such as material resources and social connections . The aim of this study is to examine whether some aspects of doctor-patient relations exacerbate or compress health inequalities. Existing research points in this direction. Firstly, it has been shown that doctor-patient communication is affected by patient's SES in the sense that doctors convey less information and directions to patients from the lower social strata of a society, deploy a more directive consulting style, and give patients fewer positive socio-emotional responses . Secondly, aspects of doctor-patient relationships have been shown to be related to outcomes such as patient satisfaction, coping with disease, and recall and understanding of information received from the physician , as well as to self-reported and objective health measures . However, there has been little research connecting the two issues of the SEShealth gradient and SES differences in the doctor-patient relationship. Our study fills this gap by explicitly addressing the role of doctor-patient relations in relation to educational inequalities in selfrated health. The current drastic transformations in the doctor-patient relationship make it a particularly interesting factor in the study of inequalities in health. There are strong tendencies towards a more egalitarian relationship between physicians and their patients, in part due to the growing awareness of the crucial role of doctor-patient relations for health outcomes. The traditional 'doctor-centered' model , with its dominance of the doctor's role, is gradually being replaced with the 'patientcentered' model that gives greater weight to patients' needs and preferences, and aims at a partnership between the doctor and patient. For our analyses, we rely on survey data from 24 European countries. As there are no empirical data on doctor-patient relations over a longer period of time, these cross-national data allow us to increase the available variation in doctor-patient relations and allow for additional validity checks of our findings. By addressing the role of patient-doctor relations in a model of the social gradient in health, we are extending existing research on health inequalities. Our central argument is that the differences in the doctor-patient relationship could affect the social gradient in health in two ways. First, a more egalitarian relationship between doctors and patients might be accompanied by a more egalitarian distribution of health outcomes. Thus, an egalitarian doctor-patient relationship could potentially compress the social gradient in health. The doctor-patient relationship could hence be a pathway factor that mediates the relationship between education and self-rated health. Second, differences in the doctor-patient relationship could potentially affect the health of different educational groups in different ways. Compared to the lower educated, higher-educated individuals might command greater psychosocial resources and have greater experience negotiating with highstatus actors such as doctors, irrespective of the doctor-patient relationship. The lower educated might thus benefit to a greater extent from a more egalitarian doctor-patient relationship. We will test whether aspects of the doctor-patient relationship moderate the relationship between education and health, i.e. testing whether educational groups are differently affected by different forms of the doctor-patient relationship. --- Theoretical background The doctor-patient relationship The relationship between doctors and patients is a complex phenomenon. It involves communication about issues of vital importance, but is often emotionally laden, sometimes involuntary, requires close cooperation, and is based upon interaction between individuals in unequal positions . Traditionally, the doctor-patient relationship is an asymmetric one, with the doctor being advantaged in several ways. First, doctors command superior knowledge with regard to health and illness. Second, doctors have a gatekeeper function with respect to many aspects of health and illness, both in a formal and an informal way ). Third, doctors have great power in steering interactions with patients, e.g. in terms of waiting times and the content of meetings. There is a long-standing tradition in sociology of research on the doctor-patient relationship and some classical works of sociology deal with the diffusion of innovations among physicians or interactions between physicians and patients . However, there are relatively few studies which focus on the relationship between the doctor-patient relationship and health outcomes. A number of studies have shown that aspects of doctor-patient relationships can impact patient wellbeing and behavior, such as the ability to understand medical information, compliance with treatment regimes, coping with illness, satisfaction with care, overall quality of life, and health status . The existing literature identified different pathways along which aspects of the doctor-patient relationship can impact health. One potential pathway is the quality of health care. A review of 139 studies revealed that the doctor-patient relationship is the most important predictor of patient-perceived quality of health care . Another possible pathway is adherence to doctors' recommendations. Stavropoulou shows that individuals' reports of adherence to medication recommendations depend on their relationships to their doctors, corroborating results of earlier studies . --- Paternalistic and egalitarian models The asymmetry between doctors and patients is fundamental for different models of the doctor-patient relationship. Two different models are treated here, namely, the 'paternalistic' or 'doctor-centered' model and the 'egalitarian' or 'patient-centered' one. The paternalistic model actively builds upon the asymmetry between doctor and patient, as the doctor's role as an expert is crucial for this model to function. The general idea of the model is straightforward: Based on their superior medical knowledge, doctors give advice to their patients, who, in turn, comply with this advice. Driven by a desire to appear competent , i.e. to make correct decisions, and aiming at behavioral confirmation , patients defer to the doctors' authority . This authority is based on the doctors' expert status and their superior skills and knowledge. As the model greatly draws upon characteristics of the doctor while the patient holds a largely passive role, this model is sometimes called the 'doctor-centered model' . In the last decades, however, the doctor-centered model has given way to the egalitarian model . This model tries to bridge the asymmetry between doctor and patient by also conceiving the patient as an expert, namely an expert for their own situation , and assigning them a more active role. This development has been fostered by campaigns encouraging patients to ask their physicians questions . Reasons for this shift of paradigm are numerous, with the most important ones being the crisis of the expert role in society , the greater availability of medical information for patients in self-help groups and internet sites , the increasing role of chronic incurable diseases , and medical progress that has created more treatment options and, thus, more opportunities for decision-making in the cure process. The general idea of the model consists of multiple aspects: Both the doctor and patient are seen as experts-the doctor for conveying medical information, and the patient for their own preferences and living situation. The doctor's task is seen less as providing the patient with binding advice, but more to communicate medical information about diseases and treatments to the patient and to discuss the information with them, such as benefits and risks of different treatment options. If decisions are to be made, they will be made jointly , taking into account not only the medical facts, but also the patient's preferences. The mechanism at the core of the patient-centered model is also different from the one underlying the doctor-centered one. Instead of satisfying the needs for competence and behavioral confirmation via complying with doctors' orders, the patient-centered model fulfills the desires for patient competence and autonomy by informing the patient and treating them as capable of deciding for themselves. According to self-determination theory , a combination of competence and autonomy will induce intrinsic motivation in patients. Feelings of competence can be achieved via the provision of information, but these feelings of competence need to be accompanied with a sense of autonomy to result in intrinsic motivation . Intrinsic motivation is potentially of crucial importance in the cure process, as individuals with intrinsic motivation have more interest, confidence, and self-esteem, and perform better in comparison to individuals who are externally controlled . --- Socioeconomic status and doctor-patient relations There are two ways doctor-patient relations can operate in the context of the educational gradient in selfrated health. Firstly, there might be an educational gradient in doctor-patient relations. Previous research has shown that doctor-patient relations vary with patients' SES. Doctor-patient communication is affected by patients' SES in the sense that doctors give lower-SES patients fewer positive socio-emotional responses, convey less information and directions, and make use of a more directive consulting style with less involvement in treatment decisions . Furthermore, doctors are able to identify the SES of their patients: Psychologists demonstrated that individuals can reliably identify the approximate income and SES of others within seconds . These differences appear to be driven both by patients' characteristics as well as the perceptions of doctors: High-SES patients communicate more actively and are able to elicit more information from their physicians, whereas doctors often misperceive the need for information of low-SES patients . Given the evidence that higher SES might be associated with a different, more egalitarian model of doctor-patient relations and the evidence suggesting that a more egalitarian model of doctor-patient relations is associated with better wellbeing and health outcomes , the relationship between education and self-rated health might be at least partially due to differences in the doctor-patient relationship. In line with this reasoning, we formulate the following expectation: H1: An egalitarian doctor-patient relationship reduces the size of the positive association between education and self-rated health. Secondly, the models of the doctor-patient relationship might have differing effects on different educational groups. On the one hand, the lower-educated might particularly benefit from a more patient-centered approach that takes their living conditions into account . In turn, the paternalistic approach might be particularly detrimental for the lower educated, as it does not pay attention to special needs they might have. Furthermore, if doctors give fewer positive socio-emotional responses, convey less information and directions to lower-educated patients than to higher-educated patients , then it follows by implication that lower-educated individuals benefit more from a more egalitarian style than higher-educated individuals. On the other hand, the higher educated can be expected to be less affected by different types of the doctor-patient relationship: The higher educated know better how to deal with directive high-status actors, as they have more selfconfidence and other important psychological resources and more experience in negotiating and asking, irrespective of the style of the other party. Based on this, we further hypothesize: H2: The doctor-patient relationship moderates the education-health relationship as follows: An egalitarian doctor-patient relationship is associated with higher self-rated health, and that this is more strongly the case for the lower educated than for the higher educated, and A paternalistic doctor-patient relationship is associated with lower self-rated health, and that is more strongly the case for the lower educated than for the higher educated. --- Data and methods --- Data This study exploits the second round of the European Social Survey , a largescale survey conducted in 26 countries in 2004/2005. This survey does not only comprise information on education and health, it also includes items on the doctor-patient relationship, making it a unique resource for our endeavor. We include 24 countries in our study, namely Austria, Belgium, Switzerland, the Czech Republic, Germany, Denmark, Estonia, Spain, Finland, Great Britain, Greece, Hungary, Italy, Ireland, Iceland, Luxembourg, The Netherlands, Norway, Poland, Portugal, Sweden, and Slovenia. Turkey and the Ukraine were excluded from the analysis, as first results revealed substantial cultural and socioeconomic differences to the other countries in the sample. The quality of the ESS data collection process is generally considered to be high , as in all countries random probability samples are drawn and face-to-face interviews are conducted, following standardized procedures and strict quality controls to ensure representativeness and comparability of the data. Response rates are on average 61.8%, ranging from 43.6% to 78.8% . More detailed descriptions of the ESS, including detailed information on data collection and response rates for all countries are available on the ESS web site . Our study excludes respondents 25 years old or younger, as a substantial number of them has not yet finished education, which is one of our focal predictors in this study. Our sample size is further reduced by respondents who have missing values on any of the study variables. Descriptive statistics of the final sample of 31,189 respondents are provided in Table 1. Health outcome: self-rated health. The outcome variable of this study is self-rated health. Self-rated health was measured on a five-point scale ranging from 'very bad' to 'very good' . Self-rated health is a general assessment of one's health status, not connected to any specific illness, but covering mental, physical, and social aspects of health . It has been shown to predict mortality and morbidity and has high test-retest reliability in a number of studies . Furthermore, this variable has been recommended by the World Health Organization for comparative research and a large number of researchers have followed this advice . Research has also shown that the associations between objective health indicators and self-perceived health are largely similar across countries . --- Predictor variables Doctor-patient relationship. The ESS questionnaire contains six items covering different aspects of the doctor-patient relationship. Respondents were asked how often statements applied to doctors in general. These six statements largely correspond to the two models of doctor-patient relations described previously and are reproduced in Table 2. All answers are recorded on a five-point scale, ranging from 'never or almost never' to 'always or almost always' . Zero-order correlations between the six items are low to moderate . The items of the egalitarian model and the paternalistic model are positively correlated among one another and negatively correlated with the items of the other model, yet a factor analysis does not yield a clear two-dimensional structure. This points to the fact that the items capture distinct features of the doctor-patient relationship, but does not, however, allow for the creation of composite scores that would be more reliable than the single-item indicators we use for our analyses. Table 2 also shows that the bivariate relationships between the indicators of the doctor-patient relationship and education are substantively small . For the items of the egalitarian model, the higher educated are more likely to agree with the statements that doctors are not difficult to understand and that doctors treat their patients as equal, yet there is no statistically significant difference between the educational groups for the statement that doctors discuss treatment decisions with their patients. The direction of these associations, however, is not always straightforward: For the items of the paternalistic model, agreement with the statement that doctors keep the whole truth from their patients is lower among the higher educated, yet for the statement that they are not willing to admit their mistakes, agreement is higher among the higher educated. Education. We rely on education as our focal indicator of socioeconomic status. Based on the International Standard Classification of Education , we distinguish between three educational groups: low , medium , and the higher educated . There are several reasons for relying on education for our purposes. First, education reflects both individuals' material and non-material resources and their social status in a broad fashion. Second, the ISCED, with its high degree of cross-national standardization, allows meaningful comparison of educational groups across countries. Third, educational attainment is usually completed in early adulthood and remains stable across the life course, thus reducing the chance of reverse causation: unlike occupational status, educational degrees do not change even when an adult experiences a health shock . Fourth, as the response rate for questions about education is usually much higher than for those about income, education is a comparatively straightforward indicator of SES. Finally, education is a meaningful measure for the SES of both men and women . --- Control variables. As individual-level control variables, the models contain marital status/cohabitation , sex , age, migrant status , being a lone mother , and number of children . Method and modeling strategy. For analyzing a clustered dataset, multilevel modeling is often recommended . Given the relatively low number of countries and our prime interest in the relationship between variables at the individual level, we present results as obtained from ordinary least squares regression models. These models account for the country variance by including country fixed effects, and correct for the clustering in the data by reporting cluster-robust standard errors. We have centered our predictor variables at their respective country means, allowing us to obtain unbiased estimates of the within-country regressions . Alternative estimation strategies are reported in the online supplement. Notes: F-tests are based on OLS regression models that regress the respective doctor-patient relationship item on two education dummies to assess the correlation between education and the item. For all F-tests, the degrees of freedom are N-1 ¼ 31,195 and K-1 ¼ 2. Item correlations are Pearson product-moment correlation coefficients. --- Results Model 1 reported in Table 3 confirms the presence of an educational gradient in health. Respondents with medium and high education report significantly better health than lower educated respondents. A 31,198 31,198 31,198 Wald test further confirms that the coefficients for medium and highly educated respondents differ from one another ¼ 196.82, p ¼ .000), indicating that there is not a single dividing line between the lower educated and anyone with higher education, but that the health benefits of education increment gradually. The control variables in the model behave as can be expected from the literature: Older respondents and females report worse health, and those with a partner report better health. A comparison of the coefficients shows that apart from age, education is the most important predictor of health. The indicators of an egalitarian doctor-patient relationship are entered in Model 2 of Table 3. Reporting that doctors are easy to understand and that they treat their patients as equals is associated with significantly better health, whereas the third indicator-whether doctors discuss their decisions with patients-is unrelated to self-reported health. Hypothesis 1, which had posited that a more egalitarian doctor-patient relationship attenuates the educational gradient in health, has to be rejected. When comparing the coefficients for education as obtained from Model 1 and Model 2, barely any difference can be detected, indicating that by and large, more egalitarian doctor-patient relations do not reduce educational differences in self-rated health. A reason for this is that, as shown before, the educational differences in the doctor-patient relationship are comparatively small. In order to assess differences between countries for this finding, we have run separate regression models by country and report the education coefficients before and after adjusting for the egalitarian doctorpatient relationship in Figure 1. The size of the educational gradient in health varies considerably across countries, but closer inspection reveals that differences between the adjusted and unadjusted coefficients Unadjusted high ed. Adjusted medium ed. Adjusted high ed. Figure 1. Educational gradient in health by country, before and after adjusting for egalitarian doctorpatient relationship, based on country-specific ordinary least squares regression models. Note: All models control for age, sex, marital status, migrant status, single motherhood, and number of children. are negligible, indicating that the finding of no meaningful reduction in the educational gradient in health in case of a more egalitarian doctor-patient relationship holds for all countries under study. For the indicators of a paternalistic doctor-patient relationship, Model 3 of Table 3 indicates direct negative associations with self-rated health for one of the three indicators. When patients report that doctors keep the truth from patients, it goes along with statistically significant poorer self-rated health. Reporting about a reluctance to ask questions and about doctors denying mistakes have negative effects that are not statistically different from zero. The size of the educational gradient in health is, however, unaffected by including indicators of a paternalistic doctor-patient relationship, as can be seen from comparing the coefficients for education between Models 1 and 3. Table 4 reports the interactions between indicators of the doctor-patient relationship and education. For two of the six indicators, we can detect a statistically significant interaction with education, namely when respondents report doctors to be easy to understand and when patients are reluctant to ask questions . To facilitate interpretation, we have plotted the interactions in Figure 2. The upper panel of Figure 2 plots the interaction of Model 4 from Table 4. It shows that whether doctors are easy to understand plays virtually no role for the self-rated health of the higher and medium educated; the slopes of their predicted lines are virtually flat over the continuum of how easy doctors are to understand. For the group of the lower educated, however, we can see a marked increase in self-rated health as the reports of doctors being easy to understand increase. The plot provides evidence that it is the lower educated who benefit from a more egalitarian doctor-patient relationship, giving support for Hypothesis 2a. The lower panel of Figure 2 plots the interaction of Model 7 of Table 4. Again, for the higher educated and for those with medium education, it shows that there is virtually no association of self-rated health with reporting that patients are reluctant to ask questions, as the slopes of the predictions are flat. However, for the lower educated, we can see a falling slope as the agreement to the statement that patients are reluctant to ask questions gets stronger. Substantively, this means that a reluctance to answer questions is irrelevant for those with educational degrees classified as medium and high, but for the lower educated, such a reported reluctance is associated with worse self-rated health. This gives support for Hypothesis 2b. Sensitivity analyses. The online supplement to this article shows the following robustness checks: We were able to replicate our findings from pooled OLS models with country dummies and robust standard errors by using random effects models without obtaining substantially different results. We also ran analyses making use of the weights calculated by the data providers, leading to substantively identical findings. We furthermore replicated our analyses dropping one country at a time or dropping all Eastern European countries altogether, yielding substantively similar findings. --- Discussion How doctor-patient relations affect the educational gradient in health is largely unknown. Drawing on a large European general population survey, our study generated several new insights into this field. First, our analyses showed that two different models of the doctor-patient relationship, namely the traditional paternalistic model and the more egalitarian 'patient-centered' model, are directly associated with self-rated health. Reports that doctors are easy to understand and treat their patients as equals go along with better self-rated health. Conversely, reports that doctors do not tell the whole truth are associated with worse self-ratings of health. This confirms findings from previous studies, which have suggested that a more egalitarian doctor-patient relationship is associated with better wellbeing and health outcomes . Second, the different models of the doctor-patient relationship do not mediate the educational gradient in self-rated health. Despite the evidence for differential treatment of patients by their doctors according to their SES in the literature , our data show that differences in doctor-patient relations are substantively small. Accounting for those differences does not decrease the education-self-rated health association. Third, when doctors are more paternalistic and less egalitarian, educational differences in self-rated health can indeed be larger. We found a number of interactions between education and indicators of the different models of doctor-patient relations. For the lower educated, reporting that doctors are easy to understand-an indicator of the egalitarian model of doctor-patient relations-is associated with better self-rated health, while for those with medium and high levels of education this does not translate into better health. Also, reports that patients are reluctant to ask questions-an indicator of the paternalistic model of the doctor-patient relationship-is negatively associated with self-rated health for the lower educated, but for the medium and higher educated there is no negative association. This finding has important policy implications, as measures to improve doctor-patient relations can have a positive impact on overall population health, and can be expected to compress the social gradient in health. As has been argued prominently by Marmot , doctors should engage with the social background of their patients. Marmot encourages doctors to recognize the social context in which poor health develops when treating patients, and suggests that doctors should cooperate with other community actors such as social workers, and engage in advocacy to improve the living conditions of their patients. Yet rather than expanding the responsibilities of doctors, our findings build, however, on the core business of physicians and show how doctors' bedside manners go along with particular health effects and the educational gradient in health. We believe that this can have an easier and greater impact on population health and health inequalities rather than overburdening doctors with new tasks and new arenas for action. Furthermore, future research should engage with the effects different health care systems have on doctor-patient relations. While our study focused on the micro-level processes between doctors and patients that govern the educational gradient in health, considering differences between health care systems would be a promising next step , which, however, needs 31,198 31,198 31,198 31,198 31,198 31,198 Robust standard errors in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001 careful theorizing to disentangle the mechanisms that govern the macro-micro link between health care systems and doctor-patient relations. As a first step, our study already documented important country differences in the strength of the education-self-rated health relationship, yet future research is necessary. Exploiting differences in health care system financing would also allow further investigation of any effects of a marketization of health care. We acknowledge that our conclusions are based on a cross-sectional, observational study. Our results are associational and cannot make any claims about causality. Despite these caveats, we believe that our study gives important impulses for future research under improved methodological conditions. Drawing on longitudinal data would help to rule out causality issues and could yield greater insights as to how the change in doctor-patient relations occurred over time. Also, using validated and more comprehensive indicators of the different models of the doctor-patient relationship could strengthen the case for the arguments laid out in this article. Additional health outcomes that do not rely on self-reports would help overcome any biases that might stem from cross-national or individual heterogeneity . However, our results proved to be robust to a multitude of model specifications and were not sensitive to the selection of countries, making it unlikely that institutional or cultural specificities are driving our findings. Notwithstanding these limitations, our study was the first to show that in a large European population sample the move towards a more egalitarian model of doctor-patient relations can have positive effects on reducing social inequality in health, which opens up avenues for further research on the social gradient in health. --- Article The educational gradient in self-rated health in Europe: Does the doctor-patient relationship make a difference? --- Supplemental material The online supplement is available at http://asj.sagepub.com/supplemental

Research suggests that doctor-patient relations have evolved from a doctor-centered, paternalistic approach towards a more patient-centered, egalitarian model of interactions between physicians and their patients. Given the long-running debate on the positive relationship between education and health, the question arises how this development in doctorpatient relations affects social inequalities in health. First, we test to what extent egalitarian (e.g. discussing treatment decisions with patients) doctor-patient relations are underlying the education-self-reported health association. Second, we test whether egalitarian and paternalistic (e.g. withholding some information from patients) doctor-patient relations show differential effects on self-reported health across educational groups. Analyses of the European Social Survey (ESS) 2004/2005 for 24 countries demonstrate that a more egalitarian doctorpatient relationship does not substantially reduce educational inequalities in self-reported health. However, some direct positive effects of egalitarian and direct negative effects of paternalistic doctor-patient relations on health ratings can be found. Finally, results show how the health status of the lower educated can improve with a more egalitarian and less paternalistic doctor-patient relationship.

Introduction A policy agenda at the European level is the reduction of social exclusion [1][2][3], defined as a "process whereby certain individuals are pushed to the edge of society and prevented from participating fully" [4]. Social exclusion is associated with reduced quality of life and health deterioration [5,6]. Older people have been identified as a population subgroup at particular risk of social exclusion . Every second older person in England is experiencing social exclusion [8] and in deprived urban areas in England two-thirds of older people experience social exclusion [9]. While social exclusion has been explored in relation to a number of factors, including income, health, and disability, it has rarely been considered in relation to informal care-receipt. This is despite the fact that with increasing age older people require greater amounts of care and support and that informal care comprises the majority of this care [10]. Furthermore, research has shown that the transition to requiring care can reinforce social exclusion . This paper considers how dimensions of, as well as risk factors for, social exclusion are linked to informal care-receipt in older people. Social exclusion is a multifaceted concept covering dimensions such as production activity, financial activities, social relations, social activity, and political/civic activity [12,13]. Lack of participation in or access to activities/services specified by these dimensions can be seen as indicators of social exclusion. Within social exclusion research concerning older people Scharf and colleagues [6] propose that older people may experience social exclusion on the following dimensions: social relationships; civic activities; basic services; neighbourhoods; and material resources. By comparison a United Kingdom government report [8] defines social exclusion of older people in terms of exclusion on the following dimensions: social relationships; cultural and leisure activities; civic activities; basic services; neighbourhoods; financial products; and material goods. As can be seen, although there are different operationalisations there is consensus around core dimensions of social exclusion for older people. What is apparent is that in social exclusion research on older people there is less emphasis on production activity and more emphasis on the importance of neighbourhood, also often discussed in terms of community. Many older people have spent a substantial period of their lives in a particular neighbourhood, have strong emotional investments in the surrounding community, increasingly rely on neighbourhood relationships for support in old age, and also tend to spend more time than younger people in the immediate neighbourhood [14]. Research exploring risk factors for social exclusion among older people has identified an enhanced risk of social exclusion with increasing age and ethnic minority origin and, among those living alone, having no children, and being on low income [6,8]. Poor health, long-standing illness, and depression have also been found to increase the risk of social exclusion . Negative consequences of poor health can to some extent be compensated for by access to social and health care [17], thereby helping to prevent social exclusion. A few studies have focused on access to formal care in relation to social exclusion [11,[18][19][20], but research on access to informal care in the context of social exclusion is very limited. This is surprising, since as noted previously the majority of all care for older people in the community is informal and recent research indicates that the level of informal care is increasing [21,22]. As people age, a reduction in functional capacity occurs that increases their need for help with activities of everyday life [23]. While the level of care received by an older person might be expected to map onto their need for care , there are many factors that can disrupt this mapping resulting in unmet need for care in a significant minority of older people [24][25][26]. Unmet need can have serious consequences for an older person including increased mortality risk [27] and also prevent them from participating fully in society. Thus, unmet need for care could arguably place an older person at risk of social exclusion. The relationship between care-receipt and social exclusion could be hypothesised to take different forms. It could be argued that the level of care-receipt if directly mapping onto the level of need in the older person would be a marker for his/her level of social exclusion, given that higher levels of care received would be in response to higher levels of illness or frailty, that is, risk factors for social exclusion. Yet carereceipt is linked to indicators of social inclusion such as the availability of, as well as access to, social networks, so an alternative hypothesis is that higher levels of contact with family and friends as a result of being a care-receiver might serve to reduce social exclusion. There is also the critical issue of where the level of care-receipt is not appropriate to need: where need for care exists but care is not received, is the degree of social exclusion particularly high? This paper aims to examine the relationship between social exclusion and the receipt of informal care and reports the analysis of relevant data from the Barnsley Social Exclusion in Old Age Study, which sought to explore social exclusion among older people via a survey of over 1,000 respondents . --- Materials and Methods --- Design and Sampling. A questionnaire survey was carried out in the metropolitan area of Barnsley, England, UK. Barnsley was selected as the study site since it encompasses both urban and rural areas, allowing the exploration of how social exclusion processes might differ in such contrasting areas. To ensure adequate cell size at subgroup level for specific analyses, a sample of 𝑛 = 600 for each of the two areas was proposed, providing good statistical power for the analyses reported in this paper. Sampling occurred from seven electoral wards with an urban profile and from 16 electoral wards with a rural profile. Within each electoral ward households were randomly selected via local electoral registers. With oversampling of households required in order to obtain sufficient participants, a total of 11,035 households were sampled. --- Participants. Potential participants were ineligible if they were under 65 years of age and were excluded from the study if their physical and/or mental health was too poor to allow them to complete an interview or respond to questions reliably; 59 individuals were excluded from the study as a result of this latter criterion. Only one older person was recruited per household regardless of whether more than one older person resided at a given address. In total 1,255 older people participated in the study, of whom 6.5% were recruited from supported accommodation. The response rate was 68.1% and did not differ significantly between urban and rural areas. --- Materials. A questionnaire was developed that addressed a range of indicators of, as well as risk factors for, social exclusion, together with the topic of care-receipt. Given the potential frailty of some respondents, the need to keep the questionnaire concise meant that brevity was a key criterion during instrument selection. On occasion items and scales were adapted to more precisely address the study population or agenda. An item used in the EUROFAMCARE study [29] to identify informal carers was adapted to produce a categorical variable of informal care-receipt: "Do you rely on a friend or relative to provide you with care or support for four hours per week or more?" . Those participants responding "yes" were categorised as carereceivers. Participants who responded "no" were asked: "Do you have someone who looks in on you to see if 'everything is all right'?" . Those participants responding "yes" were categorised as assurance-receivers; those responding "no, no need" were categorised as nonreceivers without need; those responding "no, despite need" were categorised as nonreceivers with need. This procedure therefore produced a four-category variable of care-receipt. In this article, the following dimensions of social exclusion were considered: financial resources; social relationships; community; and social engagement. Financial resources were operationalised via the following item measuring income discomfort: "Which of these descriptions comes closest to how you feel about your household's income nowadays?" with response options ranging from very comfortable on present income to very difficult on present income [31]. Social relationships were measured via items on social contacts, informal caregiving, and loneliness. For data on social contacts a question asked "how often do you meet and spend time with any of the following people?" Independent responses were required for family members , neighbours, and friends [32,33]. For family member contact responses were coded as weekly contact or more or less than weekly contact for each category and items summed to indicate overall level of contact . For the two categories of nonfamily members responses were combined and coded for analysis as no contact , at least twice weekly contact with friends or neighbours , or at least twice weekly contact with friends and neighbours . The informal care item in the EUROFAMCARE study mentioned above was used in its original form for measuring informal caregiving, that is, "Have you a friend or relative who relies on you to provide them with care or support for four hours per week or more?" . Loneliness was measured by the De Jong Gierveld Loneliness Scale, in which respondents indicate the extent to which 11 statements relating to loneliness apply to their situation and the way they feel now, with response options yes, more or less, and no. The items are scored in relation to two subscales: Emotional Loneliness and Social Loneliness [34]. Respondents were asked to what extent they agreed with each of 13 statements about their local community, defined as "within 20 minutes' walk or about a mile from home." Example items are "I feel really part of this area"; "Vandalism and graffiti are a big problem in this area" to strongly disagree [7,8,35]). For data on social engagement participants were asked, first, to consider for "how often, if at all, do you engage in the following activities?" and record a response for twenty different activities. Example activities are "Go out for a meal"; "Attend leisure activities "; "Tend to the garden or allotment" to never ) and, second, to indicate for each applicable activity if it is performed usually with somebody , sometimes with somebody , or usually alone [32,33]. The product of participants' engagement scores and socializing scores resulted in a single "social engagement" score for each activity. This study considered the following risk factors for social exclusion: sociodemographic characteristics; health; and well-being. The questionnaire contained standard items addressing sociodemographic characteristics: age; gender; marital and coresident status; ethnicity; and duration of local residence. Education was measured by an item with six response categories, merged into two broader categories for analysis: "low education" and "medium to high education" . Self-reported health was assessed via the item: "In general, would you say your health is. . ." [31] measured on a fivepoint scale to very poor ). Psychological wellbeing was measured using the World Health Organisation-5 Well-being Index [36] ; sample Cronbach 𝛼 = .87). Finally information on formal care-receipt was gathered via items asking if the participant had in the last month received medical care, personal care , or practical support ; response categories for all items were yes/no. 2.4. Procedure. In order to ensure standardisation of data collection procedures and maximise interviewer sensitivity to reliability issues , interviewers were provided with training commensurate with their prior experience. Training therefore amounted to a few hours to two days . Upon selection, a household was sent a letter presenting the purpose of the study. An interviewer subsequently visited the address to establish whether anybody in the household was 65 years or older and, if so, whether this person was willing to participate in an interview. Each interview lasted on average 50 minutes. Seventy-eight percent of the interviews were conducted alone with the interviewee, while the rest were carried out with the interviewee accompanied, usually by a family member. Interviewers completed several items at the end of the questionnaire as a quality check addressing whether or not the respondent had tried to answer the questions to the best of his or her ability; whether the respondent understood the questions; and whether anyone was present during the interview that could have interfered with the interview. There was also an option for the interviewer to write additional information about the interview. Where data drawn from these items indicated a problem with the interview, researchers discussed the problem with the interviewer, and if there was a suggestion that the reliability of the data could be suspected, the interview was excluded from the study. Bivariate analyses identified significant associations between the dependent variable and independent variables . One-way ANOVA was performed for continuous IVs, with Hochberg's GT2 and Games-Howell tests as appropriate for post hoc analysis of group differences; Chi-square procedures were used for categorical IVs. A multinomial logistic regression was then performed to determine those IVs that predicted membership of the categories of care-receipt in a multivariable model. No adjustment to experimental alpha was made for multiple testing; significance for each test was set at 𝑝 < .05. Given also the substantial sample size significant tests should be regarded cautiously and with thought to effect size. --- Results Care-receipt status and sample characteristics are summarised in Table 1. Regarding care-receipt 21.8 percent of the respondents were care-receivers and 24.1 percent were assurance-receivers, while 5.5 percent were nonreceivers with need and 48.6 percent were nonreceivers without need. --- Bivariate Analysis. A brief summary of the bivariate analyses is described below, with a full presentation of the results of the analyses in Table 2. For the sake of concision IVs with nonsignificant associations with the DV are not presented. At the top of the table bivariate associations between carereceipt status and social exclusion dimensions are presented. In post hoc tests nonreceivers without need had significantly lower dissatisfaction with household income than carereceivers and nonreceivers with need. Post hoc tests also showed that family contact was significantly lower, and Social Loneliness and Emotional Loneliness was significantly higher in nonreceivers with need compared to the three care-receipt categories, and the lowest proportion of older people with contact with friends and neighbours was also found in this group. Post hoc tests also indicated that nonreceivers without need were significantly lower on both Social and Emotional Loneliness than care-receivers and lower on Emotional Loneliness than assurance-receivers. Other post hoc tests indicated that nonreceivers with need scored significantly lower on Perceived Community Integration than older people in the other three care-receipt categories. For Perceived Community Trust and Perceived Community Security post hoc tests indicated significantly lower scores for nonreceivers with need and care-receivers compared to assurance-receivers and nonreceivers without need. Finally, post hoc tests indicated that nonreceivers without need had significantly higher scores on Social, Cultural and Leisure Activity and Sport and Outdoor Activity than older people in the other three care-receipt categories. In the next part of Table 2 analyses of risk factors of social exclusion are presented. Of the categorical IVs gender, coresident status, education, and place of residence were all significantly associated with care-receipt. Of the continuous IVs, in post hoc tests, nonreceivers without need were significantly younger than those in the other three categories. Post hoc tests determined that duration of local residence was significantly greater in care-receivers and nonreceivers with need than among assurance-receivers and nonreceivers without need. For measures of self-rated health and wellbeing post hoc tests indicated that nonreceivers without need differed significantly from those in the other three categories, having better self-reported health and better well-being. In addition, post hoc tests indicated that care-receivers had poorer self-reported health and well-being than assurancereceivers. At the bottom of Table 2 associations between carereceipt and formal care use are presented. The highest level of medical care receipt was found among nonreceivers with need, with the lowest level in nonreceivers without need. This pattern differed in comparison with formal personal care-receipt and formal practical support-receipt as carereceivers and assurance-receivers were those with the highest proportion in receipt of these formal services. --- Multivariable Analysis. The nonreceivers without need category was used as the reference category for the DV in the multinomial logistic regression. IVs entered into the regression consisted of those established through bivariate analysis to be significantly associated with care-receipt. The IV Social, Cultural and Leisure Activity was recoded from a 5-point to a 3-point scale to reduce the number of cells in the model. Following trial runs best model fit statistics were obtained when nonfamily contact and Perceived Community Security were deleted. Due to missing data model 𝑛 = 980 . In comparison to a constant-only model the model was reliable = 488.68, 𝑝 < .001), with good model fit = 2905.15 𝑝 > .05, Nagelkerke 𝑅 2 = .43). Table 3 presents for all variables in the model the Wald test for significance of each coefficient and the odds ratio with 95% confidence intervals for each of the three comparisons between the three care-receipt categories and the reference category. We consider first the prediction of membership in the care-receiver category relative to being in the reference category. Starting with social exclusion variables, a unit increase in Perceived Community Trust and Social, Cultural and Leisure Activity corresponded to 0.87 and 0.98 odds of being a care-receiver, and participants at the lowest and middle levels of Sport and Outdoor Activity had, respectively, 4.79 and 1.93 odds of being a care-receiver compared with participants at the highest level. Regarding risk factors of social exclusion, a unit increase in age corresponded to a 1.06 increase in the likelihood of being a care-receiver; participants in the rural group had 0.38 odds of being a carereceiver compared to the urban group; a unit increase in self-reported health corresponded to a 2.54 increase in the likelihood of being a care-receiver. Regarding formal care use, participants who had received medical care had 2.81 odds of being a care-receiver compared to those who had not received medical care. Next we consider prediction of membership in the assurance-receiver category relative to the reference category. Regarding social exclusion variables, one-unit increase in family contact and Emotional Loneliness corresponded, respectively, to 1.21 and 1.16 increases in the likelihood of being assurance-receivers, while participants at the lowest and middle levels of Sport and Outdoor Activity had, respectively, 1.85 and 1.71 odds of being assurance-receivers compared with participants at the highest level. Regarding social exclusion risk factors, a unit increase in age corresponded to a 1.05 increase in the likelihood of being an assurancereceiver; male participants had 0.63 odds of being assurancereceivers compared to female participants; participants in the rural group had 0.49 odds of being assurance-receivers compared to the urban group; a unit increase in duration of local residence corresponded to 0.99 odds of being assurance-receivers; a unit increase in self-rated health corresponded to a 1.30 increase in the likelihood of being assurance-receivers. Finally, participants who had received medical care had 1.51 odds of being assurance-receivers compared to those who had not received medical care. Lastly, we considered prediction of membership in the nonreceivers with need category relative to being in the category nonreceiver without need. Regarding social exclusion dimensions, analyses show that a unit increase in Social Loneliness corresponded to a 1.39 increase in the likelihood of being nonreceivers with need. Regarding risk factors of social exclusion, a unit increase in age corresponded to a 1.07 increase in the likelihood of being nonreceivers with need; participants in the rural group had 0.26 odds of being nonreceivers with need compared to the urban group. Finally, participants who had received medical care had 5.82 odds of being nonreceivers with need compared to those who had not received medical care. --- Discussion Our multivariable analysis produced a significant model that predicted older people's membership of care-receipt categories on the basis of social exclusion dimensions and a range of risk factors for social exclusion. The pattern of associations in the model had face validity. Greater age, recent contact with medical care, and poorer health, all being meaningful indicators of frailty and/or need for support in an older person, increased participants' likelihood of being an assurance-receiver and were associated with an even greater likelihood of being a care-receiver, relative to being a nonreceiver with no need. Greater age and poorer health are established risk factors for social exclusion [8]. Another variable significant in the model, Sport and Outdoor Activity, is included in the social exclusion dimension social engagement. The picture that emerges from the model is that older people with higher scores on dimensions of and risk factors for social exclusion were significantly more likely to be in receipt of assurance and care. Membership of the nonreceivers with need category was also predicted by indicators of frailty and an indicator of social exclusion . If one examines the pattern of bivariate associations between the IVs and the care-receipt variable, nonreceivers with need in comparison to older people in the other categories demonstrated the highest levels of Social and Emotional Loneliness and the lowest levels of income comfort, social contact (with family members, friends and However, other variables uniquely contributed to membership of different care-receipt categories, suggesting that nonreceivers with need, care-receivers, and assurancereceivers might differ as social positions or roles not just in terms of degree. A continuum model of care-receipt might therefore be insufficient. Lower Perceived Community Trust and Social, Cultural and Leisure Activity increased the likelihood that a participant was a care-receiver; being female, greater family contact, and greater Emotional Loneliness increased the likelihood that a participant was an assurance-receiver; and greater Social Loneliness increased the likelihood that a participant was a nonreceiver with need. When seeking to conceptualise care-receipt in older people, therefore, one model would be that the experience of carereceipt is a progression through increasing levels of support and care until in many cases the informal care network breaks down. A different model would be that whether one is a care-receiver or an assurance-receiver or a nonreceiver with need is not about the current occupation of one role within a progression of care-receipt, but rather an outcome of that role being more likely when other factors are true, for example, as where being female and having recent contact with one's family increase the odds that one is an assurance-receiver. Our findings suggest that both models coexist in later life. --- Social Exclusion and Care- Receipt. An Irish study of carereceipt among older people [37] reported that 49% of older people received care over a 12-month period. The study used a broad definition of care-receipt, including people receiving care once weekly or less. This corresponds in our study to those participants in the care-receipt and assurance-receipt categories combined, which comprised 46% of our sample, a similar figure. The present study's focus on the concept of assurance-receipt is relatively unique: most studies explore the issue of caregiving among burdened carers, that is, those providing many hours of care each week, with few focusing on "light" carers, and by extension older people in receipt of only a few hours of care . A question posed in the introduction to this paper was whether being a care-receiver might be associated with lower levels of social exclusion, as care-receipt is almost by definition linked to contact with friends and family, an indicator of social inclusion. Our findings offered little evidence to support this conjecture. Being a care-receiver was not significantly associated in the multivariable model with greater contact with friends or greater contact with family. Furthermore care-receipt was predicted by lower Perceived Community Trust, suggesting that care-receipt might be linked to a poorer relationship with one's neighbourhood, another indicator of social exclusion. While one predictor of being an assurance-receiver was greater contact with one's family, this indicator of inclusion has a counterbalance in the association between being an assurance-receiver and greater Emotional Loneliness. Similar results have been found in research on formal care-receipt. Barrett et al. [11] argue that home-based formal care contributes to a disconnectedness of the care-receiver from self, family, home, and the broader community, thereby contributing to social exclusion . In the multivariable analysis one of the key predictors of membership in the different care-receipt categories was place of residence with care-receipt being more common in urban areas. There is a lack of research on rural/urban patterns of informal care-receipt, and the findings from the existing research are contradictory. For example, in line with our findings, it has been suggested that American urban older residents are more likely to have informal support from kin and to have children living nearby, as there is a general pattern of outmigration of children from rural areas and an in-migration of older people to rural areas [39]. On the other hand, Canadian research indicates that informal care-receipt is more common among rural residents [40]. With regard to our findings, since several demographic, health, and social variables were controlled for in our analysis, we suggest it is likely a complex interaction of personal, demographic, and social factors related to residence in urban environments that produced high levels of care-receipt and an unmet need for assurance, relative to residence in rural areas. More research is needed in order to obtain a deeper understanding of the mechanisms behind rural/urban care-receipt patterns. --- Policy and Practice Implications. One Spanish study on older people in need of help with activities of daily living identified a group of older people who did not receive help [26], and a Canadian study [25] found that approximately 2% of older people outside institutions experienced unmet care needs. In our study such a group was represented by the nonreceivers with need, which comprised 5.5% of our sample. This group was found to score highly on several dimensions of and risk factors for social exclusion . Of interest is the fact that we found that older people with higher levels of care-receipt received more formal personal care and formal practical support. The exception to the pattern was the nonreceivers with need, where despite a lack of informal care and assurance there were also low levels of formal personal care and formal practical support. However, this group had the highest level of recent contact with medical services, a relationship also found in the Spanish study mentioned above [26]. Similarly, recent research has found that lonely older people have a higher use of health care services than nonlonely people [41] and that loneliness and social isolation are a common nonmedical problem presented by noncritically ill older people in emergency departments . One interpretation of this finding is that, in the absence of informal care and in the presence of need, older people will turn to medical services in order to address their unmet need. Whatever the contributory factors, an opportunity exists during contact with medical care providers whereby identification and assessment could offer a route into referral for appropriate formal personal care and/or formal practical support. This recommendation is in line with previous calls for integrated care and joined-up services for older people [44], and programmes have been targeted at referral of medical care patients to social services . Potential savings in medical care services that follow such referrals might compensate for increased costs to the social care budget. There is therefore a need for more emphasis on the issue of unmet need in older people within the education of medical and nursing practitioners and their own significant role in appropriate onward referral. --- Conclusions. This study provides valuable insights into the personal, interpersonal, and social factors related to informal care-receipt in older people, a topic that has received relatively little attention. The experience of care-receipt can last for a substantial amount of an older person's life and its nature and content impact significantly on quality of life. The present study indicates that the need for and receipt of informal care in an older person is connected to dimensions of, as well as risk factors for, social exclusion. In our model, as older people's scores on dimensions of and risk factors for social exclusion increased so did the odds of them being care-receivers, or being nonreceivers with need. Very little evidence emerged that the receipt of care or assurance was associated with social inclusion. Our findings suggest that access to informal care is important to consider when studying social exclusion processes, in that the receipt of informal care may be a marker of social exclusion, that is, that older people relying on informal care may not be integrated into society and may have unmet social needs. --- Ethical Approval SHU Faculty Research Ethics Committee , Faculty of Development and Society, provided ethical approval. --- Disclosure This research was undertaken in collaboration between Age UK, Sheffield Hallam University , Dalarna Research Institute, and Dalarna University. Phil Wormald at Age UK and Tamsin-Bowers Brown, Maria Burton, Rachel Ibbotson, Dot Biggin, Martin Hyde, and interviewers at SHU contributed to the research. --- Competing Interests The authors declare that there is no conflict of interests regarding the publication of this article. Submit your manuscripts at http://www.hindawi.com --- Stem Cells International Hindawi

Older people have been identified as being at risk of social exclusion. However, despite the fact that care is commonly required in later life and the majority of that care is provided by informal carers, a connection between social exclusion and informal carereceipt has rarely been considered. The aim of this study was to examine how informal care-receipt is related to social exclusion. A face-to-face questionnaire survey on social exclusion and informal care-receipt was carried out among older people (𝑛 = 1255) living in Barnsley, United Kingdom. Multivariable analyses examined the association between social exclusion and categories of informal care-receipt: care-receiver; assurance-receiver; nonreceiver with no need; and nonreceiver with need. Compared to being a nonreceiver with no need, participants were more likely to be care-receivers or assurance-receivers if they had higher levels of social exclusion. The highest level of social exclusion, however, was found in nonreceivers with need. Despite a lack of informal care and support, formal practical support and personal care were also low in this latter group. Findings are discussed in relation to the conceptualisation of care-receipt and how contact with medical services could be an opportunity for identification and appropriate referral of nonreceivers with need.

Introduction The Ice Bucket Challenge, a campaign to raise awareness and support for people with amyotrophic lateral sclerosis, 1 is widely recognised to have gone socially viral. Although there are variations to the trial, participants typically complete the challenge and nominate three others to undertake it. 2 As of 1 September 2014, more than 17 million videos related to the Ice Bucket Challenge were shared on Facebook alone, and these were viewed more than 10 billion times by more than 440 million people. 3 Socially viral phenomena include videos that are made popular by sharing on the internet, and a video is said to have gone viral if it spreads rapidly as a result of frequent sharing. 4 5 However, the infectious disease modelling framework has seldom been used to quantify the transmissibility of such socially viral phenomena. 6 The most commonly used metric of transmissibility is the basic reproduction number , defined as the number of secondary cases generated by a single index in a fully susceptible population. 7 The value of R 0 is a major determinant of the size of an epidemic, and an infection can only be self sustaining if R 0 is greater than 1. The R 0 also provides a measure of the effort required to control the epidemic. 7 8 We estimated the transmissibility of the Ice Bucket Challenge among globally influential celebrities and identified the associated risk factors. Global influence was defined by the criteria: listed in TIME 100: The Most Influential People in the World 9 or TIME: Great People of the 20th Century, 10 or having at least five million view counts for the Ice Bucket Challenge on YouTube. Among the small pool of potentially eligible participants we arbitrarily chose David Beckham, Cristiano Ronaldo, Benedict Cumberbatch, Stephen Hawking, Mark Zuckerberg, Oprah Winfrey, Homer Simpson, and Kermit the Frog as index cases . We also included successful nominations up to the fifth generation seeded from each index case. --- Data collection We completed a marathon viewing of 145 Ice Bucket Challenges up to 13 September 2014. For each case we recorded the age, sex, occupation, net worth, popularity , number of successful nominations, and total number of nominations. We obtained personal details of the celebrities from Wikipedia and specialised websites, although the validity of such information cannot be reliably ascertained. 11 Wikipedia has been used in previous studies 12 13 and its comparative accuracy has been reported. 14 From verified accounts we obtained the number of Facebook likes and Twitter followers. Where these were not available, we used the unofficial accounts with the most likes or followers. We excluded unofficial accounts with fewer than 1000 likes or followers. We traced contacts through Google, YouTube, Facebook, Twitter, and Instagram. Online written or video responses by the nominated contacts were used to determine completion of the challenge. If we found no evidence for a response to the challenge on social media, it was assumed that the contact had declined. We included all successful contacts. Two of the authors independently collected the data. After the fifth generation seeded from each index case they stopped data collection owing to investigator fatigue. We did not seek previous consent from participants as they presumably participated in the Ice Bucket Challenge without coercion, albeit with varying degrees of peer and social pressure. --- Statistical analysis We used fully observed nomination chains to estimate the R 0 , a measure of transmissibility, for a celebrities based cohort, up to the fifth generation seeded from each index case . R 0 was estimated as the product of the mean number of nominations and the probability of successful nomination , assuming a branching process in which each participant of the Ice Bucket Challenge generates a random number of next generation participants. To account for the fact that the challenge had already taken off in early August 2014, 15 we excluded nominees in the calculation of p who had previously completed the challenge but included them in the calculation of n so that the estimated R 0 would reflect the transmissibility in a fully "susceptible" population. We obtained 95% confidence intervals of the R 0 estimate by bootstrap with 1000 replications. Mean serial interval was calculated by assuming at least a half day delay in accepting the challenge, based on all nominator-nominee pairs in our dataset. We used multivariable ordinal logistic regression to examine the association between number of successful nominations and personal factors or popularity . To handle missing data we carried out sensitivity analysis with multiple imputation. All analyses were done using R version 3.0.2 and SAS university edition, 2014. --- Results Eight index cases complied with the inclusion criteria. In total we included 91 nominees up to the fifth generation seeded from each index case, and a total of 99 participants were enrolled into the cohort. Overall, 24.2% of participants had zero successful nominations, 32.3% had one, 26.3% had two, and 17.2% had three. Twelve out of the 247 nominations had already participated in the Ice Bucket Challenge and were therefore "immune" from future participation. Among those observed were a Nobel laureate, a university vice chancellor, "heart throbs", and Muppets . Among the index cases, Mark Zuckerberg's nomination chain produced the most successful number of contacts up to the fifth generation . No serious adverse events arising from the Ice Bucket Challenge act were observed in this series, but adverse events have been reported elsewhere, 16 including falls, head injuries, a temporomandibular joint dislocation, cuts, and at least one fatality. We estimated the measure of transmissibility, R 0 , to be 1.43 and a mean serial interval of accepting the challenge of 2.1 days . We excluded Homer Simpson and Kermit the Frog in the regression models because of difficulty in ascertaining their personal characteristics. Participants with a higher log net worth were more likely to spread the Ice Bucket challenge , adjusted for age and sex. One unit change on a log scale is equivalent to a 10-fold change in net worth. We tested the proportional odds assumption for the regression model and found no indication of violation. Age, sex, occupation, number of Facebook likes, or number of Twitter followers were not associated with transmissibility . Data were missing for 24.0% of net worth, 15.6% of Facebook likes, and 13.5% of Twitter followers. Sensitivity analysis by multiple imputation yielded similar results . --- Discussion The Ice Bucket Challenge-a campaign to raise awareness and support for people with amyotrophic lateral sclerosis by dousing oneself with a bucket of iced water or giving a donation-was moderately transmissible among a group of globally influential celebrities. The celebrities took a mean 2.1 days to accept the challenge . Our finding of an R 0 value greater than unity , indicating sustained spread, in a celebrities based cohort is consistent with observations reported in the press. Indeed this social viral pandemic shares a similar R 0 with pandemic A/H1N1 2009 influenza, or the high end of the Middle East respiratory coronavirus since 2012 . Nine out of the 99 participants completed the challenge but did not nominate anyone, analogous to self isolation. We included these participants in our analyses and assumed all losses to follow-up to have declined the challenge. As with other microbial pathogens, transmissibility of the Ice Bucket Challenge may also exhibit seasonal variation, 17 where it might be anticipated that the R 0 value would decrease during winter. Possible factors accounting for the speed and extent of the Ice Bucket Challenge pandemic may include the online social media mode of transmission and the short serial interval for taking up the challenge . Despite a lower R 0 than for measles or smallpox, the Ice Bucket Challenge spread quickly across the world as nominees became immediately "infectious" once nominated. A previous social network experiment found that individual uptake improved with reinforcing signals from clustered social ties. 18 Therefore, specified nominations within No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe a social network, such as celebrities, may be more successful in spreading promotional messages for public health interventions than are generic nominations. However, whether such social networks should be leveraged for health communication depends on the nature of the intervention. 19 Finally, our findings suggest that the R 0 of a social media campaign may increase with the support from people of a high net worth, independent of age or sex. --- Limitations of this study We had limited power to detect differences in R 0 for personal factors as the inclusion criteria restricted the number of eligible index cases. Reporting of the Ice Bucket Challenge might have been socially patterned, contributing to potentially biased associations between transmissibility and net worth. The epidemiological assessment is based on a non-representative sample, and generalisability to other population groups is limited. We did not account for the likely broader influence 19 of our cohort of global celebrities on the much larger population of fans and followers worldwide , thus underestimating the overall R 0 , analogous to the "super spreader" concept. 20 We thank Y Yuan, CKL Yiu, and T Li for their contributions to the data collection and the illustration of the Christmas trees. Contributors: MYN conceived the study, wrote the initial protocol, collected data, and wrote the first draft of the manuscript. BHYC developed the protocol and collected data. EHYL, HP, MYN, and GML analysed and interpreted the data. All authors critically revised the final manuscript, had access to the full dataset, take responsibility for the integrity of the data and the accuracy of the dataset, and gave final approval for the submission of this version for consideration of publication. EHYL and HP contributed equally to the work. MYN and BHYC are guarantors for the study. Funding: This study received no external funding. Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work. Ethical approval: Not required. Data sharing: The dataset is available at the Dryad Digital Repository . Transparency: The lead author affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained. --- Figure Tree --- What is already known on the topic Numerous reports have suggested that the Ice Bucket Challenge has gone "viral" Until now formal analysis of its transmissibility has been limited --- What this study adds The Ice Bucket Challenge was moderately transmissible among global celebrities and their nominees Higher net worth may be associated with greater transmissibility --- Tables

Objectives To estimate the transmissibility of the Ice Bucket Challenge among globally influential celebrities and to identify associated risk factors.

country are Mexican, and of that number, more than half are undocumented. "Overall, Mexicans comprise about sixin-ten of the estimated 11.9 million unauthorized immigrants in the U.S." . As citizens, we must be concerned about those in our midst who may need health and human services. This is especially true in the matter of health and should be of special concern with respect to Latinos. While new Latino immigrants tend to be healthier than the general population , their health deteriorates in time with acculturation . This is often attributed to the fact that Latino immigrants and their children often lack health insurance and access to health care services . In addition, it has been found that many immigrants are unaware of the health and community resources available to them . The purpose of this study is: to examine patterns of emergency room use among undocumented Mexican immigrants who choose this rather than usual sources of care; and to describe sources of health care information among these immigrants, as the lack of such information constitutes a major barrier to accessing appropriate care. --- Conceptual Framework and Literature Review As we consider health-seeking behavior of undocumented Mexicans which results in use of an emergency room rather than a clinic or private physician, we use here an ecosystems perspective as a theoretical framework for our analysis . In this framework, behavior is seen as resulting from reciprocal transactions between the continually developing person and a multivariate, multidimensional environment. Bronfenbrenner's four transactional levels between person and environment-macro, meso, exo, and micro-here frame the context for health-seeking behavior of Mexican immigrants. While it is not the purpose of this paper to explore each of such factors at each level in depth, they are presented here to facilitate a better understanding of immigrant healthseeking behavior within the context of a reciprocal ecosystems environment. --- Macro and Exo Level Factors At the macro, or societal/cultural level, and the exo, or distal institutional level, there are many factors that may be viewed as influencing immigrant health behavior. These include economic factors, such as the desire for work, which often results in immigrants working at jobs that have no health benefits , and immigration policies. The Immigration Control Act of 1996, among others, denied some benefits to legal immigrants, not just undocumented ones, and precluded legal immigrants from receiving public benefits for five years after they had become legal residents . Other critical factors are social welfare policies, such as The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 , which has limited access to health care in some states for undocumented immigrants except for treatment of an emergency medical condition and immunizations and the structure of the health care system. Those who do not have the resources to purchase health care insurance are, in most cases, excluded from accessing this system . Other factors include societal attitudes towards immigrants, with many viewing immigrants as "undeserving outsiders" and a "problem" ; institutional discrimination, as "stereotyping, biases, and uncertainty on the part of healthcare providers can all contribute to unequal treatment" ; and cultural attitudes regarding health and illness. As an example of the latter, Borrayo and Jenkins have asserted that the concept of prevention is meaningless in traditional Mexican culture. There, in order to receive attention, one's health problem must show symptoms. --- Meso and Micro Level Factors At the meso, or proximate institutional level, and at the micro, or individual and familial levels, there are also numerous factors affecting immigrant healthcare behavior, including accessibility of healthcare, social support networks, demography, knowledge of available health care resources, individual/ familial resources, and perceptions/attitudes regarding care. In addition to the structural barriers encountered at the macro and meso levels, the immigrant in this country may face other issues relating to accessibility, for example, lack of transportation to healthcare facilities and lack of childcare, which would deny a parent the opportunity to use such facilities. New immigrants often do not know of health and community resources that may be available to them . Newcomers have to learn how to navigate a complex healthcare system from which they are often excluded . New immigrants are often better able to access health systems with social support . Individual demographic characteristics, such as length of residence in the United States, immigration status, and language ability, all contribute to access to care, or lack thereof, for Latino immigrants. Thamer, Richard, Casebeer, and Ray found that the longer one resides in the United States, the more likely one is to have health insurance. Similarly, LeClere, Jensen, and Biddlecom found length of residence to be a significant factor in whether and how often a person saw a physician in the last year. Language ability can also be a significant barrier to accessing care, as many facilities have no Spanish-speaking translators and many new immigrants are not bilingual in English . Ku and Matani state that language problems are cited by Latino parents as the leading barrier to child health services. In addition, many Latino patients, even those with proficiency in English, have difficulty communicating in English when they are ill . Studies have shown that immigrants are more likely to be uninsured than U.S. born persons . This is particularly true for Hispanics, who have the highest percentage of uninsured of any immigrant group . In particular, immigrants who are not U.S. citizens are more likely to lack health insurance, although the number decreases with length of time in U.S. In the case of immigrant families, U.S.-born children are eligible to receive SCHIP , but often undocumented parents are less likely to enroll their children in such programs . Undocumented immigrants not only tend to lack health insurance, but also report more difficulty in obtaining health care than others . In many hospitals, doctors' offices, and clinics, eligibility requirements often screen out undocumented immigrants . Some studies have shown that immigrants are fearful to access health services because they are afraid they will be asked their immigration status and denied service, or even worse, be deported . In addition, some immigrants may mistrust the medical system and their treatment by healthcare providers. Many Latinos feel that they receive lesser care, have been discriminated against in the health care system, and have been treated with disrespect based on their ethnicity and English-speaking ability . Given these myriad, interacting forces affecting immigrants' lives, it is readily apparent that decisions to seek primary medical care, or routine care from other usual sources of care, are not isolated events but rather take place in a dynamic, sociopolitical/cultural context. Within this context, we wanted to know where Mexican immigrants choose to receive health care, particularly at what rate they would choose an ER as the source of medical care, rather than other sources of primary care or other types of care. The authors focused on emergency room use because health care obtained in this way is a public concern due to its high cost and also due to the consequent overcrowding of emergency rooms with non-emergency issues, placing great strains on hospital systems. A secondary aim of the study is to describe where immigrants obtain information about health care resources and how such information sources may impact the choice of care setting. --- Methods This study is a secondary analysis of the Mexican Migrant Worker Survey conducted by the Pew Hispanic Center . The survey comprised a convenience sample of nearly 5,000 Mexican migrants in the United States who were interviewed while applying for a "Matricula Consular," an identity document issued by Mexican diplomatic missions. As the immigrants had no legal papers authorizing them to be in the United States, they were assumed to be undocumented or unauthorized. Data were collected at consulates in seven large American cities known to have high numbers of Mexicans: Los Angeles, New York, Chicago, Atlanta, Dallas, Raleigh, andFresno, from July 12, 2004, to Jan. 28, 2005. In each location, data collection was conducted for five to ten business days, depending on the estimated size of the target population in each city. Potential participants were identified in the consulate waiting rooms where they received announcements regarding the survey and its content, the nature of the questions and the length of time needed to fill out the survey, as well as a detailed explanation of the anonymity and confidentiality of their responses. Data were collected with a self-administered questionnaire in Spanish. Because the targeted sample is characterized by a high rate of illiteracy, special attention was paid to the potentially illiterate or semi-literate people in the sample by emphasizing that reading and writing were not a prerequisite to participation and that interviewers were available to provide assistance and to conduct as much of the survey as necessary. In return for filling out the questionnaire, respondents received a telephone card to be used to call Mexico. The Pew Hispanic Center made the data publicly available after the surveys were completed. These data were downloaded from the Center's website. The dependent variable in this study was the binary variable of choosing ER for medical care. The independent variables were age, sex, education, income, employment status, length of residence in U.S., number of children, and having family in the same town. In order to learn more about the health-seeking behavior of this undocumented population, researchers here utilized descriptive statistics to ascertain demographic characteristics of the sample. Inferential statistics were used to determine the correlates and predictors of choice for health care and sources of health care information. We also described the sources of information regarding health care and investigated whether information sources independently affect the choice of setting for medical care. The sample characteristics in this study are comparable to estimated characteristics of undocumented Mexican migrants living in the United States . The survey captured a distinctively young and recently arrived segment of the Mexican-born population living in the United States: nearly half of the sample was aged between 18 and 29 years, or had been in the country for five years or less. In addition, respondents showed a higher level of educational achievement than the adult population of Mexico at large . Of the sample of 4, 836, the majority were male and 41.3% were female. Slightly over 50% of the sample were 29 years of age or younger. About a third had not completed high school, while 59.5% were either secondary/technical school graduates or high school graduates. About six percent were college graduates. These findings are similar to those of Cuecuecha , who found that Mexico was losing its middlerange educated workers, not its uneducated campesinos. Almost fifty percent of the immigrants had been in the country for five years or less, while twenty percent had been in the country for more than fifteen years. Related to this residency is fluency in English, with over fifty percent of the respondents able to speak "only a little" or no English, as compared to 13.6% who said they spoke "a lot" of English. A majority of the respondents were married and had three or more children in the United States . Most of them lived in a town or city where they had relatives . Of those who worked, 76% worked full-time. And unlike findings in other studies, the majority of them stated that they had health insurance. Respondents were asked, "Which of these would you be most likely to use if you were seeking medical care?" Responses included: hospital emergency room; clinic; private doctor; and consult with a friend. Multiple responses were allowed. The dichotomous dependent variable of the current study, emergency room use for medical care, indicated all respondents that choose hospital ER for medical care. Potential predictors of ER use included gender, age, education, level of English, marital status, number of children in the U.S., presence of relatives in town/city, working full-time, time in the U.S. and the city where the data was collected. In addition, analyses controlled for presence of health insurance. Respondents were asked about the two most important sources from which they receive medical and health information. Analyses included a binary variable indicating that the respondent received such information from church, a second binary indicator for receiving information from medical care providers, and two other binary indicators for receiving information from Mexican and U.S. newspapers. For purposes of analysis, an acculturation index was created from three variables found to be associated with acculturation: English-speaking ability, time in the U.S., and level of education. A number of acculturation scales have used these proxy measures and these scales have been tested and validated to some degree . Those with better facility in English , longer residence in the U.S., and higher levels of education were assumed to be more acculturated. The proportion of immigrants choosing ER for medical care was tabulated across covariates, and chi-square tests were used to identify bivariate associations between the outcome measure and the potential predictors. Multivariate logistical regressions were used to model the relationship between ER use and explanatory variables. Estimates were converted to odds ratios. Because the acculturation index is a composite of time in the U.S., education, and level of English, two separate models were estimated. The first model included the acculturation index and did not control for items that make it up, in order to avoid multicollinearity. The second model controlled for the items that compose the acculturation index individually, and excluded the index. Data management and statistical analyses were conducted using Stata software . --- Results Thirty-eight percent of the respondents reported that they would use hospital ER for medical care . Males were more likely to use ER for care, compared to females . Those youngest reported higher ER use rates compared to those aged 30-39 or older . Those with one or two children reported the Contrary to expectations, and the literature studying nonimmigrant or documented immigrant populations, those with some insurance coverage reported slightly higher ER use rates compared to the respondents that lack health insurance coverage . It should be noted that respondents were simply asked "Do you have health insurance?" and there were no inquiries on quality or source of coverage. Of the 4,836 respondents, 345 of them reported that they receive health care related information from churches; 1,415 reported receiving information from medical care providers, 162 respondents reported Mexican newspapers as a source, and 511 of them reported U.S. newspapers as a source of information for medical care . Respondents who receive information from a church reported less ER use, compared to all others , whereas respondents receiving information from U.S. newspapers reported higher ER use rates . Observed variations in age remained in the multivariate analyses: the odds of using ER declined with age . However, the gender difference was no longer significant once all other covariates were controlled for. The findings on number of children, insurance, education or employment status remained in the multivariate analyses. Variations by location also remained after other characteristics were controlled for: compared to the respondents surveyed in Los Angeles, those surveyed in New York, Fresno, Atlanta or Raleigh had significantly higher odds of reporting ER use. Respondents that use churches or medical providers as a source of information had lower odds of reporting ER use for medical care, whereas those that use Mexican or U.S. newspapers for information were no less likely to report ER use, compared to those that do not rely on newspapers as an information source. The acculturation measure was insignificant in multivariate analyses . In the second model, where components of acculturation were entered independently to the estimation, we observed that level of English was associated with ER use, at a borderline significance level. Those who rated their level of English "some" or "a little" were more likely to report ER use compared to the respondents that reported speaking "a lot" of English. An important deviation from the bivariate findings was related to time in the U.S.: spending 15 or more years in the U.S. was no longer associated with ER use, once all other covariates were controlled for. --- Discussion and Conclusion For all population groups in this country, some conditions do require ER care, including such urgent care as that required for a heart attack, in which case primary care physicians send the patient immediately to the ER. Other conditions may need immediate attention at night or on weekends. If a person, although having a source of routine care, does not have access to it , she or he may be channeled to the ER. However, some ER visits are "preventable," resulting from untreated or under-treated conditions that may lead to acute events requiring an emergency visit; limited access to routine health care is likely to be the root cause of such visits. In other cases, however, the ER is the immigrant's "preference" because he or she fears discrimination in doctors' offices and clinics, or simply cannot afford such services and perceives ER care as free. While the survey does not inquire as to why the person chooses ER for medical care, the questionnaire is most likely to capture the preference-based visits, thus reflecting the problem of access to care. One of the limitations of this study, however, was the questionnaire instrument itself. While the question was asked, "Do you have health insurance?" no follow-up question was asked as to type or source of the insurance. Contrary to the literature, a high percentage of the undocumented immigrants responding to the survey reported having health insurance. As Mexico has universal health care coverage, it is possible that some of those who responded affirmatively, when interviewed in the consulates prior to receiving documentation, may have been referring to coverage in their former homeland. Given that this is a retrospective secondary analysis of an existing survey, it is beyond our capacity to investigate the validity of the insurance coverage variable. Still, we opt to report the findings-related insurance variable, since the literature shows that it is the most influential factor that impacts health care access in the United States. Another limitation of the study was that no question was asked about the health status of the respondents or whether they had a regular health care provider. A recent study has found that undocumented Mexicans in New York City who reported having health insurance, but not a regular source of care, were more likely to use the emergency room if they reported more health care needs . ER use declines by age, even after controlling for acculturation and other factors. It is likely that older people with chronic conditions are more connected to usual sources of care, and younger people with acute problems are more likely to use ER rather than other providers. Families with one to two children were more likely to use the ER, compared to families with three or more children. It is also possible that families with more children were more informed about the outpatient health care services available for children, or more likely to insure their children with SCHIP, thus less likely rely on the ER for medical care. The finding that younger immigrants tend to use emergency rooms in higher numbers is consistent with the literature, as is the finding that ER use declines with time in the U.S. Younger immigrants may not have as much knowledge about other health resources available, and they are less likely to be working at jobs that offer health care coverage. Therefore, they would be more likely to seek care at a hospital emergency room. One of the interesting findings had to do with location. As indicated above, this survey was conducted in seven major cities known to have high numbers of Mexicans: Los Angeles, New York, Chicago, Atlanta, Dallas, Raleigh, and Fresno. The highest rates of ER use were found in New York City, Atlanta, GA, and Raleigh, NC. We can only conjecture what the reasons for the differences by area may be. Some of these may be due to issues discussed previously, such as access, fear, and lack of knowledge of resources, particularly in locations where this population has only recently migrated. The areas with the lowest rates of ER use, Los Angeles, Fresno, Dallas, and Chicago, have very large Mexican populations of long standing. The Mexican neighborhoods in these areas may have developed better formal and informal systems of communication regarding health care resources. In some cases, the Mexican population may have become, over time, better accepted by the population at large and thus be better positioned to access available resources and services than are Mexican immigrants in cities like Atlanta and Raleigh, where this population has much more recently begun to settle. Another reason for the difference in reliance on ER use may be the structure of health care systems from location to location. Large numbers of Latinos are now moving into rural areas and small towns that often have limited healthcare facilities and personnel, especially bilingual ones . Some dentists and physicians in these areas may be reluctant to participate in Medicaid and in the State Children's Health Insurance Program . Another factor related to location is transportation. Cities in the U.S., including the ones used in this study, have varying degrees of quality public transportation. Even such public transportation as is available may be difficult to navigate for an immigrant with limited English-speaking ability. And whatever the locale, emergency rooms, as components and functions of hospitals, are often more visible than are neighborhood offices or clinics. The differential use of emergency rooms by location is an area for future study. While our data do not allow us to investigate the reasons for regional variations, our findings suggest that macro and meso factors, such as health care structure and transportation, are highly significant predictors of ER use by Mexican immigrants for health care, and explain a larger proportion of the variation compared to the variables we controlled at the micro level . Another area for future research is the question of sources of health care information. It was found that those immigrants who received information from churches were less likely to use emergency rooms, perhaps because they were provided abundant information on other health care sources, and possibly even assistance in accessing such sources. The majority of Mexican innigrants are Catholics. Catholic social service organizations in the United States have long provided social services to this immigrant group as well as many others. It is possible that these organizations provide the local parishes with information and assistance regarding other sources of health care for immigrants. As to that percentage most likely to use the ER according to source of information, those persons relying on U.S. newspapers as a source for information on health care, one can perhaps speculate that the information presented through this medium was either not sufficiently specific to resources, or that those reading the information did not feel those resources would be accessible, but more research in this area is needed. This paper looked at a specific population in the United States, undocumented Mexican immigrants, who are of concern for many reasons. Because of their undocumented status and the realistic fear it may engender among them, it can be a difficult population to find or even to persuade to participate in research studies. Further, the use and overuse of emergency rooms is a complex issue, and one not confined to this one group. Also, in the survey utilized for this study, the survey instrument had some shortcomings, at least for our purposes, including the lack of follow-up questions to those immigrants indicating they had health insurance, and not including other factors of interest at both the macro and micro levels. Even given the shortcomings of the data, the survey did open a window onto a crucial policy question: that is, how many undocumented immigrants use emergency room services and what may be some predictors of that use? Regardless of immigrant status, and whether one is documented or undocumented, health care access is critical to each individual. And at the societal level this issue confronts both human rights and public health considerations. When public policies leave undocumented immigrants with few choices for health care, they may find that the emergency room becomes their only choice. Further research may help solve the dilemma of how to help ameliorate health care problems of immigrant populations caused by disparate policies and health care systems across the country.

This study examined emergency room use by undocumented Mexican immigrants and their sources of health care information. Thirty-eight percent of the respondents reported that they would use a hospital emergency room (ER) for primary medical care. ER use rates declined with time spent in the United States. Emergency room use rates varied significantly by region. Respondents receiving information from a church reported less ER use, compared to all others; respondents receiving information from U.S. newspapers reported higher ER use rates. Lack of health care access for undocumented immigrants remains a public health issue as well as a social justice concern.

Background Sexual violence is a major public health problem globally. The prevalence of partner and non-partner sexual violence in women is as high as 59 and 6% respectively, in some regions [1][2][3]. Twenty-two percent of females aged 15-49 years in Uganda have experienced sexual violence [4]. Sexual violence was common among human immunodeficiency virus positive women at 28.3% [5]. Trauma-informed care or services is a strength-based framework. It increases providers' awareness and understanding of the impact of trauma, thus guiding and monitoring their interventions, actions, and behaviors in working with clients and creating opportunities for survivors to rebuild a sense of control and empowerment, and minimizes any risks of re-traumatizing clients [6,7]. Trauma-informed services ensure safety, empowerment, choice, and meeting the unique needs of trauma survivors, so they are more active participants in their care and recovery [8,9]. Studies have shown that women with a history of sexual violence desire care and treatment of their unique needs from trauma-informed health care providers [10,11]. The trauma-informed care framework has been implemented successfully for survivors of sexual violence in primary and mental healthcare settings [11][12][13], women clinics [14], and perinatal care settings [15]. The use of trauma-informed services for survivors of sexual violence has been documented in Kenya [16]. The effects of sexual violence range from psychological, physical, social, and reproductive health problems [17][18][19][20]. All these negatively impact the health of individuals and so needs to be addressed to improve livelihoods. Uganda has formulated policies and laws against sexual violence and has been evaluated over time to see their effectiveness [20,21]. A package of trauma-informed services should be provided to survivors of sexual violence so as not to re-traumatize the survivors. The components of trauma-informed services for survivors of sexual violence in Uganda include HIV testing, emergency antiretroviral therapy, emergency contraception, sexually transmitted infections testing, counseling services, tetanus toxoid, provision of prophylactic antibiotics, medico-legal examination, and provision of a medico-legal certificate, and referrals to other organizations for social support [22]. These components are adapted from the World Health Organization guidelines and should be available at all health facilities from health center three and above. The survivors of sexual violence go back to their residence, where they may not get a safe shelter. Some health service providers at these health facilities were trained on trauma-informed services to survivors of sexual violence [20]. These services are provided in a way that avoids exacerbating existing trauma, prevents creating additional trauma, empowers the client, promotes recovery, and respects rights. Few survivors of sexual violence can access or even complete the necessary health care services [23]. According to Uganda's demographic and health survey 2011, up to 42% of sexual violence victims did not seek medical assistance [24]. Also, most survivors of sexual violence did not go to health facilities first [24]. Even when survivors of sexual violence assessed health facilities, a big percentage did so late after 24 h after the act of sexual violence [25]. This delay exposes them to an increased likelihood of developing different health conditions. There is a paucity of studies on trauma-informed services in Uganda. The study explored the knowledge, attitudes, and perceptions of females, and factors associated with the accessibility of trauma-informed services for survivors of sexual violence. Understanding these factors in a resource-constrained setting, like Uganda, is crucial in influencing policy recommendations and creating public awareness about improved trauma-informed service delivery. --- Methods --- Study design This study was a descriptive cross-sectional design employing a qualitative approach conducted in April 2021. --- Study setting The study was carried out at selected health facilities of Kyangyenyi health center III and Kigarama health center III in Kyangyenyi and Kigarama sub-county respectively in Sheema district. In Uganda, health center IIIs do not have general doctors but have clinical officers, nurses, and midwives. Sheema district is bordered by Buhweju district to the north, Mbarara district to the east, Ntungamo district to the south, Mitooma district to the southwest, and Bushenyi district to the west. The district headquarters are approximately 33 km west of Mbarara city along Mbarara Kasese road. The projected district population was 220,500 whom 114,400 were females, and 106,100 were males [4]. Subsistence agriculture, animal husbandry, and petty trading are the main economic activities in the area [26]. People have water and food insecurity [27]. Among the challenges faced in the area like other areas in Uganda is the limited number of health workers [28] and the fact that health providers are well known, or members of the community, and confidentiality and providers' fear of retaliation by alleged perpetrators or their families poses a challenge [29]. --- Study population and sampling Our study population was females attending Kyangyenyi health center III and Kigarama health center III in the Sheema district. The study employed purposive random sampling to ensure that participants represented all age categories; adolescents aged 15-19 years, young adults aged 20-30, middle-aged women 31-49, and the elderly above 50 years. Females were categorized into different age groups because we wanted to get a wide range of views from all age groups since each category faces challenges regarding sexual violence. At each health facility, we recruited four participants from each age category. In total, there were 32 participants for in-depth interviews. --- Interviewers Two social research assistants conducted the interviews. They had experience in qualitative data collection. They underwent training on how to use the research tools for 1 day. The interviewers had not established any relationships with the study participants before data collection. --- Data collection tool We specifically developed interview guides to collect data for this study based on literature. The interview guides were pre-tested among eight participants, and the results were used to adjust and validate the tools. Interview guides were written in English, translated from English into the local language , and then back-translated to verify fidelity to the original wording. Data were collected through in-depth face-to-face interviews using open-ended questions, followed by probes with the study objectives. The interview guides collected data on the socio-demographic characteristics of the participants, understanding of sexual violence, awareness about sexual violence in the community, knowledge about trauma-informed services, and the perception of these services. --- Data collection procedure We obtained administrative approval from the district health officer and in-charge of health facilities. Different females attending the facilities on days of data collection were approached, the study was explained to them and invited to participate. The health provider at the health facility introduced the interviewers to potential participants. The potential participants were informed that we wanted their perceptions about trauma-informed services for survivors of sexual violence in the community. We did not specifically target survivors of sexual violence through some of the participants had experienced sexual violence. All those we asked to participate, accepted. Participants were 18 and above years old. We obtained written consent from participants. Research assistants interviewed participants face to face, using a semi-structured interview guide. The interviews were conducted in a quiet place at a health facility. The in-depth interviews were conducted in the local language . The interviews were audio-recorded and backed by field notes with the participant's consent. Each interview, lasting on average 30 min, was conducted until no new information was being generated. --- Quality control We piloted our data collection tools at a separate health center for clarity and usability. The research assistants were trained on study objectives, informed consent, sexual violence, trauma-informed services, and data collection procedures. There were feedback meetings between research assistants and VM at the end of each day of data collection, to check for completeness of the data. RO and VM attended some of the interviews to ensure that interviews were conducted according to the study objectives. Transcribing was done in the local language. --- Data management and analysis Interviews were audio-recorded on a digital voice recorder notes were taken during the interviews. At the end of the data collection, recorded interviews were transcribed verbatim into Microsoft word documents. The transcripts in the local language were translated into English by a different person who was not involved in interviewing the participants to ensure quality. The transcripts were cleaned and anonymized. Processed data were entered into Atlas. ti version 7.5.7 for coding and analysis. After internalizing data a coding frame was developed, and agreed upon by three researchers . Data were coded. Data were analyzed using thematic content analysis. --- Ethical considerations --- Mbarara University of Science and Technology Research Ethics Committee gave ethical approval . The Uganda National Council for Science and Technology granted the study clearance in line with national guidelines. We also obtained administrative approval from the district health officer and health center administrators. All participants were provided with explanations about the study and requested to provide written informed consent. In the case of illiterate participants, they used a thumbprint, and an independent witness signed to confirm that they consented freely. Participants provided written consent. Participation was voluntary, and a participant was free to withdraw from the study at any time without any penalties. Privacy and confidentiality were ensured by conducting the interviews in private and allocating non-identifiable field codes to each participant. --- Results Three themes emerged during the analysis: a) knowledge of females about sexual violence, b) knowledge about trauma-informed services, and c) attitudes towards trauma-informed services. The study employed purposive random sampling to ensure that respondents represented all age categories. A total of 32 in-depth interviews using an interview guide were conducted. The mean age of participants was 32.9 years. Twenty-five of these were married, three were widowed, two were divorced, and two were single. Seventeen of the participants did not complete primary level of education, seven completed primary level, three completed senior four, and only four completed senior six of whom two were diploma holders, and one was a bachelor's degree holder. --- Knowledge of participants about sexual violence This section of the manuscript describes the understanding of the participants about sexual violence and its different forms, awareness about the situation of sexual violence in the communities, and the effects of sexual violence. Understanding sexual violence demystifies the trauma-informed services for survivors of sexual violence in the community. Also, understanding what sexual violence and its various forms have a bearing on whether survivors of sexual violence will seek the services. --- Understanding of sexual violence by participants Participants viewed sexual violence as an act of men forcing them into sexual acts. Other participants understood sexual violence as children raped by their fathers, relatives, and sometimes strangers. A husband forcing his wife into having sex was sexual violence. Participants also understood sexual violence with examples like someone touching their buttocks and breasts when someone has not consented. --- "Sexual violence means that me as a woman, I am not interested in being engaged in sexual activities, and yet you find a man is seriously using force to make me get involved when I am not interested. " P33: IDI_KYA_12 "Sexual violence is engaging someone in sexual acts without their consent, or asking someone to have sex with you, and when she refuses, you force her and do whatever you wanted. " P12: IDI_KIG_30 Participants mentioned diverse forms of sexual violence in the community. These included intimate partner sexual violence, fathers/ guardians raping and defiling daughters, and women raped. Another form of sexual violence was men touching a lady's private parts such as breasts without her consent which was more pronounced among superiors to their subordinates. "Husbands forcefully want to have sex with us, and when we refuse, they fight with us and even end up beating us and hurting us very badly. These men even catch our children and rape them. You find our girl children pregnant and even do not know who the person responsible for this is. At times you find this pregnancy is even for the father… sincerely speaking men are on the rampage. They touch the breasts of girls, squeeze girls tightly without their consent, romancing them forcefully. " P28: IDI_KYA_07 In contrast, others viewed sexual violence as the way men keep looking at females inappropriately. Females being bothered with unsolicited gifts, love letters, phone calls, and love messages were perceived as a form of sexual violence. Men demanding sex for having given basic things of life to women mentioned as being sexual violence. --- "Some men write love letters to women to woo them just because you are there and yet you are not interested. Secondly, when he meets you, he wants to always sweet talk, so that you end up liking him and yet you do not like him. Then with time, he will want to bring you a nice gift to drift you from your decision and influence you into doing what he wants. " P15: IDI_KIG_31 "He keeps on looking at you for long, he keeps on coercing you as he says so many things, and he mixes in some other good words. Sometimes he makes several phone calls, disorganizes you… sometimes he sends you bad messages with an intention of forcing you into sexual activity. " P 1: IDI_KIG_19 "There are times when someone you do not even know starts calling you on phone, he keeps calling and sending you messages that you do not want, but that person insists and shows you that he knows you. Or you find a man is touching your breasts or your butt, so such things happen even this side, someone keeps looking at you as if you owe them something and yet you owe him nothing. " P22: IDI_KYA_01 --- Effects of sexual violence mentioned by participants Participants had different views about the effects of sexual violence. These views ranged from physical, biological, social, and mental effects. Respondents pointed out that survivors of sexual violence might become infected with sexually transmitted diseases such as HIV/AIDS. The survivors may conceive unplanned pregnancies with their untoward complications. They pointed out that in cases where force is used, survivors sustain physical trauma like damage to the uterus and private parts. Participants mentioned stigmatization, stress, and depression amongst the survivors of sexual violence that can lead to suicide. Participants pointed out marriage breakup, isolation, failure to associate with others, and low self-esteem as other effects of sexual violence. "You find a man infecting his children with HIV, impregnating her, the dangers are so many, now if this girl gets impregnated by his father, which man is going to marry her, yet she is even HIV positive. So, you find that girl getting stressed, lacking peace and getting stigmatized because her father made her pregnant, so she will not get someone to marry her", P24: IDI_KYA_03 "Now, for example, a person who has been raped … some of them get so much disturbed to the extent that they refuse completely to get married because they completely hate anything that is called a man in all her life just because of being raped. " P27: IDI_ KYA_06 "She might get pregnant, contract diseases, sometimes they might hurt her and might end up losing her uterus. Sometimes when she gets pregnant, she might want to abort and might end up dying in the process of aborting, or you find her loitering a long road without any form of care, you find her crying, in the end, she commits suicide. " P25: IDI_KYA_04 --- Awareness about the existence of sexual violence in the community When the community is not aware that sexual violence occurs, they will not advocate for the services intended to address the problem. Participants stated that sexual violence was in their communities, and it was a big concern. Some of the participants had even witnessed the people suffering from the effects of sexual violence. "Sexual violence happens especially in my home areas , there is a man that chased his wife away from home, after that, he started to violate his children sexually, and by the way, he was arrested by the police. " P24: IDI_ KYA_03 "You find a man having sex with his daughters sometimes they go ahead to encourage them to use family planning so that they do not get pregnant and give birth to babies so that he can continue having sex with them. These things are happening in our community. We see them. " P10: IDI_KIG_28 "You find a lot of young girls in our communities getting pregnant when they were not willing and ready for their pregnancies. If you are to investigate, you find that they were raped. " P33: IDI_ KYA_12 "That child, her colleagues kept on harassing her saying that, "look at this small child who was raped" and she finally developed self-hatred when they kept on saying that "that small child was raped" she ended up drinking poison and committed suicide. " P 4: IDI_KIG_22 Some of them shared their lived experiences with sexual violence. Although this was not the primary focus of the study, some of the responses ended being personal. --- "Sexual violence occurs because there is a badly behaved man who usually sends me messages when I am not even willing. We even have a child that I know, a badly behaved man met her there at the well and raped her. " P 4: IDI_KIG_22 "It occurs because personally, a certain man forcefully raped me and made me pregnant. " P 6: IDI_ KIG_24 Contrary views were that sexual violence did not occur in the community, and or if it occurred, it was not common. A few people held this view. --- "Others do not know anything about sexual violence. " P33: IDI_KYA_12 "No, in my village it is not so common. " P25: IDI_ KYA_04 --- Knowledge about trauma-informed services for survivors of sexual violence This section of the manuscript describes participants' knowledge about whether survivors of sexual violence knew where to seek help and if they did seek help. It also intends to find out if the participants knew about different trauma-informed services available. This was done bearing in mind that trauma-informed services are services provided with prior knowledge and experience of the violence so as not to re-traumatize the survivor, and know that health providers had been trained in offering these services. This is important since if services are available but the community is not aware of their presence and/ or does not use them, then they become white elephants. --- Knowing where to seek trauma-informed services Participants said that survivors of sexual violence sought help. In most cases, this help was sought from both health facilities and police for treatment and justice respectively. There were mixed views on whether the survivors of sexual violence go first to a health facility or a police station. Those who advocated for the survivors of sexual violence going to the health facility first reasoned that the survivor needed medical attention to address injuries sustained or to prevent infections. On the other hand, proponents of survivors going to the police first were more concerned that delay to report may aid the perpetrator to run away from the place. They mentioned that thorough investigations are made to identify real and fictitious cases of sexual violence. The prosecution is then done and justice is provided. --- "A person who has been sexually assaulted should go and report to the police immediately and then they serve justice. After coming from the police, and if you find you have some health issues then you go to the hospital and you get the treatment that you deserve, and then after you can go back to the police to go ahead with your case so that it can be investigated". P 8: IDI_KIG_26 "They are supposed to go to the hospital both in the private and government hospital, and if it is a child and the case is defilement, the perpetrator will be taken to the people who are fighting for children's rights, if it is where they chose to go first, they will do what they can and then they forward them ahead for further services. " P10: IDI_KIG_28 "They should run to the hospital to carry out a checkup, to see what harm has been caused, in case she has not yet contracted any diseases, they give her medicine to block and stop all possible diseases that she could contract. Some run to the police and report what has happened, so they carry out investigations at the police. " P35: IDI_KYA_14 In contrast, some participants revealed that some survivors opted not to seek any services in case they are victimized due to the fear of the perpetrator, bribing of the family of the affected and authorities, or even fear of family breakup and stigmatization. To some survivors of sexual violence, it was unheard of that the husband can sexually violate his wife. Some participants reported that they feared being identified as rape survivors as this was very shameful and embarrassing so they would rather not report it anywhere. --- "….you have nothing to do because the whole world knows that he is your husband, you decide to keep quiet. And again you get shy to talk about it, you get fear for disclosure… you let him do what he wants…" P 3: IDI_KIG_21 Also, participants stated that other survivors of sexual violence did not have someone to share with their experience and ended up not seeking help. Others did not even know where to seek help. On the other hand, some of the survivors of sexual violence lived in fear of death since they were warned that they would be killed should they disclose that they were sexually violated. --- "When it happens to some, they don't get who to share with or others fear to talk, they decide to keep quiet and keep it at "heart". Others don't know who to talk to or where to go, she doesn't know what she should have done. Some keep silent and they look at it as a shameful act… they think that if they say it they are going to be embarrassed. And they think that if they tell some, they are also going to tell oth-ers… and they end up ashamed and bothered. " P 1: IDI_KIG_19 Some are being warned by their father saying "the moment you say anything about it, I will kill you". He instead makes her keep silent and quiet by giving her some gifts. In most cases, you find such children are not fitting properly in their homes just because of that. P27: IDI_KYA_06 --- Knowledge about available trauma-informed services Participants were asked to mention different traumainformed services. Participants pointed out that survivors should report immediately to the health facilities so that they can be examined to see if they have contracted illnesses or injuries and then report to the police to apprehend the perpetrator. A section of the participants mentioned that the services to be received included testing for HIV and other infections. Receiving the prophylaxis treatment for HIV and other infections was also cited. Other participants talked about pills that are given to prevent pregnancy. Participants talked about counseling that is done to comfort the survivors. Respondents stated that those with physical injuries are also provided with the services. Participants cited the legal services like having medico-legal forms completed and assistance obtained from police as one the services. Most of the participants mentioned only the services related to HIV leaving out other services. "A person who has been sexually assaulted should go and report to the police immediately and then they serve justice. After coming from the police, and if you find you have some health issues then you go to the hospital and you get the treatment that you deserve and then after you can go back to the police to go ahead with your case so that it can be investigated", P 8: IDI_KIG_26 "At the hospital, the victim is checked to see the extent to which they have been affected, tested to see if they have been infected with HIV, checked to see if they are pregnant, and given treatment to curb any diseases that have been transmitted to the victim. They also provide counseling services to the victim to reduce the trauma the crime has caused. " P 7: IDI_ KIG_25 --- Attitudes towards trauma-informed services for victims of sexual violence This section of the manuscript describes the perceptions of participants about the quality of trauma-informed services. Quality is viewed as the extent to which a service satisfies a patient. Quality appreciation is key to the willingness to utilize services by clients. The participants described the quality of trauma-informed services. Quality service is comprehensive, accessible, relevant, efficient, and equitable. --- Comprehensiveness of available trauma-informed services A section of participants mentioned that the services are of good quality. Participants said that the services are available. The survivors are examined and given appropriate medication like ARVs, counseling. The services offered gave confidence and hope to survivors of sexual violence. However, most participants did not know about testing and treatment for other sexually transmitted infections and counseling. The majority focused mainly on HIV. The police investigate cases of sexual violence leading to the prosecution of perpetrators. In this way, survivors of sexual violence get justice. They stated that if one goes to any of the facilities, one will get help. Health workers sometimes interrogated survivors and blamed them for being victimized. Some participants mentioned that these services are not all available at the health facilities and survivors get some but not all of them. --- Accessibility of services and equity of services provided Participants reported that trauma-informed services were available and accessible. Services were free since they were in public health facilities. Most participants considered geographical accessibility to be good. "So, when you reach the hospital, and you look for your doctor immediately you tell him or her what has happened, she or will understand the situation quickly and he knows what to do to you. Then he will first give you the treatment that you deserve according to how you have explained to him. Sometimes it may be a very young child and then he will have to first check to find out whether her uterus is okay, then after they test for HIV and they also find out if she is already infected. From there then he or she makes a medical form for you to take to the authorities like police, and police help you to hunt for the perpetrator", P19: IDI_KIG_34 In another perspective, some participants reported that the services were not that perfect due to some factors. Participants mentioned events of bribery to the police survivors and not getting justice. The poor road network delays reporting to the respective facilities by the survivors of sexual violence. --- Discussion This paper explored the knowledge, attitudes towards trauma-informed services for survivors of sexual violence and factors associated with the utilization of these services by females. There is no consensus on the definition of sexual violence. However, most participants perceived the meaning of sexual violence as forcing someone into acts of sexual activities against the will or when someone has not consented. Different forms of sexual violence were mentioned included rape and defilement, forced sex among married women by husbands, unauthorized touching of body parts like breasts and buttocks, verbal violence, and inappropriate looks. The understanding of sexual violence by our study participants was in agreement with the definition of sexual violence by the World Health Organization. Sexual violence is defined as "any sexual act, attempt to obtain a sexual act, unwanted sexual comments or advances, or acts to traffic, or otherwise directed, against a person's sexuality using coercion, by any person regardless of their relationship to the victim, in any setting, including but not limited to home and work" by World Health Organization [30]. In this study, most of the participants mostly mentioned contracting HIV as the effect of sexual violence. These could have been due to a high prevalence of HIV in the area at 7.9% compared to Uganda's national average of 6% [4]. The different effects of sexual violence cited in this study were in line with those found in other studies [17][18][19][20]. Participants showed varying knowledge about trauma-informed services for survivors of sexual violence. Most participants mentioned different traumainformed services like HIV testing and treatment and counseling that were in line with the services mentioned by the Uganda Clinical Guidelines, 2016 [22]. Participants mostly mentioned services related to HIV and left out other necessary services like antibiotic prophylaxis, tetanus toxoid, and this pointed out a gap in the knowledge about these services. In this study, we found out that some survivors of sexual violence did not seek help. This finding was similar to the results of other studies done in low and middleincome countries [24,[31][32][33]. Women sexually violated by their husbands would not even try to seek traumainformed services, because it is unheard of in the communities that a husband can sexually abuse his wife. There was a perception that sexual violence was acceptable. This finding was in agreement with other studies done in Uganda [34,35]. Survivors of sexual violence did not seek help due to fear, stigmatization from the community, and intimidation by the perpetrator. These findings are similar to those found in Uganda's demographic and health survey, 2011 [24]. Although some participants mentioned that services offered were of good quality and that all the services were accessible from the respective facilities, gaps in services were noted. The highlighted gaps included bribery, limited availability at the health facilities, survivors being interrogated by health workers while seeking the services, survivors not being given the attention and care, and delayed service delivery may be due to the uniqueness of the communities. In health care, actual practices in response to survivors of sexual violence may vary despite the existence of both international and local guidelines. A systemic review done by Gatuguta and others found that health care workers were uncomfortable examining survivors of sexual violence due to confidentiality and lack of clarity of role [23]. Other studies have shown flaws in delivering trauma-informed services by a shortage of doctors [21,36]. The utilization of trauma-informed services by sexual survivors of violence was taking place. However, it was negatively affected by stigma at the community level, being judged negatively by the community, poor road network, long-distance from health facilities, unreliable or unavailable services, and healthcare professionals' attitudes. These findings are similar to other studies [37,38]. These point out the need for sensitization about traumainformed services and sexual violence for both the community and victims of sexual violence. --- Strengths and limitations The study participants were females attending a health facility. They might not have experienced sexual violence and thus may not be having the information regarding trauma-informed services for survivors of sexual violence. We counteracted this by including participants of different age groups who provided us with views about trauma-informed services for sexual violence. Participants were encouraged to share with us what was taking place in the community. Also, some of the participants shared lived experiences of sexual violence. In addition, sexual violence is a sensitive topic in the community. Thus, the respondents may have felt compelled to answer questions in a particular manner whether or not it represented their opinions despite being assured of confidentiality and that data would be anonymized. We limited this by using interviewers who had no established contacts with participants before. Nevertheless, this study provides results that can be used with adaption in similar settings to improve the lives of survivors of sexual violence. --- Conclusions Generally, the participants showed a knowledge gap about trauma-informed services for survivors of sexual violence. Sexual violence was present in the community. The study highlighted barriers to trauma-informed services. There is a need for thorough sensitization about these services and evaluation of different service providers by authorities to improve service delivery. The knowledge, attitudes, and factors associated with the utilization of trauma-informed services should be considered while formulating interventions for survivors of sexual violence. Further studies are needed to document the lived experiences of survivors of sexual violence. --- Abbreviations ARVs: Antiretroviral therapy; AIDS: Acquired immune deficiency syndrome; HEPI-HEPI-TUITAH: Health professional education partnership initiative -transforming Ugandan institutions training against HIV/AIDS; HIV: Human immune virus; IDI: In-depth interview; KIG: Kigarama; KYA: Kyangyenyi; MUST: Mbarara University of Science and Technology. --- --- Authors' information EA and IO are HEPI-TUITAH micro-research grantees, bachelor of nursing students at Mbarara University of Science and Technology . RO is a HEPI-TUITAH micro-research grantee, a bachelor of medicine and surgery student at MUST. MA is a HEPI-TUITAH micro-research grantee, a third-year bachelor of medical laboratory science student at MUST. BA is a HEPI-TUITAH micro-research grantee, a bachelor of physiotherapy student at MUST. VN is a nurse, Lecturer at MUST. VM is a family physician, Lecturer at MUST. --- Funding --- --- --- --- Competing interests None declared . ---

Background: Sexual violence is a public health concern globally and locally, and trauma-informed services are put in place to mitigate its consequences. A few studies have evaluated the quality and uptake of these trauma-informed services for sexual violence. This study aimed at; i) assessing the knowledge of participants about trauma-informed services, ii) exploring the attitudes of females about trauma-informed services, and iii) assessing different factors associated with the utilization of trauma-informed services. Methods: This study employed a descriptive cross-sectional qualitative design. The participants were females at Kyangyenyi health center III and Kigarama health center III in Sheema district, southwestern Uganda. We used a purposive sampling procedure for all participants and then a consecutive sampling of females. Data about; knowledge of trauma-informed services for sexual violence, attitudes towards trauma-informed services, and factors influencing the utilization of trauma-informed services were collected using an in-depth interview guide. Data were analyzed using thematic content analysis.We interviewed 32 participants. There was a high prevalence of sexual violence, and it was a big concern in the community. Many of the respondents were not sensitized about trauma-informed services. Most of them knew only about HIV treatment. Our study shows that a good section of females did not seek the services after being sexually violated due to the fear of the perpetrator, bribing of the family of the affected and authorities, or even fear of family breakup and stigmatization. There were a lot of bribery, poor road networks, and inadequate health care services. These hindered survivors of sexual violence from utilizing trauma-informed services. Conclusions: There was a knowledge gap about trauma-informed services for survivors of sexual violence. There was sexual violence in the community. Sensitization needs to be done among the populations by respective authorities to iron out issues of ignorance about the services and health workers evaluated for competence in offering the traumainformed services.

INTRODUCTION Most children living with HIV reside in sub-Saharan Africa . Of these, an overwhelming majority are unaware of their HIV-positive diagnosis . With meticulous adherence to antiretroviral therapy , children infected with HIV can survive into adulthood and assume responsibility for their own care . The issue of how and when to inform children of their status has received attention in recent years. Evidence to support disclosure is robust, and both the American Academy of Pediatrics and the World Health Organization support status disclosure in an age-appropriate fashion and recommend that disclosure be integrated into clinical care . Natasi and colleagues describe a "culturally constructed clinical practice model," in which illness narratives guide practice . 'Illness narratives' are stories which express understanding of the etiology and outcome of illness within a broader socio-structural environment . In pediatric oncology, which has served as an exemplar for the benefit of disclosure, illness narratives of children with terminal cancer produce a framework on which the argument for disclosure was based . Like pediatric cancer patients, children infected with HIV have likely developed explanations for the lived experience of their illness . In the 1980's Bibace and Walsh suggested that illness perceptions in children progress analogously to Piaget's stages of cognitive development, maturing from illness as an external phenomenon to a psychologic and physiologic one . Yet, it is unclear how a child living with HIV perceives the illness experience when the child has not been informed that he or she has HIV. Illness narratives among children living with HIV have been explored to a limited extent by others . In the United States, a grounded-theory analysis of drawings showed that children who do not know they are HIV-positive exhibit emotional distress and disturbed self-image . In another study in Brazil, children living in an AIDS institution who did not know their status exhibited confusion regarding the need to take medication . Semistructured interviews with Congolese youth living with HIV revealed that, before knowing their diagnosis, youth reported non-adherence to medications . One participant recalled: "Before, I was refusing to take my medicine, but since they announced to me that I was infected with that disease, I started taking medicines without any problems" . In this study, we explored the illness perceptions of children enrolled in a disclosure intervention study in Ghana . To our knowledge, this is the first qualitative study exploring illness perceptions of children infected with HIV in SSA prior to disclosure. --- METHODS --- Design This study examined qualitatively the illness perceptions of children living with HIV who had not yet been informed of their HIV diagnosis. The data were collected in interviews at baseline as part of the Sankofa study, a randomized controlled trial that tested a pediatric HIV disclosure intervention . Study procedures were approved by the Institutional Review Boards of University of Ghana Medical School, Komfo Anokye Teaching Hospital, and Yale University Human Investigation Committee. --- Study Setting and Recruitment Full details of the Sankofa protocol have been described elsewhere . In brief, participants were enrolled in the Sankofa disclosure intervention study at Komfo Anokye Teaching Hospital in Kumasi, Ghana, and the Korle-Bu Teaching Hospital in Accra, Ghana, between January 2013, and June 2016. Together, these sites provide care to 3500 children living with HIV. The children routinely attend clinic every three to four months and more frequently as needed. The study was introduced to attendees by designated site project staff members during regular clinic visits . The study was explained, and written consent was obtained from caregivers while verbal assent was obtained from children. Non-disclosure was ascertained from screening questions administered to caregivers and medical records. Children aged 7-18 who were enrolled in clinical care who did not know their HIV status were eligible to participate. Exclusion criteria for the children included: HIV-infected children less than 7 years of age and children with congenital or developmental disorders, comorbidities such as sickle cell disease or diabetes that require frequent clinic visits or hospitalizations, clinically identified AIDS-defining illness present in the GHS chart at entry, or end-stage AIDS clinically identified in the GHS chart at entry. This sub-study included 435 out of a total 446 Sankofa child participants. Qualitative interview data were missing for eleven of the child participants. --- Interview Procedures Interviews were conducted with the child participants through individual face-to-face interviews at baseline following enrollment to the study. Four study staff with backgrounds in clinical psychology conducted the interviews in a private setting in the clinic. They were actively employed by each hospital and responsible for HIV and sexually transmitted infection-related counselling. They have experience in qualitative research and the data was analyzed by team members who have experience in qualitative research. In order to avoid potential accidental disclosure and to foster trust, children were interviewed in the presence of their caregivers. Interview guides with semi-structured, open-ended questions and prompts were used to explore the child's illness perceptions. Interviews lasted approximately 20 minutes. Interviews were conducted in either English or Twi/Ga and responses were recorded. The child was asked "Can you tell me why you are at the clinic today?" The interviewer would use whatever cognition the child used in his or her response for the rest of the discussion. This included queries such as: "How long have you had [this condition/ illness]? How did you learn that you have [this condition/illness]? Do you know anyone else who has [the same condition/illness or have to take medicines] like you? Does having [this condition/illness or taking medicines] affect your life very much compared to other kids? Have you ever talked about having [your condition/illness or taking medicines] with anyone? If yes, the interviewer asked who had been told and "What kind of things did you talk about?" Who knows about [condition/illness or that you have to take medicines]? Do you ever have any questions about [your condition/illness or medicines] or things that you'd like to talk about with someone but haven't been able to? Responses to open-ended questions were recorded in the first person and with interviewer's observations . Responses ranged from 1-2 words to longer phrases , with an average length of 7 words per record in answer to the first open-ended question and 5 words per record in answer to the second open-ended question. --- Data and Analytic Procedures The transcribed interviews were coded and thematically analyzed using an inductive technique . Two researchers initially developed a framework of key issues and themes. Initial codes were generated using open coding . The first data were compared for similarities and differences and coded into categories that reflected the children's perceptions. Codes were compared, and similar codes were clustered and given an initial label, thus forming a category. The data were analyzed for relationships among categories. For example, in response to the question, "can you tell me why you are at the clinic today," the codes for appointment, for laboratory testing, for injection were collated into the theme "Routine experiences of clinical care." Many responses included multiple reasons for the visit, thus themes are not mutually exclusive. Themes which characterize more than 20 responses per question are summarized here; these are considered to have reached thematic saturation . Further analysis produced other categories. A comprehensive data chart was constructed by extracting segments of data from the transcripts and arranging these in a matrix according to emerging themes. After verification, the corpus of extracts was managed using a password-protected Microsoft Excel spreadsheet . Draft copies of the written report were shared with each member of the research team for review, verification, and comment. Children's ages are reported in brackets following quotes. In response to dichotomous questions, frequency of response is reported. --- RESULTS --- Participant Characteristics Sociodemographic characteristics of participant caregivers and children have been published . In brief, 78% of caregivers reported self-employment and 63% reported a monthly household income of ≤300 Ghana Cedis . A majority of caregivers had completed elementary education or less . Characteristics of participant children are reported in Table I. Children ranged from 6 to 18 years of age . There were approximately equal numbers of males and females . Most children acquired HIV via maternal to child transmission . Of the children with available CD4+ counts results , 23% had a CD4+ count of <500 cells/mL, meeting WHO classification for 'mild to severe HIV-associated immunodeficiency'; 9.4% had a CD4+ count below 200 cells/mL, meeting criteria for AIDS . --- Reason for The Clinical Visit Interviewers asked children "Can you tell me why you are at the clinic today?" This was followed by further prompts, as needed. Responses were available for 335 children. The following themes emerged: The need to collect medication.-The most cited reason for clinic attendance was to collect medication. Some children who mentioned medication linked the need to take it to specific health concerns or symptoms: "I have been having stomach and headaches occasionally, whenever that happens I come here for medication." [12 years of age] Others associated the need to take medicine to unspecified illness. "I came to collect medicine because I have a disease in my body." [11 years of age]. Other children expressed that they took medication but did not know why: "My aunty said my medicine is finished so I came to collect more. I don't know why I take the medicine." [9 years of age]. Others indicated that though they took medication, they were specifically not sick: "I came for a check-up and I will later be given medicine. There is nothing wrong with me." [10 years of age] Specific ailment or physical malady.-Children expressed that they were attending clinic to address specific physical concerns. Common ailments included stomach pains, cough, headache, rashes, and ear problems. Other children said that they attended the clinic due to chronic conditions other than HIV/AIDS: "I [am] coming for[a] check-up and medicine because I am sick. Mother told me something hit my heart when I was young. ." [12 years of age] Ignorance.-Many children could not offer any explanation for why they were at the clinic. "I just followed my dad here and I have no idea why I'm here." [10 years of age] Routine lived experiences of clinical care.-For many children, clinic visits appeared to be routine . This included visits for: review , blood work, or injections. "I came for review. It means seeing the doctor and talking with him about your problems. I am not sick." --- Disclosure to Others Children were asked: "Have you ever talked about having your 'condition' or 'illness' [child's terminology used] or taking medicines with anyone?" Of 320 respondents, a minority indicated 'Yes,' while a majority of respondents indicated 'No.' Those children who had never told anyone were asked to extrapolate why they had not; 187 responses were recorded. Answers to the question "What brings you to the clinic today" are included in double brackets. Other Reasons.-Less common explanations included that nobody had asked , others wouldn't believe them , or that they were feeling well : "Am feeling better because of the medicine. That is why I don't discuss with others." [[I came because of sickness.]] [14 years of age]. --- DISCUSSION In this study, we explored the illness perceptions of children living with HIV who had not been informed of their diagnosis. None of the children indicated awareness of having an HIV diagnosis when asked why they were attending the outpatient [HIV] clinic. Many children mentioned physical manifestations of illness as the reason for clinic attendance. This included both symptoms possibly linked to HIV-infection and those not necessarily linked to their HIV-infection . We previously reported that the overall disclosure rate among school-aged children in Ghana is low as 21% . Reasons for non-disclosure include caregiver fears that it will cause undue psychologic injury to children and that informed children living with HIV will subsequently disclose to others . In contrast to the common caregiver concern that children who had been disclosed to would be unable to keep their status secret, a majority of children interviewed reported that they did not talk to others about their illness/ unwellness. Many children reasoned that they maintained secrecy because their caregiver had asked them to. Other children attributed their secrecy to fears of adverse repercussions despite presumably not knowing their HIV status. Interestingly, these findings are similar to those from a qualitative study conducted by Abubakar and colleagues, which showed that among children living with HIV who knew their status in Kenya, secrecy was maintained because of fears of rejection, isolation, or stigmatization . HIV-related stigma has been well-extrapolated and incorporates negative attitudes towards persons living with HIV/AIDS, fear of disclosure , and internalized stigma . Children in our cohort exhibited signs of both internalized and anticipated stigma. Although nondisclosure may act as a protective factor with respect to avoidance of enacted stigma, internalized stigma has been shown to diminish quality of life among persons living with chronic illnesses and is associated with anxiety and depression . This suggests that as these children age into adulthood, interventions to encourage adherence to medications, disclosure to sexual partners, and psychosocial well-being should build upon an understanding of lifelong experiences with stigma as well. Caregivers also express concerns that disclosing to children will cause them distress. In pediatric oncology, social scientists find that even among children who "do not know" their condition, partial awareness along with a state of unknowing is present . The state of unknowing was mentioned often by children in our cohort: taking medication , coming for check-ups . Many children explained clinical care in a distinctly pre-operational manner, positioning caregivers as external phenomenon, driving decision-making. Other children went so far as to reiterate wellness , drawing attention to the dissonance between attending a medical visit and internalized notions of wellness. Some children expressed a knowledge that there was an explanatory framework that had not yet been revealed to them. As disclosure protocols are designed, in-depth interviews before and after disclosure might expound upon some of the themes identified here. Before disclosing to children living with HIV, specialists should seek to determine the child's current beliefs about illness and the source of those beliefs. Specifically, future work might extrapolate on how children's understanding of illness relates to their understanding of the importance of medication and their overall adherence behaviors. A majority of children reported that they learned about their illness from a family member. Clinic-based disclosure interventions must be careful not to alter the trust between caregiver and child while also providing reliable, accurate information. Our unique study population, very large sample size, and topic of interest make our findings unique. However, our study was in part limited by the cross-sectional design. Data regarding the illness perceptions of the children living with HIV were only captured at baseline. It would be valuable to capture the children's illness perceptions over time and post-disclosure in future studies. In the Sankofa study, over the course of nearly three years, specialists met with children-caregiver dyads and built the skills necessary to facilitate disclosure. However, illness perceptions were not captured systematically over time. Future studies might also report objective measures of cognitive development using an adapted Piagetian assessment such as the Southern California Ordinal Scales of Development . This type of analysis would likely yield more information about beliefs, stigma, fears and developmental changes of children. --- CONCLUSION Children living with HIV who do not yet know their status have nevertheless formed explanations for their lived experiences with HIV. Held beliefs among children engaged in HIV-related care in SSA is of particular interest as national and regional disclosure protocols are standardized and disseminated.

Despite available guidelines for disclosure of HIV status to children, most children living with HIV are unaware of their diagnosis. We sought to characterize the concepts of illness and treatment among children living with HIV who do not know their status. As part of the Sankofa trial we interviewed 435 children aged 6-18 enrolled in clinical care at pediatric HIV clinics at two teaching hospitals in Ghana. Theoretic thematic analysis generated themes among responses. Terms of use and reuse: academic research for non-commercial purposes, see here for full terms. https://www.springer.com/aamterms-v1

Introduction Primary care is the first point of contact for most young people seeking nonemergency healthcare in Aotearoa New Zealand . In line with the common NZ use, the term "rainbow" is used in this paper to describe people of diverse genders, sexualities, and variations in sex characteristics, and is inclusive of indigenous genders, such as takatāpui 1 and we define young people or youth as those age 15-25 years old. Rainbow young people seek the same general healthcare as their cisgender heterosexual peers but may also present with sexuality-or gender-specific health needs . RYP experience higher rates of mental health distress than their cisgender heterosexual peers, 2 which is another reason they may access healthcare. International and local research shows that RYP face a range of unique barriers in accessing appropriate primary healthcare, and report negative experiences accessing primary healthcare. 3,4 These barriers include heteronormative and cisnormative environments such as lack of gender identity options on forms, gendered pamphlets and incorrect use of names and pronouns, as well as providers who lack knowledge of appropriate care such as gender-affirming healthcare. 5 One group of RYP who face particular barriers is transgender and nonbinary young people. Almost half of the 1,178 respondents in the 2018 Counting Ourselves national survey of transgender people in NZ reported that a general healthcare provider knowingly referred to them by the wrong gender, and a significant number of youth respondents reported not accessing healthcare or feeling uncomfortable discussing their gender with their general practitioner . 2 Transgender young people often report feeling the need to educate their GP or nurse and that this can lead to distress. 6 Conversely, experiencing respectful interactions and feeling comfortable with their GP has been shown to have a positive effect on the mental health of transgender patients. 7 Many individuals who belong to a sexual minority report being apprehensive toward disclosing their sexuality to their healthcare provider, leading to the specific health needs of this group being poorly served. 8 A NZ study found that only half of gay men had disclosed their sexuality to their GP, and this was especially true for younger patients. 9 This is of particular concern because this population may have higher rates of sexually transmitted infections 10 and are at higher risk of acquiring HIV infection. Sexual minority youth are at higher risk of experiencing mental distress, including depression, anxiety, suicidal ideation, and nonsuicidal self-injury, 11 and are more likely to smoke cigarettes, use cannabis, and/or take multiple substances than their heterosexual peers. 12 Sexual minorities are also at greater risk of experiencing sexual violence and intimate partner violence, with bisexual women facing disproportionate risk. 13 RYP overall face increased minority stress linked to discrimination, exclusion, and lack of visibility in heteronormative and cisnormative landscapes. 14 These effects are compounded for Māori who face the additional barriers created by racism and colonization, and whose conceptualizations of gender and sexuality may differ from those understood by western systems. [15][16][17] Past negative experiences with a GP can impact on a RYP's level of comfort disclosing their sexual orientation and the likelihood of accessing or returning for healthcare. 8 This in turn can further contribute to existing health disparities and result in missed opportunities for appropriate healthcare. 18 Lack of education on gender and sexuality diversity at undergraduate level, and few subsequent training opportunities, can be a barrier to providing rainbow-affirming care, using inclusive language, and awareness of the sexual and general health needs of rainbow patients. 19 There is little research describing the views and education needs of general practice staff providing care to RYP in NZ. This study sought to explore general practice clinician and staff views, experiences, and learning needs related to the provision of care to RYP. --- Methods --- Participants and recruitment Three focus groups were held during April and May 2021. Focus groups were chosen because of the potential to generate rich data from a range of practices and staff perspectives over a short period of time. 20 Practices were purposively selected for invitation based on the research team's knowledge of their practice population and reputation within the rainbow community. Five practices were invited to attend, and all accepted the invitation. Two of the practices who accepted the invitation were known to have 1 or more "rainbow-friendly" staff, a good reputation amongst rainbow patients or more experience with RYP, and 3 practices had no reputation or focus on caring for this population. The research team includes clinical expertise in rainbow healthcare and 2 members of the team are from the rainbow community. An invitation was emailed to practice managers and several individuals known to research team members, with an information sheet and consent form. All staff were invited to attend, including administrators, receptionists, and practice managers, as well as clinicians . One practice team took part in a single-practice face-to-face focus group during a lunchtime at the clinic. Four practices attended a multipractice session held at a community hall in the evening. This larger group was divided into 2 focus groups, deliberately composed of interprofessional and mixed practice groupings and a 45-min rainbow health education session was given at the end of the focus group for all attendees. --- Data collection A semistructured discussion schedule was developed from relevant topics and the facilitators took an appreciative enquiry approach. 21 The focus group questions were tested for understandability by the research team, which includes primary care professionals. Participant views were sought on current provision of rainbow-friendly care and what services, approaches, and environment they provide to increase comfort for RYP attending. They were also asked about education received or needed to support their provision of care and what resources they currently use. --- Analysis Focus groups were audio-recorded and transcribed verbatim. Transcripts managed in NVivo were analysed using thematic analysis. 22 Firstly, 2 researchers used an inductive process to review and code transcripts, to identify categories relevant to the research question. All data collated to each category was then reviewed to identify salient patterns and themes in the data. Provisional themes were then documented, reviewed, and discussed by all members of the research team until consensus on the final overarching themes was reached. 1. There were 25 participants in total including 2 Practice Managers , 1 Primary Care Assistant , 15 Practice Nurses/Nurse Managers , and 7 General Practitioners . The descriptor in brackets is used to describe the role of participants following quotes presented in the results. --- Results --- Characteristics of participating practices and staff are shown in Table Four main themes were identified from the data: Practice experience, Knowledge and training, Feeling awkward, and Systems limitations. --- Key messages • Gender and sexual minority youth report negative healthcare experiences. • GP staff often feel awkward in consultations with GSM youth. • Systems limitations contribute to awkward clinical encounters. • Levels of confidence and knowledge vary greatly between GP staff. • GP staff are keen for more education in this area. Staff from practices 1 and 3 considered that their practices had built positive reputations amongst the rainbow community through word of mouth, and perceived they saw higher numbers of rainbow patients than other practices as a result. Participants from these practices described their staff as reasonably experienced and confident in delivering care to RYP. …we've got a really good reputation with the rainbow community, we have a higher-than-average trans community here because we have a good reputation, it stems from a couple of patients and then spread by word of mouth… …so, I think, you know, on the grapevine, people you know, know that is a good safe place to come. Staff from other practices described more limited experiences and confidence in working with RYP. Some perceived that the number of RYP attending their practice was low and they described having less experience and knowledge. this mum rang up and she was like, "I need to book in my son for their puberty blocker," and everyone's like, "what the heck is a puberty blocker?" I had no idea I've referred one patient [for gender-affirming healthcare] and this was about two years ago… The importance of strong progressive practice leadership and knowledge of rainbow healthcare was viewed as key to providing rainbow-friendly care by staff from both practices who reported higher levels of experience. One practice had a manager who championed rainbow health and inclusivity, and notably, both practices employed GPs who also worked in youth specialist services . It was clear from comments that these knowledgeable staff members took on leadership and supportive roles with other clinical staff in the practice, as well as influencing the culture of the practice with intentional hiring of staff aligned with this open-minded philosophy. We're quite lucky that we have a couple of GPs that have come from , so they've got heaps of experience and kind of guide us a little bit. I think that the thing that makes it easiest for us to be hopefully as good as we think we might be in this area, is that we've got really strong leadership on the issue. Although participants from the more experienced practices claimed all their staff were supportive and confident managing RYP, it was indirectly acknowledged that some staff were more experienced than others, and that RYP tend to gravitate to particular clinicians. at reception, when you enrol a new patient … you would kind of direct them to the ones that you know would be most supportive --- Theme 2. Knowledge and training Clinical staff from most practices commented they had received little or no previous formal training on rainbow healthcare provision. The topic had not been covered in their preregistration training , and instead rainbow healthcare was something they had learnt about "on-the-job." Yeah, it wasn't something, it was like over 20 years ago [I trained], so it wasn't really something that was taught. it's one of the many things that um, comes up and you learn as you go. Since registration, most clinical staff mentioned they had not received any ongoing training or professional development specifically related to rainbow healthcare, nor could most recall being offered any such educational opportunities. Clinical staff from all participating practices noted gaps in their knowledge around caring for RYP, of specific rainbow health interventions , and/or where to find information for themselves or for patients or their parents. I've felt my way all the way through this, it would be great if there was a bit more out there for upskilling and upknowledging Sometimes too though in your role, you haven't got time. A lot of it's time critical so you want something at your fingertips to be able to just go, so it's easier to pick up the phone and ring somebody isn't it and talk to them, rather than spend hours on the net fossicking through. Clinical staff expressed feeling "behind the game" in terms of their knowledge compared to RYP and described themselves as often being led by RYP in consultations. Examples of this included discussion of binding and PrEP prescribing. …I didn't know about binding and so then I bring it up and they know all about it, so oh yeah, we've been doing it for ages, move on. Right, you know, you just feel like you know, offering as much as you could 'cause you're already behind the game. … It felt like … they were kind of dragging me through. I think kind of worryingly it's the patients themselves who have pointed me to resources like the fellow I prescribed PrEP to… Training needs reflected the gaps in knowledge identified by general practice staff. Participants from all practices expressed a desire for more education to upskill and improve their knowledge of both rainbow healthcare issues, and how to work more effectively with RYP. Whole-practice sessions where information and experiences could be shared between all staff were considered an effective form of training. Ensuring reception/admin staff were included in training was also viewed as important. I think having somebody coming into a practice to do a session would be beneficial because if you get people to do things online or read this in your own time, if you get everybody together and then it's an open conversation, I think that's the best way to have it. … it drums up conversation between the nurses, the reception, the doctors so everybody sort of communicates how they deal with things and gets a better idea of how to manage it or what can be an awkward situation for the person as well as the staff. --- Theme 3. Feeling awkward A common theme identified in many participants' accounts was a sense of discomfort when talking with and addressing RYPs' health needs. The term "awkward" was used repeatedly by different staff and in separate focus groups. This appeared to be due to personal discomfort, concern about saying or doing the wrong thing, perceived discomfort of the patient, general unease around not knowing how to refer to RYP or lacking knowledge about a specific topic. Participants from the less experienced practices commented they did not know how to have "these sorts of conversations" with RYP. I'm sort of looking through the pathway which talks about discussing things like binding the breasts and you know, suppressing the [menstrual] period and these are all things I don't know about, so it was a very awkward consultation because I didn't know… …. what can be an awkward situation for the person as well as the staff Undertaking cervical smears with RYP was identified as particularly uncomfortable. Clinical staff from a range of practices raised this as an area where they lacked confidence, including practices with more experience seeing RYP. Smears are extremely awkward, you know, they still have a cervix, we still need a smear, but they don't want it because that's what women have and so that's really awkward How do you have that discussion? [about smears with transgender males] Discomfort was also evident in examples staff shared where they "got it wrong." They expressed the need to be mindful and to take particular care when addressing RYP so as not to cause offense. it's knowing how to deal with that [misgendering someone] 'cause some things that you are not so used to doing, you know and that's where I think we need more support, so we know how to do these things, you know, otherwise it doesn't come across too good. I'm a smear taker and I certainly have early on in my smear taking career made the mistake of assuming that someone has a male partner. --- Theme 4. Systems limitations Limitations of patient management systems for recording gender and preferred names and pronouns were a key challenge identified by all practice staff . Participants noted that the field for recording gender within the PMS was based on sex assigned at birth and there was no built-in field for recording where a person's gender was different to that assigned at birth. Practices with more experience were more likely to Downloaded from https://academic.oup.com/fampra/advance-article/doi/10.1093/fampra/cmad024/7076994 by Hochschule Luzern user on 20 February 2024 have systems in place to record this information such as a free text "alert" on the patient's file. However, all agreed the alert method of identifying RYP was inadequate, as alerts could easily be missed increasing the likelihood of addressing RYP inappropriately. An IT system which made this easier could improve staff comfort by getting the practice encounter, including reception phone calls and the consultation, off to a better start and reducing fears about misgendering. So, we do have alerts for pronouns on the PMS, it'll say "uses they/them pronouns" Sometimes they tend to be a bit more like wallpaper, you don't really read them. Due to reliance on gender being entered into the PMS as male or female to recall patients for screening or other interventions, participants described situations where RYP were recalled erroneously or missed from recalls . Furthermore, inflexibility within the PMS also meant referral letters or other documents generated from this system would misgender patients or use their deadname . I don't think our computer systems are very transgenderfriendly, either. Because if you change their sex, then they come up with smears and they don't have a cervix to take smears from……. Yeah, cos it's really embarrassing … the other issue with MedTech [the PMS] is that if they go to a male, it removes all the recalls of smears that, [lab provider] cancels them all out, you can't order them. Accessing mental healthcare either as part of gender-affirming intervention criteria or to access timely affordable mental health support for RYP was also raised as a barrier for young people and a challenge for practice staff, as demand for affordable services resulted in long waiting times and resulting mental distress. There are some quite significant steps to be taken before starting medication that aren't easily accessible, the psychologists particularly, they're not easily accessible in primary care Although all the participating practices were keen to meet the needs of RYP, differences were observed in health provider knowledge, comfort, and experience in providing primary care for RYP, particularly when comparing practices who had more interactions with RYP with those who had few interactions. This observation is depicted in Fig. 1. Practices experienced with seeing RYP, attracted more RYP and therefore became even more experienced and confident. By contrast, practices with less experience, knowledge, and confidence attracted fewer RYP, in turn limiting their opportunities to learn and improve their provision of care. --- Discussion Discussion with general practice staff in this study identified 4 main themes. The experience of practice staff, feelings of awkwardness, knowledge, and training, and systems barriers were identified as key factors impacting on the frequency and quality of care provided to RYP. Participants who recognized their own limitations were eager to upskill to gain knowledge and confidence. A self-perpetuating cycle of practices with knowledge and experience attracting more RYP through reputation and practices with less knowledge and experience attracting less RYP was observed. All general practice staff need to provide welcoming, inclusive and safe environments, regardless of how many RYP attend the practice. 23 Studies in this area repeatedly demonstrate that relatively small changes in general practice, such as correct use of names and pronouns, can impact positively on a RYP's level of comfort and willingness to attend a clinic to seek healthcare. 5 Visual signs of inclusivity such as rainbow flags and relevant pamphlets can play crucial roles in facilitating an environment where disclosure can happen, thereby ensuring the health needs of this population are effectively met. 24 When healthcare providers treat RYP with respect and learn to overcome their own discomfort, it enables the development of trusting relationships and improves communication. This a vital first step which is achievable by all general practice providers. The NZ Counting Ourselves survey findings indicated that the most common supportive interactions with general practice doctors, which had a positive impact on transgender patients' wellbeing, included being treated the same as any other patient, demonstrating support to patients seeking gender-affirming care, using correct names and pronouns, and health providers being willing to educate themselves when necessary. 7 This suggests that healthcare providers do not need in-depth specialized knowledge to provide supportive interactions with their transgender patients and can be reassured that implementing these basic first steps will make a positive difference. Healthcare providers in this study described a lack of training in rainbow-specific healthcare contributing to their lack of knowledge and confidence in this area. Most participants had been working in clinical practice for many years, which may be why undergraduate education in this area was lacking, however more recent surveys of NZ and Australian medical school curricula have shown an ongoing issue with limited content on rainbow health. 25 A recent Australian survey of 188 primary healthcare providers found that lack of knowledge and feelings of significant discomfort were common if providers were asked to provide care to transgender patients, both of which improved after an education session which included a community member with lived experience. 26 The introduction of rainbow-specific teaching as standard to all health providers is likely to improve knowledge and reduce feelings of awkwardness among healthcare providers, and will likely support young people to disclose their rainbow identity when relevant. Education and resources which focus on consultation skills and practical examples of how to phrase questions could improve health provider comfort, and therefore allow RYP to engage more easily. Developing a whole-of-practice rainbowfriendly healthcare environment requires not only clinical staff to upskill, as reception and administration have a key role to play in welcoming the RYP to the practice. It has been shown that reception and administration staff are aware of the importance of their role and were keen to learn how to meet the needs of RYP. 27 Many of the systems barriers identified relate to inflexible IT used to store patient information. System limitations include the inability to accommodate gender diversity, compounding opportunities for RYP to be misgendered or called by a name they no longer use. This further contributes to potentially "awkward" clinical encounters as mistakes are more likely to occur. Some practices had found ways around this while others had not. Implementation of a standardized IT system with the ability to accurately record a person's name, gender identity, and pronouns across all areas of the health system is fundamental to the provision of appropriate healthcare to RYP. --- Strengths and limitations This is one of the first studies to describe the experiences, challenges, and needs of general practice staff who provide healthcare to RYP in NZ. A key strength of our study is the inclusion of a range of practices and participants from a mix of disciplines. Use of focus group methodology allowed participants to exchange anecdotes about their experiences and challenges they had faced. Participants engaged in open and frank discussion about their own perceived shortcomings and identified where they needed further information and education. The addition of the teaching session following the multipractice focus groups provided a forum for upskilling in some of the areas identified by participants as knowledge gaps. Our results may not be generalizable to the experiences of staff in other practices or regions. Participation in the focus groups by nonclinical, patient-facing staff such as receptionists and telephonists were sought but few took part which may be because professional development for nonclinical practice staff is often not funded or that they did not see the relevance of the topic to their role. These staff members might have shared unique or contrasting views and issues to those raised by participants in this study. As the study took place in a city viewed as progressive and rainbow-friendly, results may be different in other regions of NZ. A rainbow health education session was provided at the end of the multipractice session in an attempt to draw in more participants, however this could have influenced the staff who chose to attend. They may have been staff who lacked knowledge and wanted to upskill, or it may have drawn in those who already had an interest in this topic. --- Conclusions It is essential that all primary care practice staff including healthcare professionals, reception, and administration staff, receive education on rainbow healthcare. A focus on interpersonal communication strategies and consultation skills with examples of inclusive language, particularly relating to sensitive topics, is likely to have a positive impact on consultations for both health providers and RYP. Further work is needed to develop and test effective methods to teach these skills to all members of the primary care team in order to improve staff confidence with RYP. Providing safe environments for RYP in all practices regardless of the numbers of RYP currently seen, may lead to more exposure and learning opportunities for less experienced staff. National clinical IT systems require a unified approach to ensure appropriate recording of name, pronouns, and gender identity across all clinical settings. By addressing the education and experience needs of primary care providers, care for RYP can be improved which may result in improved health outcomes, less distress, and improved access to general practice. --- Data availability The data underlying this article cannot be shared publicly due for the privacy of individuals that participated in the study. --- Ethical approval Ethical approval was obtained from the University of Otago, Human Ethics Committee . ---

Background: Rainbow young people (RYP; people of diverse genders, sexualities, and variations in sex characteristics) face barriers accessing primary care, often report negative experiences and the need to educate clinicians on rainbow-specific issues. This study explores general practice staff views and experiences of providing care to RYP. Methods: Three focus groups were held with staff from 5 general practices in an urban area of Aotearoa New Zealand (25 participants in total). Practices were purposively selected to include some known to be more "rainbow-friendly" seeing larger numbers of RYP and some with no particular focus on RYP. All members of each practice were invited to participate, including administration and reception staff. Focus groups were audio-recorded, transcribed, and analysed in NVivo using inductive thematic analysis. Results: Four main themes were identified: (i) Practice experience, (ii) Feeling awkward, (iii) Knowledge and training, and (iv) Structural barriers. Differences were apparent in health provider knowledge, comfort, and experience in providing care to RYP. Participants identified a lack of knowledge and training and wanted more resources and education. Systems limitations were common (e.g. with IT systems for recording gender) and often contributed to awkward clinical encounters.Participants recognized shortcomings in their training, knowledge, and level of confidence providing care to RYP and expressed a desire to improve their competency. Further work is needed to identify and trial practical strategies that help improve communication skills, knowledge, and the delivery of more equitable healthcare to RYP.

INTRODUCTION Across all levels of public education, academic achievement gaps exist between students of color and White students. Such gaps may be interpreted to mean that public schools are failing to meet the needs of students of color . Because students of color compose 41% of the United States' public school student population, it is imperative to remediate existing academic achievement gaps between ethnic/racial groups of students . In Texas, the state of interest for this article, public school student enrollment continues to increase as well as becoming more ethnically/racially diverse . During the 2018-2019 school year Texas had a high school enrollment of 1,563,774 students . For the same school year Texas public high schools had 196,335 Black students, 12.55% of the state's high school population; 808,204 Hispanic students, 51.6% of the state's student population; and 448,630 White students, 28.76% of the state's public school student population . During that time Texas reported having 358,450 public school teachers, of which 10.6% were Black, 27.7% were Hispanic, and 58.4% were White . The statistical disparities between the ethnic/racial composition of students and teachers in Texas public schools, documented in the statistics presented above, have drawn scrutiny from previous researchers . In earlier studies conducted by Bone and Khan , the ethnic/racial composition of Texas public school students and the teacher workforce were analyzed. Bone examined the trends of student and teacher demographics at the high school level from the 1999-2000 school year through the 2009-2010 school year. In a subsequent investigation, Khan analyzed the same trends but for the 2002-2003 school year through the 2012-2013 school year. Both Bone and Khan documented the presence of trends in the ethnic/racial population of Texas public school students and teachers. As such, they generated concerns regarding Texas public school hiring practices as the diversity of the state's public school student population increases. For the 11-school years of data that were examined by Bone , a larger percentage of White teachers was employed at Texas public high schools than at Texas public elementary schools or at middle schools. Bone established that the percentage of Black Texas public high school teachers increased from 0.30% in the 1999-2000 school year to 2.59% in the 2009-2010 school year. During the same time span, the average percentage of Black students in Texas public high schools increased from 5.00% to 6.9% . Although both the percentages of Black teachers and of Black students were low, the percentage of Black students was more than twice the percentage of Black teachers. The numbers and percentages of Hispanic teachers and Hispanic students both grew during the Bone study, with the Hispanic student population increasing by a much greater percentage than the percentage of Hispanic teachers. The average percentage of Texas public high school Hispanic teachers increased over the 11year study from 3.4% to 8.75%, whereas the average Hispanic student average percentages increased from 22.20% to 40.20 % over the school years examined by Bone . Accordingly, the percentage of Hispanic high school students was more than four times the percentage of Hispanic high school teachers. The average percentage of White teachers decreased from 91% to 78%, and the average percentage of White students decreased from 56.60% to 33.30% over the 11-school years. In contrast to the underrepresentation of Black teachers and of Hispanic teachers in comparison URL: www.onlinesciencepublishing.com | April, 2023 to the percentage of Black students and Hispanic students, the percentage of White teachers was more than two times the percentage of White students. In a subsequent study by Khan , which spanned from the 2002-2003 school year through the 2012-2013 school year, similar trends were described for the ethnic/racial diversity of Texas's public high school teaching population. The percentage of Black teachers at Texas public high schools ranged from 7.9% to 9.3% . Black student percentages grew from 5.00% to 6.90% over the course of Khan study. The percentage of Black teachers at Texas public high schools was consistently one and a half times higher than that of Black students. Within the same study, the Hispanic teacher percentages grew from 14.6% in the 2002-2003 school year to 20.6% in the 2012-2013 school year. However, the average percentage of high school Hispanic teachers was the lowest for Hispanic teachers across all three Texas public school levels . Though the Hispanic student percentages composed nearly 50% of the student population at the high school level, the percentage of Hispanic teachers fell well behind the percentage of White teachers. Khan documented that the percentage of White teachers in Texas public high schools decreased from 76.7% to 68.5%. Both Bone and Khan were able to describe, through analyses of archival data, Texas public high school leaders failed to employ teaching staffs that were ethnically/racially representative of the state's evolving ethnic/racial student population. The efforts of supporting the shifting student demographics by hiring more teachers of color often presents as a challenge for school district leaders and campus administrators. Across the United States, school leaders are tasked with cultivating a teacher population that is representative of the student population . A consistent gap in the demography between teachers and students, and the difference between the two populations has increased focus on the issue of equity in education . In published studies, researchers have described how ethnic/racial demographic congruence between teachers and students leads to better academic outcomes and decreased discipline referrals for students of color . Congruence of demography between teachers and students also has benefits for teachers as it facilitates more comradery within the workplace . Due to the inability of Texas to provide their public high students of color with teachers of color, students of color are potentially not receiving many benefits which ethnoracial teacher-student matching provides. Plachowski explained that a diverse teacher workforce has benefits to minority students by providing harmony in teacherstudent relationships and by preparing students for successful participation as a global citizen. As noted previously, over a span of 11-school years in Texas public high schools, the average percentage of both Black and Hispanic students was over four times greater than the average percentage of employed Black and Hispanic teachers . Redding stated that the benefits of ethnoracial matching between teachers and students are potentially more discernable given that Black and Hispanic students are assigned to teachers who share their ethnicity/race at low rates. One of the advantages of student-teacher ethnic/racial matching can be to ease the transition and matriculation of students of color as that work through the final stages of their public grade school education . Transition disruptions are constant with the life course theory in which physical, emotional, social, cognitive, and physical development are interrelated . Because of the complexities of the high school process and the growing/diversifying student population in Texas, an examination of the effects of matching students of color with teachers of color is essential. Hill and Jones suggested that comfort levels with students, because of ethnic/racial similarities, may be an important factor in judging the influence of teachers on students. Gershenson, Holt, and Papageorge analyzed the effects of a student-teacher demographic mismatch on teacher expectations of educational achievement of students. In their study, teachers of color displayed a higher level of educational expectations for students of color URL: www.onlinesciencepublishing.com | April, 2023 than were displayed by White teachers . If biased teacher expectations were to be communicated with students, negative outcomes for current and future student educational decisions could be possible . In addition, biased feelings towards students of color by non-ethnic/racial minority teachers has also been expressed in low numbers of Gifted and Talented program referrals and higher numbers of discipline referrals for students of color . In a 2018 report by the Government Accountability Office, exclusionary discipline consequences were assigned at disproportionate rates to students of color . When student academic successes are not celebrated, and questionable behaviors are pointed out at unfair rates in comparison to other ethnic/racial groups, an educational environment is created in which students of color have little interest. When teachers can maintain the interest of students, academic success often follows. In schools where the number of Hispanic teachers has increased, a higher number of Hispanic students participating in Advanced Placement classes has occurred . From 1960 to 2016 the Hispanic population of Texas raised from 16% to 48%; and in that same time period, the state's Hispanic student population swelled from 20% to 48% . Through these statistics, Kauffman explained the failure of Texas public schools to retain and educate the state's Hispanic students. Being mindful that high school dropouts rarely attend college or have substantial positive financial effects on their communities , research studies need to be conducted, and necessary actions initiated that will improve the quality, the equity, and the relationships within educational settings. --- Statement of the Problem The lack of consistent academic success of students of color has relevant implications that threaten the future prosperity of the United States; and may have substantial negative influences on individuals to function as productive citizens in society . Low high school graduation rates and increasing dropout rates in Texas public high schools can be interpreted to represent an education system that is not meeting the needs of its students of color . The Texas Education Agency has self-published data which delineates Texas's struggle with students of color. Based on the available statistics in the Texas Education Agency publication, more students of color drop out each school year as opposed to other ethnic/racial demographic groups . Because over 60% of Texas's high school students are Hispanic or Black, and only 38.3% of the state's teacher are of that same ethnic/racial group, this issue warrants research . Texas's demographic changes are generating the most varied student population that Texas has seen. As schools navigate the state's new ethnical/racial public education demography, they must confront the challenges of creating and implementing teaching practices which influence demographic populations which they serve . By determining the relationship between the ethnicity and race of Texas public high school teachers and students, school campus and district administrators can begin work toward ensuring that the state's high school population has a strong academic and social foundation cultivated by teachers who have had, or are having, similar social and academic experiences due their ethnicity/race. --- Purpose of the Study The overall purpose of this study was to examine the racial/ethnic demographic characteristics of students and teachers in Texas public high schools. The first specific purpose was to describe the ethnic/racial demographic characteristics of teachers in Texas public high schools for the 2010-2011 through the 2018-2019 school years. A second purpose was to examine the ethnic/racial diversity of students in URL: www.onlinesciencepublishing.com | April, 2023 Texas public high schools for each school year from the 2010-2011 school year through the 2018-2019 school year. A third purpose was to identify any trends that were present in the ethnic/racial diversity of teachers in Texas public high schools from the 2010-2011 to the 2018-2019 school year. A fourth and final purpose was to determine the extent to which trends were present in the ethnic/racial diversity of students in Texas public high schools from the 2010-2011 school year through the 2018-2019 school year. --- Significance of the Study Though the ability of students to adjust into middle school has been more heavily investigated, the high school educational environment has the ability to affect students' well-being as they learn to navigate larger schools, interact with new social groups, and prepare for adult life following grade school . Assimilation into high school can be eased by campus goals that promote student-teacher relationships which create a level of comfort for students, account for students culturally, grow students socially, and stretch their possibilities . When students are matched with teachers who have the same ethnic/racial demography, a greater chance exist that the teachers will offer a fairer review of the student's school behavior and/or academic performance than teachers who do not share a cultural understanding with the student . Because of the importance of a high school education, Texas educational leaders and policymakers can benefit from the results identified in this empirical multiyear analysis in their efforts to aid campuses and districts across the state to review teacher-hiring practices to ensure that Texas is meeting the needs of its expanding ethnic/racial demographic high school student population. --- Research Questions The following research questions were addressed in this investigation: What is the ethnic/racial diversity --- METHOD --- Research Design To answer the research questions previously described, a longitudinal descriptive research design was present . In this study, archival data were downloaded from the Texas Academic Performance Report. Given that archival data that had already occurred were analyzed herein, the independent and dependent variables were not able to be manipulated . --- Participants and Instrumentation Two sets of participants were present in this investigation. The first set of participants were all Texas teachers who were employed in Texas public high schools for the 2010-2011 through the 2018-2019 school years. The second set of participants were students who were enrolled in a Texas public high schools for 2010-2011 through the 2018-Academic Performance Reports website. The Texas Education Agency annually collects and archives data from Texas public school districts concerning a myriad of demographic and academic accountability categories. The collected data were disaggregated for public consumption by the Texas Education Agency and divided by individual school district and made available for analysis, for state/local district comparisons, and for the general information of the public. The acquired data for this study were representative of 100% of the student and teacher population in the Texas public schools. --- RESULTS The first research question about the ethnic/racial composition of the Texas public high school teaching population was answered through descriptive statistics. Specifically calculated were the M, Mdn, and SD . For Asian, Black, Hispanic, and White Texas public high school teachers, descriptive statistics are respectively depicted in Tables 1, 2, 3, and 4. In determining the presence of a trend in the ethnic/racial composition of Texas public high school students, descriptive statistics for the 2010-2011 school year through the 2018-2019 school year were reviewed. A 6.49% decrease was observed in the average percentage of White students during this study. Hispanic students had an increase of 5.81% during the 9 years examined. Black student percentages were inconsistent in small increments during this study resulting in a near flat line. Asian students were also represented by a near flat line as their percentages remained minimal and fluctuated across the duration of the 9-school years investigated. Depicted in Figure 4 is a difference between the rising average percentages of Hispanic students and the deteriorating average percentages of White students, a slowly fading gap between the decreasing percentages of White students and the slight wavering average percentages of Black students, and a consistently small and sporadically increasing average percentage of Asian students. Descriptive statistics were examined to ascertain the extent to which trends were present between the ethnic/racial diversity of teachers and the ethnic/racial diversity of students in Texas public high schools. Both White teacher and White student percentages decreased over the 9-school years of this study. White teacher percentages were an average of 31.01% higher than that of White students. Hispanic teachers and Hispanic students each had average percentage increases. Hispanic student percentages were on average 26.94% greater than the average percentages of Hispanic teachers at the public high school level in Texas. Even with the percentages of Black students varying coupled with documented consistent growth of Black teacher percentages, Black student average percentages were an average of 3.11% greater than the average percentages of Black public high school teachers in the state for the school years examined. Asian student percentages delineated consistent, but small, growth while the minimal percentages of Asian Teachers fluctuated for the 9-school years. On average, Asian student percentages were 0.83% URL: www.onlinesciencepublishing.com | April, 2023 higher than the average percentages of their teaching counterparts. Figures 5,6, 7, and 8 contain the average percentages of the public high school students and teachers in Texas. --- DISCUSSION For this multi-year analysis, Texas public high school teacher and student ethnic/racial characteristics were examined for the 2010-2011 school year through the 2018-2019 school year The only ethnic/racial group included in this study whose average percentages were higher than their student comparison ethnic/racial group for each school year examined were White teachers. Even with persistent decreases in the percentages of White teachers and students in this analysis, White teacher percentages were an average of 31.01% higher than the percentages of White students. White teacher percentages decreased from 71.91% in 2010-2011 to 65.69% in 2018-2019, and White student percentages declined from 41.28% in 2010-2011 to 34.79% in 2018-2019. While recognizing the small growth of Black and Asian public high school teachers in the state, readers should also focus on the slowly decreasing gap between the percentages of White and Hispanic teachers. The average percentages for Hispanic, Black, and Asian students increased each school year from 2010-2011 through 2018-2019. In Texas public high schools, Hispanic students represented the largest student percentage for each of the 9 analyzed school years. --- Connections to Existing Literature The outcomes of this multi-year analysis were consistent with the results of earlier studies that were conducted by researchers who analyzed the ethnoracial diversity of public high schools in the State of Texas. Through their examinations of Texas public school diversity, Bone and Khan established the presence of consistent decreases in the percentages of White middle school teachers and students. Also documented by Bone and Khan were the consistently increasing average percentages of Hispanic teachers and Hispanic students, as well as incremental percentage increases for Black and Asian students and teachers. Egalite et al. determined that the gap between students and teachers has heightened the focus on educational equity, and the results of this study may be viewed as supportive of the need for an increased concentration on employing teachers of color who match URL: www.onlinesciencepublishing.com | April, 2023 the growing diversity of the Texas public high school student population. Detailed in this study is the increase of Texas public high school diversity; the data which were analyzed revealed swelling average percentages of ethnic/racial minority students in public high schools throughout the state. Ethnic/racial demographic similarity between teachers and students leads to improved academic outcomes and less discipline referrals for students of color . The results of this study highlight that, with Hispanic students being the largest portion of the Texas public high school student population, those students are potentially not receiving the best opportunities to improve their academic performance or receiving the benefit of fair judgment in regard to discipline referrals due to the documented inequities between the percentages of Hispanic students and Hispanic teachers. At the public high school level an emphasis should be placed on the transition of students of color as they prepare for life after grade school in which physical, emotional, social, and cognitive skills are necessary . The percentages of teachers of color delineated herein do not evoke confidence that an adequate emphasis is being placed on postsecondary life preparation for students of color. In addition to preparation as productive citizens, academic advancement has also been potentially compromised for students of color as evidenced by low numbers of Gifted and Talented program referrals and increased discipline referrals for students of color . The combination of the results of this study, and the research which has been conducted on the development and preparation of ethnic/racial minority students, reflects a need for Texas public high schools to increase the percentages of teachers of color who are employed. It is necessary to utilize the theoretical frameworks of this study to analyze the consistently high average percentages of White teachers which were identified in this examination of Texas public high school teacher and student diversity. The teacher percentages which are detailed in this study highlight the effects of Homosocial Theory as those individuals who are responsible for hiring in Texas public high schools remain more at ease with employing teachers who share ethnic/racial characteristics similar to themselves . Given that 60.12% of Texas public school principals are White, the availability of teaching opportunities for teachers of color at Texas's public high schools remains limited and also supports the presence of Social Closure theory . Derived from the statistics detailed in this study, ethnoracial characteristics seem to be less of a hindrance to employment for Texas public high school teachers of color as percentages incrementally increased over the 9-school years examined. However, gaps between the percentages of teachers of color and students of color must continue to be decreased through concentrated practices so that the needs of the increasingly diversifying Texas high school student populace can be met. --- Implications for Policy and Practice In consideration of the results of this multi-year study, several recommendations can be made to be made to improve policy and practice within public education in Texas. The hope is that, through reviewing this descriptive analysis, productive conversations will be had the educational leadership level which target the improvement of hiring practices of Texas public education teachers. The data revealed in this study should encourage school districts to begin internal reviews of their individual teacher and student ethnic/racial diversity and hiring practices. In addition, individual campus should see the see the academic and lifelong benefits of public high school students of color having the opportunity to be educated by classroom instructors of color. Plachowski emphasized that a teacher population, which is diverse in ethnicity/race, provides benefits for students of color by establishing continuity in teacher-student relationships and preparing ethnic/racial minority students for success as global citizens. URL: www.onlinesciencepublishing.com | April, 2023 Following a review of the statistical information detailed herein, accountability standards should be created with established goals of holding those in charge of campus and district hiring accountable. Prioritizing ethnic/racial diversity in hiring practices conveys a message to district/campus stakeholders that cultural diversity is of significant importance to their local education agency. The importance placed on ethnic racial/racial hiring practices, in turn, may assist in the cultivation of a quality relationship between school districts/campuses and their community members of color. Also, by taking advantage of the information provided in this analysis, state and local education agencies can develop a rigorous system of checks and balances that can assist in developing procedures for the recruitment and hiring of teachers of color by ensuring that protocols are reliably monitored and implemented. Coffey and Farinde-Wu ; Benner et al. ; Yarnell and Bohrnstedt and Plachowski each explained the how the presence of teachers of color in public high schools can affect the academic and postsecondary success of students of color. By becoming more aware of individual campus demographics Texas public high school administrators can create the most productive learning environment for the majority of their students, who in this study were ethnic/racial minorities. Of substantial importance is the necessity of educational policy developers to focus on students who are not provided the opportunity to be educated by teachers of similar ethnic/racial characteristics thus affecting their ability to reach their full academic potential, and not on other issues which may accompany their political acquired positions. --- Recommendations for Future Research Several recommendations for future research can be made based on the results identified in this study. Considering the potential academic and postsecondary benefits for Texas public high school students, teacher and student ethnic/racial matching should be analyzed further. Researchers are encouraged to extend this study to other grade levels across Texas. An identical study, but with a focus on public elementary and middle schools, would greatly benefit state, district, and individual campus leaders. Further studies would make available critical information which would assist educational leaders in making the most informed decisions possible for the benefit of Texas' diversifying and growing student population and the employment of teachers of color to support them. The degree to which the outcome of this longitudinal analysis can be applied to private and charter high schools is not known. Accordingly, another recommendation is that this study be performed for private and charter high schools at the elementary, middle, and high school level. This study also has potential benefits for educational agencies in other states. Researchers are encouraged to extend this study to her states because it is unknown how the results of this examination of Texas public high school teacher and student diversity would compare. In efforts to comprehend fully the concerns surrounding the recruitment and hiring of teachers of color, an analysis of a potential existence of a relationship between the ethnic/racial demography of administrative leaders, teaching staffs, and students at the elementary, middle, and high school level would provide a robust insight. The degree to which the demographic composition of Texas public high school administrators effects the employment of ethnic/racial minority teachers, and in turn the academic and post-secondary success of high school students of color, is not known. To develop a more complete understanding researchers encouraged to examine the relationships of the ethnic/racial composition of administrators, teachers, and students in private and charter schools at all grade levels as well. In addition, academic and postsecondary benefits associated with public high school teacher and student gender matching is not common knowledge. Also, of value to educational leaders would be to conduct research investigation on the gender diversity between teachers and students in Texas public elementary, middle, and high schools. Future researchers are encouraged to study the effects of gender diversity matching between students and teachers on the URL: www.onlinesciencepublishing.com | April, 2023 academic success of Texas school students, as well as students in other states. Similarly, the influence of gender matching should also be reviewed between teachers and students in private and charter schools at elementary, middle, and high schools. --- CONCLUSION This longitudinal statewide descriptive analysis spanned the 2010-2011 school year through the 2018-2019 school year and was about the ethnic/racial diversity of teachers and students at the Texas public high school level. Though the average percentages of White teachers and students decreased over the 9 school years of data that were analyzed, White teachers, with percentages decreasing from 71.91% to 65.59%, remained the largest among the four ethnic/racial teacher groups which were reviewed. Consistent growth was identified in the average percentages for both Hispanic teachers and students, as each had consistent increases in their average percentages over the 9-school years. With percentages growing from 43.69% to 49.50%, Hispanic students gradually increased the gap between their average percentages and the average percentages of White, Black, and Asian students. Black student average percentages varied for the 9-school years that were reviewed. However, the average percentages of Black teachers reflected consistent small growth from 7.46% to 8.57%. Asian students represented average percentages which grew from 1.92% to 2.63%, and Asian teachers had minimal inconsistent growth of their average percentages for the 9school years. More efforts to increase the Texas teacher workforce are clearly warranted by the findings of this multiyear analysis.

The purpose of this longitudinal, statewide, descriptive analysis was to examine the high school teacher and student ethnic/racial diversity for the 2010-2011 school year through the 2018-2019 school year. Data were obtained from the Texas Education Agency on all student enrollment and on all teacher employment. As such, a causal-comparative longitudinal descriptive research design was present due to the use of archival data. Consistent decreases were observed in the average percentages of both White teachers and White students over the 9 school years of data that were examined. Consistent increases were documented in the average percentages of both Hispanic teachers and students over the 9 school years. Slight but consistent increases were established in the average percentages of Asian teachers and students and of Black teachers. The average percentage for Black students decreased in all school year examined, with the exception of two school years. Inequities existed as several school years of data analyzed revealed greater average percentages for ethnic/racial minority students when compared to ethnic/racial minority teachers. Also provided within this study are recommendations for future research along with implications for local and statewide policy and practice.

The burden of serious mental illness is enormous to both the individual and the family. In a given year, an estimated 13 million American adults have a seriously debilitating mental illness . Mental health disorders are the leading cause of disability in the United States, accounting for 25 percent of all years of life lost to disability and premature mortality . Since relatively few persons with SMI marry and public mental health services are often fragmented and underfunded , parents often provide ongoing support until they no longer have the capacity to continue in this role. These parents face many challenges in coping with the long-term care of an adult child with SMI, which takes a progressive toll on their health and psychological well-being . One such challenge is balancing work responsibilities with those of helping to support their son or daughter with SMI . Much of the research on families of persons with SMI has focused on how the burdens of caregiving affect the world of work . There is a growing literature finding that stress at work has a major impact on family life in the general population and that work-family conflicts are a significant predictor of poor health and well-being . Although the great majority of caregivers of persons with SMI work , no study has yet examined how work-related stress that spillovers into family life affects the health and well-being of parents providing support to an adult child with SMI. In this study we examine how stress related to the work-family interface affects the well-being of parents of adults with SMI as well as whether greater work schedule flexibility is a particularly important resource for these parents. There is a growing body of research investigating the adverse effects of work and family conflicts on individual well-being. Specifically, work-family conflicts have been associated with poor physical health , higher cholesterol levels, higher body mass index, lower physical stamina , and increased alcohol consumption . Studies find that work-family conflicts are frequently experienced by families of persons with SMI. In a focus group interview of 41 employed parents whose children had mental health diagnoses, Rosenzweig, Brenna, and Orilvie found that these parents experienced significant conflicts between work and family, including problems in work performance and interference of family demands during the workday. Although there are various job characteristics that may help parents cope with their adult child's mental illness, we focus in the present analysis on flexibility in work schedule, which has been associated with better physical health , fewer physical symptoms , and less stress among working adults. Having a flexible work schedule may contribute to health and well-being by enabling individuals to balance competing demands from multiple life domains and consequently reducing conflicts with other roles . Work schedule flexibility also enhances health by increasing perceived job control, which has been consistently linked to better physical health , and by promoting positive health behaviors such as physical activity and more sleep . Additionally, one of the major stressors for families of persons with SMI is the overall higher level of uncertainty regarding how their son or daughter will function from day-today. Even when the individual with SMI has been quite stable over time, parents may still experience feelings of heightened uncertainty because of past experiences seeing their child quickly decompensate after a long period of stability . Having a flexible work schedule allows parents to respond to crises in their child's life while still be able to meet their obligations to their employer. Finally, many of the ways that parents help an adult child with SMI, such as transportation to medical appointments or making calls to clarify public benefits, must be completed during work hours. Job flexibility allows the parent to make work accommodations when tasks for their child can only be completed during work hours. Our examination of the effects of work schedule flexibility and work-family conflicts on parental health is guided by ecological systems theory , which posits that the environment surrounding an individual is a multilevel system. The different levels of an individual's environment interact with one another as well as with the individuals within the systems, and these interactions influence developmental outcomes. Ecological systems theory has guided past research investigating the work-family interface and its impact on health behaviors in adulthood . Within an ecological systems theory, the microsystem refers to the immediate physical and psychological environment as well as an individual's behavioral patterns and roles ; in the current study, parenting a child with SMI is conceptualized as a factor in the family microsystem, and work schedule flexibility is conceptualized as a factor in the work microsystem. The mesosystem consists of the interface between microsystems; in this study, the negative spillover from work to family is one dimension of the mesosystem. Other subsystems in ecological systems theory could not be adequately operationalized by the available measures in the data set and thus could not be modeled in our analysis. Based on the above review of the literature and ecological systems theory, we investigated the following hypotheses by analyzing data from parents of children with SMI and a closely matched sample of parents who had children without disabilities: 1. Mothers and fathers who have an adult child with SMI will have poorer health profiles than mothers and fathers who have unaffected adult children. --- 2. Mothers and fathers who do not have flexibility in their work schedules and who experience higher levels of negative spillover from work to family will report poorer health than mothers and fathers who have work schedule flexibility and report lower levels of negative spillover from work to family, regardless of parent status. 3. The association between work flexibility, work to family spillover, and health of mothers and fathers will be more salient among the mothers and fathers of adult children with SMI than among comparison group mothers and fathers. --- Method Data and Sample The Wisconsin Longitudinal Study is a long-term study of a random sample of 10,317 women and men who graduated from Wisconsin high schools in 1957; the sample also includes 5,823 randomly selected siblings of the graduates. The original sample members were surveyed in 1957, 1975, 1992, and 2004, and their siblings were surveyed in 1977, 1994, and 2006. Most respondents were White, which reflects Wisconsin's population in the mid-twentieth century. The current survey uses the sample from the 1992/1994 survey, which encompasses both original sample members and their siblings. We focus on the 1992/1994 survey because at this time respondents averaged 53 years of age and the majority was still in the workforce, whereas by the subsequent round of data collection, approximately half of the WLS respondents were retired. In total, 10,353 respondents completed both a phone interview and a self-administered questionnaire in 1992/1994 . In the present study, the original sample members and their siblings are treated as a single group of respondents from which a group of parents of adult children with SMI and a comparison group were selected. In cases in which both the original respondent and his/her sibling met the criteria for selection into the comparison group, we randomly selected one to avoid dependency in the data. There were no such sibling pairs among the parents of children with SMI. In 1992/94, telephone interviews and mail-back questionnaires were administered to collect information about various dimensions of respondents' lives, including family, work, and health. Parents of individuals with SMI were initially identified through the review of an array of variables about caregiving and children's education and disabilities in the 1975/77 and 1992/94 WLS surveys, and the sample was confirmed and expanded through a series of direct screener questions asked of all parents in the 2004/06 survey. The screener consisted of a maximum of 31 questions that began by asking parents if any of their children had SMI, and the specific diagnosis. The parent was included in the sample if he/ she indicated that the son or daughter had schizophrenia or bipolar disorder, or a major depression that required hospitalization or interfered with his/her ability to work or live independently. We include a case in the SMI sample only if the child was the biological or adoptive child of the WLS respondent. Using these procedures, we identified 333 WLS respondents who had a child with SMI, of whom 175 parents reported that their child's symptoms began before the 1992/94 interview and the child with SMI was still alive at that point of data collection. The analytic sample for this study included two groups. The first group consisted of parents who were working full-time at the time of the 1992/1994 survey , had completed both the phone interview and the mailback questionnaire in 1992/1994, and had children with SMI whose conditions started before 1992/1994 . The second was a comparison group of WLS respondents who were working full-time at the time of the 1992/1994 survey, had completed both the phone interview and the SAQ in 1992/1994, and did not have any children with SMI, deceased children, or children with either developmental disabilities or serious health problems. Next, stratified random sampling was used to select a comparison group matched to the SMI group on age and education . To increase statistical power, rather than selecting one comparison group case for each SMI case, we selected at a ratio of 5:1, resulting in a comparison group of 500 parents . --- Measures --- Dependent variables Self-rated health: Overall self-rated health was measured by an item asking "How would you rate your health at the present time?" . Numerous studies have shown that this single item is significantly associated with mortality as well as morbidity . --- Number of diagnosed illnesses: Respondents answered questions asking whether a medical professional ever diagnosed them with one of 16 illnesses, including anemia, asthma, arthritis, bronchitis, cancer, chronic liver trouble, diabetes, serious back trouble, heart trouble, high blood pressure, circulation problems, kidney/bladder problems, ulcer, allergies, multiple sclerosis, or colitis . A count of the number of diagnosed illnesses was computed for each respondent. --- Cardiovascular problems: Respondents were asked whether they had experienced chest pain or shortness of breath in the past six months. If they answered yes, two more questions were asked regarding how often they had experienced these symptoms and how much discomfort the symptoms had caused the respondent in the past six months . Respondents also were asked whether they had been diagnosed with heart problems or high blood pressure by a medical professional and how much these conditions interfered with what they like to do . Following procedures developed by Warren et al. , the items were scaled 0 to 10 . --- Musculoskeletal health problems: Respondents were asked whether they had experienced aching muscles, stiff/swollen joints, or back pain/strain in the past six months. If they answered yes, two more questions were asked regarding how often they had experienced these symptoms and how much discomfort the symptoms had caused in the past six months . Respondents also were asked whether they had been diagnosed with serious back trouble by a medical professional and how much the back trouble interfered with what they like to do . Following procedures developed by Warren et al. , the items were scaled 0 to 10 . --- Independent variables Parenting status: Parenting status was coded as a dichotomous variable depending on whether the respondent had any child with a SMI defined as bipolar disorder, schizophrenia, or major depression if the depression was accompanied by a hospitalization or resulted in functional impairments . Work schedule flexibility: Schedule flexibility at the workplace was assessed by an item asking the respondent whether he/she could decide when to come to work and when to leave . --- Negative work-to-family spillover: Respondents were asked to rate the following three items to assess negative spillover from work to family life: "To what extent do you agree that your job reduces the amount of time you can spend with the family?", "To what extent do you agree that problems at work make you irritable at home?", and "To what extent do you agree that your job takes so much energy you don't feel up to doing things that need attention at home?" . The mean of the three items was computed, with a higher score indicating a greater level of negative spillover from work to family. --- Control variables: Background variables associated with the physical health of adults were controlled for in all analyses. These included the respondent's age and socioeconomic status , marital status , and the number of children in a family . Also, we controlled for whether the respondent lived with the "target" child. For respondents with children with SMI, the target child was the child with SMI. For the comparison group, as part of the WLS survey methodology, the researchers randomly selected one child from among the children of the WLS respondent and a series of questions were asked about this child including whether the child lived with the respondent. This child was designated the "target" child for the comparison group. --- Data Analysis Multiple regression models were estimated to examine the effects of work schedule flexibility and negative spillover from work to family life on each physical health outcome, controlling for age, SES, marital status, number of children, and co-residence status with the target child. In each model, interactions between parent status and work flexibility and work-to-family spillover were added to investigate the moderating effects of parenting status on the associations between work flexibility and work-to-family spillover and the physical health of parents. All models were estimated separately for mothers and fathers. Since the WLS represents a simple random sample of one-third of the 1957 graduating class and randomly selected siblings, no weights are applied to the data. --- Results Table 1 presents descriptive statistics for the parents of children with SMI and parents in the comparison group; results for mothers and fathers are shown separately. There were no differences in age, but there were significant differences in the number of children, marital status, and co-residence status between mothers of adult children with SMI and mothers in the comparison group. The mean age of the mothers was 54 in both groups. Mothers of children with SMI had more children , were less likely to be married , and were more likely to be living with the target child compared to the mothers of children without SMI. Occupational SES, work schedule flexibility, and the level of negative work-to-family spillover were comparable between the two groups of mothers. Regarding the health outcomes, there was a significant difference in the number of diagnosed illnesses: mothers of adult children with SMI self-reported more diagnosed illnesses than mothers with unaffected adult children. The other health outcomes were not significantly different between the two groups. With regard to the fathers of adult children with SMI and comparison fathers with nondisabled adult children, there were no differences in either demographic characteristics or work-related characteristics. Overall, fathers were about 53 years old and had 3 children, approximately 90% were married, somewhat fewer than 20% were living with the target child, and a little under 60% had flexibility in their work schedule. The physical health of the fathers did not differ between the two groups, with the exception of musculoskeletal health problems; fathers of children with SMI reported more musculoskeletal health problems than the fathers of unaffected children. --- Predicting Maternal Health Table 2 presents the results of regression analyses examining the effects of perceived work flexibility and work-family spillover on mothers' physical health, with the moderating influence of parenting status . Model 1 shows the main effects of parenting status, work flexibility, and negative work-to-family spillover. Although we hypothesized that mothers of adult children with SMI would report poorer health outcomes than the comparison group, there were no associations between parenting status and the health variables. Similarly, the main effect of work flexibility was largely unrelated to the health outcomes, with one exception; unexpectedly, mothers with higher levels of work flexibility reported more diagnosed illnesses. In contrast, negative work-to-family spillover was significantly associated with all health outcomes such that mothers with greater work-to-family spillover reported poorer self-rated health, a greater number of diagnosed illnesses, more cardiovascular problems, and more musculoskeletal problems. The results in Model 2 portray two interaction effects on mothers' musculoskeletal health problems and cardiovascular health problems. As shown in Figure 1, mothers of adult children with SMI who had schedule flexibility at work reported lower levels of musculoskeletal health problems compared to mothers who had adult children with SMI and no work schedule flexibility, but there was no association between work schedule flexibility and musculoskeletal health problems among mothers who did not have children with SMI. The results also showed a significant interaction effect of work-to-family spillover and parenting status on one aspect of mothers' health. As shown in Figure 2, for mothers of an adult child with SMI, higher levels of negative work-to-family spillover were significantly associated with a greater number of cardiovascular health problems as compared to those reporting low levels of work-to-family spillover. There were no significant group differences in the effects of work flexibility or work-family spillover on self-rated health and number of diagnosed illnesses. Regarding the control variables, in general, the results indicated that mothers who had higher occupational SES had better health profiles: these mothers had better self-rated health, fewer cardiovascular health problems, and fewer musculoskeletal health problems than mothers with lower occupational SES. Married mothers had fewer diagnosed illnesses than mothers not currently married. None of the other control variables were significantly associated with health outcomes in working mothers. --- Predicting Paternal Health Table 3 presents the results of regression analyses examining the effects of family characteristics, work flexibility, and work-family spillover on fathers' physical health, with the moderating influence of parenting status. Model 1 presents the main effects of parenting status, work flexibility, and work-family spillover. As hypothesized, fathers of adult children with SMI reported more musculoskeletal health problems than comparison fathers, but did not differ with respect to the other health outcomes. Work flexibility was not associated with any health outcomes as a main effect, but negative work-to-family spillover was significantly associated with self-rated health of working fathers. Consistent with hypothesis 2, all predicted interaction effects between work schedule flexibility and parenting status on fathers' health were significant, including working fathers' self-rated health, number of diagnosed illnesses, cardiovascular health problems, and musculoskeletal health problems. Figure 3 illustrates that fathers of adult children with SMI who had flexible work schedules reported better self-rated health than their peers who had adult children with SMI but whose work schedules were not flexible. In contrast, although comparison group fathers without work flexibility reported slightly better health than comparison group fathers with work flexibility, this difference was not significant. Figure 4 shows that fathers of adult children with SMI who reported greater work schedule flexibility indicated that they had fewer diagnosed illnesses than fathers whose adult children had SMI and whose work schedules were not flexible. Among the comparison group fathers who had nondisabled adult children, the number of diagnosed illnesses was comparable regardless of flexibility at work. Figure 5 illustrates that having schedule flexibility at work was also associated with reports of fewer cardiovascular health problems among fathers of adult children with SMI, but not among fathers with unaffected adult children. Figure 6 shows that fathers whose children had SMI and who had flexible work schedules reported lower levels of musculoskeletal health problems compared to fathers of children with SMI who did not have flexible schedules. In contrast, the musculoskeletal health of comparison fathers did not differ based on whether or not they had flexible work schedules. The results also indicate that the interaction between negative work-to-family spillover and parenting status was significantly associated with fathers' health, specifically number of diagnosed illnesses and cardiovascular health problems. Figure 7 shows that higher levels of negative work-to-family spillover was associated with a greater number of diagnosed illnesses in fathers of adult children with SMI as compared to such fathers reporting lower levels of work-to-family spillover. In contrast, there was no difference in the numbers of diagnosed illnesses among comparison fathers who experienced higher or lower levels of negative work-to-family spillover. Figure 8 illustrates that fathers of adult children with SMI who reported higher levels of negative work-to-family spillover indicated a greater number of cardiovascular problems compared to their counterparts with lower levels of negative work-to-family spillover. There was no association between the levels of work-to-family spillover and cardiovascular health problems among comparison fathers. Regarding the control variables, the results showed that older fathers were less healthy than their younger counterparts: older fathers had poorer self-rated health and more diagnosed illness than younger fathers. Fathers with higher occupational SES reported better self-rated health than fathers with lower occupational SES. --- Discussion Guided by ecological systems theory, the present study examined the effects of work flexibility and work-to-family spillover on the health of parents of adult children with SMI. In general, work-to-family spillover and working in a position without flexibility in one's schedule had a more detrimental impact on health among parents of adult children with SMI than among parents in the comparison group, which supports our hypothesis. Work schedule flexibility was associated with better health outcomes for both fathers and mothers of adult children with SMI. Among fathers of adult children with SMI, those who reported flexible work schedules had better health profiles than those who did not perceive their work schedules as flexible . Among mothers of adult children with SMI, those with flexible work schedules reported fewer musculoskeletal problems than those without flexible schedules. At higher levels of negative work-to-family spillover, mothers of adult children with SMI reported significantly more cardiovascular health problems than mothers in the comparison group. Fathers of adult children with SMI who experienced more negative work-to-family spillover indicated significantly more diagnosed illnesses and more cardiovascular problems than fathers in the comparison group. These results provide evidence that the physical health of parents of adult children with SMI is more vulnerable to the adverse influence of the lack of work schedule flexibility and to the spillover of stress from work to family life than parents of adult children without disabilities. The results revealing the health effects of both microsystem and mesosystem factors also support the assertion of ecological systems theory that various subsystems in which individuals are embedded influence individual developmental outcomes not only via direct impacts but through their interactions with one another . Unexpectedly, work schedule flexibility was significantly associated with more diagnosed illnesses among mothers regardless of parenting status. Supplemental analysis showed that mothers who had work schedule flexibility also had higher occupational SES compared to the mothers whose work did not allow a flexible work schedule. It is possible that working women in higher status positions are exposed to greater job stress due to increased demands and responsibilities at work. Given empirical evidence linking stress and health problems and findings that working women are more vulnerable to the adverse effects of job strain than working men , working mothers in higher SES occupations, who are more likely to have flexible work schedules, might be at an increased risk of health problems. Alternatively, given the cross-sectional nature of the study, it is quite plausible that working mothers with higher SES are able to negotiate more flexible work schedules when they experience health problems, because employers might be willing to accommodate these mothers' needs in order to retain qualified employees. Future studies using longitudinal data would improve the understanding of the causal nature of these associations. In addition, although we predicted poorer health profiles of parents of adult children with SMI than comparison parents, the results showed no group differences in health outcomes. Considering that previous empirical evidence has shown poorer health and well-being among parents of adult children with SMI, these results were unexpected. Mothers who had an adult child with SMI and remained in the work force into their mid-50s may be a unique subgroup; in fact earlier studies have shown that parents of a child with SMI frequently change their work or career patterns or quit their jobs and stay home to accommodate the needs of their child with SMI . Furthermore, whereas approximately 91% of fathers of adults with SMI were married, only 62.7% of the mothers of adults with SMI were married. One criterion for this study was that the respondent had to be working full time. Whereas in the WLS marital status has little affect on whether men are working full time or not, this is not the case for women. Whereas only 44% of married women were working full-time, 70% of divorced women were working full-time in 1992/94. The pressures to work full-time may be even great for divorced women who have a child with SMI because of the additional costs of caring for a child with a condition. Overall, the findings show that parents of adult children with SMI are more susceptible to the impacts of work-related stress spilling over into family life than the norm. Family clinicians are increasingly realizing the benefits of offering families psychoeducation to help them learn about serious mental illness and strategies to cope with their child's behavior problems and needs for assistance. Though these programs are successful in teaching families successful strategies to better cope with their child's problems, there is very little emphasis on how the world of work may be contributing to the stress of caregiving. Our findings suggest that work related stress may increase stress at home, which takes a toll not only on parental caregivers but it may also affect the quality of life of the individual with SMI. It would be desirable for psychoeducation programs for families of persons with SMI to be strengthened to include a focus on work place stress management. With regard to potential policy changes, our findings suggest that increasing work schedule flexibility would be especially beneficial to working parents who have an adult child with SMI. In previous studies, parents of children with SMI have noted the importance of work flexibility for balancing work and caregiving demands. These parents reported choosing jobs with fewer benefits in order to secure the time and flexibility necessary to manage care duties . Given the previous empirical evidence that increased work schedule flexibility is associated with less work-family conflict , the findings from the current study suggest that providing work schedule flexibility would be especially beneficial for working parents who have a child with SMI. Some limitations of this study must be noted. The WLS sample consists of mostly non-Hispanic Whites who have obtained at least a high school education. Thus, although the sample was representative of the Wisconsin population in the mid-twentieth century, it does not represent the current population. In addition, although the WLS is a longitudinal study that has been conducted over 40 years, the present analysis used a cross-sectional approach because complete information about children's SMI status was not available in the previous wave , and because the majority of parents had retired by the following wave . The use of cross-sectional data, however, entails inevitable problems for interpreting causality. Further, the analysis of data collected in the 1990s might limit the generalizability and applicability of the results to current work and family situations of parents of children with SMI. The relatively small sample of parents of children with SMI also warrants caution with regard to the generalization of the results. Finally, the data were collected as part of large scale longitudinal survey. There are inherent limitations to surveys which are based on respondent self-reports. Future studies that might address some of these limitations by incorporating biomarkers into the measurement of health. In addition, an examination of the effects of other subsystems in ecological systems theory would help increase the theoretical interpretation of the health predictors of working parents of adult children with SMI. Despite these limitations, the present study contributes to the literature on the health of parents of children with SMI by using a systems approach to examine the influence of work flexibility and work-family conflicts on the physical health of these parents, and by using a probability sample to eliminate the potential limitations of selection bias. The results have significant implications for family policy and family life practitioners. Social policies that encourage work schedule flexibility for families caring for children with disabilities, and interventions that reduce the spillover of stress from the work place to the home could help mitigate the burdens experienced by parents who have children with SMI and would consequently improve the well-being of these parents.

This study examined the effects of work schedule flexibility and the spillover of work stress to family life on the health of parents of adult children with serious mental illness (SMI). We compared 100 parents of adult children with SMI to 500 parents with nondisabled adult children using data from the Wisconsin Longitudinal Study. The detrimental impact on health of a lack of work flexibility and of higher levels of negative work-to-family spillover were more pronounced among parents of adult children with SMI than parents with non-disabled adult children. The results have significant implications for developing interventions to help midlife families of persons with SMI cope with work-related stress and for policies that provide for greater work schedule flexibility.

Background: Receiving a breast cancer diagnosis and treatment is both a physical and emotional journey. Previous studies using single-source data have revealed common and culture-specific emotional experiences of patients living with breast cancer. However, few studies have combined such data from multiple sources. Thus, using a variety of data sources, the current study sought to explore the emotional experiences of women in China newly diagnosed, post-operative, or undergoing chemotherapy. We posited that even though women living with breast cancer in China have multiple channels through which they can express these emotional experiences, little variance would be found in their emotional expressivity and the themes they want to express due to cultural inhibitions. Methods: Text data from female patients newly diagnosed, post-operative, or undergoing chemotherapy were collected between June and January via a Python web crawler, semi-structured interviews, and an expressive writing intervention. Data were transcribed and subjected to thematic analysis. Reporting followed the consolidated criteria for reporting qualitative studies guidelines. --- . Introduction Breast cancer is the most common cancer in women worldwide with an estimated 2.26 million new cancer cases diagnosed in 2020 . Accounting for 15.5% of female cancer deaths worldwide, it has also become a heavy global public health burden . BC diagnosis and treatment are stressful experiences that can evoke a variety of negative emotions and lead to broader emotional experiences . "Emotional experience" encompasses an individual's emotional reactions and associated cognitions regarding relevant events, including thoughts, feelings, worries, and concerns . A BC diagnosis usually occurs when women are at a life stage often characterized by childrearing and being mid-career, and when they may be less able to cope with the risk of lifelong cancer recurrence or potential death . Treatment side effects and prognostic uncertainties cause women to suffer negative emotional experiences, including stress to their significant relationships , body-image disturbance , sexuality and fertility anxieties , social stigma , confusion about recurrence and hereditability , and negative quality-of-life . Despite this, previous studies have also reported some positive emotional experiences, including post-traumatic growth and self-compassion . BC diagnosis and treatment is both a physical and emotional journey. One study exploring individuals' psychosocial wellbeing across disease phases indicated that distress peaks occur after diagnosis, during chemotherapy, after mastectomy, at the end of adjuvant therapy, at diagnosis of recurrence, and when the disease is declared to be terminal, highlighting that patients across different BC diagnostic and treatment phases can experience a marked range of emotions and cognitions . However, most research has focused exclusively on patients who are newly diagnosed, those receiving chemotherapy, or across a variety of durations postdiagnosis , which may neglect treatment continuity and integrity. Although previous studies provide valuable insights into Chinese women's culture-related views on BC, most have focused on Chinese immigrants in other countries . Those have found that, as Chinese immigrants acculturate to the mainstream culture of other countries, they may incorporate new post-immigration cognitions . Specifically, those in Western cultures tend to value free open emotional expression; in contrast, those in Eastern cultures tend to value emotional suppression, or an individual's ability to consciously control their emotional expressions . From a Chinese cultural perspective, free expression of emotions, especially negative emotions, may temporarily disrupt group harmony . Chinese women especially are conflicted about disclosing emotional distress and experience high levels of ambivalence about doing so . In Chinese culture, cancer is often perceived as misfortune and contagious, which has been associated with self-stigmatization . Fear of cancer being discovered and bringing shame or burden to the family leads patients to prioritize emotional suppression to preserve social harmony . Meanwhile, socially constrained responses may be negatively associated with relationship satisfaction, aggravate self-stigmatization, and result in persistent emotional distress and reduced self-efficacy in coping with stress . Consequently, this context is not conducive to healthcare professionals and caregivers' timely identification of patients' emotions or their ability to provide corresponding emotional support. Numerous published studies have shown that expressive writing and interviewing are effective ways to record patients' experiences through hearing their voices . Furthermore, social media has been recognized as another potential source of important patient data. Platforms such as Facebook, Twitter, and Weibo, are all based on the sharing, dissemination, and acquisition of user relationship information, where short, real-time information can be obtained through attention mechanisms . In the Internet age, social media has emerged as a rich and largely untapped resource for understanding patients' frank communications about their experiences and treatments . Despite the advantages of these data, their limitations also warrant mention as they may introduce bias in identified experiences. First, face-to-face interviews may skew the sample toward participants who find it easy to talk about their experiences with strangers, while those in serious emotional distress may refuse to participate . Second, although using EW as a method to collect the experiences of patients with BC is effective, such patients must be physically able to write by hand for more than 20 min; those with physical disorders like lymphedema or those whose surgery occurred on the same side as the hand used to write are thus ineligible . Our previous research has also identified that not all patients with BC like expressing their emotions in writing; there are individual differences in how feelings are expressed in different ways . While emotional expressions are made on social media in the real world and in real-time, the patients' demographic and disease characteristics are incomplete or unavailable on these platforms . Arguably, therefore, it is difficult to capture the full range of nuances in patients' disclosures of their emotional experiences from a single data source, especially concerning women living with BC in China, where cultural norms strongly discourage emotional disclosure . Use of a single-source method would likely skew results and fail to reflect the target population's comprehensive range of experiences. To date, few studies have explored patients' experiences using multiple sources. Supporting and accompanying female patients with BC throughout their cancer journey requires a deep understanding of their emotional experiences. Therefore, we conducted this qualitative text analysis based on data collected from patients during three BC phases: when they were newly diagnosed, postoperatively, and while undergoing chemotherapy. The study goal was to help healthcare professionals provide individualized culturally appropriate emotional support and management for female Chinese patients living with BC. --- . Methods . . Study design The aim of this study was to explore the emotional experiences of women newly diagnosed, post-operative, or while undergoing chemotherapy for BC in China. Qualitative data was collected from Weibo, semi-structured interviews, and an expressive writing intervention. This qualitative study uses the consolidated criteria for reporting qualitative studies , and was analyzed based on the data collected from Weibo, EW, and semi-structured interviews . --- . . Participants and settings EW participants were recruited from the breast surgery departments of six tertiary hospitals in four cities in China. Semistructured interview participants were selected from one of the above hospitals and they did not participate in EW. Weibo participants were selected from among network users within the column of the super topic #Breast cancer#. The inclusion criteria for patients with BC included in the EW and semi-structured interviews are presented in Table 1. Consistent with the requirement to include maximal variation during purposive sampling for qualitative text analysis, we recruited 50 EW participants undergoing each of three BC phases: newly diagnosed, postoperative, or undergoing chemotherapy . Furthermore, with respect to the potential differences in phase specificity and acuteness of patient emotions, newly diagnosed and postoperative were defined as within 1 month of a new diagnosis and 1 month post-surgery, respectively, consistent with previous studies . Since participants from Weibo are anonymous and we could thus only analyze the texts they disclosed on the Internet, we were unable to apply inclusion criteria to this sample. The EW and semi-structured interview participants' demographic characteristics are shown in Table 2. --- . . Data collection EW and semi-structured interview participants were informed about the standard EW instructions and interview guide, respectively . Patients were approached in their hospital rooms by four trained female nurse researchers who are experienced in qualitative methods and were part of the BC department staff of their respective hospitals. EW participants were asked to write about their most prominent stressrelated, upsetting, or traumatic feelings caused by BC for at least 20 min over four consecutive days at their convenience in accordance with the standard EW by Pennebaker and Beall . If their EW materials were not received within 1 week, the research assistant at each hospital asked the participants whether they had encountered difficulties in the writing process and whether they were able to complete the writing task. The 50 EW texts from participants in the chemotherapy phase were randomly selected from our previous study, a multicenter randomized controlled trial on the effect of prolonged EW on patients receiving chemotherapy . The semistructured, face-to-face interviews were conducted by an experienced female researcher at the patients' convenience in a quiet conference room. Each interview lasted 30-40 min and the audio was digitally recorded. Semi-structured interview participant recruitment ended when data saturation was achieved, which was when no new information emerged . The data generation and collection were synchronized. With the exception of the EW texts by patients in the chemotherapy phase being obtained from our previous research data, the rest of the EW, interviews, and Weibo texts data were generated and collected from June 2021 to January 2022. We conducted 17 interviews and transcribed the recordings immediately following each interview. In the process of EW data collection, we identified some common invalid contents. For example, some participants did not finish the four-day continuous writing, or their written content was almost an expression of gratitude to and praise of the medical staff. Therefore, every time a document was collected, we would check its content validity so that we could recruit new participants in time if we found that the content was unqualified within the study period. Thus, EW data collection was synchronized for completion from June 2021 to January 2022. Posts and comments published by Weibo participants from June 2021 to January 2022 were obtained using a Python web crawler. --- . . Data analysis Thematic analysis, as outlined by Braun and Clarke, was used to identify themes, similarities, and differences between participant experiences . Analyses of these large qualitative datasets were carried out using a structured approach . Electronic texts from EW, interviews, and Weibo were regarded as three separate datasets. We performed double screening to manually exclude --- Pennebaker's expressive writing instructions --- Interview guide Advise your patients to: • Find a place and time where and when you will not be disturbed. • Write about what you are worrying about or what you have been avoiding thinking about that is affecting your life in an unhealthy way. • Write for a minimum of 20 min a day for at least 4 consecutive days. • Write continuously without worrying about proper Chinese or censoring content. • Decide whether to write about the same issue each time or different issues. • Consider reviewing what you are written over time to see how your thinking or emotions have changed. • The information obtained will only be used for this research and will not have any influence on you. • "Please tell me about your feelings during this stage of your diagnosis. Did you have any thoughts or confusions?" • "Please tell me about your feelings before and after the operation. What adverse symptoms or experiences did the operation result in for you?" • "Please tell me about your feelings when you began to receive or during the chemotherapy. What adverse symptoms or experiences did chemotherapy result in for you?" • "What was your feeling during the whole period of illness or treatment?" • "Do you have any new experiences, feelings, or views recently? What things or symptoms have changed your mood?" contents that were obviously not self-reported by patients with BC . Each dataset was imported into NVivo to allow for electronic coding. First, one researcher read and coded each of the three datasets. Then, a second researcher familiarized herself with a random sample from each dataset, after which the first and second researchers discussed and agreed on coding decisions. Next, three authors compared, contrasted, consolidated, and grouped these codes into categories according to their similarities, organized these categories into themes, and identified quotations that represented each theme. Discrepancies and disagreements were discussed by the team and adjusted as needed. In the final step, all authors evaluated the findings and agreed upon the most salient themes to be highlighted herein. --- . . Ethical considerations Ethical approval for this study was obtained from Nanfang Hospital of Southern Medical University in China , and a standardized informed-consent form was administered. Eligible EW and semi-structured interview participants were informed about the study and provided written informed consent prior to participation. During the writing and interview processes, an experienced psychologist was available for patient consultation upon request. Codes were assigned to each participant instead of their real names . --- . . Rigor To establish study credibility, all researchers engaged long-term with and established appropriate relationships with the participants. To enhance study dependability, any disagreements on the design, methods, data analysis, or results were discussed by the research team until consensus was reached. To ensure confirmability, the researchers used field notes and memos to support the connections between the data and findings. To guarantee authenticity, all major identified themes and sub-themes were supported by participant quotes. Transferability was accomplished by including the widest participant diversity possible, and the authors sought to describe all parts of the study methods in detail to guide future studies. --- . Results Materials from 17 semi-structured interviews and 150 EW texts from female patients with BC were collected from across the three BC phases. After removing the contents unrelated to the self-report of BC patients, a total of 5,675 Weibo posts and comments published by 448 posters and 1,842 commenters were obtained. The texts from these three sources represented a total of 461,348 Chinese characters. During the EW and interview processes, no participants sought counseling or psychotherapy. Three major themes comprising eight sub-themes were identified from the texts: conflicting emotions after diagnosis; long-term suffering and treatment concerns; and benefit finding and cognitive reappraisal. Themes, sub-themes, and representative participant quotes are shown in Supplementary material 2, along with the number and proportion of participants who mentioned each theme and sub-theme. The proportion of themes and sub-themes mentioned in each source is shown in Supplementary material 3. --- . . Conflicting emotions after diagnosis Receiving a BC diagnosis changed patients' views of the future, as it required that they consider the potential for losses, unrealized goals, and death, while they simultaneously considered the feelings of their loved ones; this immediately increased the risk of stress-related disorders for patients. The participants also shared experiences of diagnosis-specific distresses, such as concerns about diagnosis disclosure , disbelief and escapism , and treatment decision-making distress ; they were also concerned that these experiences would persist and have implications for their mental health and adherence to followup treatment. --- . . . Concerns about disclosing cancer diagnosis Disclosing one's inner pain to a trusted other may help relieve stress and ease anxiety. While bewildered by their inner conflicts over disclosure, several newly diagnosed participants tended to withhold their cancer-related distress for fear of burdening others, especially their older parents. One participant said that: I dare not disclose my cancer diagnosis to my parents. I'm worried that they can't accept this shocking news, and that their bodies can't bear this pain. [PD18] In addition, some participants reported an unwillingness to disclose because they did not want to receive excessive attention or suffer being stared at or gossiped about by others. As one stated: If everyone knows the news, their various concerns and sympathies will be overwhelming to me, and I am afraid that I will become more vulnerable instead. [PW31] . . . Disbelief and an escape from reality Disbelief of the diagnosis was particularly common among participants who had annual physical examinations and who felt well prior to their diagnosis, as they could not reconcile a cancer diagnosis with their healthy identity. Strong disbeliefs even led some participants to attribute their cancer to having done something bad: What? Cancer? I was scared because I never thought cancer would have anything to do with me. I was a timid and kind-hearted person and I had never killed a chicken or a cockroach. I couldn't believe it! [PD2] Moreover, some participants deliberately turned their attention to things other than the disease, such as traveling, shopping, and other interests to escape the reality of being a cancer patient who may die. One participant said: I have been pretending to be strong since diagnosis. I want to paralyze myself [sic] by going shopping, going to the movies . . . but in vain, my heart is always full of fear, anxiety, sadness, and hopelessness... [PC4] Some participants mentioned that when some warning signs of BC appeared in their bodies, they did not pay attention to them, and because of their blind belief that they did not have cancer, they delayed medical treatment for careful examination and regular checkup. There was a lump in my left breast, and I did not feel anything abnormal, so I did not believe it would turn into BC. However, when I went to the hospital for reexamination after a long time, the results made me feel hopeless. [PI2] --- . . . Distress over rapid treatment decision-making Surgery type, especially mastectomy, is considered a "preference decision" for newly diagnosed patients who are clinically eligible for either option . These patients must weigh risks and benefits of treatment options on multiple grounds. Concerns regarding cancer recurrence, prognosis, personal costs, benefits, and other salient aspects of surgery play crucial roles in the decision-making process, which can cause intense distress. Mastectomy posed a greater threat to participants who were young and unmarried or married but childless and did not want to be deprived of their rights as a partner, wife, or even mother. They were afraid of being a "monster" in others' eyes. One single participant complained that: My sister has been persuading me to receive mastectomy, but I have no boyfriend, no husband, and no children. What will happen to me in the future? Certainly, no man wants to spend his whole life with an incomplete woman. [PO45] Most participants who had originally preferred breast-conserving surgery eventually chose the mastectomy based on fear of prognosis, recurrence, financial burden, long follow-up treatment, and a strong will to live, with the accompanying belief that "nothing matters except my life". One participant said: I'm afraid of recurrence after breast-conserving surgery. Life is more important for me, so I will directly choose [sic] the mastectomy. If my husband minds my incomplete breast and abandons me, I will also not care. I just want to live for myself. [PI7] Interestingly, we found that some participants felt confused about where to receive treatment and from whom to receive it. Some tended to look for the hospital with the greatest expertise in more developed cities to ensure a cure. One participant who traveled from afar stated: . . Long-term su ering and treatment concerns I Diagnosis is often followed by surgery or chemotherapy. The realization that they must begin treatment and/or the impacts of treatment was extraordinarily stressful for most participants. The special treatment-related concerns that caused suffering for female patients living with BC throughout treatment included body-image disturbances and sense of stigma , anxiety over sexuality and fertility changes , and guilt and powerlessness over loss and conflict in their gender roles . --- . . . Body image disturbance and sense of stigma Body-image changes like breast loss or disfigurement, hair loss, scars, and increased weight undoubtedly resulted in a sense of bodyimage disturbances, which can have devastating effects on women's identity, attractiveness, self-esteem, sexuality, and social adjustment. Hair loss can be a traumatic experience for some women as it is suggestive of cancer, which can be stigmatizing, and may result in feelings like embarrassment, sadness, shame, helplessness, and a sense of impending death. Several participants declared that they preferred to cut off all their hair to avoid the dreadful process of hair loss: I keep losing hair! Is my physical condition getting worse and worse, and am I going to die? It's horrible! So it's better to shave it off to avoid it falling all over the floor. [PC17] Breast loss, scarring, and deformity associated with surgery may lead participants to feel that they are no longer normal women, but instead embarrassed, ugly, and self-conscious. They were afraid of looking at themselves in mirrors and of being looked at as peculiar and being discarded by their intimate partner. One participant who had a mastectomy said: I was desperate when I saw my surgical scar in the shower. I felt I was no longer a perfect woman, but a defective product. I didn't want my husband to see my ugly appearance, I hate myself. [PO42] . . . Guilt and powerlessness over gender role loss and conflict Gender roles changing from participants' original identities to patient was a great pressure. Participants mentioned a sense of guilt over losing their original role, the crisis of self-identity, and low self-efficacy over the conflict between their original role and new patient role. Participants with older parents and younger children experienced more obvious distress, because they had changed from the main family caregiver to the cared for, a change that made them feel deeply guilty for making their families suffer and led them to label themselves as "unfilial", "useless", and "a burden". One participant said: My children are still young, and my parents are old. I'm so unfilial, I failed to give them a good life, and even have this disease, which will only make them worry. [PO23] Breastfeeding and caring for a newborn are conducive to establishing a close mother-child relationship. Yet accomplishing this can be impossible for these women, some of whom were in the lactation period. A participant who had just given birth said: My child is only 3 months old, but I can't breastfeed him, accompany him out, and take care of him like other children's mothers. I can't help crying when I miss my child. [PD45] Career impacts caused by BC, such as delayed work schedules, job changes, and more seriously, unemployment, also bothered participants. Being away from work for long periods made them feel that life was aimless and hopeless. One participant expressed: Now I don't have to go to work because of illness, I'm just like a gyroscope losing its inertia or a mobile phone without electricity. --- It is extremely miserable for me, who relies on work to gain a great existence [sic]! [PC17] Notably, regardless of whether the real cause was discrimination, dismissal of an employee suffering from BC undoubtedly caused selfidentity crisis. A participant who was dismissed from their company for no apparent reason recalled that: I was dismissed, which was too hurtful. I asked my boss the reason, they bluntly said it was just because the company laid off employees, and I was ill, so they fired me. . --- . [PW1170] That patients are physically and psychologically vulnerable during treatment, and that they are unable to take on the same responsibilities as mothers, wives, and daughters as before deserves special attention. If they perceive inadequate family support or cannot be understood patiently, they form a sense of self-perceived burden and even consider suicide. One participant said: When I quarreled with my husband, he scolded me, "Why did you get this disease? Didn't you know that you couldn't get sick?" He is right! I have no right to get sick! I should die! [PC19] --- . . . Anxiety about sexuality and fertility changes Women with BC experience a range of negative emotional changes because of disturbances to their sexuality, including fear of relationship deterioration, loss of fertility, sexual unattractiveness, loss of femininity, depression, and anxiety, as well as alterations to their sexual self. A few participants mentioned that, since beginning treatment, both their sexual desire and frequency of sexual intimacy had obviously decreased; some even felt indifferent to their sexual life: Since the operation and chemotherapy, I have been indifferent to sex. I usually have sex two or three times a month, but now once is more than enough. [PC20] Despite a low sexual desire, some participants still wanted information related to having sex. Yet there were also indications that they forewent counseling for fear of shame. One participant stated: I don't know if I can have sex after surgery. I feel that my sexual desire is gone. Is it really that I can't have sex for several years? I really want to ask the doctor, but it's too embarrassing to mention it. [PO36] Sexual dysfunction from treatment is a difficult hurdle for participants who want to get married or have children. They are also confused about the answers to questions like "Can I get pregnant after treatment?" and "When can I get pregnant?". Yet, interestingly, the detailed aspects of fertility anxiety varied among participants with different marital statuses. Those who were unmarried and childless became utterly dispirited and depressed over future relationships and marriage. A very young participant complained that: I don't have a boyfriend and I'm not married. How can I accept the fact that I am a young girl who has no breasts? What should others think of me? Who would choose a crippled woman? I'm a waste.. --- . I'm going crazy because of my body disfigurement. [PO25] Those who were married but childless had worries concerning fertility and guilt about their husbands, as they thought they had not fulfilled their responsibilities as a wife and feared being abandoned. Similarly, influenced by the traditional Chinese construct of "son preference", these negative emotions were more serious among participants who were married with only a daughter. Despite going to painstaking efforts to conceive, BC diagnosis during pregnancy means that patients may also face a serious choice between the fetus and themself. A participant reported that: The only dream that will come true soon is giving birth to a boy. But now I have to terminate my three-month pregnancy. It is simply unacceptable! I think I may not be able to give birth to him and I may die soon. So suffocated [sic]! [PC29] --- . . Benefit finding and cognitive reappraisal In addition to painful memories during diagnosis and treatment, participants also recorded expressions of benefit finding and cognitive reappraisal regarding their BC. They realized the importance of their health and planned to reestablish their life goals to really live for themselves. They also noticed that there were many around them who wholeheartedly accompanied them, which made them feel grateful and inspired to put effort into surviving their painful journey. --- . . . Post-traumatic growth Once they had overcome the initial shock of diagnosis and painful treatment-related side effects, participants realized the importance of treatment to survival and the process of healing was initiated. Many expressed that they had also experienced positive psychological growth following this traumatic event, and tended to change their negative mindset, cooperate with treatment, and strive for their "new birth", instead of being afraid and wanting to escape. A few participants regarded the cancer as a rest and welcomed being given a chance to reflect on their priorities and focus on themselves. Some also stated the importance of a correct understanding of their new roles and offered suggestions that reflected both their growth through the illness process and a newly significant way of conveying positive emotions. One participant suggested: Don't think of yourself as a patient. You can cook some delicious food for your family, clean up your home, grow some flowers . . . treat yourself as being on a holiday and spend more time with your family. [PW35] Some also began to establish more positive cognitions about diseases and attach greater importance to the subtle changes in their bodies. One participant said: Regret is useless. I would like to appeal to those who seem to be very healthy. Once you find anything wrong with your body, you should go to the hospital for examination immediately. Do not hesitate. [PC19] --- . . . Perception of social support Many participants acknowledged that suffering can be reduced with sufficiently supportive national policies, family, friends, healthcare professionals, and peer patients. Participants were extremely grateful for their company, encouragement, and sound guidance. Such support was often viewed as a behavioral manifestation of emotional support, which largely strengthened their confidence in defeating cancer. Some participants, particularly those with heavy family burdens, indicated that they were grateful for the national health insurance reimbursement policy. One said: It's not affordable for an ordinary worker to have the extra treatments. Fortunately, under the guarantee of national medical policy [sic], my financial burden seems to have been reduced a lot. [PW226] Family support was deemed as the most important source of social support. Participants' main caregivers were almost always their husbands, mothers, or children. Many realized that the tireless care and intimate company of family gave them strong confidence and motivation to persist. In their view, only by adhering to treatment and recovering early could they live up to their family's efforts. One participant said: I have become a fragile baby in my family since I was ill. My son will not make me angry. Besides, my husband tries every means to make me happy! I think, no matter how hard the process is, I must persist! [PC35] The effectiveness of peer patient support was also mentioned frequently. Participants admitted that they could better understand each other's feelings and share experiences about fighting side effects to help reduce the misery of those who were in treatment. One participant stated that: All the patients were full of warmth. It is easier for us to understand and sympathize with each other. We often chat together to encourage and take care of each other, as if we have become a family. [PO15] . --- Discussion Prior studies addressing the emotional experiences of patients living with BC used single-source data . This study adds to the literature by providing insights from textual data collected from multiple sources and across different treatment phases. Despite gathering information from multiple sources, there is little variance within the theme of self-reported emotional experience presented by these women with BC in China. We contend that cultural inhibitions play a significant role here given women in China are encouraged to suppress their emotions . The multifaceted experiences were identified as representing three consistent themes that comprise eight sub-themes: conflicting emotions after diagnosis; longterm suffering and treatment concerns; and benefit finding and cognitive reappraisal. Our findings indicate that patients suffer painful emotional conflicts regarding both disclosing their cancer diagnosis and rapid treatment decision-making. In China, cancer is predominantly considered a result of karma or bad luck caused by the immoral behavior of individuals or their ancestors . The value orientation of Chinese traditional culture focuses on neither spiritual sustenance nor material wealth, but on coordinating interpersonal relationships and creating a harmonious society . Within a cultural atmosphere of "faithfulness and honesty" that advocates ethics, feminism-which emphasizes individuality-is doomed; this cultural system that stresses integrity and emphasizes balance further ignores women's personal needs . Interestingly, and consistent with previous studies , many participants herein shared that they prioritized emotion suppression to preserve social harmony and were tempted to keep their disease secret as they were conscious of the burden and shame that cancer would impose on their families, and were afraid of the staring, gossip, and excessive sympathy they would receive . Within such a cultural milieu, patients feel that having BC is unimportant to others and unnecessary to share. Thus, some participants stated their unwillingness to disclose their cancer-related issues and true thoughts in public. Nevertheless, more frequent suppression of emotions and emotional expression tends to be associated with greater psychological distress and poorer interpersonal functioning . Consistent with other studies , we found that most participants ultimately chose the relatively simple surgical option based on greater perceived survival expectations and faster hospital discharge, rather than physical appearance. Similar to Mazaheri's study , in rapid treatment decision-making, the participants' primary worries were not about themselves, but about who would take care of their families and how soon they could be cured and discharged . Thus, healthcare professionals who spend time with patients and provide comprehensive care during hospitalization, especially nurses, must be properly trained to provide consultation and empathetic interactions that will encourage participants to bravely disclose their inner emotions. These participants were also discovered to grapple with a great deal of confusion in their decision-making between mastectomy and breast-conserving surgery. Clearly, more accurate and clearer education and BC-related treatment information from healthcare professionals is greatly needed to guide appropriate decision-making. Herein, participants frequently mentioned body-image disturbances. Consistent with previous studies , we found that among these participants, losing a breast was a terrible experience that made them think of themselves as fragmented. Likewise, among chemotherapy-related side effects, hair loss was considered among the most traumatizing and distressing experiences . Many participants herein also felt uncomfortable in public because they were worried that others might recognize them as cancer patients from their asymmetrical breasts or hair loss . Consequently, several participants were even distressed by dressing to go out and thus limited daily activities like exercise and shopping. It is unsurprising that body changes significantly impair body-image satisfaction, though this also appears to be a potential lifelong source of mental distress, including feeling less attractive and reduced self-esteem . Therefore, healthcare professionals need to pay greater attention to body-image alterations of patients living with BC during both chemotherapy and surgical procedures; and should provide culturally appropriate psychosocial support or self-care strategy interventions regarding hair loss and altered body image. Notably, we found that participants herein expressed significant concerns about fertility changes. These concerns may heighten guilt and lead to a self-identity crisis over the loss of gender roles. In China, the child is considered an important link between husband and wife, and a family without a new child is seen as incomplete . Interestingly, we summarized the different concerns over fertility of participants with different marital statuses influenced by traditional Chinese culture. Those who were married and childless had doubts about whether and when they could get pregnant after treatment. Those who invested a lot of energy and time into getting pregnant and were then diagnosed with BC were confronted with two simultaneous critical life events: one symbolizing the beginning of a new life, the other an existential threat to survival . Those who were married with a daughter felt great sorrow and self-blame over being unable to carry on the family line because the traditional Chinese concept of fertility-that a son is the status source of a married woman with her in-laws-persists in some areas of China . The three invisible shackles of "good wife, good mother, and filial daughter" within traditional Chinese culture mean that participants' emotional distress over role changes carries great significance. In China today, being married to a good wife remains a husband's most common psychological expectation, and being "a good wife" is also Chinese society's evaluation of a woman's competence . In traditional Chinese culture, the master-slave nature of the husband-wife relationship dictates that the wife's duty and responsibility are meeting her husband's needs . Moreover, a filial daughter must fulfill a series of obligations to her family, making the necessary sacrifices to cultivate and perfect her virtue . In many people's opinion, being able to marry a good husband and have children are also regarded as a daughter's filial piety to her parents and parents-in-law . Furthermore, a loving mother is considered the standard of a good mother. Yet such unselfish, great love also fetters women's freedom and the choices they can make for themselves. When participants became sick, they had to set aside their employment and housework, shifting from a service provider to served. Such a changed gender role did not lead patients to realize "I am sick, I should rest and care for myself first"; instead, it strengthened their role consciousness of being "idle", "a liability", and "incompetent" . Given their role in caring for patients, it is suitable for nurses to assess patient's concerns about sex and fertility, and to open a dialogue with them regarding these topics. This rapport can make it easier for patients to feel comfortable discussing their concerns about sensitive, intimate topics, and nurses can, in turn, provide patients with resources to help them better understand and address these anxieties. The traditional Chinese virtues of bearing hardships and withstanding hard work can often be found in Chinese women , so it is easy to understand the positive phenomenon of post-traumatic growth and cognitive reappraisal discovered herein. Examples of participants' post-traumatic growth include gains in self-esteem, self-confidence, and beliefs in their own potential, which might be carried over to future events . The use of cognitive reappraisal led participants to face negative emotions instead of suppressing and escaping them. They became aware of a meaningfulness in maintaining an optimistic attitude. Moreover, they began to develop a renewed life with a new attitude, facing cancer and others' judgements bravely and confidently, and identifying themselves as achievers to enhance other's confidence in their ability to overcome cancer . Looking back, many participants realized that "Life is very precious. We should take good care of ourselves" . They encouraged benefit finding and cognitive reappraisal recommending that other patients reassess their priorities and participate in activities that hold greater significance. Consistent with previous studies , participants herein acknowledged that sufficient social support from their nation, family, friends, and peer patients were powerfully important buffers for coping with stress and helping reduce the harmful effects of negative events on their physical health and emotional wellbeing. Most participants expressed that the meticulous care from and constant company of their families gave them motivation and confidence to persist. This, in turn, strengthened their belief in cancer survival, for themselves and their families . Thus, instead of considering this stressful event as an inescapable challenge, healthcare professionals may guide patients to assess it as a challenge that offers growth opportunity. One of the impacts of cancer was that some participants were treated as "patients" and had great difficulties fulfilling their responsibilities maintaining their children's wellbeing and their families' stability . Consistent with Lu's report , several participants stated that they seldom disclosed their feelings to avoid placing emotional burdens on their family members. Others expressed that having no one understand them, and feeling so physically and mentally distressed and exhausted, was driving them to the point of insanity. Indeed, social support quality depends on an individual's ability to communicate their needs, while a lack of emotional disclosures may actually prevent others from detecting their support needs and reduce life satisfaction , further challenging early identification of their emotions. It is important to consider that participants may also feel misunderstood by others because they are bound by a traditional culture. Importantly, cultural proscriptions may stifle one's ability to seek support not only from friends and professionals, but from family members . The Chinese perspective that there is a personal or intrinsically moral cancer cause discourages seeking medical advice and social support outside one's immediate circle because of social stigma . Chinese women are taught to be self-sufficient and to seek help only from within the family if they cannot cope with situations on their own . Seeking help outside the family indicates character weakness and is considered a shameful family disgrace . Patients are both eager for targeted social support and afraid of being seen as weak and incompetent. This contradiction between inner needs and traditional culture leads to significant pain. Healthcare professionals should be aware of the conflicts that may arise in gender identity to fashion suitable programs to help and support patients as they transition between the roles of mother, wife, and daughter, to that of patient. Public health campaigns and psychoeducation about appropriate family support methods are also essential for patients' caregivers, so they can help patients transition from sickness to resilience. By analyzing the interviews, EW, and Weibo text data, we found that although the three major themes and eight subthemes are presented across these sources, there were also some differences between them. The Weibo content tended to reflect a daily record of patients' lives and encouragement of themselves or others. In contrast, during the semi-structured interviews, participants tentatively asked questions in relation to issues about which they felt confused and were eager to address via direct communication . However, most noteworthy was the EW content. In both EW and the interviews, we encouraged patients to express the emotional experiences that bothered them most; however, EW seemed to promote patients' emotional disclosure and the EW contents were also richer than those of the interviews or Weibo. Although we identified the theme of body-image disturbance and sense of stigma in the Weibo texts, we also found that the cause for this experience was often isolated to chemotherapy-related alopecia. In their EW texts, participants directly expressed a wider range of authentic feelings about their breast loss, such as, "I can't look at my breasts", "I am no longer a real woman", and "I dare not look in the mirror". Furthermore, Weibo participants' expressions of worry about their role changes mainly centered on being unable to live normally because of physical dysfunction from surgery. In contrast, in EW texts, participants expressed guilt over role changes and focused on being unable to fulfill their responsibilities as wives, mothers, or daughters due to breast loss and long-term treatment. Participants' disbelief and an escape from reality after diagnosis could be found in all three data sources, but only in the interviews did we hear the voice of patients' regrets over delaying medical treatment due to a lack of disease awareness. Moreover, during those interviews, they encouraged women who feel abnormal in their bodies to have an early check-up. Importantly, our study identified the different concerns of participants relating to fertility being impacted by breast cancer treatment. These differed according to marital status and, given cultural expectations, may indicate a need to provide targeted emotional support to female Chinese patients with different marriage and childbearing concerns. As for the emotional trends in these data sources: Weibo texts tended to be more positive; it was obvious that participants preferred to remain calm during the interviews; and in the EW texts, negative emotions were very strong and clear. That some cultural-specific emotional experiences identified via the three data sources herein reiterated those in previous studies with single-source data supports our hypothesis that there are few differences in self-report-based themes of emotional experiences by Chinese women, who tend to suppress emotions across methods of expression. Furthermore, considering that the participants herein were primarily experiencing an acute phase, when emotional distress and emotional fluctuation are most common, their reports likely reflect their strongest emotional experiences. That many emotional experiences reported herein were similar to those from previous studies that did not focus on disease phases emphasizes the long-term, persistent emotional distress common among patients with similar experiences and suggests that the emotional management and support they receive throughout the entire BC diagnosis and treatment process is insufficient or ineffective. Emotional communication modes preferred by Chinese women are non-verbal, indirect behaviors rather than direct verbalizations . Coincidentally, for patients with emotional distress and emotional needs, these data also show that EW and interviews are useful for both data collection and for patients to express their feelings and realize self-healing through emotional self-report and self-disclosure, further confirming the effectiveness of non-verbal, indirect behaviors. Thus, care providers should attend to levels of acculturation among patients living with BC to maintain clearer communication about treatment expectations and appropriate social support. --- . Strengths and limitations This qualitative text analysis of data from three sources highlights the culture-based emotional experiences of female Chinese patients living with BC. We also analyzed the causes of these emotional experiences based on traditional Chinese culture and advanced corresponding care suggestions for healthcare professionals. Importantly, the theme and sub-themes summarized herein can nearly all be found within Weibo data alone, illustrating the validity and usefulness of this type of cost-effective Internet data for health research and providing a novel data source for future research analyzing patients' emotional experiences. Despite the novel and exciting nature of this study, we recognize several challenges with this analysis type. First, participants' sociodemographic characteristics were limited to the EW and interview participants, because personal information on Weibo network users is not publicly available. Thus, we were also unable to compare similarities or test differences in emotional experiences based on participant backgrounds. Second, most comments in Weibo are generally short and concise, and, although most of them are full of emotion, they are not conducive to theme coding. Third, some participants experience diagnosis, surgery, and chemotherapy within a short period, so their texts reflect their emotional experiences across phases. Due to the difficulty analyzing their emotional experiences during the three different phases separately, we included all the patients in all three phases as a whole and the results thus reflect patients' emotions overall. --- . Conclusions The emotional distress suffered by patients with BC can be varied, long-term, and extremely challenging. However, when women engaged actively in benefit finding and cognitive reappraisal with strong social support, patients were able to find ways to adapt and reported post-traumatic growth. Strong social support was an important facilitator in this growth. Therefore, healthcare professionals need to understand and respect these unique experiences and must provide these patients with psychosocial counseling and timely, tailored, culturally appropriate emotional support. This is a potential target for nursing interventions. Moreover, female Chinese patients with BC are generally restrained in their emotional expressions and generally reluctant to reveal their feelings. It is therefore valuable for healthcare professionals to increase cultural sensitivity when providing services to female patients with BC who are from diverse cultural backgrounds. Furthermore, it is essential to screen for distress symptoms in patients at all BC diagnostic and treatment phases as part of the standard nursing care routine. Additionally, these data support the provision of different channels to encourage female patients with BC to express their emotions, and they show that exploring validated, reliable measures of emotion expression, body image, sexuality, and fertility is warranted for healthcare professionals. --- Data availability statement The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author. --- Ethics statement The research has been conducted in accordance with the Declaration of Helsinki and following ethical principles and guidelines. Ethics approval for the study was granted by the Medical Ethics Committee of Nanfang Hospital of Southern Medical University and a standardized informed consent form had been established . Eligible participants were informed about the study and provided written informed consent for participation prior to the study. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. --- Supplementary material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2023. 1114139/full#supplementary-material

Results: Analyses were based on , Weibo posts and comments published by posters and , commenters, transcription texts from semi-structured interviews, and expressive writing texts. From this total collection of , Chinese characters, three major themes emerged: (i) conflicting emotions after diagnosis; (ii) long-term su ering and treatment concerns; and (iii) benefit finding and cognitive reappraisal. Conclusions: Despite gathering information from various sources, we found that distress from body-image disturbances, gender role loss and conflict, and changes in sexuality and fertility, were consistent among this sample of female Chinese patients with breast cancer. However, when women engaged actively in benefit finding and Frontiers in Public Health frontiersin.org Li et al. . /fpubh. . cognitive reappraisal with strong social support, patients were able to find ways to adapt and reported post-traumatic growth. Strong social support was an important facilitator in this growth. These study findings emphasize that healthcare professionals ought to increase cultural sensitivity, provide multiple channels to encourage patients to express their emotions, and incorporate screening for patients' emotional distress at all diagnostic and treatment phases as part of routine nursing care.

One definition of socioeconomically vulnerable women is the female head of the household who is usually poor due to low education; limited access to jobs, social services, sources of production, capital, credit, and land; and several kinship networks that support it . Divorced women, who live or die, can immediately become poor. Therefore, efforts to empower women are needed to improve their skills so they can work and earn a decent income so that they can help the family economy . One of the goals of empowering women is to create social mobility. Take the example of women in rural areas who carry out social mobility from casual farmers to factory workers. This condition is similar to Sorokin's theory of social mobility, especially horizontal mobility. Sorokin in Coser defines social mobility broadly as the movement of people in social space. Sorokin distinguished two types of social mobility, namely horizontal mobility and vertical mobility. Horizontal mobility is a movement from one social position to another at the same level . For example, a move from a farmer to a factory worker is a move from one position to another with the same degree. Meanwhile, vertical mobility is when someone moves to increase their status or level. For example, someone who was just a peddler becomes a trader by owning his shop and having subordinates help him . The goal of empowering women is to raise women's awareness about gender equality so that they can realize their full potential and contribute to family well-being. There are three important reasons why women must be empowered in the context of poverty alleviation, namely : women have the same interest in the development and are also users of development results who have the same rights as men ; women have special interests for women themselves and children, which are not optimal if initiated by men because they require special sensitivity, related to daily life, existing socio-cultural ; and empowering involving women in development, will indirectly empower and transmit a positive spirit to the next generation, who are generally very attached to the mother figure in their daily lives . In the context of the above, many contemporary researchers have shown that women working in poor families are always the redemptive outlet of the family economy. Women are considered to be the lifeblood of the household economy as poor women play various roles in the household , including: first, as managers of household finances; second, as heads of all household chores; third, as the economic pillar of the family; fourth, as one of the important nodes of social networks in terms of social transfer, especially in times of crisis and crisis . Housewives have the right to work and earn income outside the home, which is mobility or a change of status from the domestic to the public sphere. This social mobility is related to the socioeconomic conditions of the family, as mobility is movement from one social class to another . This means that socioeconomically disadvantaged women who receive more decent jobs and incomes, these women have the potential to contribute to meeting their needs while improving the social welfare conditions of their families . In countries with a low-end economy, many mothers are the savior of the family economy. They also seek additional family income, as their father's income is insufficient to support the family . Women in poor families are mainly women of working age. Another characteristic of women in poor households is their low level of education and limited land ownership . Efforts made by women to alleviate poverty by diversifying businesses and handling expenses related to fulfilling food, clothing, and social interests . Based on the above background, the overall question of this research is how to empower and mobilize women from poor families from the domestic sphere into the public sphere to improve family well-being. In particular, the question this study seeks to answer is: What are the demographic characteristics of socioeconomically disadvantaged women? What forms of empowerment are given to socioeconomically disadvantaged women? How mobile are women in increasing household income through productive enterprises? What constraints and support do socioeconomically disadvantaged women face in developing a productive economy to improve family well-being? The objective of this research is to identify of demographic characteristics of socio-economic vulnerable women, to determine the forms of empowerment given to socioeconomically vulnerable women, to assess women's mobility to increase family income through productive economic endeavors, to explain constraints and support faced by socio-economic vulnerable women in developing a productive economy to improve family welfare in urban areas. --- Literature review --- Empowerment Empowerment is providing resources, opportunities, knowledge, and skills to citizens to increase their ability to determine their future and participate in efforts to influence the lives of their groups . Still, according to Jim Ife, the concept of empowerment has a close relationship with two main concepts, namely: the concept of power and the concept of disadvantage . There are at least five principles of community empowerment, starting from ecological principles, social justice, local wisdom, and process, local and global . There are three important reasons why women should be empowered in the context of poverty alleviation , namely: women have the same interests in development, as well as users of development results who have the same rights as men; women, have special interests towards women themselves and children, which are not optimal if initiated by men because they require special sensitivity, related to everyday life, existing social culture ; and empowering women in development, indirectly empowering and transmitting positive enthusiasm to the next generation who are generally very attached to the mother figure in everyday life . In particular, five aspects need attention in empowering women in rural areas, including financial assistance, infrastructure related to access to technology, business development, advocacy, and assistance; the formation of behavior is associated with legal and regulatory aspects . Women's empowerment programs in family life will be able to become the driving force for creative economic enterprises that can create increased family welfare . Women are required to be able to support family resilience which can sustain family economic resilience. The number of poor people in rural areas is still higher than in urban areas. In total, the number of poor people was recorded at 26.5 million people as of September 2021 . The results of the study show that socio-economically vulnerable women have problems with the lack of access that can be obtained, such as education and services that affect their economic conditions, causing them to become poor. The role of assistance to empower is quite effective because the companion becomes a mediator who connects women's needs to help overcome social welfare problems . According to Jim Ife's theory of community empowerment in his book Community Development, Creating Community Alternatives-Vision, Analysis, and Practice , empowerment is defined as "providing citizens with resources, opportunities, knowledge, and skills to increase their abilities to determine their future and participate in efforts to influence the life of the group". Contextually community empowerment is defined as a concept of economic development that contains social values . This concept reflects a new paradigm of development: people-centered, participatory, empowering, and sustainable . Empowerment driving schemes and network patterns in women's empowerment are built through government participation and responsibility with regulations, policies, and facilitation. Non-government consists of private institutions/NGOs. Society as an actor empowered by the government and non-governmental organizations. The people involved in this case are socioeconomically vulnerable women . Women are involved in various sources, knowledge, opportunities, and skills to improve their abilities and skills so they can prosper . Empowering women through education and raising awareness through increasing knowledge and skills to increase their strength. Community empowerment in Actors' theory, society is considered a subject capable of making changes if it is released from strong control and gets the freedom to be responsible for its ideas, decisions, and actions. That perspective is by the acronym Actors, namely authority; confidence and competence; trust; opportunity; responsibility; and support . --- Gender Equality Today, gender equality is important in social relations, which is important for the development of society. As a result of the special attention paid to increasing women's participation in society, especially in society, there have been major changes in both social relations and legislation . Gender equality is the main issue of development goals which has its value. Gender equality is one of the goals to be achieved in the Sustainable Development Goals agenda; there are still various problems in the form of gender-based discrimination and inequality in society . Gender equality is one of the priority issues in the vision of sustainable development , including realizing gender equality and justice, fulfilling women's rights, and eliminating all forms of violence against women. Gender equality will enable women to do anything to increase their family income. He can work outside the house anywhere and return to his home . Social mobility is one of the activities that always exist in human life. Everyone always carries out social mobility because social mobility is one of the efforts to achieve primary, secondary, or even tertiary needs . Ideally, every human being has the same right to do a job. Recruitment or assignment must be given based on one's abilities . In practice, there is still discrimination against several parties, including women. Companies need resources that have good competence so that the company's productivity will increase. Women are also able to give good performances if given the same opportunity to work. In this case, leadership that gives equal rights to women needs to be done so that women can show their competence. Various theories and scientific developments on human resources also emphasize gender equality and science-based applications should pay attention to that as well . --- Social Welfare Social Welfare in Law 11 of 2009 Social Welfare is a condition of fulfilling the material, spiritual and social needs of citizens so that they can live properly and be able to develop themselves so that they can carry out their social functions . Article 34, paragraph of the 1945 Constitution of the Republic of Indonesia, mandates the state's obligation to care for the poor and neglected children. The implementation of social welfare according to Law 11 of 2009 concerning Social Welfare , demands the widest possible role of the community, both individuals, families, religious organizations, community organizations, non-governmental organizations, professional organizations, business entities, social welfare institutions, as well as foreign social welfare institutions. Social workers help families is about creating a harmonious environment that meets the physical and social needs of family members without creating serious obstacles in the family . Family welfare is the creation of a harmonious situation and the fulfillment of physical and social needs for family members without experiencing serious obstacles in the family, and in dealing with family problems, it will be easy for family members to solve them together, so standard family life can be achieved . --- Legal Foundation Regulation of the Minister of Social Affairs of the Republic of Indonesia Number 08 of 2012 Guidelines concerning Data Collection and Management of Data on Persons with Social Welfare Problems, Potential, and Sources of Social Welfare . Socioeconomic Vulnerable Women are adult women who are married, single, or widowed and do not have enough income to be able to meet their daily basic needs. Criteria: a. women aged 18 years to 59 years; b. a wife whose husband left without clarity; c. being the main breadwinner of the family; and d. income less or not sufficient for decent living needs . Law Number 13 of 2011 concerning Handling the Poor is explained in Article 2 that Handling the poor is a directed, integrated, and continuous effort carried out by the government, local government, and or the community in the form of policies, programs, and empowerment activities, assistance, and facilitation to meet the basic needs of every citizen . Article 34 paragraph of the 1945 Constitution of the Republic of Indonesia, mandates the state's obligation to care for the poor and neglected children. --- Relevant Research According to Khoirudin et al, Vulnerable women are a problem that needs attention because one of the factors that cause poverty is socio-economically vulnerable women whose income is below average. Low income can result in social inequality . Research on the Role of Social Welfare Workers in Kendari City Regency in Empowering the Potential of Socio-Economic Vulnerable Women. The results of the study show that there are several roles for District Social Welfare Workers in empowering the potential of Socio-Economic Vulnerable Women . Research on the Empowerment of Socio-Economic Vulnerable Women was conducted in Triwidadi Village, Pajangan District, Bantul Regency. The results showed that after getting empowerment through coaching, motivation guidance, social guidance, entrepreneurial skills, and innovation showed changes in behavior, and there was an increase in family welfare The results of A'yun and Faidati's research , Empowering Socio-Economically Vulnerable Women and identifying factors that Influence the empowerment of socio-economically vulnerable women in Bantul Regency. The results showed that the process of empowering socio-economically vulnerable women in improving family welfare in Bantul Regency was quite successful. This is done through training in cooking skills, making snacks, making crafts, and trading. These empowerment activities contribute to reducing the poverty rate. --- Research Methods The approach used in this research is qualitative, and the use of descriptive methods is to the objectives of this research which seeks to explore various opinions and answers from socio-economic vulnerable women, families, community leaders, and stakeholders . Research on the empowerment and mobilization of vulnerable women from the domestic sphere to the public sphere is carried out to improve the welfare of their families . Making it possible to understand the mobility of this more comprehensively and in-depth . The location of the research was determined purposively in West Bandung Regency, West Java, Malang Regency, East Java, and Pekanbaru City, Riau, with the consideration that the location has a high number of socio-economically vulnerable women and they have received empowerment . Secondary data is data obtained using the documentation method, which collects data from books, newspapers, and magazines that are relevant to the research . Subject targets were determined purposively, namely, women who are vulnerable to socio-economic problems and who are already empowered. The object of research in this study is the socio-economic conditions of women who are vulnerable to socio-economic conditions before and after empowerment in increasing welfare; supporting and inhibiting factors for socio-economically vulnerable women in developing a productive economy . Data collection techniques used interviews, this technique used interview guidelines to reveal the socio-economic conditions of socio-economic vulnerable women, strategies to maintain the socio-economic life of the family, and the contribution of socio-economic vulnerable women in improving family welfare. Also explored the supports and obstacles in developing a productive economy . Analysis Method; Data analysis technique using descriptive-qualitative-interpretative analysis. Qualitative data will be analyzed descriptively and then given meaning interpretative by reducing data selection, centralization, simplification, and abstraction of raw data by the research objectives . The analysis model used refers to the Miles and Huberman model . According to Sugiyono, data analysis was carried out using the Miles and Huberman model on the Socioeconomic Answers for Vulnerable Women sheet and interview results. Data analysis is carried out through data reduction, data presentation, and conclusion drawing . --- Results and Discussion The number of socioeconomically vulnerable women in Malang Regency is 5,026 people, while in Pagak District there are 102 socioeconomically vulnerable women in . Gampingan Village and Pagak Village, Pagak Sub-District, are villages where there are quite a lot of socio-economically vulnerable women, so these villages are locations for residents to participate in empowerment to improve their welfare. The following analysis will present the results of research in these two villages. The number of socioeconomically vulnerable women in West Bandung Regency is 827,530 women, with a population aged between 15-29 years 194,948 people and those aged 45-59 years 148,265 people, in Lembang District the most is in the exact location in Cibogo Village 14, 08% of 3,856 women 827,530, or around 543 people . In Pekanbaru City. The selection of informants was carried out using the snowball sampling technique. In terms of location of residence, the selection of informants was not limited to one particular sub-district or sub-district but came from various sub-districts in Pekanbaru City. The distribution of informants provides an overview of empowerment in various sub-districts in Pekanbaru. Figure 1, it will be known about the age of the informants in this study, most of whom are socio-economically vulnerable women in the three locations, and some of whom are stakeholders, sub-district social welfare workers, and also village assistants who assist socio-economically vulnerable women who receive the empowerment. Figure 3 shows the age of the informants in three locations, namely Malang Regency, West Bandung Regency, and Pekanbaru City . It can be seen that the highest age in Malang Regency is 58%, West Bandung Regency is 55%, and Pekanbaru City is 60 % which is in the age range of 31 years to 40 years. This age group is still a productive age group for working and improving the family economy. In the age range of 41 years to 50 years in Pekanbaru City, it is still quite high at 40%; in West Bandung Regency it is 15%, while in Malang Regency it is 11%. In the age range of 21 to 30 years, Malang Regency is 26% higher than West Bandung Regency, which is 25%. This age group is an age group that is still very productive, so their knowledge and skills deserve to be improved because they still have great potential. Figure 4 below will show the level of education of this research informant. Figure 4 can be seen the level of education of the informants and stakeholders in the three locations. In Malang Regency, most of them did not complete elementary school education with a total of 42%, in West Bandung Regency, most of them had senior high school education as much as 72%, in Malang Regency with high school education there were 11%. Most of the City of Pekanbaru had diploma 3 education, as much as 76%, and for Malang Regency and West Bandung Regency, each as much as 5%. There are 24% of those with bachelor's degrees in Pekanbaru City, 5% in Malang Regency, and 6% in West Bandung Regency. They are stakeholders, SubDistrict Social Welfare Workers, and village assistants. The low level of education and the incompatibility of the expertise and skills possessed are very influential for job seekers with the qualifications required by the company. Figure 5 below shows the various types of jobs owned by women who have participated in empowerment. Figure 5 can be seen that most of the informants work as food sellers in Pekanbaru City as much as 55%, in Malang Regency as much as 45%, and in West Bandung Regency only 10%. Jobs as wholesalers in Malang Regency 30%, New Feed City 10%, and West Bandung Regency 5%. There are 25% of handicraft craftsmen in Malang Regency, 10% in West Bandung Regency, and 5% in Pekanbaru City. Tofu makers and dairy farmers only exist in West Bandung Regency. They get this type of work after empowering local governments such as the Office of Social Affairs, the Office of Cooperatives, micro, small and medium enterprises, and corporate social responsibility that care about women's empowerment. It can be seen from Figure 6 that the number of informant family members is 2 or 35% in Malang Regency, as high as 32% in West Bandung Regency, and as high as 20% in Pekanbaru City. For informants with a family of 3 members, Malang Regency accounted for 15%, West Bandung Regency 26%, and Pekanbaru City 15%. Most of the respondents were a family of four and lived in West Bandung 20% in Malang, and 15% in Pekanbaru. Figure 6 can be interpreted that all informants who received empowerment have family dependencies and tend to be in large numbers. This will add to the family's economic difficulties . Figure 7 below will show family income in three locations, namely in Malang Regency, West Bandung Regency, and Pekanbaru City. Figure 7 it can be seen about the income of informants in three locations, namely in Malang district, for those who earn < Rp. 1,000,000, there are as many as 25%, in West Bandung Regency there are as many as 10%, and in the City of Pekanbaru, there are as many as 40%. There 35% of informants who earn Rp. 1,000,000 to 2,000,000 in Malang Regency, 40% in West Bandung Regency, and 20% in Pekanbaru City. Most of the informants earn Rp. 2,000,000 to 3,000,000 were in West Bandung Regency with a total of 40%, while in Malang Regency and Pekanbaru City, each only amounted to 10%. The highest income of 3,000 to 4,000,000 is in Malang Regency, with a total of 30%, while West Bandung Regency has only 15% and Pekanbaru City has 10%. There is a disparity in income between the rural areas of Malang and the urban areas of West Bandung and Pekanbaru. The mobility of disadvantaged women in urban areas affects incomes compared to rural areas, where better transport and infrastructure can increase incomes. In rural areas, the mobility of disadvantaged women is limited due to poor access to transport, infrastructure, and culture. Rural women's beliefs and social norms limit mobility, so they affect income inequality, thereby hindering the economic advancement of disadvantaged rural women . --- Supporting factors There is support from the local government, Corporate Social Responsibility, the private sector, the family, and the community. There is stimulant assistance for business and business development, as well as assistance for informants who are women who receive empowerment and social assistance. --- Obstacle factors Empowering women is difficult to do on an ongoing basis because the implementation of activities is not complete, and assistance is not complete until they can become independent. There is a lack of commitment from the government and the community that the activity cannot be continued because the organizers only do it once. Obstacles in coordination between village officials and institutions that have not been well established, women's empowerment activities are often only carried out once without being continuous, thus disappointing women who want to get up to gain skills to be independent. --- Conclusions Research on the Empowerment and Mobility of Women from Domestic to Public Spaces in Improving Family Welfare has important implications for the impact of women's empowerment. Some of the implications of empowering economically vulnerable women include: o Women become more socioeconomically empowered so that they can improve socio-economic welfare in terms of income and access to resources. o Women's empowerment has an impact on social and economic life to improve the family economy because they are trained and given skills so that they can strive to improve the economy. o Women's empowerment increases the participation of vulnerable women in the labor market, as well as empowers women to make decisions. o Empowerment of socio-economically vulnerable women can reduce poverty because it can prosper the family. o Women's empowerment can transform socioeconomically vulnerable women who previously did not have access to become more empowered which strengthens social and economic inclusion into a productive part of society. --- Recommendations The existence of novelty research, namely the differences in social empowerment and women's mobility in rural areas and urban areas, requires collaboration between the local government and the business world to empower women. Women's empowerment has a very positive effect on improving family welfare, for local governments, so that they are given more intensive assistance so that socio-economic vulnerable women will be more empowered in developing the economy or socializing with the environment. So it is necessary to make regulations or policies that bind the involvement of all parties in women's empowerment activities so that women can develop and be independent to improve welfare. --- Ethical considerations Not applicable. --- Conflict of Interest The authors declare that they have no conflict of interest.

Likewise in the informal sector, men also dominate, although the differences are not striking. In February 2022, the female Labor Force Participation Rate is still far below that of men, with the male Labor Force Participation Rate at 83.6 percent and women at only 54.2 percent (Indonesia's central bureau of statistics 2022). Furthermore, the Central Statistics Agency reports that the gender wage gap is the largest in the service business workforce group, namely 53.34 percent in 2021. This wage gap indicates that male business workers earn wages that on average, are 64 percent higher than women in 2021. The gap has narrowed slightly from 51.89 percent in 2019 (Central Bureau of Statistics 2021). We still often encounter inequality with women, not a few of those who work earn lower salaries/income than men even though they have the same workload (Ferulano 2021). The minimum income earned by female workers can cause various problems, especially economic problems in the family, especially for women who have to bear all household needs or become the backbone of the family. This makes women vulnerable to socioeconomic problems (Ready 2018). The 2018 National Socioeconomic Survey shows that the literacy rate for men is still higher than for women, namely 97.33 percent for men and 93.99 percent for women. In the health sector, the number of women who have access to health insurance is 64.47 percent. In addition, the Labor Force Participation Rate, the Figure 1, based on 2018, Sakernas shows a fairly high difference, namely 82.69 percent for men and 51.88 percent for women (Amini et al 2020).

INTRODUCTION Cigarette smoking continues to be one of the main preventable causes of cancer, heart disease and premature death [1][2][3][4]. During adolescence many youngsters experiment with smoking, which may lead to regular smoking [5]. Numerous studies report smoking behaviour to be similar among friends [6][7][8][9]. Although early research argued predominantly that similarity was caused by peer influence, recent studies have provided evidence that similarity is also caused by selection of similar friends [10][11][12][13][14][15][16][17]. Friendships are an important aspect of an adolescent's life. Children tend to associate mainly with same-sex others and most close friends are of the same gender [18]. Although cross-gender friendships begin to develop during adolescence, gender similarity among friends is still marked [19,20]. Same-gender peer relationships are an important socializing context that influences the development of gender differences in social interaction styles. Differences between female-female and male-male friendships may lead to differences in smoking-based selection and influence processes. This study aims to examine selection and influence processes, differentiated by gender, in the context of smoking behaviour among adolescent friendship networks. Researchers have demonstrated that friendship patterns differ for adolescent males and females. Females have more intimate friendships [18,21,22], report higher intimacy levels [22] and are more likely to turn to peers for support [23], which could result in more opportunities for influence. Furthermore, previous research has found that females perceived more social pressure to smoke [24,25] and were more susceptible to social influences [26]. However, to our knowledge, no study has examined specifically differences between adolescent male and female friendship networks regarding smokingbased selection and influence processes. The present study will examine selection and influence processes in male and female networks by using new methods of social network analysis [27,28] which, contrary to conventional methods [10,12], allow the current state of continuously changing friendship networks to act as a dynamic constraint for changes in smoking behaviour, while allowing smoking behaviour to be simultaneously a dynamic constraint for friendship changes. The possibility of unobserved changes in friendships and smoking between two observations is considered and as well as including other possible determinants of smoking behaviour, such as age, parental and sibling smoking [29], these methods allow controlling for other determinants of friendship selection besides smokingbased selection of friends. Some of these determinants are reciprocation of friendship; becoming a friend of somebody who is a friend of a friend [30,31]; and selection based on alcohol consumption [27], age, ethnicity and education [20]. We hypothesize that smoking-based selection will be present in males and females and that females will be more susceptible to influence. Furthermore, we will explore the impact of reciprocity of friendship on the strength of smoking-based selection and influence. --- METHODS --- --- Procedure Self-administered questionnaires were distributed among all 7th graders of participating schools. Similar questionnaires were distributed during follow-up among 8th and 9th graders [32,33]. Students present on the days of data collection completed the questionnaire. It was explained that responses would be treated confidentially. Students could refuse to participate, and returned questionnaires in sealed envelopes to guarantee anonymity. At baseline, the proportion of refusals was 0.03% [32]. --- Questionnaire The ESFA questionnaire was based upon earlier studies about adolescent smoking [33][34][35][36][37][38]. • Friendship ties were assessed by asking adolescents to name up to five best friends inside and/or outside school [39]. Only best friends inside the same school grade are included, as only they also completed the questionnaire. • Smoking behaviour of adolescents was assessed by one question: 'On average, how many cigarettes do you smoke during a week ?' . • Parental smoking behaviour was measured by two questions: 'Does your father smoke?' and 'Does your mother smoke?', and was recoded into one variable . • Sibling smoking behaviour was measured by two questions: 'Do one or more of your brother smoke?' and 'Do one or more of your sister smoke?', and was recoded into one variable . • School achievement was assessed by one question: 'Last year, how well did you do in school, compared to the others in your class?' . • Alcohol consumption ; age was also recorded. --- Analysis plan A stochastic actor-based model [28,[40][41][42] was constructed to represent realistically mutual dependencies between friendship changes and changes in smoking behaviour by a simulation model. This model takes into account the mutual feedback processes between friendship and smoking occurring between observations. This new approach was applied successfully to delinquency [31] and alcohol consumption [43]. All respondents were included and allowed to enter the study later or leave earlier [44]. Missing values on adolescents' attributes and smoking behaviour were allowed and treated as noninformative in the estimation procedure and imputed by the mean for the start of the simulations [45]. The model encompasses two parts: one part models friendship network changes , the other part models smoking behaviour changes . Both parts are integrated as one internally dependent process. In this manner, selection and influence processes can be examined simultaneously while controlling either process for the other one. Mathematical specifications are given by [28,40] and a sketch of the model is described in the next section. --- Friendship network changes: selection processes The friendship network evolution part of the model specifies the preferred direction of friendship change by including effects that determine probabilities of changes in friendship status, such as current network structure and adolescent's attributes. Four main smoking-based friendship selection components were included: the effect of adolescent's smoking behaviour on number of friends chosen ; the effect of potential friends' smoking behaviour on choosing them ; and the interaction between smoking behaviour of adolescents and potential friends, to test that adolescents who smoke more also prefer friends who smoke more . We included the raw and squared value of alters' smoking to control for possible curvilinear dependence of the attractiveness of potential friends on their smoking behaviour. Because friendship choices depend strongly upon characteristics of the current network [20,40,46], the effects of number of friends chosen , number of reciprocal friends chosen and number of friends chosen who are also a friend-of-afriend were included. The selection model controlled for alcohol consumption, age, and school achievement of adolescents and potential friends. All included effects are presented in the upper part of Table 1. --- Smoking behaviour changes: influence processes The smoking behaviour evolution part of the model specifies the preferred direction of change in smoking behaviour by including a list of functions of network, smoking behaviour and other attributes upon which changes in smoking behaviour may depend. Included effects are described in the lower part of Table 1. Three main friendship network-related influence components were modelled: the effect of friends' average smoking behaviour, effect of the number of received friendship nominations and the number of outgoing friendship nominations on adolescent smoking behaviour . Included control effects were the tendency to smoke, a feedback effect of own previous smoking behaviour to control for non-linearities in smoking , parental and sibling smoking and adolescents' alcohol consumption, age and school achievement. --- Analysis For each wave, a female and male network were constructed within each school. All female adolescents would be members of the female network, all males formed the male network. As the focus of the present paper was upon same-gender friendships, cross-gender friendships were excluded. For each network, the dynamic actor-based model was analysed using SIENA software [44]. Effects were tested on the basis of t-ratios defined as estimate divided by standard error, which follow an approximate standard normal distribution [40]. Subsequently, results of all network analyses were combined for males and females in two metaanalyses. The null hypothesis that the effect is 0 in all networks was tested twice by Fisher's combination procedure [47], once for the right-sided and once for the left-sided test. The right-sided test, for example, examines the null hypothesis that in all networks the coefficient of this effect is non-positive, while the alternative hypothesis is that in at least one school the coefficient is positive. To control for multiple testing, there was deemed to be significant support for an effect if either of these combination tests were significant at the 0.025 level. This Fisher's combination procedure [47] is preferred over the Snijders-Baerveldt method [30], as it does not make the assumption that estimated standard errors and parameter values are uncorrelated, nor the assumption that the networks are a sample of a population. The null hypothesis, that effect parameters are constant across schools, was tested by Cochran's method [48], adapted for network dynamics by Snijders & Baerveldt [30,48]. To examine differences between male and female networks, within each school parameter results of the male and female networks were compared with an independent-sample t-test. The results of these nine t-tests were then combined using Fisher's combination procedure [47], assuming a significance level of 0.025. --- Differences between reciprocal and non-reciprocal friendships To explore whether the strength of smoking-based selection of friends differs when selecting non-reciprocal or reciprocal friends, and whether influence of friends differs within non-reciprocated and reciprocated friendships, interaction effects of smoking-based selection and influence of friends with reciprocity were tested with score tests [31, M. Schweinberger, unpublished data]. --- The relative contribution of smoking-based selection and influence As a similarity measure of individuals linked in a network we used Moran's I, a spatial autocorrelation coefficient [49]. By calculating the average similarity of linked individuals in simulated models with coefficients estimated under different model specifications, the relative contributions of selection, influence and control effects to observed smoking similarity can be expressed. This method is explained in detail elsewhere [42,50]. Two male networks were excluded due to very low smoking rates , which might bias the results of these simulations. For comparative purposes the two female networks within these schools were also excluded . We will represent graphically average proportions of similarity allocated to smoking-based selection, peer influence, selection and influence mechanisms not based upon links between friendship and smoking , and general trend effects . --- RESULTS --- Descriptives Table 2 presents the average network structure within male and female networks, the average smoking behaviour in each wave and baseline characteristics. Males and females smoked at similar rates. Females nominated more friends at wave 2 and reported more often to have smoking parents and siblings. --- Differences between male and female networks: selection processes The results for the friendship evolution submodel are reported in the upper part of Table 3. Males and females tended to nominate more smoking friends when their own smoking behaviour was higher, as indicated by the significant 'smoking behaviour ego ¥ alter' effects. There were no significant effects of adolescent own smoking and smoking of potential friends on friendship selection. Results for the control effects indicate that males and females tended to choose relatively few friends to reciprocate friendships , and to become friends with friends of their friends . Adolescents did not select friends based upon similar alcohol consumption, age or school achievement. Among males, a propensity to select low-achieving friends was found, while females tended to select friends drinking alcohol at medium level . Only transitivity differed significantly between males and females , implying that females showed a stronger tendency to select friends of their friends. --- Differences between male and female networks: influence processes Results of the smoking behaviour evolution submodel are reported in the lower part of Table 3. Females were influenced by their friends' smoking behaviour. Although there was evidence that the magnitude of effect of friend's smoking behaviour differed across the nine included female networks [c 2 = 25.43, df = 8, P = 0.001, estimated true standard deviation = 0.000], the effect was found to be consistently positive . Although males and females did not differ significantly, males did not adjust their smoking behaviour significantly to their friends' smoking behaviour. However, there was evidence for some variation in the effects across the male networks . Both males and females who nominated fewer friends tended to smoke more. The control effects indicated that adolescents had a significant overall tendency not to smoke, but smoking behaviour tended to be self-reinforcing, as indicated by the significantly positive 'tendency to smoke squared' effect. Male and female adolescents smoked more when their parents smoked and when they drank less alcohol. Females smoked significantly more when they were low achievers. High-achieving males had a higher tendency to smoke, but this effect was not significant. None of these effects differed significantly between males and females. --- Differences between reciprocal and non-reciprocal friendships The score test of the interaction between smoking of adolescents, friends and reciprocity indicated that in male and female networks, the tendency to select reciprocal or non-reciprocal friends who are similar in smoking behaviour did not differ . The score test of the interaction between friends' smoking and reciprocity in its effect on smoking dynam-ics showed that among males and females influence of friends did not differ between reciprocal and nonreciprocal friendships . --- The relative contribution of smoking-based selection and influence Figure 1 shows that the mean proportion of similarity in smoking behaviour between friends attributed to smoking-based friendship selection was 22% for males and 18% for females. The mean proportion attributed to influence of friendship networks was 15% for males and 21% for females. Trend effects accounted for 42% in male and female adolescents, and other determinants of friendship and smoking played only a small role in the explanation of smoking behaviour similarity. --- DISCUSSION The main goal of this study was to examine differences between adolescent male and female friendship networks regarding smoking-based selection and influence processes using newly developed tools of social network analysis. We found evidence that males and females formed friendships based on similarities in smoking behaviour. This process was similar for adolescent males and females. These results are in line with our hypothesis and earlier research reporting peer selection to be important in explaining similarities between adolescent friends with respect to smoking [11,[13][14][15][16][17]. However, no study examined smoking-based selection processes specifically in gender-segregated friendship networks. We found empirical support for influence of friends with respect to smoking behaviour within female networks. This effect was not significant among males. Although our findings seem in line with previous studies, suggesting that girls perceive more social pressure [25] and are more susceptible to social influences [26], the difference between males and females was not significant and parameter estimates were such that a relatively large amount of the smoking behaviour similarity among male friends was explained by influence . The lack of strong support for influence processes within male networks might be explained by males fostering friendship intimacy through sharing activities with friends [51], which takes place mainly outside school. Out-of-school friends may exert a stronger influence compared to within-school friends. Further research with a larger database is needed to give more unequivocal results about male-female differences in this respect, and to study dependencies of these differences on school context. There was no evidence that reciprocation of friendship modified the effect of smoking similarity on friendship choice. In contrast with previous studies reporting stronger support for influence processes among reciprocal friends [52,53], the strength of influence processes did not differ significantly between reciprocal and nonreciprocal friendships in the present study. However, researchers found support for influence among reciprocal [12,52,53] and non-reciprocal friends [10,54] among different populations. More research is needed to clarify the specific role of friendship reciprocity. Besides smoking-based selection, we controlled for several alternative processes explaining peer selection. Only transitivity differed significantly between males and females, indicating that females showed a higher tendency to select friends who were friends of their friends. Males and females were similar in their tendency to select arbitrary friends and reciprocate friendships. Furthermore, females preferred to select females who reported medium scores on alcohol consumption. In contrast with findings of previous studies [27], no support was found for selection based upon similar alcohol consumption. However, as findings on the complete Finnish sample showed support for adolescents selecting their friends based on similar alcohol consumption [50], lack of support may be caused by the reduced sample size due to restricting the networks to solely males or females. In line with earlier studies arguing that boys' culture is less study-orientated [55], and school achievement is not considered to be 'cool' among boys [56,57], males preferred to select males scoring low on school achievement. Regarding alternative influence mechanisms, no significant gender differences were found. Females smoked more when they scored low on school achievement. In line with previous studies, males and females were influenced by parental smoking [17,25,29]. Although previous research has demonstrated that tobacco use predicts subsequent alcohol use more effectively than the converse [58], the negative effects of alcohol consumption on smoking were counterintuitive. Complete-sample in-depth analyses confirmed the significant negative effect during the first wave, but this effect lost significance during the second wave and became positive, although not significant, during the last wave. Alcohol is very expensive in Finland. Lack of sufficient resources at a very young age may have forced youngsters' to choose between smoking or alcohol. The following limitations of this study can be reported. First, self-reported smoking behaviour was not validated biochemically. However, self-reported smoking can correspond well with biological indicators under measurement conditions assuring anonymity [59]. The ESFA project optimized measurement conditions by guaranteeing confidentiality [32]. Secondly, data were gathered from the Helsinki area only. Further research should illuminate differences between urban and rural areas. Thirdly, we included only friendships within the same school grade. Although these friends represent an important part of adolescents' social environment, future studies should include out-of-school friendships. Fourthly, previous research has demonstrated that parents can also have an effect on the types of friends that adolescents select [60]. The role of parenting styles and practices in smoking-based selection processes should also be studied using actor-based modelling techniques. Fifthly, several included constructs were measured with one item. Future studies should include variables based on multiple questions to increase reliability and validity. Sixthly, although adolescents can report parental smoking reliably [61] our results might be biased, as no direct measures of parental and sibling smoking were available. Finally, we did not consider possible differences between the successive data waves. However, a previous study on the same data set indicated no differences in smoking-based selection and influence effects between the three waves [50]. This study has several practical implications. First, both male and female adolescents selected friends based upon similar smoking behaviour, implying that prevention programmes for adolescent males as well as females should focus upon selection processes. Previous studies have already demonstrated the importance of peer networks besides the focus on social influences [62,63]. Although more research is needed on the relation between attitudes and selection processes, reinforcing non-smoking attitudes among males and females might be a way to increase the likelihood that they select nonsmoking peers. Secondly, only females were influenced significantly by their friends to take up smoking and should therefore benefit from learning skills to resist influence of friends. Although support for influence of friends among males was weak, the effects between males and females did not differ significantly and we therefore cannot conclude that prevention programmes will benefit from targeting adolescent males and females in a different way. Thirdly, as both males and females were influenced by parental smoking, the nature of this link should be investigated in more detail to learn how parents should be involved in youth smoking prevention programmes. --- Declarations of interest None.

The main goal of this study was to examine differences between adolescent male and female friendship networks regarding smoking-based selection and influence processes using newly developed social network analysis methods that allow the current state of continuously changing friendship networks to act as a dynamic constraint for changes in smoking behaviour, while allowing current smoking behaviour to be simultaneously a dynamic constraint for changes in friendship networks. Design Longitudinal design with four measurements. Setting Nine junior high schools in Finland. Participants A total of 1163 adolescents (mean age = 13.6 years) who participated in the control group of the ESFA (European Smoking prevention Framework Approach) study, including 605 males and 558 females. Measurements Smoking behaviour of adolescents, parents, siblings and friendship ties. Findings Smoking-based selection of friends was found in male as well as female networks. However, support for influence among friends was found only in female networks. Furthermore, females and males were both influenced by parental smoking behaviour. Conclusions In Finnish adolescents, both male and female smokers tend to select other smokers as friends but it appears that only females are influenced to smoke by their peer group. This suggests that prevention campaigns targeting resisting peer pressure may be more effective in adolescent girls than boys.

Introduction The decade following high school is pivotal in setting the stage for adulthood functioning and adjustment. In descriptions of the nature of the transition to adulthood in contemporary Western societies, psychological models tend to emphasize self-directed exploration , demographic traditions highlight the impact of social structure on the life course , and interdisciplinary approaches focus on the interaction of historical, cultural, and structural constraints with personal agency in the creation of individualized life courses . Despite some differences, dominant models of the decade following the completion of mandatory education share three important areas of agreement. First, it is broadly agreed that the last 50 years have heralded later median ages at the initiation of traditional adult roles ; greater withinand between-country heterogeneity in the timing, order, and resulting patterns at any given point in time ; and increased individualization of the life course, resulting in more choices and a correspondingly greater need for personal agency and soft skills such as teamwork and networking . Second, dominant models note that formal social institutions and informal social norms provide less structure after the end of secondary school, with individuals able to take on a diverse set of potential life changes, including school completion, employment, residential independence, intimate cohabitation or marriage, and parenthood . Third, scholars agree that the 20s are dense with transitions to adult work and family roles, and are pivotal for developing healthy and fulfilling lifestyles . As a result, Settersten concluded that this decade is both formative and risk-laden, a period that contributes to the accumulation of advantage and disadvantage over the life course . In this paper, we focus on role configurations of early adults in their mid-20s. We explore the heterogeneity of social roles and how these are combined in the lives of participants in two nationally representative samples of American youth in their last year of secondary education, who were followed longitudinally into adulthood. Data are drawn from the ongoing Monitoring the Future study, focusing here on two cohorts who graduated from high school 12 years apart. We focus on the big five social role domains of early adulthood : education, residential status, employment, cohabitation/marriage, and parenthood. We focus on ages 25 to 26 because most young adults have completed formal education by this point, freeing them from the constraints that the pursuit of formal education places on marriage, parenthood, and employment, because considerable heterogeneity in configurations of roles is both possible and likely , and because this age focus facilitates comparability with three related studies . Our three aims are to: identify heterogeneity in social role configuration during early adulthood using latent class analysis; examine whether demographic and educational aspirations and performance assessed during the senior year of high school predict social role configurations at ages 25/26; and examine concurrent health and adjustment correlates of the configurations, including substance use, economic independence, and life satisfaction. The introduction focuses on the background to each of the three aims in turn. --- Role patterns in young adulthood What is known about prototypic role configurations for young adults? Osgood and colleagues classified young adults in a longitudinal study of a regional sample of white working-and middle-class youth, based on their role involvement in five adult social roles at age 24 , identifying six paths to adulthood. "Fast starters" tended to be partners, parents, living independently, and in long-term employment, with a low likelihood of having completed a college degree. "Parents without careers" had similar spousal and parental roles as well as independence from their own parents and low education, but differed from Fast starters primarily by having lower levels of employment. "Educated partners" lived with partners rather than parents, but had not become parents themselves. They tended to be employed and to have higher levels of education. "Educated singles" did not live with romantic partners, but rather with parents or in rented housing. They had the highest levels of education. "Working singles" were not yet parents or partners, and tended to live with parents and be employed full-time. Their typical level of education was to have attained some college, but not a degree. Finally, "Slow starters" tended to have made the fewest transitions, live at home, not be partnered, have low education, and be unemployed or in shortterm jobs. Thirty percent of this group were parents, compared to 20% in the full sample. Osgood et al provide an important foundation for the present analyses due to its person-centered approach, identification of cohesive role-based configurations in early adulthood, and examination of correlates of role combinations. In Aim 1, we seek to extend their analyses in a longitudinal communitybased sample by using similar indicators of social roles with the addition of current student status, employing a nationally representative US sample of high school seniors followed longitudinally, examining contemporaneous correlates in the mid-20s, and replicating results among young adults in similar and earlier historical time periods. --- Predictors of young adult role patterns Who is likely to inhabit specific configurations of social roles in young adulthood? Past research suggests the key importance of education in shaping the timing and sequence of adult role transitions . Specifically, those who pursue higher education tend to initiate the adult roles of partner, parent, and employee later; during this time many receive significant assistance in the form of parental financial support and subsidies . Education tends to confer significantly higher incomes , and has arguably become more important historically as careers have become less secure and more variable and self-directed . In this way, the more educated group may receive multiple benefits not experienced by all. Of course, opportunities to pursue higher education are, in turn, not randomly distributed in the population . Due to the importance of education in shaping the transition to adulthood , we focused on educational as well as demographic predictors, assessed at age 18, of age 25/26 role configurations. Specifically, for aim 2, we focus on two types of predictors: sociodemographics and educational background . We ask the research question: to what extent do socio-demographic and educational background indicators assessed at the end of high school predict the role configurations participants inhabit at ages 25 to 26? --- Wellbeing correlates of role patterns There are several concerns regarding how wellbeing relates to the role patterns. For example, is one role configuration ideally conducive to health and wellbeing? Or, do different configurations coincide with complex patterns of risk and benefits? We focus on three domains of wellbeing in young adulthood: life satisfaction, economic independence, and substance use. The mid-20s are a period of normative developmental changes toward greater health and independence. For example, subjective markers of adjustment and life satisfaction increase as does economic independence . On average, harmful substance use declines , particularly among those who are presently partnered . Although the normative pattern is toward increased healthy lifestyles and greater wellbeing, this is not true for all. In the US, the third decade of life is when many substance use disorders and psychopathology manifest, unemployment is high, and access to health insurance is the lowest in the lifespan . Historic declines in the availability of living wages and affordable housing have added to the challenges of launching adult independence. Thus, even though wellbeing tends to increase, so do serious difficulties highlighting the increased heterogeneity or "fanning" of individual differences during the third decade of life . Recent depictions of the transition to adulthood and the young adult period have highlighted both the increased importance of exploration of roles before settling on permanent decisions and the significant challenges faced in recent decades of negotiating the transition to adult roles . In the present study, we examine whether indicators of contemporaneous wellbeing-such as substance use, economic independence, and life satisfaction-differ among our observed role configurations. By doing so, we seek to explore whether alternate social role configurations represent diverse pathways to equivalent levels of health and happiness, or whether some configurations appear intrinsically more challenging or risk-laden. --- Method Sample Longitudinal data were obtained from the ongoing Monitoring the Future project, which has recruited nationally representative samples of about 16,000 students in final year of secondary school annually since 1975 . Approximately 2,400 of these participants are selected each year for biennial follow-ups using mailed surveys. Illegal drug users are over-sampled for follow-up, thus sampling weights are used in analyses to best approximate population estimates. These longitudinal follow-ups begin one year after the baseline for a random half of the participants, and two years after the baseline for the other half; these two halves are combined in these analyses. Thus, longitudinal data collected at age 18 and at ages 25/26 are used in analyses reported here. Retention rates through the mid-20s typically exceed 60% for cohorts used here, and differential attrition analyses reveal that those retained are more likely to be female, to be white, to report higher high school grades/marks, to have more educated parents, and to report lower senior year drug use , and the cohort described as the 90s cohort represented the high school graduating classes of 1988 and 1989 who were at modal ages 25 or 26 in 1995, 1996, or 1997 . The 80s cohort roughly parallels the National Child Development Study in Britain and the Jyväskylä Longitudinal Study of Personality and Social Development in Finland , and the 90s cohort roughly parallels the British Cohort Study and the Northern Finland Birth Cohort 1966 . . --- Measures --- Demographic and educational predictors Demographic indicators were assessed at age 18. These included gender, race, two-parent family, and parent education. Gender was coded such that male=0 and female=1. Race was coded with two dummy variables with White as the reference group. Living with two biological parents was distinguished from other family forms . Parental education was assessed with two questions, "What is the highest level of schooling your [mother/father] completed?" Responses were 1=Completed grade school or less, 2=Some high school, 3=Completed high school, 4=Some college, 5=Completed college, and 6=Graduate or professional school after college. If both parents responded to the question, the responses were summed and multiplied by 5. If only one parent responded, his/her answer was multiplied by 10, so that responses ranged from 10 to 60. Educational background was assessed as high school grade point average and educational aspirations. High school GPA was measured with the question, "Which of the following best describes your average grade so far in high school?" based on a range of responses from 1=D to 9=A . Grades of A are the highest possible in the US system and D is the lowest passing grade; one letter grade is obtained in each course each year and averaged. Educational aspirations were based on the question, "How likely is it that you will do each of the following things after high school?" Responses included "Attend a technical or vocational school," "Serve in the armed forces," "Graduate from a two-year college program," "Graduate from college ," and "Attend graduate or professional school after college" with a response choices for each of Definitely won't, Probably won't, Probably will, and Definitely will. College aspirations were coded to reflect clear intentions to pursue a 4-year college degree or attend graduate school. Individuals who responded that they definitely would graduate from a four-year college degree or attend graduate school were coded as 1 and all others were coded as 0. Among those who did not have college aspirations, trade school aspirations were coded as 1 for those who reported that they would definitely do technical school, armed forces, or graduate from a two-year college program and all others were coded as 0. --- Social roles in early adulthood Participation in the big five social roles of adulthood was assessed via self-report at age 25 or 26 using categorical indicators. The percents in each role are reported in the first columns of Tables 1 and2 . Student status was distinguished in three categories: Full-time student, Part-time student, and Non-student. Educational attainment was coded at four categorical levels: Completed high school, Some college, Associate's degree, and Bachelor's degree or higher. Four categories of employment were distinguished: Not employed, Homemaker, Parttime employee, and Full-time employee. Regarding residential independence, we distinguished cohort members who lived with parents from those who lived away from parents. Three categories of partnership were distinguished: Single, Cohabiting , and Married. Regarding parenthood, we distinguished cohort members who had no children, 1 child, and 2 or more children. --- Wellbeing at ages 25 to 26 To assess wellbeing and adjustment in the mid-20s, we focused on perceived global life satisfaction, economic independence, and lifestyle indicators of legal and illegal substance use. Life satisfaction was based on the question, "How satisfied are you with life as a whole?" Responses ranged from 1=completely dissatisfied to 7=completely satisfied. Economic independence. Participants were asked, "During all of last year , how much of your financial support came from each of the following sources?" Responses were on a scale of 0=None to 6=All for a range of income sources . The proportion of income originating from the respondent and his or her spouse were summed to indicate economic independence, with responses ranging from 0% to 100%. Cigarette use was measured with the question, "Which best describes your cigarette smoking in the last 12 months?" Response options were 1=Have not smoked at all, 2=Smoked once or twice, 3=Smoked occasionally but not regularly, 4=Smoked regularly but stopped or cut back, and 5=Smoke regularly now. Alcohol use . Participants were asked, "On how many occasions have you had any alcoholic beverage to drink-more than just a few sips…[in your lifetime/during the last 12 months]?" Response options for each timeframe were 1=0 occasions, 2=1-2 occasions, 3=3-5 occasions, 4=6-9 occasions, 5=10-19 occasions, 6=20-39 occasions, and 7=40 or more. Heavy drinking was reported by the question, "During the last two weeks, how many times have you had five or more drinks in a row?" Response options were 1=None, 2=Once, 3=Twice, 4=3 to 5 times, 5=6 to 9 times, and 6=10 or more times. Marijuana use . Participants were asked, "On how many occasions have you used marijuana/hashish…[in your lifetime/during the last 12 months]?" Response options for each questions were 1=0 occasions, 2=1-2 occasions, 3=3-5 occasions, 4=6-9 occasions, 5=10-19 occasions, 6=20-39 occasions, and 7=40 or more. --- Use of illegal drugs other than marijuana . Using similar questions for lifetime and past 12 month use of a series of other illegal drugs, participants were coded as 1=used an illicit drug other than marijuana, or 0=did not use an illicit drug other than marijuana. Note that this measure does not include marijuana use. Both those coded as "0" and "1" may or may not have used marijuana. --- Plan of analysis Latent class analyses are designed to empirically identify homogeneous subsets of a larger population . When solutions are clear, their utility is strongest if identified subgroups represent meaningful types or classes that have strong face validity, clear inter-group discrimination, and useful predictive validity . Here, confidence in generalizability beyond the specific sample is enhanced by replication of the observed classes across varied samples, as well as the observation of similar antecedents and correlates. All analyses were conducted within Mplus, Version 5.2 , and used a maximum likelihood estimator that is robust to nonnormality. All multigroup model comparisons involving non-categorical dependent variables were based on χ 2 difference tests , and those involving categorical dependent variables were based on log likelihood ratio tests . --- Results --- Latent class selection and description In latent class analyses of both the 80s and 90s cohorts, a four-class solution was selected as the best fitting model. The optimal number of latent classes was based upon three criteria as outlined by Muthén , Muthén and Muthén , and Nylund et al : the Bayesian information criterion statistic ; the classification quality, which can be determined by examining both the posterior probabilities and entropy values ; and solution interpretability . Because MTF oversamples drug users for longitudinal follow-up, sampling weights were used in the latent class analyses to best approximate population estimates. Preliminary analyses indicated that the inclusion of sampling weights did not alter the latent class structure . Collectively, the different decision criteria clearly indicated that the optimal number of latent classes for both cohorts was four . For both cohorts, the BIC values decreased steadily through the 4-class solution, but then increased between the 4-and 5-class solutions. For the 90s cohort, entropy increased through the 4-class solution but then decreased between the 4-and 5-class solutions, and though the average posterior probability decreased across the 1-through 5-class solutions, it decreased most sharply between the 4-class and 5-class solutions. For the 80s cohort, entropy increased through the 4-class solution but then decreased between the 4-and 5-class solutions, and the average posterior probability decreased most sharply between the 4-and 5-class solutions. Finally, as we cover in more detail below, for both the 80s and 90s cohorts, the 4-class solution resulted in a set of social role configurations that were both interpretable and practical. Examination of the frequencies of the six indicators of adult social roles across the 4-class solutions revealed strong similarities in the patterns observed across the two cohorts. The order of presentation of the classes is based on two considerations: a rough conceptual ordering from fewer adult roles to more adult roles and the ordering that emerged in the latent class analyses. As shown in Tables 1 and2, the first two classes inhabited fewer adult role statuses, but differed in which adult roles they were engaged in. Members of the first class, Educated Students without Children , were typically full-time students, had little involvement in full-time employment, and did not have children. Over fourfifths lived independently from parents, and about two-thirds were single. The level of education they had already achieved was the highest of all the classes. All participants in both cohorts who were classified as Working Singles Living with Parents reported living with their parents, almost all were single, and very few had children. Full-time employment was the norm, with small numbers in part-time education. Levels of education achieved were similar to sample levels in their respective cohorts. The remaining two classes were engaged in more adult roles than the first two classes, but also differed from each other in important ways. Among both classes nearly all were living independently from parents and were not enrolled in education. In other respects, they evidenced different patterns. Those described as Educated Workers without Children were working full-time, had no children, and had relatively high levels of achieved education. The proportions married, cohabiting, and single were similar to sample levels in both cohorts. In contrast the large majorities of those described as Married Workers with Children were married or cohabiting and had at least one child. The majority were working full-time, with an additional onequarter/one-sixth being full-time homemakers. This group had the lowest levels of achieved education in both cohorts, with the modal response being a high school education. --- Late adolescent predictors of early adult latent class membership The second objective was to examine age 18 demographic and educational predictors of class membership at age 25 to 26. To test for latent class differences in age 18 demographic and educational predictors, for each cohort the indicator of latent class membership was saved using the "Save data" command within Mplus, and used as a grouping variable in subsequent multigroup analyses. Results are shown in Tables 4 and5. Odds /]) are listed for all categorical age 18 demographic and educational predictors. In these and all subsequent tables, superscripted letters indicate rank order mean differences in each indicator across a cohort's four latent classes . Also in these and all subsequent tables, each significant complex or pair-wise group difference among the latent classes and the model comparison carried out to test for that difference share the same superscripted number. For example, in Table 4 the first notable group comparison "4>1,2,3", which compares the odds of being female for Married Workers with Children to the odds of being female of the other three latent classes, has a superscripted "1". The model comparison at the bottom of the table that shares the superscripted "1" = 40.05, p < .01) provides the results for this particular notable group comparison. Observed results were similar across the two cohorts with some notable differences. In terms of age 18 demographics, females in both cohorts were more likely than males to be classified as being Married Workers with Children. In the 80s cohort, participants who were not from two-parent families were more likely to be classified as Married Workers with Children or Working Singles Living with Parents than classified as Educated Students without Children or Educated Workers without Children, and in the 90s cohort, they were more likely to be classified as Married Workers with Children than any other group. Turning to the educational background predictors, in both cohorts, the four latent classes were ordered similarly on their levels of parent education, student high school GPA, and aspirations to pursue a college degree. In their late adolescence, those classified as Educated Students without Children in early adulthood had the highest levels of these three predictors, followed by Educated Workers without Children, Working Singles Living with Parents, and then Married Workers with Children, in that order. Plans to pursue trade school were ordered differently, that is, Married Workers with Children had the highest odds of having said they aimed to pursue trade school, compared to the other three groups. These age 18 rankings for education were similar to the age 25/26 levels of education attained by the four classes, that is, individuals with higher parental education, high school GPA, and college aspirations at age 18 tended to be members of classes who achieved more education by ages 25/26. This pattern was observed in both cohorts, except that the differences between adjacently ranked groups were more distinct in the 90s cohort. That is, more pairwise comparisons were significant than in the 80s cohort. --- Latent class differences in young adult correlates To test for latent class differences in young adult correlates, for each cohort we used the saved indicator of latent class membership as a grouping variable and conducted multiple-group analyses. Results are shown in Tables 6 and7, which list odds /]) for all categorical young adult correlates. In both cohorts, the two classes who were engaged in more adult social roles evidenced the highest economic independence as well as high life satisfaction. These groups reported annual cigarette use rates that were similar to the sample averages. In contrast, these two classes were the most dissimilar in terms of alcohol, marijuana, and other drug use. Specifically, the Educated Workers without Children class reported relatively high levels of use of these substances, and the Married Workers with Children class reported the lowest rates. The Educated Students without Children and Working Singles Living with Parents classes also evidenced similar patterns across cohorts in terms of their economic independence and life satisfaction. Educated Students without Children in both cohorts had the lowest levels of economic independence, yet their life satisfaction was as high as both faster track groups. Working Singles Living with Parents had medium economic independence but the lowest life satisfaction in both cohorts. Patterns of substance use were less consistent across cohorts for Educated Students without Children and Working Singles Living with Parents, except that the Working Singles had higher annual smoking prevalence and Educated Students had the lowest. Levels of alcohol, marijuana, and other drug use for these two classes tended to fall in between the relatively high drug use of the Educated Workers without Children class and the relatively low drug use of the Married Workers with Children. --- Discussion The transition to adulthood in recent decades has lengthened and diversified , not least in the United States which has great wealth but also high social inequality in income, health, and education coupled with relatively limited social safety nets in comparison with Europe . For example, in the US in 1984, the median age of first marriage was 25.4 for men and 23.0 for women and the mean age of mother at the birth of her first child was 23.5. In 1996 by contrast, the median age of first marriage was 27.1 for men and 24.8 for women and the mean age of mother at the birth of her first child was 24.6 . Across many nations, the passage to adulthood has become also become marked by more heterogeneity in the timing and content of social role changes. However, despite these changes and challenges, we identified four clear cross-sectional configurations of social roles that were remarkably similar in two cohorts separated by 12 years, as well as similar to those observed in two cohorts roughly matched in historical time in Britain and Finland . The patterns are also relatively consistent with prior studies using latent class analyses . Why would social roles tend to cluster together in cohesive patterns? First, at a particular point in time, the successful pursuit of specific combinations of social roles may be mutually facilitative or inhibitive. For example, the income earned through longer-term full-time employment makes residential independence more possible. The support of a stable romantic partnership may promote a desire or practical support to become a parent. Conversely, the demands of parenthood or full-time education may make employment more challenging, though a source of income would facilitate the pursuit of both these roles. Second, over time, as Bynner et al argued, prior achievements and already-adopted roles may impose constraints or at least mount challenges to the unbridled pursuit of new role adoptions and achievements. For example, in the mid-20s, it may be practically easier for a person with a college degree to return to full-time education than a person with a high school diploma, due to financial, academic, and other resources. In this way, betweenperson heterogeneity may increase over time. Educated Workers without Children and Married Workers with Children can both be characterized as having taken faster tracks to adult roles, in that they were not involved in education, they lived independently from parents, and they evidenced high levels of full-time employment or homemaker status. At the same time, the Educated Workers without Children class came from more advantaged family backgrounds and had achieved more higher education. Those in the Married Workers with Children class, as implied by their class label, were almost all married and had children by their mid-20s, whereas there was great diversity in the partnership status of the Educated Workers without Children group, almost none of whom were parents. Alcohol and marijuana use in the two groups were very different, as previously shown in research examining links between marriage and parenthood and reduced substance use during the 20s through midlife . Yet despite the many differences in the daily realities of their lives, the groups reported similar levels of life satisfaction and economic independence. As Settersten noted, "Young adulthood is a time when people begin to sort out commitments to persons, jobs, values, and ideologies. The realization that these commitments not only matter, but that they can actually be associated with happiness and life satisfaction even when they limit autonomy, is a symbol of maturity" . It would appear that these young adults may have found some fulfillment in these adult roles. The Educated Students without Children and the Working Singles Living with Parents could be seen as having taken slower tracks to adulthood. Again similarities and differences can be noted. Both groups tended to be single and few had children. However, Educated Students were typically in full-time education and tended to be more highly educated , while Working Singles were very seldom enrolled in any schooling and had more average levels of achieved education. Educated Students were likely to live away from parents while Working Singles lived in the parental home. Educated Students had the lowest levels of economic independence while Working Singles had medium economic independence. Life satisfaction of Educated Students was as high as both faster track groups, whereas Working Singles had the lowest levels of life satisfaction. Turning to substance use, Working Singles had higher annual smoking prevalence and Educated Students had the lowest; levels of other substance use for these two groups tended to fall in between the relatively high alcohol, marijuana, and other drug use of the Educated Workers without Children class and the relatively low use of these substances of the Married Workers with Children class. Overall, those classified as Educated Students were the most economically dependent, but it did not appear to negatively impact their life satisfaction, perhaps because their economic dependence is understood to be temporary and/or due to their own volition . Working Singles, who were the most likely to still live with their parents and who reported the lowest levels of life satisfaction, may have had the sense that they were stuck somewhere between adolescence and adulthood. Finally, we also compare the Educated Students without Children with the Educated Workers without Children classes. These groups came from more advantaged backgrounds, achieved higher levels of education, were very unlikely to be parents by their mid-20s, and had similar partnering profiles. Given the consistent documentation of higher rates of substance use among college students in the 18-22 year old age range , it may appear surprising that those who were working had higher rates of some substance use than the Student group. We suggest two possible explanations. First, we speculate a selection effect that draws attention to the age and relatively high achieved education level of the Students. Given the relatively small size of the Student group , we speculate that many of these may be more academically-inclined and currently means-limited graduate students, whereas those childless, single/cohabiting workers within the Educated Workers without Children group likely have greater time and resources to continue their youthful social life beyond their college years . Second, some prior comparisons of the substance use of students and non-students have not distinguished among the other social roles of the non-students, as we do here. What configuration of social roles could best be described as thriving? There is no immediately apparent answer to this question. On the basis of our subjective and most general indicator, self-reported life satisfaction, in both cohorts three groups report similarly high life satisfaction, distinguished only from the Working Singles Living with Parents who were least satisfied with their lives. These three more satisfied groups were very different in their combinations of social roles, suggesting that there is clearly more than one pattern of role configuration that may support satisfaction. These three groups were all actively engaged in some meaningful social roles or goal pursuits, with Educated Students pursuing higher education and the Educated Workers and Married Workers groups engaged in employment and/or parenting. Consistent with age norms for the mid-20s , these three more satisfied groups had higher rates of independent living, unlike the working singles who all lived with their parents. It is tempting to attribute the lower life satisfaction of this latter group to their lack of residential independence, but the specific mechanisms underlying such an association require further elaboration. It is unknown whether it is the circumstances that led to not leaving or to returning to the parental home that are challenging or the realities of sharing the home . --- Limitations and strengths Of course, our study does have some limitations. First, because the sample included only those who completed high school, generalizability to those who dropped out of high school is limited. High school dropout rates among 16-24 year olds in 1976-1977 and 1988-1989 ranged from 12.6% to 14.1% nationally, and differed by race/ethnicity ranging from roughly one-tenth among Whites, one-sixth to one-fifth among Blacks, and one-third among Hispanics . Individuals who leave school prior to high school graduation may combine roles in different ways and experience additional challenges. Second, role configurations at only one point in time were examined. Future research may fruitfully explore patterns of change in role configurations across the transition to adulthood. Third, role configurations may also differ in complex ways by gender and by additional demographic variables such as family social class background or race/ethnicity. For example some recent work has shown that gender differences in the timing of marriage and residential parenthood may underlie differences in the likelihood of inhabiting specific configurations . An important strength of this study is the use of panel data collected from two nationally representative cohorts of US 12 th grade high school students followed for an eight year period into the mid-20s, allowing us more confidence about the generalizability of the findings, as well as the sufficient coverage of time during which high school plans about the future unfold. The inclusion of two different senior year cohorts, both with the exact same sets of procedures and measures, allows us greater confidence about the replicability over history. Results were very similar for those who passed through their 20s 12 years apart, in the 1980s and the 1990s. For example, although there were historical differences in the timing of marriage between the cohorts, the combination of marriage with other social role changes was similar. As noted at the outset of this paper, scholars agree that the transition to adulthood has lengthened and diversified since the 1950s. In our comparison of these two cohorts who came of age in the 1980s and 1990s, however, we see clear similarities in the patterns and prevalence of how young people in their mid-20s combine social roles. --- Endnote i More detailed information about the MTF design and methods is available in Johnston et al and at www.monitoringthefuture.org

The decade following secondary school is pivotal in setting the stage for adulthood functioning and adjustment. We identify four social role configurations of early adults in their mid-20s using latent class analyses in two nationally representative samples of American youth in their last year of secondary education (modal age 18) who were followed longitudinally into adulthood (age 25/26). We focus on the big five social role domains of early adulthood: education, residential status, employment, cohabitation/marriage, and parenthood. Aims were to identify latent classes of social role configurations in early adulthood, examine demographic and late adolescent educational predictors of these classes, and explore contemporaneous health and adjustment correlates focusing on life satisfaction, economic independence, and substance use. Four classes with very similar characteristics and prevalence were identified in the two cohorts who were born 12 years apart: Educated Students without Children (8% in 80s cohort/9% in 90s cohort); Working Singles Living with Parents (16%/18%); Educated Workers without Children (45%/46%); and Married Workers with Children (31%/27%). Late adolescent demographic and educational variables and mid-20s variables were related to class membership. Results evidenced notable similarities (and some differences) across cohorts. Discussion focuses on how roles facilitate or inhibit each other and the potential diversity of optimal patterns of transitions to adulthood.

Farhat, Iannotti, & Caccavale, 2014; Lundberg et al., 2011; Reina et al., 2013) . Some variables have been studied as potential factors associated with body image perception among adolescents. Firstly, body image perception and its correlates tend to be designed and constructed differently between genders . Sociodemographic characteristics, such as low socioeconomic status and opportunities for health education, as well as limited resources and access to health services; behavioral factors such as physical activity, eating habits, sexual behaviors, alcohol, and tobacco use; and individual aspects, for example, nutritional status and self-rated health, have been associated with difficulties in dealing with perception of body image among adolescents . Most previous studies were conducted in developed countries , while studies involving adolescents from socially vulnerable areas are scarce. Areas with high social vulnerability are common in low-and middle-income countries; this vulnerability is usually considered when the Human Development Index is up to 0.699 . The HDI is a composite index ranging from 0 to 1 based on three dimensions of human development: a long and healthy life, access to knowledge, and a decent standard of living. It is the geometric mean of normalized indices for each of the three dimensions . Inhabitants in these areas tend to live in poor and highrisk conditions and have difficulty in accessing education and good nutrition . Thus, it is important to understand socioeconomic, behavioral, and individual variables that can have a negative impact on health indicators, for example, body image perception, in adolescents from these vulnerable areas. A study exploring body image perception among adolescents can provide information to aid in developing strategies to promote health and body image satisfaction among adolescents from vulnerable areas and alleviate risk conditions for mental health issues. Thus, the present study aimed to evaluate sociodemographic, behavioral, and individual factors associated with body image perception in a sample of adolescents from schools in low HDI areas in Brazil. --- Method --- Study Design and Participants This is a cross-sectional study using the baseline data from the "Fortalec ¸a sua Sau´de"/"Strengthen Your Health" project, a cluster-randomized controlled trial carried out in public schools in Fortaleza, northeastern Brazil . The National Research Ethics System approved the study protocol . Fortaleza is the capital of the state of Ceara ´, northeastern Brazil, and has a population of 2,452,185 inhabitants, which makes it the fifth largest city in Brazil in terms of population. This municipality has an area of 314 km 2 and an HDI of 0.754 . This study included a randomized sample of 1,182 students , aged 11-17 years. They represent 92.9% of the total students enrolled in Grades 7-9 from all six full-time public schools, representing one school in each administrative region of the municipality in Fortaleza, northeastern Brazil, during the 2014 school year. The neighborhood HDI where the six schools were located was 0.170, 0.215, 0.341, 0.377, 0.443, and 0.491 . The closer the HDI score to 1, the better the level of human development; and the closer to 0, the worse the level of development . Consent for students' participation was obtained from a parent/guardian by signing an informed consent form, according to the ethical considerations advocated in Brazil. A detailed description of the sample selection and methods can be found in a previous publication . --- Measures and Procedures Body image was evaluated using the nine-silhouettes scale . The ninesilhouettes scale included illustrations that range from extremely thin to extremely obese . Students were asked to select the figure that best represented their current body shape, being the one they "think they look" most like, and the figure that best represented their desired body shape, which would be the one they "wish they looked" most like. The body image perception score was estimated by calculating the difference between real and ideal body images. Positive scores indicated that the subject is dissatisfied with being heavier than ideal or a desire to reduce silhouette. On the other hand, negative scores indicate that the subject is dissatisfied with being lighter than ideal, demonstrating a desire to increase the silhouette. Scores equal to 0 indicated body image satisfaction. This measure has acceptable reliability . Sociodemographic, behavioral, and individual variables are potentially associated factors with body image and were evaluated using a self-reported instrument . The factors used were age-groups, economic class, physical activity level, frequency of condom use, eating habits, alcohol and tobacco use, body mass index , self-rated health, and healthy body image perception. Age was reported and organized into early, aged 11-14 years, and late adolescence, aged 15-17 years . Economic class was evaluated with the questionnaire from the Brazilian Association of Research Companies, which groups subjects into best and worst economic classes based on a score combining ownership of assets, parents' schooling, and number of employees in the household . Age and economic class have acceptable reliability . Physical activity was assessed using a list of 24 types of activities that were validated for Brazilian adolescents . The weekly volume of physical activity was calculated as the product of the weekly frequency and daily duration of each activity that they performed in a typical week. Thus, the physical activity level, classified as insufficiently active or sufficiently active, was estimated based on physical activity guidelines . Eating habits were evaluated using 2 items related to daily consumption of fresh fruit and soft drinks in a typical week . Two items on current tobacco and alcohol use in the month preceding the survey, classified as yes or no, and one question on condom use in the year preceding the survey, classified as always use/never had sex and sometimes use/never use, were considered . These questions had an ICC ranging from 0.71 to 0.99 . BMI was calculated and classified according to international cutoff points . Self-rated health was measured using the question, "Overall, how would you rate your health?," and the answer scale was dichotomized as very good/good or regular/bad. Finally, the nine-silhouettes scale was used to identify the students' perception of a healthy body image. The difference between real and healthy body image was calculated in order to identify whether students perceived their body image as healthy or unhealthy . These measures had an ICC higher than .80 . The dichotomization of the outcome is due to the focus of the study on the prevalence of adolescents' perception of body image and possible associated factors. This is the reason why dichotomization was necessary, since it allows identifying the subgroups of adolescents more prone to this outcome . Data collection was conducted in July 2014. The questionnaire was administered by evaluators to students in the classroom. Data were computed by scanning using the SPHYNX1 software , with correction for error and/or inconsistencies. --- Data Analysis Absolute and relative frequencies were used for description of categorical variables. The w 2 test for linear trends was used in order to compare variables between boys and girls. Multinomial logistic regression analysis was used to estimate odds ratio and a confidence interval of 95% CI in order to identify factors associated with body image in boys and girls separately. The dependent variable was body image perception . The inclusion of independent variables in the regression model was performed using a hierarchical approach , considering three levels: sociodemographic, behavioral, and individual variables. We initially used the adjustment of the first-level variables. Analyses of subsequent levels controlled for the variables from the same level and those from the previous level. The final significance level was 5% . All analyses were conducted using the SPSS 23.0 statistical software . --- Results The final sample included 609 boys and 573 girls. Higher proportions of students who had the best economic condition , who were sufficiently active , who reported condom use sometimes/never , and who reported current tobacco use were observed in boys compared to girls . Inversely, a higher proportion of overweight/obesity and negative health perception were observed in girls . All other variables were similar between boys and girls . Most boys and girls were dissatisfied with their body image mainly because they desired to increase their silhouette . There was no statistically significant difference between boys and girls for body image perception . Age-groups and economic class were not associated with body image dissatisfaction among boys . Boys who reported daily fruit consumption had a 43% lower chance of desiring to increase their silhouette in comparison to peers who did not eat fruit daily . All other behavioral variables were not associated with body image in boys. BMI status was an individual variable associated with body image dissatisfaction in boys . Overweight and obese boys had a higher OR of desiring to reduce their silhouette when compared to normal weight boys. Also, a lower OR of desiring to increase their silhouette was observed in overweight and obese boys in comparison to normal weight boys . Boys who perceived their body as an unhealthy body had higher OR of desiring to reduce or increase their silhouette in comparison to peers who perceived having a healthy body . In girls, sociodemographic and behavioral variables were not significantly associated with body image dissatisfaction . However, BMI status and healthy body image perception were significantly associated with body image dissatisfaction . Overweight and obese girls had a higher OR of desiring to reduce their silhouette in comparison to normal weight girls. Also, students from these groups had a lower OR of desiring to increase their silhouette compared to normal weight girls. Girls who perceived their body as an unhealthy body had a higher OR of desiring to reduce their silhouette or to increase their silhouette in comparison to peers who perceived they had a healthy body . --- Discussion The present study found that approximately 8 of the 10 students were dissatisfied with their body image, with similar estimates between boys and girls. A surprising result was that most students were dissatisfied with their body image because they desired to increase their silhouette . This dissatisfaction with body image is alarming due to adolescents' vulnerability for eating disorders such as anorexia and bulimia and can be explained by the desirable silhouette promoted by the media, peer standards, and models . Studies have revealed the presence of eating disorders in adolescents because of the influence from family, for example, strict dietary rules, exposure to media and nutritional problems, stunted growth and weight gain, and the induction of vomit or intake of laxatives as procedures to achieve the ideal body image . Elevated BMI was associated with body image dissatisfaction in both genders, which was confirmed in previous a Odds ratio , confidence interval of 95%, and p value were estimated using multinomial logistic regression. The boldface values above show the significance p < 0.05. studies . This dissatisfaction may be related to a constant "thin and beautiful" body reference used in the media, which can affect how adolescents view themselves, and thus suffer from the influence of the society of women must be thin, men must be strong . Media and social dissemination of obesity-related health problems is important in health promotion and disease prevention in public policies, but when not satisfactorily achieved by the individual, can increase body image dissatisfaction. The present study found a strong association between the perception of a healthy body and body image among boys and girls. As mentioned earlier, the cognitive process of building a concept of "ideal" body image and its discrepancies with the "real" body image can be influenced by social relationships and impositions. The adolescent's concept of a "healthy" body image is not different , and dissatisfaction with healthy body image evaluation seems to be closely related to body image dissatisfaction. This evidence is important because health education actions can promote the understanding that patterns of a "healthy body" are not corresponding to a holistic definition of health including social and mental aspects; combating body image dissatisfaction in adolescence seems to be important. This study showed no association between body image and sociodemographic variables, which corroborates previous studies . Similarly, most behavioral variables were not associated with body image perception among boys and girls in the present study, which is different from other studies that observed a negative influence of risk behaviors, for example, physical inactivity, tobacco use, and alcohol use, on body image perception in adolescents . These differences may be due to the peculiarity of the sample in terms of behavior, for example, most adolescents did not use tobacco or initiate sexual behaviors, and socioeconomic aspects, as most adolescents were from a low economic class and lived in low HDI areas. Daily consumption of fruit was the only behavioral factor significantly and positively associated with body image perception in this study, but only among boys, that is, daily fruit consumption had a 43% lower chance of desiring to increase their silhouette in comparison to peers who did not eat fruit daily. This result shows that healthy eating habits can be important to both physical and psychological health including body image satisfaction . --- Strengths and Limitations To our knowledge, this was one of the first studies to focus on evaluating body image dissatisfaction and associated factors in a sample of adolescents from low HDI areas. Still, the composition of potentially associated factors including sociodemographic, behavioral, and individual variables was important to show how different variables are associated with body image perception; this can guide intervention strategies focused on body image satisfaction among young people. One limitation was the self-report measurement of some variables, especially physical activity and eating habits. Although reliable instruments were used , there are possible recall limitations for these items. Also, although the measure of body image using the nine-silhouettes scale is acceptable , there was no breakdown of dissatisfaction with body image for body parts, which could indicate different results, for example, the relationship between physical activity and body image is significant for specific body parts. In addition, due to its cross-sectional nature, it was possible to examine associations between risk factors, that is, sociodemographic, behavioral, and individual variables, and body image perception. However, it's not possible to determine their causal relationships. Finally, another limitation of this study is that it was performed locally. --- Implications for Nursing Research and Practice Education and health care focused on body image must pay special attention to young people from vulnerable areas with high BMI and focus on strategies that enable improving the perception of a healthy body and a healthy eating, especially boys. These results suggest attention to the perception of healthy body image evaluation and eating behaviors during educational activities in the family, as well as consultations or home visits, and school contexts with health education lessons. The results of this study can help health professionals, sectors, and institutions to identify which individual and social aspects can be the focus of effective strategies for health promotion and education among adolescents. --- Authors' Note Tha ´byta Silva de Araujo contributed to conception, design, acquisition of data, analysis, and interpretation of data; drafted the article; and gave final approval of the version to be published. Valter Cordeiro Barbosa Filho contributed to analysis and interpretation of data, critically revised the manuscript for important intellectual content, and gave final approval of the version to be published. Fabiane do Amaral Gubert gave final approval of the version to be published. Paulo Ce ´sar de Almeida critically revised the manuscript for important intellectual content and gave final approval of the version to be published. Mariana Cavalcante Martins critically revised the manuscript for important intellectual content and gave final approval of the version to be published. Queliane Gomes da Silva Carvalho critically revised the manuscript for important intellectual content and gave final approval of the version to be published. Ana Cristina Pereira de Jesus Costa critically revised the manuscript for important intellectual content and gave final approval of the version to be published. Neiva Francenely Cunha Vieira contributed to analysis and interpretation of data, critically revised the manuscript for important intellectual content, and gave final approval of the version to be published. --- Declaration of Conflicting Interests The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. --- Author Biographies

This study aimed to evaluate sociodemographic, behavioral, and individual factors associated with body image perception in a sample of adolescents from schools in low Human Development Index areas in Brazil. This cross-sectional study included 609 boys and 573 girls (aged 11-17 years). Body image perception (nine-silhouettes scale) and sociodemographic, behavioral, and individual variables were included. Multinomial logistic regression analysis was used. Most boys (76.9%) and girls (77.5%) were dissatisfied with their body image. Body mass index status and healthy body image evaluation were significantly associated with body image dissatisfaction in both boys and girls (p < .001), and daily fruit consumption was associated with body image dissatisfaction only in boys (p ¼ .035). Education and health care focused on body image can pay special attention to young people from vulnerable areas with unhealthy nutritional status and focus on strategies that enable improving the perception of a healthy body and a healthy diet.

Introduction Social media has deeply penetrated people's lives in many aspects. In developing and developed societies, social media has played a significant role in health management and disease control [1]. Social media is integrated into empirical examinations of the prevention and control of various types of diseases, including emerging, infectious, and chronic diseases [2][3][4]. Social media has been employed to study health phenomena among different populations or social groups such as children, pregnant women, and older adults [5,6]; different genders [7,8]; and individuals in various social classes [9]. Agencies widely use social media to fulfill different health purposes. For the general public, social media is used to satisfy its orientation for health information, linking with health services and communication with others who share the same health interests [10]. For public health professionals and organizations, social media serves as a multifunctional tool to launch interventions to reach a wide array of the population efficiently [11][12][13][14]. The large volume of mobility and discourse data on social media can be conducive for public health management, including disease surveillance, assessment, and control [15][16][17]. These studies have demonstrated the increase of scholarly interest in empirical research conducted on social media platforms with public health goals, including the social media-based public health research in this study [18][19][20][21]. Although many scholars in social science and public health have contributed to this field, the overview about how social media has been integrated into public health research remains limited. Prior systematic reviews on social media-based public health research often focused on certain domains or topics. Many reviews systematically investigated the effectiveness of social media interventions for varied specific health outcomes, such as the promotion of safe sexual health behaviors [22], vaccine uptake [23], noncommunicable disease management [24], and HIV prevention [19]. Scholars are often dedicated to one or two domains, neglecting the fact that using social media in one field may shed light on another. Meanwhile, focusing on one particular area, these reviews often face challenges to identify similar patterns across domains and capture an integrated picture about social media-based public health research. In addition, most existing reviews included a limited number of original articles [25,26]. Even a review of systematic reviews only extracted few studies [18]. In this emerging and fast-growing subject field, the limited literature being included may fail to provide a panoramic description of social media-based public health research. Furthermore, most prior systematic reviews have adopted a top-down approach and therefore may have narrowed the view by overlooking certain nuances and novelties that have been emerging. This study adopts a bottom-up approach [27] to understand the growth of social media-based public health research and remain open to map the intellectual landscape in this area. Specifically, the study aims to address the following research questions: --- Methods --- Data Collection To examine how social media has been adopted and integrated into public health research, a list of terms was identified and the Web of Science and PubMed databases were searched . Lists of keywords about social media and disease were established. This study focused on emerging, infectious, and chronic diseases. Specifically, 14 diseases were selected that are of high prevalence among the population or pose major public health threats according to the World Health Organization [28]: influenza, HIV, hepatitis A, hepatitis B, hepatitis E, dengue, Ebola, Middle East respiratory syndrome , asthma, diabetes, obesity, cancer, oral disease, and alcohol use. A list of keywords was constructed. Then a list of social media keywords, including the general social media categories and specific social media platforms, was established. The lists of disease and social media keywords were combined pairwise and submitted to search titles, abstracts, and keywords of published studies in the databases of the Web of Science and PubMed. The publication period was limited to between 2000 and 2018. The article language was limited to English, and document type was limited to scholarly journal articles. Considering the PubMed database includes numerous medical and clinical studies that are beyond the research scope of this study, the search results were refined by setting the broad subject terms as related categories, such as public health and medical informatics to reduce the noise in the search results. This search strategy led to the identification of 5271 articles in the two databases. Then, another round of data checking to remove unqualified studies such as non-social media-related articles, duplicate records, and clinical studies was implemented. Finally, a dataset of 3419 articles was collected for further analysis. Ultimately, document level information of the 3419 articles from the databases, including authors, article title, journal title, abstract, author keywords, and cited references, was retrieved. --- Data Analysis An automatic text-mining approach was adopted to extract research themes in the field of social media-based public health research. As the abstracts of published studies conveyed the themes or foci of the articles [29], the article abstracts were mined through latent Dirichlet allocation topic modeling, which is a popular unsupervised text-mining technique in computational social science. LDA topic modeling helps recognize the structure of research development, current trends, and interdisciplinary landscapes of research [27]. The LDA topic modeling [30] was implemented with the tm package in the R software . Data preprocessing, such as removal of stop words and numbers, was performed before the LDA topic modeling. Numerous LDA topic models were estimated with various numbers of topics. These models were evaluated on the basis of three main criteria: a substantial proportion of articles exists under each topic, themes show independence with one another and the lists of top terms of topics are not highly overlapped or not relevant, models differing in theme number are compared to identify the nuanced differences and determine the best theme extraction by assuring that each term list is coherent. Finally, a topic model with 25 topics, which presented adequate discrimination between topics and convergence within a topic, was selected. The articles were classified into the research theme with which they had the greatest probability scores. To understand how social media has been integrated into public health research with different thematic foci, a manual content analysis was conducted among a randomly selected sample of 500 abstracts to understand the role of social media. A coding scheme was developed by two authors of this study. The two authors first separately coded a subsample of 60 randomly selected abstracts to construct the coding scheme. After several rounds of exploration and discussion, they achieved a satisfying intercoder reliability . The role of social media is categorized into two main types. First, social media provides a substantial research interest for public health research, which includes the use of social media for intervention, as human-computer interaction characteristics, as platforms of social influence, and for risk assessment or disease prevention. Second, social media is employed as a context in public health research with social media as mere reference, as participant recruitment tool, or as data source. Categories and their definitions are further illustrated in the analytical findings section . Finally, the research methods adopted by social media-based public health research were identified by searching a list of keywords associated with various research methods among the titles, abstracts, and keywords of the retrieved studies. Figure 1 summarizes the study workflow. --- Results --- Publication Trends in Social Media-based Public Health Research Publication trends in the field of social media-based public health research in the past two decades were presented in 3 dimensions: growth of overall publications, growth of publications by specific diseases, and growth of journal outlets. Empirical studies in this area were relatively limited in the first decade , which demonstrated a minimal increase as shown in Figure 2. A significant annual increase was observed from 2011 to 2018. Such trends intersected with the advancement of the internet, especially that of social media. Although popular social media platforms, such as Facebook, Twitter, and Instagram, were launched before 2010, they have been widely accepted worldwide since 2010. This implies that social media-based public health research is a study area responsive to technological development. Prior research also demonstrated similar findings that internet research evolved along with technological development [29]. Social media has been increasingly incorporated into the studies of certain types of diseases in the past decades. Dramatic increases in research occurred on cancer, HIV, diabetes, obesity, and alcohol use after 2010. Other diseases, such as influenza, hepatitis A, hepatitis B, dengue, Ebola, MERS, asthma, and oral disease, showed a relatively slow growth rate that remained quite stable from 2000 to 2018. For the journal outlets, a total of 799 journals published studies in these areas . Table 1 reports the 15 most visible journals in this area. Among them, Journal of Medical Internet Research, which published 331 articles, is the most visible one accounting for 9.68% of the total publications. Moreover, JMIR sister journals, such as JMIR mHealth and uHealth , JMIR Research Protocols , and JMIR Public Health and Surveillance also showed great interest in this domain. Other high-ranked journals included PLoS One, BMC Public Health, Studies in Health Technology and Informatics, AIDS and Behavior, and BMJ Open, each occupy more than 1.5% of publication in this field. --- Research Themes in Social Media-Based Public Health Research The 25 extracted research themes were labeled on the basis of the top 15 most frequently used terms associated with each theme and the articles assigned to the theme. Table 2 presents the lists of terms under each theme. A network graph of theme-word probability of the 25 research themes is provided in Multimedia Appendix 2. Multimedia Appendix 3 displays a typical study under each theme. The percentages in Table 2 reveal the article distribution across research themes. The articles under each theme varied greatly from 2.22% to 7.93% , with men and HIV occupying the largest number of articles and reproductive cancers the least. Among them, the mHealth family, the themes about mHealth , contained a large body of 653 articles. Themes about substance use comprised 409 articles. Another big cluster was cancer , which consisted of 385 articles. The 25 research themes were further grouped into 6 research clusters on the basis of similar concerns and associations. The first cluster was on health education, which comprises 4 themes: health education-school and students, health education-family and oral/dental health, mHealth and medical decisions, and pregnancy. Health education aims to prevent diseases through improving people's knowledge and health efficacy. School and family, as the main scenes for the students to learn health beliefs and behaviors, have been the foci of health education. Sexual health education on condom use and pregnancy have also attracted increasing scholarly attention. The second cluster was on health management with the help of mHealth. The themes mHealth and weight control, mHealth and diabetes management, digital campaigns in targeted populations, social media and alcohol drinking, substance usage and cessation, food and asthma, and vaccination and immunization all fell into this cluster. This indicates the functional attributes of social media to help manage health problems. Social media use is used to intervene in certain unhealthy behaviors and promote healthy behavior adoption. The third cluster, cancer studies, includes women's cancer, reproductive cancer, cancer survivor, and caregiving on social media. Cancer is one of the world's largest health problems and a significant cause of death. Thus, continuous attention has been paid to cancer studies. The fourth cluster, infectious diseases, includes HIV as a key topic: mHealth and HIV; men and HIV; and infectious disease, health campaign, and stigma belong to this cluster. In this line of research, social media provides breakthrough channels to reach risky subgroups and focuses more attention on campaigns to reduce the stigma surrounding infectious diseases. The fifth cluster was on mental health issues. This cluster consists of two themes: mental health and substance use and mental health-depression and digital technology. Mental health problems have been prominent in modern society. Digital technology is considered a cause and a solution to mental health issues. The sixth cluster was on extended health research empowered by social media: health and human mobility, health marketing, surveillance and Twitter, and eHealth-miscellaneous. These research areas have flourished due to the availability of geographical information, mass user behavioral data, and extensive online discourse on social media platforms. --- Roles of Social Media in Public Health Research --- Social Media as Research Context or Substantial Interest Social media is integrated into public health research by providing a new research context or producing new substantial interest in public health research. When social media was adopted as a research context, social media was specifically considered as a mere reference, a platform for participant recruitment, and as a data source. When social media was adopted as a mere reference, research mostly used social media as a tool to offer intervention and facilitate the health management of individuals. For the role of a platform for recruitment, research either recruited participants through distributing questionnaires or posting participant recruitment announcements on social media or employed users of certain social media platforms as the study target group . For the role of data source, social media could contribute to collecting data in text, image, video, and app interface formats and collecting published posts and articles for meta-analysis or scope review. When social media produced substantial interests for public health research, social media was used for intervention; employed to study human-computer interaction characteristics; used as a platform of social influence; and used for disease surveillance, risk assessment, or prevention. Under these 4 broad categories, the role of social media is described as follows: --- Intervention For public health intervention, the 4 subroles of social media in the published studies are as follows: interactive intervention tool targeted at changing personal and environmental risky health factors, intervention information-distributing tool , source for health information seeking, such as YouTube and other platforms, and usability test of social media platforms as intervention instruments. --- Human-Computer Interaction Characteristics Under this role, social media was used to serve the goal of revealing the public's attitudes toward technology and social media for health use, characteristics and behaviors of social media users and groups, factors affecting the health behaviors or attitudes of users on social media, and consequences/influences on health behaviors caused by social media. --- Platform of Social Influence Social networking and interaction between different individuals and groups on social media could facilitate the change of health behaviors through the following approaches: building online groups for patients, such as cancer patient groups on Facebook, promoting physician-patient communication or information seeker-provider communication, enhancing health-related marketing, such as precision advertising, and changing public health behavior at a macro level. All of these approaches are representations of social influence in online communities. --- Disease Surveillance, Risk Assessment, or Prevention The digital traces of online behaviors and massive online discourse granted opportunities to understand health conditions at a population level. For instance, Google trends and search query records could grant references to predict the possibility of a flu outbreak at an early stage. Figure 4 presents the percentages of articles in each type of social media role. The results showed that among substantial interest, "an interactive intervention tool targeted at changing personal and environmental risky health factors" accounted for the largest percentage of 19.2%, followed by "usability test of social media platforms as intervention instruments" , and "a source for health information seeking" . Four of other types, "characteristics and behaviors of social media users and groups" , "consequences/influences on health behaviors caused by social media" , "enhancing health-related marketing" , and "changing public health behavior as macro influence" , also occupied a relatively larger proportion of more than 6%. Among the dimension of social media as research context , "as a mere reference," which took social media as a research background or a research environment, played the dominant role . The second most frequent role that social media plays was content data source . The other three were "social media as article search platform for meta-analysis or literature review" , "for participant recruitment" , and "as platforms to recruit their users" . --- Research Methods in Social Media-Based Public Health Research Public health research with social media data was dominated by traditional quantitative research methods, whereas cutting edge computational methods played a minor role. Among all the articles, 30.6% employed survey method, 24.0% employed experiment design, 22.7% employed qualitative methods , 8.3% included employed digital methods , and 5.6% employed traditional content analysis. Figure 6 demonstrates that the method distributions under the 25 research themes were similar to the general distribution among the whole body of the studies. Survey and experiment were the two most adopted methods, whereas review article number was relatively small among all themes. --- Discussion Principal Findings --- When Public Health Research Meets Social Media: From New Phenomena to New Questions Social media has infiltrated almost all health-related processes and domains with the rapid advancement of social and mobile media. This dramatic change has entailed many new phenomena to be explored in public health research. The findings of this study are consistent with previous studies that found that almost one-third of internet studies have focused on eHealth and mHealth since 2009 [29] and a trend toward digitization exists in health care [31]. Many traditional public health activities, such as health education, health promotion, and disease surveillance, have taken advantage of social media technologies to become digitized [32][33][34]. Social media has substantially altered how individuals seek and share health information, discuss health issues, and engage in health behaviors [35]. Social media also provides innovative ways to change health behaviors in various domains, such as smoking cessation, substance use, weight control, HIV prevention, and cancer screening [36][37][38]. Consistent with previous reviews on social media and public health studies, this study concludes that social media contributes to these public health domains by broadening the reach of health education, providing accessible online professional consultation, and improving the efficacy of access to care and medication uptake, etc [19,39]. Moreover, an upward trend of integrating social media in various public health campaigns exists due to the instrumental benefits of social media technologies, such as lower intervention cost, higher user engagement, higher efficiency, and better documentation of the process [40]. When public health research meets social media, new topics have emerged and attracted the attention of public health scholars [41]: mHealth and social media-empowered health research. "Digital campaigns in targeted populations" and "surveillance and Twitter" are typical new topics where researchers frequently examine new research questions [42,43]. For example, researchers discuss how to employ user-generated content together with geolocation information to predict an outbreak of an emerging disease or visually map their diffusion routes and locate the risky population [44]. The digital trace on social and mobile media offers many possibilities to study online health behaviors such as online health information-seeking, online social support, and online medical consultation behaviors [45]. In addition, some health topics have attracted burgeoning attention in the era of social media. For instance, mental health problems have been identified as significant concerns among the 25 themes in this study. However, no conclusion has yet been reached whether and how the adoption and use of social media alleviates or exacerbates mental health problems [46]. --- When Public Health Research Meets Social Media: Unequal Status With Detached Concerns? Social media-based public health research lies in the crossroad between public health studies and social science studies on information and communication technologies [60] and benefits from both perspectives. In the cooperative process, social media-based public health research reaches various levels in elaborating on the two perspectives. Taking the initial perspective of public health interest, many acceptability studies and randomized controlled trials have been documented to examine the effectiveness of social media to reach different public health goals [40,61]. In these studies, social media is often considered a new functional tool to improve public health. Meanwhile, in research that further examines the influence of ICTs on public health, who used what social media content targeted at whom through which social media platforms with what health effects is the core concern [62]. In this line of research, studies typically focus on the transmission of health information, communication between health agencies, the uses of health apps, and so on [63]. The inherent concerns of these studies seem to be detached though not in conflict in that social media facilitates the public health promotion process, and public health outcomes add value to the communication through social media. From an overview of social media-based public health research, the dominating approach of these published studies considers public health issues as the substantial interests and ultimate outcomes rather than regard social media as an equally important area of concern. Many articles used limited space to describe the use of social media in health promotion campaigns or projects [64,65]. The subordinate role of social media suggests that the potential of ICTs has not been fully realized in the domain of public health [41]. Empirical studies should not only focus on what social media can contribute to public health research but should also examine how and why social media can make an impact in various contexts of public health research. This can substantially improve the understanding of the intended as well as unintended consequences social media can exert on health attitudes and behaviors. This can also enable public health researchers to integrate social media into their research design further. --- Limitations Despite the strengths and contributions, this study has certain limitations. First, the study may suffer from the file drawer effect given that only studies indexed in Web of Science and PubMed were included. Empirical studies published in other outlets were not considered here. Future studies are warranted to expand the pools to conference proceedings and articles indexed in other databases. Second, this study used numerous diseases as search terms in the initial search, but the list remains incomplete. Some important diseases, such as mental disorders, were not incorporated. Despite this, the topic modeling captured mental health as a major theme. Further research is suggested to include mental health keywords as search terms. Third, LDA topic modeling is a well-recognized method to identify related themes through document-word matrices. However, the results of the topic modeling were not as neat as expected. No standard and quantitative thresholds exist for researchers to choose the optimal number of topics. Future studies are encouraged to replicate this study and examine the reliability of such themes. --- Conclusions This --- Conflicts of Interest None declared. --- Multimedia Appendix 1 Social media roles and definitions. [DOCX File , 20 KB-Multimedia Appendix 1] --- Multimedia Appendix 2

Background: Social media has substantially changed how people confront health issues. However, a comprehensive understanding of how social media has altered the foci and methods in public health research remains lacking. Objective: This study aims to examine research themes, the role of social media, and research methods in social media-based public health research published from 2000 to 2018. Methods: A dataset of 3419 valid studies was developed by searching a list of relevant keywords in the Web of Science and PubMed databases. In addition, this study employs an unsupervised text-mining technique and topic modeling to extract research themes of the published studies. Moreover, the role of social media and research methods adopted in those studies were analyzed. Results: This study identifies 25 research themes, covering different diseases, various population groups, physical and mental health, and other significant issues. Social media assumes two major roles in public health research: produce substantial research interest for public health research and furnish a research context for public health research. Social media provides substantial research interest for public health research when used for health intervention, human-computer interaction, as a platform of social influence, and for disease surveillance, risk assessment, or prevention. Social media acts as a research context for public health research when it is mere reference, used as a platform to recruit participants, and as a platform for data collection. While both qualitative and quantitative methods are frequently used in this emerging area, cutting edge computational methods play a marginal role. Conclusions: Social media enables scholars to study new phenomena and propose new research questions in public health research. Meanwhile, the methodological potential of social media in public health research needs to be further explored.

Introduction Continuous improvements in wireless technologies and mobile devices have extended the reach of digital systems . This has allowed organisations across a range of domains to explore new products, services and user/consumer channels . One of the most profound changes has been the use of these technologies in mobile health initiatives; initiatives that extend health services to poorer areas that are often neglected by traditional systems, notably people living in rural areas of developing countries . Mobile technologies have diffused through developing countries far quicker than many expected . This is largely because mobile technologies do not require the same level of individual investment or supporting infrastructure . Mobile technologies do more than allowing users to make calls and receive multimedia messages; they also provide access to the web. The use of mobile technologies has become engrained in the day-to-day life of many people in developing countries . Thus, it is logical that healthcare delivery should build upon these technologies to extend care services into isolated rural areas of developing countries . This lessens the burden on individuals to travel to urban centres for care and provides an information channel between healthcare workers in rural and urban centres . Despite the potential of new mHealth systems, they seem to be stuck in a pattern of successive pilot studies that struggle for mainstream implementation . The reasons for the underwhelming implementation of pilot systems are typically because designers do not understand the full situation outside the context of the pilot, which can result in unforeseen incompatibilities with existing technical infrastructure or with the existing social norms and practices . This suggests these pilot studies are developed within a "bubble" that hides some of the complexities and interdependencies of the true environment. Indeed, local contexts are often represented as blank canvases upon which mHealth should be introduced without any great difficulty . Frequently, local perceptions and experiences are minimised to matters of user acceptance, usability and feasibility of a specific mHealth technology or "app", as part of a broader investigation of local barriers to the adoption of such an intervention . Yet, wider complexities and interdependencies are key to the eventual integration of new technologies and practices ). Thus, this study asks what existing health-related structures, properties and practices are presented by rural areas of developing countries that might inhibit the implementation of mHealth initiatives? This study addresses this question using a socio-material perspective. Socio-materiality has been heralded as a useful approach to extend traditional theorising and offer new and interesting insights into emergent practices in complex sociotechnical systems ). We apply this perspective to explore a rural area of Nigeria in the early stage of exposure to new mHealth technologies. An immersive study of this context demonstrates how existing material and social agencies may impact the implementation of mHealth in this area. The rest of the article is structured as follows. The next section reviews existing literature on mHealth in rural areas of developing countries, showing a lack of holistic socio-technical analysis and implementation-focus. We then introduce sociomateriality and discuss the unique analytical perspective it affords. Following this, the research methodology is outlined based on an exploratory case study in Enugu State, Nigeria. Finally, a thematic description of the findings is presented under the analytical headings of social, material, practice and imbrication, following the key concepts which inform the ITP 35,8 socio-materiality lens deployed, that is, the substantialist point of view by Leonardi ). The article concludes with discussions and summary. mHealth in rural areas of developing countries We began by performing a systematic literature review to examine the different streams of research that are common for mHealth in developing countries. This systematic review searched each of the leading academic publication sources that typically publish mHealth research, namely the AIS Electronic Library ; Science Direct and Web Science; JSTOR; Academic Search Complete and Scopus; OCLC FirstSearch; Google Scholar; and PubMed/MEDLINE. These publication sources were searched using an evolving set of search terms relating to mobile healthcare, specifically "mHealth", "m-Health", "mHealth Care", "mHealthcare", "Mobile Health Care" and "Mobile Healthcare". This resulted in an initial set of 329 papers. Papers published before 2010 were then excluded , as the rapid evolution of mobile technologies makes it difficult to compare devices before this period. We further removed studies that did not focus on developing countries , that focussed on non-health goals , that were not peer-reviewed , that used obscure or non-mobile technologies and that were not written in English . This presented a final set of 108 papers. The sampled research focussed on three dominant streams. The first stream focusses on the improvements in healthcare enabled by mHealth tools. In rural communities, local healthcare workers are often the first and only point of contact with the healthcare system for community members . Therefore, the ability to improve healthcare interactions between rural healthcare workers and community members is crucial . Several studies have focussed on general improvements in scope, efficiency and quality . Other studies have focussed on training for rural healthcare workers and balancing new tools with competing demands for attention and multiple priorities . The second stream is more technology-focussed, highlighting the ability of different individuals to make sense of modern technologies at an interaction and interface level. Several studies have focussed on usability and the need to design mHealth interfaces that can be used as easily and effectively as possible . Other studies have taken a slightly different approach, focussing on the reduction of errors, particularly data recording and data entry . The third stream focusses on the process of change management around the introduction of new mHealth processes. Examples include remote clinical check-ups , remote tracking of treatment and medication adherence , remote dissemination of health information for chronic diseases , remote assistance in the treatment of patients with mental disorders and participatory community healthcare reporting . These studies provide many valuable insights for mHealth in developing countries. Yet, they are collectively characterised by one noteworthy trend. Of the 108 sampled articles, 13 were purely conceptual in nature, that is, they did not actually develop or evaluate tools, and 78 relied on controlled pilot deployments for evaluation, that is, there was no evaluation of widespread implementation, ex ante or ex post. Only 17 developed and evaluated a mHealth tool in a study that did not limit participation to sampled users and areas . Thus, the vast majority of mHealth studies chose to avoid engaging with contextual issues that might A sociomaterial exploration of mHealth impact the eventual implementation, meaning the roll-out of these tools likely faces significant unresolved challenges. As an example of a valuable and promising pilot study that requires further contextual consideration, consider Littman-Quinn et al. . Those authors developed a system that coupled a camera-enabled mobile phone and an application called ClickDiagnostics. This system was used in Botswana to send digital referrals from remote areas to a specialist in a central hospital, so connecting people in resource-poor areas with remote specialists. Yet the ability to scale this system is not obvious, as the number and capacity of those specialists are limited. Neither is the long-term effect of likely delays obvious for the willingness of remote participants to send pictures. Another example was a study by Knoble and Bhusal that developed a diagnostic application called e-algo to aid remote clinical diagnosis in Nepal. Analysis of that project suggested patients were more confident when healthcare workers used e-algo in their patient care. This raises questions about the shifting power dynamics embedded in this new system and the potential delegitimising of healthcare workers. Such emerging changes often produce shifting balances of resources that can transform early enthusiasm into subsequent resistance . As a third example, Ngabo et al. developed a mobile phone SMS-based system known as RapidSMS-MCH. This system allowed community health workers to track maternal and child health records remotely in their community in Musanze, Northern Rwanda. This system has obvious practical benefits, yet also introduces the potential for intrusive monitoring practices, not only among patients but also among local healthcare workers administering treatment. Transparency is notoriously challenging to balance with the sensibilities of healthcare workers when implementing systems, many of who distrust scrutiny from others outside their profession or context . This suggests that significant contextual barriers to implementation may remain for the RapidSMS-MCH system. A socio-material view of mHealth in developing countries Socio-materiality [1] proposes that technology, people and process are mutually generative and interdependent in practice . Socio-materiality argues the use of material artefacts is shaped by social processes, understood and used within a social context, and social action is made possible by the use of material artefacts ). Socio-materiality therefore describes what happens when humans and things interact in practice without ignoring the impact of either of them on one another . This allows sociomateriality to overcome the shortcomings associated with treating the social at the expense of the material or vice versa . Conceptually, socio-materiality prefers to discuss "materiality" rather than "technology", since the latter creates the impression there are some objects, artefacts or devices out there that independently do things, and ignores that these objects, artefacts or devices only come to reality when manifested in practice . Instead, "materiality" is understood to be the fashioning of physical or digital materials into useful forms that endure across time and space . This means that we cannot drop specific types of hardware or software into different contexts and expect it to behave similarly . Equally, we cannot expect software or hardware to be completely passive and amorphous. Rather, each object contains some essential building blocks of eventual form. This quality is referred to as material agency, that is, "the way the object acts when humans provoke it" . Material agency is therefore a construction that relies partly on materiality and partly on a user's perceptions of whether that materiality affords the capabilities or constraints needed to achieve some objectives ). Thus, the material agency of some artefact affords a wide range of potential uses and actions, the nature of which depends on the context in which it is enacted ). Socio-materiality also prefers the term "social" to "people", as the former is better equipped to capture the variety of social structures involved in a system, including individuals, groups, institutions, norms and perceptions . Social agency therefore describes how diverse social actors interact differently with varied material artefacts, as they align these artefacts with different institutional structures and environmental properties . Just as artefacts have some material agency that affords some possible uses and actions, so human actors have social agency that identifies and adapts uses and actions to take advantage of those possibilities . Following this, as humans pursue different goals, they perceive technology as affording different possibilities or limitations depending on how they are enabled or constrained by the social agencies of their surrounding context ). The next important concept for socio-materiality is the concept of practice, which refers to the space in which the social and the material come together as a performance ). These practices are part of a socially fashioned ecosystem in which multiple interdependent practices are co-operatively "negotiated" . Therefore, practice is a collective activity. While social and material agencies exist in a way that transcends any one specific use or goal, it is in practice where abstract possibilities and these two agencies become realised . Thus, observable systems are considered systems of practices, co-constituted by the available social and material agencies under pressure to meet various needs and produce desired outcomes . This means designers must understand the range of practices if they are to understand the possibilities afforded by relevant social and material agencies and vice versa. Typical mHealth systems include practices that non-native developers may find intuitive, for example, end-user training , and also some that may not be obvious without local knowledge, for example, the sharing of phones and SIM cards among multiple rural users . Hence, the social and material agencies in local systems may be only partially understood unless a more extensive exploration of practices is performed. Socio-materiality brings together these notions of practice, social agency and material agency with the concept of imbrication. Imbrication refers to the emergence of structure and routine over time, as systems gradually accumulate practices and social and material agencies become more tightly intertwined . Imbrication explains the process of "organisation and technology mutually shaping nature", in which "the structure between individuals . . ., and technologies . . . evolve as a socio-material creation" . This is important to make sense of decreasing flexibility and adaptation of core structures over time, as layers of imbrication embed layers of interdependent practices and expectations into a system . Disrupting these embedded practices usually requires some key changes in the technologies, individuals or expectations that constitute a system, at which time new material and social agencies are introduced that must be reconciled with prevailing structures . An example of imbrication in mHealth for developing countries is how historic social agencies for confidence and cooperation have been enacted in practice by material agencies for service reliability and efficiency. These result in varying continuous use intentions that constrain where and how a mHealth system will be used in the future . The use of mHealth in developing countries is therefore based on the accumulated imbrication of material agencies from mobile technologies and social agencies for personal and commercial A sociomaterial exploration of mHealth purposes . This means new mHealth projects must build on the historic precedents laid out by previous mHealth projects in developing countries, as these have likely imbricated some existing material and social agencies among participants . --- Method This study adopts an exploratory case-study approach using the socio-material perspective as a guiding theoretical lens. A case study approach was selected because case studies permit the exploration and understanding of complex, loosely bounded contexts . This resonates with the needs of our study, as the empirical boundaries of exploration are not clearly pre-defined. Additionally, case studies can be useful in capturing the emergent properties of rapidly changing environments and engaging with the subtle complexities of real-life situations . Case studies help to answer "why" and "how" questions , which is especially valuable in situations where designers and developers have limited ability to control the influence of context. The selected case took place in the Nsukka Local Government Area in Enugu State , Nigeria. Nigeria is a developing country with an estimated population of more than 198m and divided into 36 states and the Federal Capital Territory, Abuja. The rate of under-fives mortality in Nigeria is the eighth highest in the world . United Nations Children's Fund and the World Health Organisation presented a set of standard operating procedures for healthcare workers in rural areas of developing countries to assess, classify and treat seriously ill children ). Despite the introduction of these SOPs, about 14 of every 1,000 live births in Enugu result in mortality, many of which are attributed to preventable medical causes . Nsukka Local Government Area is one of the 17 local governments in Enugu State, with an area of 1,810 km 2 and a population of 309,633 . An especially large proportion of the population are believed to live in abject poverty , and maternal mortality rates have been estimated at over three deaths per 100 live births . This study coincided with the introduction of an mHealth App in Nsukka to support the diagnosis and treatment of children under 5 years old in the rural community. Specifically, this app was designed to assess, classify and treat sick children under 5 years in the community for diseases such as malaria, cholera and diarrhoea using smart devices at the point of care. To do this, the app used a clinical guideline decision support rule engine with embedded classification and treatment rules to help assess sick children. This rule engine was based on existing WHO and UNICEF iCCM guidelines ) for Nigeria. The research team worked with collaborators on the ground in Enugu who were developing and evaluating the prototype Android app alongside a rural healthcare worker mHealth training programme in March 2017. Appendix 3 presents more details. This case was interesting for two main reasons. First, poverty has traditionally been high in Nsukka, and there have been few attempts to reform the infrastructure to date. This is important, as the lack of previous projects in Nsukka lowers the likelihood of propensity biases, whereby researchers are drawn to the rural contexts that are most amenable to mHealth . Second, the app was not yet in widespread use at the time of study, though discussions and demonstrations had begun with local stakeholders. This created a natural transitional period for the region that helped ITP 35,8 bring practices and agencies to light before the system has had a chance to resolve tensions and re-establish equilibrium . We present a single-case analysis of the context surrounding the introduction of a new mHealth app. Consistent with our research question, we extend our analysis beyond the actual app in question, primarily focussing on the surrounding social and material circumstances that must accommodate and enact this new app. We select a single case design for three complementary reasons. First, the researchers were involved in an ongoing funded practical project in the area that provided unusual levels of immersion and access, not least because one of the researchers is from the area under study, so understands the culture and local dialect. The researchers were not actively developing the app. Rather, they were providing support in the form of planning, requirements gathering and user training. This type of opportunistic sampling provides empirical reach and richness that is difficult to engineer by any other means . Second, a single-case design helps to bring the researchers closer to the empirical matter under investigation, allowing the data to "talk back" in a way that increases those researchers' sensitivity to emerging variables and demands re-inspection of pre-existing biases . This means a single-case design offers the greatest depth to explore an opportunistically sampled case, as it encourages the researchers to follow interesting emerging empirical insights and capture subtle quirks that may be hidden or treated as "noise" if using other methods. Third, where a sufficiently rich case can be studied, a single-case analysis helps the researchers to provide a less-reductive description of the phenomena under study . This maximises the value of the opportunistic sampling approach. Readers are presented with greater empirical detail that may add nuance when making connections with other research. A sociomaterial exploration of mHealth obtained in both the primary host institution of the researchers and a local university in Nigeria involved with recent mHealth initiatives. --- Data collection Interviews and participant observation provided the most important sources of data . We focussed on individuals that occupy key roles, participate in key binding policy decisions, have the actual power to make changes and have the important political relational power with other systems in the Enugu State healthcare system. Specifically, the authors engaged with four key groups of stakeholders that are involved in rural healthcare delivery in the Nsukka Local Government Area Parents/Guardians mothers/guardians to the children under the age of 5 in the target community whose primary tasks among others is to take care of their children's health in their homes; Rural HealthCare Workers trained healthcare workers working in the healthcare centres located in the rural communities; Developersresponsible for developing, building and maintaining the mHealth system; and Facilitatorsindividuals or bodies that expedite or enable the development, implementation and delivery of mHealth processes, for example, the public health ministry . Data gathering involved interviews , participant observation, document/ records , field notes and photographs from clinics in the rural communities. Coupled with document/records, 53 photographs and observational and reflective field notes, thirty-two interviews were conducted, with seven PGs, seven RHCWs, six Developers and six Facilitators . Interviews were conducted in Igbo or English and recorded for analysis. All recordings were transcribed verbatim into English, along with the written notes from the interviews. Participant observation involved visiting different stakeholder groups and shadowing individuals as they performed healthcare-related tasks. The two separate visits lasted a combined 7 weeks in total . In Enugu, the researchers also observed the work areas and habits of RHCWs and Parents/Guardians by visiting rural communities and accompanying individuals to healthcare centres. One researcher further visited a range of related areas that emerged as important locations during observation, such as local pharmacies and community events where individuals often discussed health-related issues, including local Church services. Plate 2. Interviewing one of the RHCWs at the health premises ITP 35,8 The focus of the interviews, observational work and field notes was to ascertain the sociomaterial factors that influence the widespread adoption and assimilation of mHealth technologies. Hence, the interviews and observations were semi-structured and evolving in nature, focussing on understanding and explaining healthcare practices relevant to the proposed mHealth tool, the individuals and groups involved, the different materials used and how each of these things had changed over time. RHCWs were interviewed at the community health clinics where they performed their duties, and Developers and Facilitators were interviewed in their offices and workplaces. ). These documents were reviewed to elicit background information about Enugu State's existing rural healthcare system and to corroborate data from interviews and observations. --- Data analysis Data analysis focusses on interview transcripts, documents/records, observations and field notes and photographs that identify situations/events in which sociomateriality was seen to be significant in healthcare practice. Our analysis focusses on identifying key types of social and material agencies, practices and signs of imbrication that have the potential to influence the widespread adoption and assimilation of mHealth in this context. Data analysis was performed using the thematic analysis method proposed by Braun and Clarke . Thematic analysis is a flexible approach to theorising, used by many IS scholars to make sense of complex systems . Thematic analysis concentrates on the identification of recurring patterns and narratives. These recurring patterns and narratives do not always maintain the clear discriminatory boundaries of construct-based variance or process theorising, though they do provide a rich foundation for subsequent studies with those goals . Braun and Clarke identify six phases for thematic analysis. The first phase demands the researchers familiarise themselves with the data. We did this by repeatedly revisiting interview transcripts, photographs and field notes during the study. The second phase involves generating initial codes. This involved listing patterns of experiences and observations in relation to the already classified categories, that is, social agencies; material agencies; practices and imbrications. Examples included "materialities for transport" and "practices of traditional healing". The third phase involves searching for themes. These themes represent the meanings attributed by the researchers to specific quotes or other pieces of data . For this study, this meant relating different material and social agencies to particular practices and imbrications. For example, a key theme in social agency was the "perceived divide between urban and rural healthcare systems", and it became evident early on that this had a clear link with the imbrication theme of "accumulated breakdowns in payment practices". The fourth phase reviewed these themes. This involved testing the ability of data to support specific themes or their underlying explanations. This resulted in some themes being abandoned and others being refined. An example of this was the separation of the material agency theme of "limited material-resources in rural healthcare centres" into two separate themes, that is, "inextensible practical utilities in rural healthcare centres" and "limited material-security apparatus at rural healthcare centres". This was done to reflect the differences in impact these two limitations had on observed practices, such as the "reliance on centrally regulated diagnosis and treatment practices" . The fifth phase requires that themes be given names. This demanded we connect all the themes that emerged from the data to provide a wide-ranging picture of the experience of the stakeholders following the works of Aronson . It also required that we commit to the "essence" of what each theme was about and thus provided a "feedback point" between the authors and the stakeholders in this analysis. An example of this was the reframing of practices in terms of "reliance", as the accounts of common practices focussed on the absence of alternatives, rather than a strong confidence in the practices themselves. Hence "practices of traditional healing" became "reliance on informal traditional healer-driven diagnosis and treatment practices". A similar change occurred for material agency, where the overarching trend "limitations" was revealed. Phase 6 required for the research to be compiled into a report. This took the form of a descriptive "theme statement" , which is presented in the research findings in this paper. Key quality markers were identified prior to data collection. The first, "reflexivity" , describes the importance of attending to researcher bias . The researchers made effort to reflect on bias wherever possible, both in the data itself and in our interpretation. The second, "ontological authenticity" , describes the extent to which shared knowledge and social action between the participants and researcher are fair and balanced. This was managed through multiple visits to Enugu to share findings with participants as part of "venting" or "member-checking" exercise, used to attest to the "truthfulness" and "trustworthiness" of the findings . The third, "internal validity" , describes the degree to which the emerging themes are logically consistent with one another . This was managed by relating themes within and across high-level categories to ensure observed material and social agencies, practices and imbrications were clearly linked. The fourth, the "interpretive validity threat" , occurs when alternative explanations, interpretations or hypotheses are not given sufficient opportunity . This was managed by maintaining an "audit trail" that included "thick, rich description" and illustrative quotes. Finally, the fifth issue "generalisability" or "external validity" describes the extent to which findings from this study are generalisable beyond this context. This was managed by linking findings with recurring socio-material categories, so creating natural parallels with socio-material studies in other contexts. --- Analysis The following sections present and discuss emerging themes for social agency , material agency , health-related practices and imbrication . Individual themes under each heading are accompanied by in-depth descriptions and illustrative extracts of data. --- Themes for social agency The first theme for social agency describes the perceived threat to existing roles and individual autonomy. This took two main forms. One threat was that the new system could diminish the A sociomaterial exploration of mHealth respect towards specific individuals. Facilitator 1 explained, "the doctors and to some extent the nurses will not be happy in that they will feel that some parts of their jobs are being taken away from them when such a tool is introduced in the healthcare system". The perceived threat of angering individual doctors, rather than nurses, seemed counterintuitive at first, given it was the general RHCWs who would require new skills and whose tasks were becoming regimented. Yet, the expansion of general RHCWs into diagnosis arguably broadened their role in a way that blurred the distinction from doctors. RHCW3 acknowledged the tension this created, lamenting "I know some doctors may be feeling we are trying to take part of their job. But I do not think the doctors should feel we are trying to take their job since it is in the interest of the poor rural communities". Another threat came from the perceived oversight from urban institutions. Developers and Facilitators viewed this ". . . we share a feeling of being left out by the system" "The aberration in healthcare distribution and healthcare provision is hurting the primary healthcare system in Enugu State" "Observed participants prefer treatment in urban than rural centres due to perceived better treatment measures" Perceived collegiality among stakeholders in rural community Social actors share a sense of common identity in addressing challenges and difficulties confronting them "I approach my friends or neighbours who may know what is happening to my child and they offer some suggestions on how to go about it in the immediate" "We interact with the villagers that come here as patients, like children with various illness, sharing in their feelings especially as most are poor and find it difficult to go for medical treatment in the urban area" as a positive, describing the benefits of monitoring and continuous interaction and oversight. However, this monitoring was viewed with scepticism in remote areas, as it would mean individuals would have to continuously answer to central institutions, explain their behaviours and possibly accept more micromanagement. The second theme for social agency concerns the constitution of social actors in rural healthcare settings, specifically the perceived limitations of skilled personnel in the rural health centres. PGs and RHCWs complain about a lack of nurses and doctors in rural health centres, which they interpret as a reduced capacity for sophisticated healthcare delivery. This is a challenging problem to address, as noted by Facilitator 6, who explained that "most welltrained personnel do not like to work in rural areas". Developer 6 agreed with this assessment, remarking that "the properly trained nurses are not available at all because every nurse that is properly trained will want to stay in the town . . . doctors, they are not also not there, even the ones that are in the rural communities are involved in their own private practice, they are not involved in the healthcare system in the State, . . . and of course . . . the consultants who should be taking decisions are not available in all the rural communities in Enugu State". This appears to result from the minimal incentives for qualified health professionals to work and live in the rural areas. Qualified healthcare workers migrate to urban areas where they can earn better wages and have their children attend better schools. Thus, the social agencies of rural health centres are discouraging skilled social healthcare professionals and therefore sophisticated healthcare delivery, through the lack of complementary skilled teachers and wealthy healthcare consumers. This creates a natural "chicken and egg" scenario, where those complementary individuals are also likely discouraged by the lack of quality healthcare. The third theme for social agency describes a perceived apathy by urban institutions towards rural healthcare systems. RHCWs and PGs feel rural healthcare systems are isolated from urban health systems and of low priority to urban actors. Developer 6 reflected "at the beginning, the rural healthcare system has no institutional base, as they were not accorded any status on which to operate in relation to the . . . healthcare delivery centres in the rest of the State". Thus, the social agency tends to produce power imbalances between the urban and rural areas. Developer 6 further explained "to compound issues, the . . . healthcare centres that would have helped support/nurture the rural/primary healthcare apparatus have broken down leaving only the tertiary institutions as the sole functional healthcare delivery structures". Thus, the prevailing social agencies of urban and rural social systems have no obvious connective institutional structures or systematic relationships to unite them in healthcare delivery. Instead, each appears to be operating within relatively distinct social configurations that present little opportunity for convergence. The fourth theme for social agency describes a perceived collegiality among actors in the rural community. This collegiality contrasts with the perceived disconnect between these actors and those in urban areas. Data suggest strong social agencies for relationship building and collaboration between PGs and RHCWs. PGs can typically reach RHCWs outside their working hours with health-related queries, suggesting the agencies of that role go beyond professional contexts. PGs also help each other to find solutions to health problems. PG2 summarised this by saying "I approach my friends or neighbours who may know what is happening to my child and they offer some suggestions on how to go about it in the immediate". This means PGs and RHCWs rely on social agencies for word of mouth and informal learning when dealing with healthcare challenges. RHCWs also noted a sense of personal responsibility for protecting PGs from third parties who may exploit their desperation or lack of understanding, for example, pharmacies selling illegitimate or overpriced drugs. RHCW2 explained "if you leave them to buy for themselves, they may buy fake drugs which is being sold out there". --- A sociomaterial exploration of mHealth Themes for material agency The first theme for material agency describes the mobility of mHealth-enabled smartphones. The introduction of the mHealth app meant the introduction of sufficiently sophisticated smartphones, and these smartphones provided a broad range of new material agencies. While the focus of the development was the app, establishing a digital connection between remote areas and urban centres created new capabilities that provoked a social reaction from either side of this historic divide. Several Facilitators saw the app as a tracer bullet for subsequent innovation, for example, Facilitator 3 noted it was not the idea of the app that was novel, the project was ". . . novel in the sense that it will be quite helpful in reaching those people in the villages and also making healthcare services available to them . . ." This reflected a common assumption that this type of approach was the only feasible way to establish connectivity and begin the larger task of bringing these remote areas back into the mainstream health system. This was mirrored by RHCWs, who described the connectivity-related material agency of the smartphones as an empowering force that would take on a momentum of its own, for example, ". . . it will help those people that are living there since it will bring healthcare to the grassroots" and "it is really for humanity that such way of services should be extended to the rural poor communities" . The app embraced this form of material agency by guiding RHCWs to collate data in summary forms and transfer these records to local government headquarters. These data are subsequently transferred to the federal health office via Enugu State's Ministry of Health. New mobile apps were being introduced to collect and transfer data at the time of writing. These apps help workers at rural health centres to collate health data more efficiently and forward this data electronically to state and federal bodies. The second theme for material agency is once again centred upon the device rather than the app, this time describing the Web connectivity of mHealth-enabled smartphones. The previous theme described how adding sophisticated smartphones added a new force that pulled remote areas into the digital sphere of urban areas. Equally importantly, these smartphones added a material agency that pulled these remote areas into the larger digital sphere of the Web. This second dimension of material agency provoked individuals to begin thinking about other, potentially unregulated health-related content on the Web. Facilitators were keenly aware of this possibility. Facilitator 1 noted how the inadequacies of the formal health system could make these capabilities more prevalent, noting "since this [android features] gives them access to Internet . . . [and] especially with the shortage of Doctors in the rural areas, they can now start looking for remedies to sickness via the Internet". Others further noted that even if they could meet remote healthcare needs, the material agency of the smartphones naturally lends individuals to searching the Web as a habit. The third theme for material agency describes the rigid existing paper-based diagnostic guidelines that are integrated into the mHealth app. The Ministry of Health introduced a set of guidelines called the standing operation procedure for use by RHCWs throughout Nigeria. Developer 5 described this as "a step by step, blow-by-blow method whereby a welltrained health officer can act in the absence of a doctor or when a doctor cannot be physically present". However, RHCWs had concerns about the rigid rules associated with the SOP and long processes involved, which often added unwelcome delays and complexity to interactions. RHCW6 commented "before they used to treat them under one page but now, they have split it into different pages, which makes it very difficult and stressful for us somehow". Some of these issues would be attenuated by the mHealth app, for example, the delays associated with finding records. Others would persist. For example, several RHCWs noted that the material agencies of these SOPs typically lead to PGs being referred for further diagnosis and treatment in urban health centres. These agencies jar with the emotional attachment many RHCWs feel towards PGs. Further, frequent referrals undermine RCHWs' clinical expertise and PGs' perceptions of the quality of treatment delivered at rural health The fourth theme for material agency describes the insufficient practical utilities in rural health centres in which the mHealth app is used. Health centres in rural villages are not equipped to anything like the standard of urban health centres . Developer 6 noted the limited material agencies further exacerbated staffing problems, remarking "no properly trained nurse would like to work in such an environment". Facilitator 6 acknowledged "there is lack of infrastructure and very few health centres are worth to be called places where any sick person can even go into". Facilitator 3 suggested this was not solely limited to rural areas, "Even in urban centres where we refer as having adequate facilities, it is not so in many cases. Doctors and nurses work with old equipment and this brings a lot of stress to them. They have protested about this, but it all came to nothing." Material agencies for water supply are a major cause for concern for the RHCWs in the communities, as there is no steady supply of clean water. This means RHCWs often rely on private supplies of water or resort to harvesting rainfall water in tanks . Another material agency from these limited utilities concerns the irregular availability of drugs at rural health centres. RHCW2 remarked "the availability of the drugs we use is also a challenge, if drugs are supplied to us in large quantity it will be a good thing, instead of having to stay and wait for the request to come through". Even where equipment, water and appropriate drugs are present, material agencies may be limited by the poor or non-existent access to electricity in health centres. RHCW2 pointed out, "In the Nigeria of today, the irregular supply of power is considered as a normal way of life. The small generators used by individuals comes as a saviour in charging of phones, those centres in the urban areas have electricity generators in their various offices while there is none at the rural health centres". As a result, RHCWs rely on oil-based lanterns in the evening at health centres and personal charging facilities to maintain the batteries of mobile devices. This raises concerns about maintaining charge in the smartphones for the proposed mHealth app. Once RHCWs become reliant on this app, a lack of charge could stop care delivery for PGs and children who have undertaken the long journey to the centre. The fifth theme for material agency describes insufficient security apparatus at the rural healthcare centres in which the mHealth app is used. Most of these centres are not wall-fenced, nor are there obvious security measures to prevent unwanted intruders. The lack of material agencies for physical security at rural health centres is a serious concern for RHCWs and PGs due to the perceived ongoing threat of attack from night marauders. This threat is further compounded by the fact that many RHCWs work in the centres during the night. RHCW4 said "When somebody knocks at the door at night you will be afraid to open because you do not know whether the person knocking is a patient or those that are coming to rob or harm you". When asked about the security issue at rural healthcare centres, Facilitator 2 explained "the resources of the state are limited and government . . . provides infrastructure as much as it can". In the absence of government-provided security resources, the material agencies of rural centres give way to community-based agencies, notably the weapons and vehicles of volunteers from local villages. These volunteers, usually groups of youths, form neighbourhood watches to guard centres and surrounding areas, so helping to prevent attacks from night marauders. The sixth theme for material agency describes the insufficient roads for transportation to and from rural healthcare centres in which the mHealth app is used. First-hand observations and feedback from interviewees suggested that roads in urban areas are better than roads in Plate 7. The entrance to the health centre, showing the strategy adopted to collect water into a tank and one of the modes of transportation Plate 8. Oil-based lanterns used for lighting health centre due to the absence of electricity A sociomaterial exploration of mHealth the areas . PG1 remarked "there are gullies and ditches on these roads and here you see, no public motor transport driver wants to work on the rural roads because of this". This lack of material agencies to enable public road transportation means rural inhabitants are forced to pay for rides on motorcycles known as "okada" or tricycles known as "keke" to travel anything other than short distances. Yet this presents new problems, as many rural inhabitants cannot pay the required fees and must instead walk to centres. Facilitator 6 summarised the limited transport-related material agencies of the area, explaining "a good number of communities are completely inaccessible, inaccessible by road, which is the major means of transportation in this part of the world, and the fact that you cannot access those places no matter how you want to look at it is disheartening". Photographs captured some of these insufficiencies in detail . This is especially problematic during the rainy season, at which time most of these villages are almost entirely cut off from other parts of Enugu State. This compounds the earlier theme concerning the rigidity of existing guidelines, as referrals to urban centres to use the mHealth app become even more frustrating when transport resources are lacking. --- Themes for practices The first practice-related theme describes the reliance on centrally regulated diagnosis and treatment practices, like those in the mHealth app. These practices are designed to advance best practice in rural areas and compensate for some of the social and material limitations. Hence, RHCWs have already been trained to use material SOPs for the assessment, classification and treatment of patients. Also, confidence has built up in rural areas around these guidelines, which are seen as an accurate representation of best practice in urban centres. One RHCW predicted, "People will be rushing to be treated with this modern technology . . . to assess and treat. Because after seeing what we are using for treatment they will tell others who will rush to be treated with an accurate instrument unlike the human assessment" . Parents echoed this excitement, as well as the positive sentiment towards technology-based assessment. However, while RHCWs, Developers and Facilitators may wish to use these SOPs as broadly as possible, this relies on PGs making their way to the health centres. This is not a reliable assumption, as earlier limitations of material and social agencies mean many PGs do not have access to the transportation resources required for such Plate 9. A typical rural road ITP 35,8 travel. This means these practices make up only one aspect of the socio-material system for healthcare in these areas. The second practice-related theme describes a reliance on informal PG-driven diagnosis and treatment practices which are not represented in the mHealth app, whereby parents bypass doctors and go straight to pharmacists for medicines. These diagnosis and treatment practices are typical in Nsukka. PG7 described "Once I notice that my child is not feeling well, I make use of some medication I have at home first before going anywhere". PG3 commented similarly "I only take my child to a health centre when I notice that the medication I have administered to my child at home is not working". This is possible because PGs often have drugs stocked at home, drugs bought from pharmacy attendants without formal prescription from a healthcare professional. These practices do not require the same level of travel and expertise as the formal SOP-driven practices at rural health centres, therefore PGs find it easier and more satisfying to enact these informal diagnosis and treatment practices instead. Many pharmacist attendants have embraced the opportunities presented by this new practice, due in part to their own limited material and social agencies for wealth and expertise. Thus, these practices effectively circumvent limitations in the material agencies of rural centres for rapid, dynamic and physically accessible diagnosis and treatment, albeit at the expense of accuracy, reliability and long-term health outcomes. "The SOP is used to diagnose illnesses, treat or refer the patient to a doctor" "we were using the paper-format before they brought an app for sending to State and federal directly" Reliance on informal PG-driven diagnosis and treatment practices, which are not represented in the mHealth app PGs circumvent RHCWs and SOPs like those in the mHealth app to buy medicine directly from pharmacists "I visit the pharmacy to get some medications I use at home before I take the person to the clinics" "If my child is sick I buy drugs that I feel is going to cure my child" "Observed Parents contact peers for treatment advice, Parents often times buy drugs/ medicine from pharmacies without prescription" Reliance on informal traditional healer-driven diagnosis and treatment practices, which are not represented in the mHealth app PGs use traditional healing practices to treat sicknesses, e.g. drinking liquid from boiled mango leafs to treat various stomach ailments "Sometimes when we have no money we make use of herbal methods of treatment within our village" "Parents do make use of traditional healing methods for treatment" Reliance on informal and clustered communication practices, which are not represented in the mHealth app PGs and RHCWs rely on informal communication channels between these groups, rather than communication with urban centres "[RHCWs] do not refer to us and we do not write back to them even if their referrals will come in a secret way" "We are not even in talking terms with [RHCWs]" Table 3. --- Health-related practice themes, descriptions and illustrative data extracts A sociomaterial exploration of mHealth The third practice-related theme describes a reliance on informal traditional healer-driven diagnosis and treatment practices, which are not represented in the mHealth app. This was seen as the next-best alternative when PGs could not secure suitable drugs from pharmacy assistants for their children. Many individuals continued to avoid rural centres due to material agencies prohibiting long journeys and community-based social agencies that value tradition and respect for one's elders. Instead, these individuals turned to local traditional African healers. PG7 explained "traditional healing remedies were handed down by our forefathers to us and it worked for them before the advent modern drugs, so, I still use traditional remedies for certain sicknesses, such as malaria, etc." These treatments were also typically cheaper than formal practices at rural centres or buying drugs from pharmacy assistants. PG3 lamented "I go to African traditional herbal homes to treat sickness with herbal remedies especially when the prescribed drugs at the health centre are too expensive for me to bear". The RHCWs found this frustrating, noting the futility of trying to convince people of little means to opt for the more expensive options. RHCW7 further linked this to limited social agencies for skilled personnel in rural areas, remarking "For those of us who work in the village, the most people we work with do not have good knowledge of healthcare systems. So, we need to boost health education for rural people". The fourth practice-related theme describes a reliance on informal and clustered communication practices in Enugu State, which are not represented in the mHealth app. A lack of formal communication practices was observed between rural health systems and urban health systems. Where cases become serious, PGs often find their way independently to urban teaching hospitals without referrals or any accompanying records. Developer 6 noted "how many references have I gotten from [rural health centres]? None, I mean zero, at best, those centres are just glorified maternity centres". The researchers witnessed this firsthand when one rural woman with an advanced illness was brought to a consulting physician by her brother without any accompanying documentation. Developer 6, the consulting physician, explained "it is very strange that I had to attend to this woman without any previous records on what my juniors in the ladder have done, what 'things' am I going to consider? How do I start?" Developer 6 further elaborated "that kind of woman cannot see a specialist like me without formal referral from where she was first treated, where a record has been established stating the history of her sickness and records of the treatments administered on her before now". These practices once again enact material agencies for limited transport, as this is part of the reason that historical paper-based records are not transported correctly. However, they also enact social agencies for the perceived divide between urban and rural healthcare systems and entangle with other informal practices that avoid those rural centres in favour of buying drugs or traditional remedies directly. --- Themes for imbrication The first imbrication-related theme describes accumulated breakdowns in payment practices. RHCWs complained that local governments routinely missed salary payments, arguing it had negatively affected their motivation to work. RHCW5 remarked "payments of our salaries is a problem, when you are not paid promptly the satisfaction and the zeal to do the job will not be there". MOH argue this has nothing to do with them, as in Nigeria, each state is responsible for providing the regulation and technical support to rural healthcare services but the local government level is responsible for rural healthcare. Those local government bodies suggest they are not given the funds to follow through on these payments, creating a circle of blame with no obvious sign of ending. Developer 5 suggested the only way to resolve this was to consolidate the payment in one place, arguing "It is just not right to leave the funding of primary healthcare systems in the hands of the local governments, it should be the primary responsibility of the Federal Government". In any case, this accumulation of missed payments has been a significant contributor for social agencies that undermine the widespread adoption and assimilation of mHealth tools. The resentment has increased the perceived divide between urban and rural healthcare systems and further discouraged skilled workers from remaining in rural areas, so contributing to perceived limitations in skilled healthcare personnel in rural areas. Those that stay must often make personal sacrifices to compensate, so deepening the perceived collegiality among RHCWs and PGs in rural communities. The second imbrication-related theme describes accumulated personal and professional phone-related practices. Many RHCWs and PGs are in the habit of carrying their personal phones. This has not been a straightforward transition, as phones are in danger of loss, theft or damage. RHCW1 explained, "I now have to carry this particular phone with me in conjunction with my personal phone, protecting them both is a challenge to me". Facilitator 1 echoed this concern, noting "we had to introduce an MOU [Memorandum of Understanding], which is once you lose your phone you have to replace it". This threat is sufficient that some RHCWs do not want to take the responsibility of carrying these additional professional phones for fear of having to replace them. For those that carry professional phones, or use personal ones instead, the benefits have been accumulating over time. PGs may also use their phones to make personal calls to RHCWs or contact friends to ask questions. PG6 remarked "I can reach my friend with my mobile phone to ask of what to do about a particular sickness I feel my child is experiencing". The accumulation of personal phones also means PGs have independent access to third-party health information, provided they have the literacy to browse the Web. Like the aforementioned accumulation of missed salary payments, this accumulation of personal and professional phone-related practices has been a significant contributor for the informal practices that undermine the widespread adoption and assimilation of mHealth tools. The ease with which rural inhabitants can access one another and spread knowledge through word of mouth fuels increasing reliance on informal and clustered communication practices. This encourages individuals to copy the informal PG-driven diagnosis and treatment practices and traditional healer-driven diagnosis and treatment practices adopted by their peers. It also creates multiple information channels that dilute the information being passed on by RHCWs and urban medical professionals. At the level of social agency, it is clear that the urban and rural healthcare environments represent separate social worlds. The lack of highly trained workers is recognised as a significant challenge to healthcare in rural communities of developing countries . In Africa, there are 2.3 healthcare workers per 1,000 population, compared with developed countries such as the USA, which have 24.8 healthcare workers per 1,000 population . Most well-trained healthcare workers prefer migrating abroad where they have better remunerations , while others often prefer to work in urban centres that are better connected and better resourced. In Enugu State, this has created a sense of isolation and neglect that binds PGs and RHCWs into systems of making do. These systems build on material and social agencies to enact new practices that bypass short-term limitations at the expense of longer-term outcomes. This tendency of rural communities in developing countries to find creative healthcare workarounds has been documented in existing literature . It has also fostered an increasing emotional attachment between RHCWs and PGs that has helped to support a group that are otherwise neglected . Thus, a sub-optimal but stable equilibrium has formed that creates challenges for mHealth initiatives in these communities. At the level of material agency, it seems that rural health centres lack the breadth and depth of appropriate complementary materiality for mHealth to be meaningfully enacted in healthcare delivery. Most RHCWs do not enjoy working in the rural health centres due to the lack of basic material utilities, transportation and security. Adequate medicines and electricity are often unavailable. Similarly, most roads in rural areas are unpaved and in disrepair, so restricting public transportation in favour of alternative modes that lend themselves to shorter journeys. The threat of intruders means RHCWs are continuously watching for signs of danger and thus leave them feeling unsafe in their workplaces. These observations of infrastructural deficiencies in rural health centres are consistent with findings from existing research in a range of developing contexts . Taken together, these limitations discourage RHCWs and PGs from using rural health centres extensively or even spending prolonged periods there, both of which are necessary for those centres to become properly entangled into the rural healthcare system. Holeman and Barrett the SOPs provides limited immediate satisfaction for PGs. Instead, those PGs gravitate towards informal diagnosis and treatment practices that can be performed more cheaply and with less delay, such as buying medicine directly from pharmacies without prescription or alternative treatments from traditional healers. These findings are also consistent with existing literature . It further appears there is little direct communication between rural and urban health centres. Instead, we noted clustered communication practices among rural stakeholders. This creates a lack of information about rural individuals in urban centres, which becomes particularly problematic given those rural dwellers may have to travel great distances to attend those centres . This breakdown in communication is not only delaying the spread of new practices and information from rural healthcare centres , it is also creating a growing threat of misinformation and malpractice within rural communities. At the level of imbrication, the lack of social and material entanglement between urban and rural systems can be attributed to historic breakdowns in practices that could otherwise have acted to strengthen these connections. This is consistent with basic assumptions of socio-materiality, which assumes that social and material elements are mutually generative . Breakdowns in payment practices have caused RHCWs to rely on other ways to earn an income in their community. This has eroded the authority of urban actors, particularly where tensions may be perceived between urban and rural interests. Indeed, before mobile technologies were made available to RHCWs for healthcare-specific reasons, many RHCWs and other rural dwellers had taken it upon themselves to acquire personal smartphones. Thus, there is no clear indication that urban actors have the authority to tell rural actors how they should use those devices. Modern mobile phones clearly have material agencies which can be enacted into centrally prescribed medical practices. For example, they can be leveraged to support on-the-spot diagnosis and treatments from a specialist somewhere outside a rural context or to facilitate referrals practices . However, they also have material agencies that lend to greater informal communication and third-party information access. This is significant for future mHealth initiatives, as the sourcing of information from this unregulated space may hamper structured healthcare delivery processes in rural areas . The isolation of rural social agencies, the limited material agencies of rural health centres and the dominance of informal practices have contributed to deepening and potential harmful local equilibria. Ignoring these equilibria by focussing on pilot studies and controlled population samples may only leave these areas further behind. Finally, we contribute to socio-materiality by providing another model or exemplar study where that perspective adds tangible value for a complex design context. This has been a concern for socio-materiality since its introduction to IS research, as the practical impact of studies is not always as obvious as other forms of mid-range theorising . One obvious source of value for the critical realist school of socio-materiality is the ability to step back and view socio-material phenomena that are missed by theories/approaches which hone in on pre-defined variance or process relationships . This study provides another example, such as in Jones andOberl€ ander et al. of this value by stepping back from application-level theorisation, such as interface design or data processing, and individual-level theorisation, such as perceptions and intentions. Instead, building on the recommendations of Cecez-Kecmanovic et al. , we pay significant attention to the material agency of the system under study, while also showing how these entangle with social agencies to form specific practices. This demonstrates how a socio-material approach can provide theoretical and empirical triangulation that lends itself to more robust and diligent future theorising . --- A sociomaterial exploration of mHealth This study further contributes to differentiating the value of socio-materiality from the preceding theories, such as actor network theory or practice theory . Although those theories are also useful for extending theorising beyond the reach of variance or process theories, yet they put less emphasis on delineating and distinguishing social and material influences . This distinction was vital in this study, as a key advantage of mHealth interventions is the ability to make small material changes that still produce significant socio-technical changes . Socio-materiality further enabled the study of not just "what is" but also "what could be". Notably, the concept of material agency illuminates the potential uses of mobile phones in the future, based on the embedded structures in that technology. Examining actor-network structures and/or practices in isolation would likely miss these insights, as they turn attention to existing patterns and tensions, rather than latent material potential. Yet, it is this analysis of future potential that makes findings actionable and allows challenges to be pre-empted. --- Summary and conclusions This study presents a detailed thematic overview of the existing socio-material structures properties, health-related practices and imbrications of rural healthcare systems that may impact the widespread adoption and assimilation of new mHealth technologies. The emerging themes combine to tell a story of a structured but minimal professional healthcare delivery system, with decentralised and peer-based practices increasingly filling in the gaps. They tell a story of isolated rural social agencies that limit the authority and value of centralised initiatives and material agencies that are not aligned for the desired enactment of mHealth tools like the one proposed. Historic imbrications mean trained healthcare professionals are rare, communities of practice are distributed and informal, and mobile phones are treated as tools to support communication among peers. Building on these, we propose four key questions to be addressed in future research targeting mHealth in rural areas of developing countries: How do we design mHealth solutions that complement the existing materiality of rural areas, for example, by minimising the need for travel where transport options are limited? This likely requires more engaged research on the deployment of mHealth tools. This study highlights how practical issues such as electricity, security and road quality can prevent the spread of systems. More engaged designers who understand these issues better may be able to create more robust designs. How do we design mHealth solutions that reinforce the connection to urban centres while still allowing rural healthcare workers the autonomy to offer immediate solutions? Our findings showed that cultural differences across rural and urban areas are inhibiting participants' interest in mHealth projects. One way to approach this question is to consider the types of "design ethnography" advocated by . This would bring social and cultural practices and values to the forefront of the design. How do we change embedded practices, particularly those that have cultural origins that go against contemporary health treatment methods? Change management is notoriously difficult, it is particularly challenging within healthcare. For this reason, Markus argued that organisations may need to align technology projects with deep structural change. The same appears true of mHealth, where the enactment of specific tools is tied to the acceptance of medical best practice. Education is therefore both a requirement for the deployment of these tools and an outcome. We therefore suggest that the design of mHealth tools should also consider more pedagogical theory. How do we avoid interference or destructive competition from unregulated information or health-related applications available from peers or on the Web? The addition of IT is not always a social good, and many of the contemporary issues with pseudoscience and "post truth" are linked with social media and Web access . This is a global issue, with many poorer countries and/or countries with low trust in their governments suffering disproportionately . Building on the previous questions, this suggests mHealth will be at the forefront of these challenges in the coming years. We call for more studies of information sharing and perceived credibility in mHealth contexts to help understand their relationship in the future. --- Limitations We acknowledge two important limitations of this study. First, our research focussed on a region in which technology-enabled guideline-driven treatment remains the priority mHealth concern. However, several other forms of mHealth initiatives exist, for example, those focussed on data gathering or those focussed on remote diagnosis and treatment . We call for similar research on those alterative topics to compare results. Second, consistent with the exploratory nature of our study, the qualitative methodology and the single-case design, we make no claims of statistical generalisability . Rather, the intention was to draw attention to important existing socio-material considerations that will add to understanding in this space . This understanding, as well as being of value in itself, can be used to underpin other forms of increasingly structured theorising . Thus, we believe the themes identified should be used to inform future theorising in the ICT4D domain that seeks to create more tightly bounded and predictive frameworks or models . A sociomaterial exploration of mHealth A sociomaterial exploration of mHealth integrated Community Case Management , which are to be adopted by individual countries basis based on their respective National Child Health Index. iCCM presents a set of meticulous and systematic guideline which enables healthcare workers to assess, classify and treat seriously ill children in rural areas ). Rural healthcare workers capture socio-demographic characteristics and clinical information regarding diseases, illness and recommend treatments, especially in malaria prevalent countries in Africa . The Nigerian healthcare system is structured as a three-tier structure comprised of federal, state and local government levels . The 36 state governments and the 774 local government areas within the states combine to assume responsibility for the provision of basic public services for Nigerians . At the federal level, the government must enact policies and provide resources. At the state level, the ministries of health must provide regulation and technical support. At the local level, local governments must deliver individual healthcare services. The distribution of resources to primary healthcare is a continuous source of tension, as "the spending priorities of states often fail to sufficiently focus on basic services" . This means rural communities are often underserved when compared with urban counterparts . A referral system is intended to extend access to urban centres for those living in rural areas; however, this system does not function effectively . Enugu state is one of the 36 states in Nigeria and located at the south-eastern part of the country . The state is positioned between latitude 50 56N-706'N and longitude 6,053E and 7,055E . Enugu is bounded to the North by the states of Kogi and Benue, to the east by the Ebonyi, to the south by Abia and to the west by Anambra states . Its capital is Enugu, and the name of the State is derived from its capital city, Enugu, means the top of the hill. Enugu state's area includes most of the Udi-Nsukka Plateau, which rises to more than 300 m and partly lies within the tropical rain forest belt to the south . Enugu state is covered by open grassland, with occasional woodlands and clusters of oil palm trees. The State was created out of the then Anambra state in the year 1991 during the Military regime of General Badamusi Babangida . The State is divided into 17 local government areas and three senatorial zones, namely Enugu East, Enugu North and Enugu West senatorial districts . The population of the state is approximately 3.3m with a land area of about 7,618 sq. km . In Enugu State, the rate of under-fives deaths is notably higher than the national average. The causes of these deaths include: for neonatal it was attributed to sepsis, birth/asphyxia and neonatal pneumonia; for 1-59-month mortality it was attribute to malaria, diarrhoea and pneumonia . The Igbo ethnic group constitutes the majority of Enugu state's population , most of which live in the rural areas . Farming plays an important role in the state's economy; yams, oil palm products, taro, corn , rice and cassava are the main crops . Enugu, the state capital, is a major centre for coal mininghence, it is referred to as "Coal City". Besides coal, iron ore also is mined, and deposits of limestone, fine clay, marble and silica sand . Industries include textile manufacturing, food processing, lumbering, soft drink bottling, brewing and furniture manufacturing. A network of roads connects Enugu town with Awgu, Ezzamgbo and Nsukka. Economically, Nsukka local government people are typically farmers . Trading occurs but mainly on agricultural products . Weaving is a traditional local craft, and coal deposits have been discovered in Obollo area east of Nsukka located on the main Onitsha and Makurdi road . Nsukka Local Government Area is one of the 17 local governments in Enugu State. The headquarters is located at the hilly sites of Nsukka town. Nsukka town lies between the geographical coordinates of latitudes 6845'N and 7800'N, and longitude 7815'E and 7830'E of the Greenwich meridian . Nsukka local government shares common border with Igbo-Etiti L.G.A on the South, Uzo-Uwani L.G.A on the West, Udenu L.G.A on the East and Igboeze-North L.G.A on the North . The local government has an area of 1,810 km 2 and a population of 309,633 . The Nsukka local government area, located within Enugu state, is subject to especially high rates of abject poverty . For example, in early year 2000, the maternal mortality rate was estimated to be more than three deaths per 100 live births in the Nsukka senatorial zone of the Enugu state . Currently, integrated community case management is being piloted in two states in Nigeria, namely Niger and Abia, with future scale-up planned in an effort to cover the basic health needs of over 300,000 children . Meanwhile, research findings in other countries have shown defects emanating from the paper-based iCCM method with documentations of poor adherence of rural healthcare workers to the guidelines, leading to poor-quality diagnosis and treatment measures . Recent research findings show that these defects or inadequacies could be remedied with the introduction of information communication technology . The current project takes place as one possible mHealth solution, developed and tested in Malawi, is being discussed and demonstrated with urban and rural actors in Nsukka. This provides a unique opportunity to explore possible influences in the earliest stages of a possible transition. --- Appendix 1 Philosophical differences in socio-material theorising The ontological and epistemological foundations for a socio-material analysis may vary . One position builds on the work of Barad ) and Latour ) to argue that social and material are inseparably related . This argument is hinged on the idea of agential realism developed by Barad, who argues "phenomena do not merely mark the epistemological inseparability of 'observer' and 'observed'; rather, phenomena are the ontological inseparability of intra-acting 'agencies'" . This was summarised by Orlikowski ) as "there is no social that is not also material, and that there is no material that is not also social". D ıaz and Urquhart, 2010, p. 353) explain "we live in a world made of both social and technical artefacts; we cannot detach society from technologyneither can we isolate technology in the abstract". This view is also closely related to Latour's work on actor-network theory , who proposed the term "actors" to avoid differentiating human and nonhuman influencers , though socio-materiality often prefers "entities" or "agents" . Collectively, this view of sociotechnical systems "makes a distinctive move away from seeing actors and objects as primary self-contained entities that influence each other . . . either through impacts . . . or interactions . . . away from discrete entities of people and technology . . . to composite and shifting assemblages" . In effect, humans or technology has no intrinsic properties, but obtains form, characteristics and abilities through constitutive entanglement . Further, this view suggests that entities, people and technology have no intrinsic boundaries but are relationally manifested in practice . An alternative view of socio-materiality is adopted in this study, namely the view proposed by Leonardi ). This view is grounded on a critical realist ontology that is substantialist in nature . The substantialist ontology "takes as its point of departure in the notion that it is substances of various kinds . . . that constitute the fundamental units . . ., self-subsistent entities, which come 'preformed,' and only then to consider the dynamic flows in which they subsequently involve themselves" . That is, entities, be it humans or things exist as separate and self-contained entities that interrelate and affect each other in practice . Building on the works of Mutch ) and Faukner andRunde , it is difficult to operationalise the empirical constructs in an agential realist approach due to the interlocking of the social and material ). Instead, the substantialist approach assumes an inherent distinction between material and social agencies, though at the same time recognises that practices and outcomes are dependent on how they are entwined in some context . Applied to a mHealth context, this means the introduction of a mHealth tool to a developing country should be treated as a change in that system's material agency. During this process of change, the essential building blocks of the mHealth tool are imbricated to fit with the goals, needs and expectations of social actors, which react to form new practices. The degree of this imbrication depends on the extent to which social actors enact the new material agencies introduced, that is, "ultimately, people decide how they will respond to a technology" . Thus, social agency determines which features are enacted and how, while the material agencies enable and constrain the possibilities. Appendix 2 Discussion of Nsukka Local Government Area, Enugu State, Nigeria Nigeria is an African country on the Gulf of Guinea located in Sub-Saharan Africa. Nigeria is bordered on the west by Benin Republic, on the east by Chad and Cameroon, on the north by Niger Republic and on the south by the Atlantic Ocean. Nigeria is a developing country with an estimated population of more than 198m , divided across 36 states and the Federal Capital Territory, Abuja. It is estimated that 120m Nigerians still live below or around the poverty line . The under-fives mortality rate in Nigeria is the eighth highest in the world , with over 100 mortalities per 1,000 births . Malaria is the leading causes of death in Nigeria , closely followed by respiratory infections . In other to address this situation, clinical guidelines for rural healthcare workers were developed by WHO and UNICEF to deliver healthcare services to children under the age of 5 in remote, hard-to-reach rural areas of developing countries . These guidelines are known as A sociomaterial exploration of mHealth --- Appendix 3 Detailed information on the app The app was proposed in early 2016 as part of the Irish Research Council competitive funded IMPACT project. As part of a patient assessment, the app users are required to complete several validated fields including the personal information of the patient , patient vital sign data, for example, respiratory rate, weight and data about the child's presenting symptoms, for example, cough, fever, diarrhoea and other symptoms. Based on these inputs, the app presents a recommended diagnosis and treatment supporting rural healthcare workers in their clinical decision-making. Depending on the data entered, treatment alternatives may include food and fluids, pain relief and/or broad-spectrum antibiotics. If the presenting child is very unwell, the clinical algorithm recommends that the child is referred to higher-level medical centre or hospital for further treatment. During the 15-month funded project, approximately 170 stakeholders were exposed to our app. These included RHCWs and other healthcare professionals , parents/ guardians of children, developers and facilitators. It was introduced to support RHCWs in their work in assessing young children in the rural community with particular focus on improving adherence to the existing clinical guidelines . Enforced form validation features means that RHCWs must complete the required form fields before they are permitted to move to the next assessment question, thus overcoming issues arising from existing paper-based data collection methods and poor data quality at the level of the state . Following patient assessment, RHCWs are provided with diagnosis and treatment recommendations, compliance with care suggestions is beyond the boundary of the app. International healthcare guidelines are updated regularly, the fixed nature of the embedded clinical algorithm means that significant additional rework is required to incorporate the latest clinical guidelines in the app. Further, our app was designed with the specific aim of assessing young children, this may be deemed a limitation in terms of supporting broader patient healthcare assessment needs including expectant mothers, other communicable diseases , escalating instances of noncommunicable diseases and general adult healthcare. --- Appendix 4 Structured interview guide Research questions for rural healthcare workers --- How do you feel about this new mHealth app? To what extent do you think that this new app would have a positive impact on your work practices? --- A sociomaterial exploration of mHealth To To what extent do you think that healthcare workers can perform their duties using this new app without outside help? What, if any, challenges do you think that rural healthcare workers would face when trying to get familiar with using this new app? Is there any reason why you think that rural healthcare workers would avoid using this new app in the future? How do you feel after using this new mHealth app? After To what extent do you think that rural healthcare workers can perform their duties using this new app without outside help? What, if any, challenges do you think that rural healthcare workers would face when trying to get familiar with using this new app? Is there any reason why you think that rural healthcare workers would avoid using this new app in the future? How do you feel after using this new mHealth app? A sociomaterial exploration of mHealth To what extent do you think that rural healthcare workers can perform their duties using this new app without outside help? What, if any, challenges do you think that rural healthcare workers would face when trying to get familiar with using this new app? Is there any reason why you think that rural healthcare workers would avoid using this new app in the future? How do you feel after using this new mHealth app?

Purpose -The implementation of mobile health (mHealth) in developing countries seems to be stuck in a pattern of successive pilot studies that struggle for mainstream implementation. This study addresses the research question: what existing health-related structures, properties and practices are presented by rural areas of developing countries that might inhibit the implementation of mHealth initiatives? Design/methodology/approach -This study was conducted using a socio-material approach, based on an exploratory case study in West Africa. Interviews and participant observation were used to gather data. A thematic analysis identified important social and material agencies, practices and imbrications which may limit the effectiveness of mHealth apps in the region. Findings -Findings show that, while urban healthcare is highly structured, best practice-led, rural healthcare relies on peer-based knowledge sharing, and community support. This has implications for the enacted materiality of mobile technologies. While urban actors see mHealth as a tool for automation and the enforcement of responsible healthcare best practice, rural actors see mHealth as a tool for greater interconnectivity and independent, decentralised care. Research limitations/implications -This study has two significant limitations. First, the study focussed on a region where technology-enabled guideline-driven treatment is the main mHealth concern. Second, consistent with the exploratory nature of this study, the qualitative methodology and the single-case design, the study makes no claim to statistical generalisability. Originality/value -To the authors' knowledge, this is the first study to adopt a socio-material view that considers existing structures and practices that may influence the widespread adoption and assimilation of a new mHealth app. This helps identify contextual challenges that are limiting the potential of mHealth to improve outcomes in rural areas of developing countries.

INTRODUCTION In the face of expanding numbers of migrants, knowledge regarding the influence of culture-specific and migrationrelated factors on mental health presents a substantial issue for research in the public health sector. In Germany, every fourth person of the population has a migration background, either having personally moved there or having at least one parent who immigrated . With 2.8 million, individuals of Turkish origin constitute the largest ethnic group in this country . Population-based health surveys in Germany , the Netherlands , and Belgium as well as clinical studies have consistently demonstrated higher prevalence rates of depression and increased levels of depressive symptoms in Turkish migrants in comparison with the majority populations and other migrant collectives . Women of Turkish origin frequently demonstrate higher levels of depressive complaints than men and also higher prevalence rates . Several sociodemographic risk factors for depression in Turkish migrants have also been identified: older age , a low socioeconomic status , unemployment , no current partnership , belonging to the first migration generation ; however, other studies found no significant differences between generations . The migration into a host country as well as the post-migration acculturation process is associated with various stressors. Acculturation is a complex, multifaceted, and long-term process of psychological and social changes resulting from continuous interaction between individuals from different cultures . The most renowned theory on acculturation is the one presented by Berry , which proposes four acculturation styles: integration, assimilation, separation, and marginalization. In the majority of studies on the association between acculturation and depression conducted among Turkish migrants living in Germany or other European countries as well as among other migrant populations, the acculturation style of integration or assimilation was often associated with a better mental health status, whereas poor outcomes were consistently predicted by marginalization . While depression represents a negative indicator for affective well-being, life satisfaction stands for the cognitive well-being . LS is defined as a person´s global evaluation of contentment with life . It involves all aspects that determine the individual quality of life, such as health-and job-related, social, financial or other factors. Previous international research on LS in migrants reveals inconsistent findings. Some population-based studies have identified lower LS among migrants in comparison with the majority population, in particular for the second generation or for the first generation migrants , however, also greater levels of LS in migrants or similar LS have been reported . A population-based survey in Germany showed lower LS in the migrants as compared with native-born Germans, however the effect size was minimal; migrants from Turkey reported the lowest LS . Also in other Western countries a lower extent of LS has been identified in Turkish migrants in relation to the majority populationso for example in the Netherlands . In a study examining migrant mothers of Turkish origin in Germany, a higher socioeconomic status was related to higher LS . Among women of Turkish origin living in Great Britain, the integration style and higher religious identity was found to be associated with higher LS . Due to the lack of gender-stratified analyses with persons of Turkish origin in Germany in population-based surveys or large samples as well as inconclusive findings, the central objective of the present study was to examine the association between acculturation and depressive symptoms among women and men of Turkish origin in Germany. Acculturation was analyzed by applying a categorization of four acculturation styles . The aims of the study were: 1. to examine the differences regarding the frequency and severity of depressive symptoms and the degree of LS between women and men of Turkish origin and the first and second migration generation ; 2. to explore the association between acculturative styles and the severity of depressive symptoms when adjusted for sociodemographic and migration-related variables separately in women and men of Turkish origin. Based on prior research, we hypothesized higher levels of depressive symptoms and lower LS to be prevalent in women and in the first generation. Furthermore, we postulated separation and marginalization to be associated with higher levels of depressive symptoms in reference to integration and no significant difference between assimilation and integration. --- MATERIALS AND METHODS --- Sample Description and Procedure Participants were recruited in the city of Essen between December 2011 and August 2012 using two recruitment methods: a community-orientated and a register-based strategy [for details see ]. Within the community-orientated approach, the representatives of the Turkish community supported the recruitment of persons of Turkish origin as key persons. Furthermore, the social networks of Turkish migrants were contacted for recruitment such as mosques, Turkish speaking general practitioners and doctors of other medical professions, Turkish parents´and teachers´associations, as well as other associations, Workers´Welfare Organization, neighbors or relatives of the participants. In addition, the investigation was widely promoted in the Turkish community of Essen by wordof-mouth recommendation. Persons interested in participation in the study contacted the study center by phone. 319 persons of Turkish origin participated in the study as a result of the community-orientated recruitment. Within the register-based approach, a random sample of 1498 potential participants with Turkish citizenship from the Essen population registry was extracted and potential participants were sent a written invitation for the study in both Turkish and German. In the case of no response within approximately two weeks a second invitation letter, after further two weeks a third letter were sent. 286 participants were recruited by the register-based method. The response rate for the register-based approach was 19.4%. Inclusion criteria for the study were: age between 20 and 69 years [due to the design of the German National Cohort Study ], agreement to participate in the study, the status of a person of migration background according to the definition applied in epidemiological research in Germany , and principal residence in Essen. The investigation consisted of different medical examinations and a set of self-report questionnaire including a self-administered questionnaire on mental health issues. Questionnaires were available in Turkish and German. Participants were asked to complete the mental health module at home and send it back to the examination office. For further details see Reiss et al. . If there was no response to the questionnaires of the mental health module within the designated time period, the participants were contacted by phone. The data have already been used for other publications focusing on somatic symptoms/somatization among Turkish migrants and the association between acculturation and health-related quality of life . There is a topical overlap with our previous published work concerning the relationship between acculturation and depressive symptoms, however in our prior work we have investigated clinical samples, whereas the present study has examined a non-clinical sample. The additional knowledge provided by this study results from the gender-stratified analyses allowing the investigation of genderspecific differences. The advantage of the present sample is also based on its composition . Therefore, a sample selection bias may be reduced. --- Ethics Statement The present study was approved by the Ethics Committee of the Medical Faculty of the University of Duisburg-Essen . Written informed consent was obtained from all participants. --- Measures --- Socio-Demographic and Migration-Specific Variables The following socio-demographic characteristics of the participants were assessed: gender, age, partnership, marital status, education level, employment status, and monthly household income. The following migration-specific variables were assessed: if born in Germany or length of residence in Germany and age at time of migration into Germany, citizenship, and language proficiency. Patient Health Questionnaire: Depression Module Depressive symptoms were assessed with the depression module of the Patient Health Questionnaire . It consists of 9 items and is based on DSM-IV criteria. The sum scores range from 0 to 27 representing mild, moderate and severe levels of depressive symptomatology with cut-off-points of ≥5, ≥10, and ≥15. The psychometric characteristics of the PHQ-9 are well documented . The internal consistency for the German version is 0.86-0.89 and for the Turkish version 0.86 . In the present sample the validated German version obtained a Cronbach´s Alpha of 0.91, the Turkish version a Cronbach´s Alpha of 0.89. --- Frankfurt Acculturation Scale The acculturation strategies were measured with the Frankfurt Acculturation Scale , a self-report questionnaire comprising 20 items rated on a seven-point Likert scale . The questionnaire consists of two indices assessing the degree of orientation towards culture of origin , and the degree of orientation towards the host culture . Higher scores indicate a higher orientation towards the CO or HC, respectively. In the present study the Cronbach's Alpha for the Turkish version was 0.75 for CO and 0.53 for HC and for the German version 0.86 for CO and 0.79 for HC. The medians of both indices of the acculturation scale were used as cut-off points to divide the sample into a group of subjects with a low CO vs. a group with a high CO, as well as a group of subjects with a low HC vs. a group with a high HC. We categorized the participants into four groups based on the schema of Berry : integration , assimilation , separation , and marginalization . --- Life Satisfaction General life satisfaction in Germany was assessed with a single item on a scale from 1 = bad to 5 = excellent. --- Statistical Analysis Data analyses were conducted with SPSS V. 21. Missing values in the questionnaires were replaced by the expectation-maximization algorithm . Descriptive statistics were computed to profile the socio-demographic and migration-specific sample characteristics. To explore the differences regarding age and gender between the non-respondents from the population registry and the participants recruited with the register-based strategy, a t-test for independent samples and a binomial test were performed, respectively. Gender differences in the frequency of being depressed and the association between the acculturative styles and the language proficiency were tested with the c 2 -test. An analysis of covariance was calculated to examine the differences in degree of depressive symptoms in relation to relevant socio-demographic and migrationrelated characteristics , controlling for age. The effect sizes were also reported . Gender-stratified multiple linear regression analyses with enter method were performed to investigate the influence of the acculturation styles as well as socio-demographic and migrationrelated characteristics on the severity of depressive symptoms in women and men. In case of missing values, the pairwise exclusion was used. Multicollinearity was checked by calculating the variance inflation factors. A level of significance of p<.05 was predetermined in all analyses. --- RESULTS A total of 605 mental health questionnaires were distributed among the study participants . Of those, 395 questionnaires were returned. Ten participants did not fulfill the inclusion criteria, and 30 respondents had to be excluded from the analyses because of too many missing values in the Frankfurt Acculturation Scale . In total, 328 individuals of Turkish origin were included in the study. Nearly half of them were recruited by the register-based approach. A comparison of the respondents and the non-respondents within the register-based approach demonstrated that the nonrespondents were significantly younger than the study participants , and less frequently women . --- Socio-Demographic and Migration-Specific Data In Table 1, the socio-demographic characteristics for women and men of Turkish origin and the total study sample are presented. Of 328 participants included in the present study, 201 were female. The average age of the participants was 41.6 years . Most of them were married, had a secondary/vocational education and a monthly household income of 1,250-2,500 € and were unemployed. Women were less frequently in a partnership, were more frequently separated/ divorced and unemployed and had more frequently a low education status in relation to men. Among men twice as many persons were job seeking than among women. Table 2 reports the migration-related data of the participants. The large part of the sample belongs to the first generation of migrants and has the Turkish citizenship. Most of the Turkish migrants came as young adults to Germany, are living in the destination country for 15-24 years and assess their proficiency of the German language as moderate [however, 93% of the second generation vs. 31.8% of the first generation report good language proficiency or even German as mother tongue ]. The distribution of migration-related characteristics is similar for both genders. --- Acculturative Styles Based on the gender-stratified median split of the FRACC, 21.9% of the women with Turkish origin were categorized as integrated, 28.4% as assimilated or separated, respectively, and 21.4% belonged to the marginalization group. In men, 26.8% were classified as integrated , 26.0% as assimilated or separated, respectively, and 21.3% were allocated to the marginalization group. --- Acculturative Styles and German Language Proficiency When regarding the association between the acculturative styles and the language proficiency , 85.2% of the separated women, 72.1% of the marginalization group, 61.0% of the integrated and 21.4% of the assimilated women reported bad to moderate language proficiency . The difference was significant =51.01, p<0.001). In men, the respective proportions were: 96.9% for separation, 53.8% for marginalization, 45.5% for integration and 24.2% for assimilation. The difference was significant =36.49, p<0.001). --- Frequency and Severity of Depressive Symptoms The cut-off-value of ≥10 for the PHQ-9 was achieved by 33.2% of the women and 26.4% of the men =1.58, p=0.209, j=0.071). The mean age-adjusted PHQ-9 score for the female migrants was M=7.81 and for the males M=6.70 . The ANCOVA with severity of depressive symptoms as dependent variable, gender as independent variable and the covariate age revealed no significant association between depressive symptoms and gender [F=2.23, p=0.137, h 2 =0.007] . Also no significant inter-generation difference was observed among women, however the significance level was only slightly failed [first generation: M=8.24, SD=6.42 vs. second generation: M=5.82, SD=6.82; F=3.57, p=0.060, h 2 =0.018]. Men who have immigrated to Germany demonstrated a significantly higher age-adjusted degree of depressive symptoms than men who were born in Germany [M=3.30, SD=7.25; F=5.75, p=0.018, h 2 =0.046]. When controlling the impact of the joblessness on depressive symptoms by exclusion of the job-seeking persons from the analysis, women showed a trend for being more frequently depressed than men [c 2 =3.68, p=0.055, j=0.117] and also demonstrated a trend for higher levels of depressive symptoms [F =3.27, p=0.072, h 2 =0.012]. --- Predictors of Severity of Depressive Symptoms To examine the influence of acculturative styles as well as sociodemographic and migration-related variables on the severity of depressive symptoms, multiple linear regression analyses were performed separately for women and men. In females as well as in males, in the unadjusted regression model separation and marginalization were significantly associated with higher levels of depressive symptoms in relation to integration. The variance explained was 5.4% for women and 12.0% for men. In the adjusted regression model , the acculturative styles and the socio-demographic variables age, partnership, education level and employment status as well as the migration-related variables migration generation and German language proficiency were included into the model as predictors and the sum score of depressive symptoms as the criterion variable. In migrant women, separation as acculturation style , having no partnership and lower education were associated with higher depression levels. The variance explained was 8.9%. In migrant men, separation as acculturation style and employment status were related to depression levels . --- Life Satisfaction Regarding the age-adjusted mean scores of LS, both genders did not differ significantly [women: M=3.03, SD=0.81 vs. men: M=3.09, SD=0.82, F=0.36, p=0.547, h 2 =0.001] . This was also the case after controlling for unemployment status. Among migrant women, the ANCOVA with degree of LS as dependent variable, migration generation as independent variable and the covariate age showed a significant association between LS and the generation of migration [F=8.58, p=0.004, h 2 =0.043]: the first generation migrants achieved a significantly lower age-adjusted level of LS in comparison with the second generation migrants . The same findings were found for migrant men: [F=8.71, p=0.004, h 2 =0.066]: the first generation migrants showed a significantly lower LS in relation to the second generation migrants . --- DISCUSSION The first aim of the present study was to examine gender-specific differences regarding the frequency and severity of depressive symptoms and the degree of LS in migrants of Turkish origin. After controlling for unemployment status , women showed a trend for being more frequently depressed than men and also demonstrated a trend for higher levels of depressive symptoms. These findings are in line with the majority of studies demonstrating increased levels of depressive symptoms in women as compared with men in general and in female in relation to male migrants . There are several possible explanations for this result. The gender gap may reflect the gender difference generally observed in the normal population showing a higher vulnerability for mental disorders in women . Furthermore, it may result from culturally influenced gender roles allowing women present more symptoms than men. Increased frequencies and severity of depressive symptoms among women may also be attributed to higher acculturative stress, higher burdens and a worse socio-economic status in comparison with men. Besides, neurobiological factors , differences in coping skills and personality characteristics as well as adverse experiences in childhood may explain the gender gap. Probably an interaction between multiple factors such as the ones mentioned above and also other factors not examined in this study could have contributed to the gender gap in depression. The second central aim of this survey was the investigation of the association between acculturative styles and the severity of depressive symptoms when adjusted for socio-demographic and migration-related variables separately in women and men of Turkish origin. A major finding was the observation that separation was related to higher symptom severity in both genders in reference to integration. These results confirm previous research showing separation to be associated with higher levels of depressive symptoms in Turkish migrants compared to integration in non-clinical as well as in clinical samples and also in other migrant collectives . This negative association between separation and mental health status was also found for other indicators of mental health, such as health-related quality of life . There are several possible explanations for this pattern. It can be assumed that separation is connected with fewer skills and resources that are required to challenge successfully the demands of living in the country of destination such as acquisition of satisfactory education or positions in the employment domain or access to healthcare institutions or social services. Thus, a low adaptation to the host society may increase acculturative stress and frustration and finally lead to elevated levels of depressive symptoms . Some separated persons may experiencein a chronically stressful waya kind of social defeat due to inadequate competences in managing effectively and satisfactory life in the new society. A prolonged experience of outsider status may contribute to the manifestation of depressive symptoms. Social defeat has been demonstrated to be a stressor inducing depression . The negative relationship between separation and mental health may also be attributed to high levels of perceived discrimination in separated persons that moderate depressive symptoms. Highest degree of perceived discrimination has been found in separated and marginalized women of Turkish origin . Perceived discrimination has been shown to be related with depressive symptoms in Turkish migrants as well as other migrants groups . The European Social Surveys from 2006/2007 and 2014 confirm the association between perceived discrimination and depressive symptoms in migrants and also non-migrants. Also other psychological or psychosocial factors may moderate the association between separation and depressive symptoms in migrants, for example self-esteem has been demonstrated to moderate the effect of perceived discrimination on depression . In some studies, an association between a higher level of acculturation and better performance on cognitive function tests has been detected . In separated and marginalized migrants, a higher proportion of persons with a low education level has been observed than in integrated or assimilated individuals . Finally, due to the cross-sectional design of the present study, a reverse relationship between separation and depression cannot be excluded. It could also be possible that depressive persons prefer separation as acculturative style. Prospective studies are needed to examine the causal direction of this association. Apart from separation that was found to be significantly associated with depressive symptoms for both sexes, also genderspecific significant predictors for the severity of depressive symptoms were identified in the present study, namely, having no partnership among women and not being employed in men. These different results for both genders may reflect gender-specific cultural expectations towards gender roles which play a crucial role especially in socio-centric societies such as the Turkish one. The traditional Turkish culture is characterized by a strict gender and generation hierarchy and a strong family cohesion/integration and close interpersonal relationships . A woman is expected to be primarily responsible for the household, child-rearing, and relationship maintenance, and a man should provide financial support for his family. Thus, having no partnership in women and unemployment in men contradicts the perception of feminity and masculinity, respectively, in the Turkish culture and may decrease the self-esteem as a woman or a man, evoke feelings of shame or guilty and finally depressive symptoms. In our investigation, the impact of assimilation on the manifestation of depressive symptoms did not differ significantly from integration. This result is consistent with some previous findings , while others have shown integration to be the most favorable acculturation style for mental health . A possible reason for the similar effect of integration and assimilation may be due to the migration policy in Germany. In the context of migration it is important to emphasize that the degree of acculturation to the host society depends not only on the migrants themselves but also on a country´s national migration policy that constitutes institutional support and acceptance of migrants or contributes to their social exclusion and barriers . In Germany, in many decades of the 20 th century a model aiming at a sociocultural assimilation of migrants was followed by German authorities. Furthermore, orientation on the host culture is more important for an adequate adaptation than maintaining the cultural identity of the heritage culture. In some contexts, assimilation may even be more beneficial, e.g., for academic achievements, while integration is favorable for indicators of psychological adaptation . In that context, the influence of perceived discrimination which seems to be higher in second generation migrants is worthy to be mentioned, especially if they try to integrate their cultural heritage after following an assimilation strategy or perceive "transparent barriers" in their career or at the level of political participation. El-Mafaalani has presented these findings as the so-called "integration paradox" and could explain new acculturative stressors if more participation is desired and could even lead to new contextual separation strategies during the life course. Contrary to our expectations, marginalization was not significantly related to higher severity of depressive symptoms as compared with integration. However, in women marginalization revealed a tendency for higher depression levels. The low explained variance in the regression analyses for depression is worth to be mentioned: 8.9% and 17.2% . It may be postulated that other essential protective and risk factors not examined in the present study, such as sense of coherence , self-esteem , social support or discrimination may also substantially influence the manifestation of depressive symptoms. In terms of inter-generational differences for LS, in the present study lower age-adjusted levels of LS were observed in the first migrant generation compared to the second. The research on LS in dependence of migration generation is inconclusive. Some surveys confirm the result found in our investigation [e.g., Fugl-Meyer et al. ], while others report contrary results [e.g., Knies et al. ]. The first generation may be faced with many substantial losses and in some cases with higher acculturative stress than it is the case for the second generation that is already socialized in the new country. In addition, in the second generation more skills and less barriers may be mostly postulated that contribute to a successful sociocultural adaptation andas a resulta higher LS. Furthermore, in the second generation the identification with the host country is probably higher than in the migrants that have left their home country. Identification with the host country have been revealed to be predictors of higher LS . --- Strengths and Limitations A strength of the present survey is the relatively large sample size that allowed stratification for gender to conduct separate analyses for women and men. Another advantage of this study was the inclusion of respondents with insufficient knowledge of the German language due to the community-orientated recruitment strategy and the bilingual questionnaire. The results should be viewed in the light of some limitations. The sample is not strictly population based because of the network sampling approach in parts of the sample and the low response rate among the register-based sample . Thus, the findings cannot be generalized to the general population of Turkish migrants in Germany or to other migrant collectives. Another limitation is the cross-sectional study design that does not allow for drawing causal conclusions regarding the influence of the variables measured. Besides, the classification of the participants into the four acculturation groups was based on the medians for the acculturation strategy orientation on the culture of origin and the host culture observed in the present sample due to the lack of cut-off or norm values. Thus, no extern criteria exist and the categorization may vary depending on the medians in different samples. Furthermore, one can call into question if the four acculturation styles proposed by Berry cover the full spectrum of all existing strategies given that people may adapt different strategies depending their personal experiences and changing societal constraints and expectations. Also the low Cronbach's Alpha for the Turkish version of the subscale orientation on the host culture may decrease the reliability of the measure of acculturation. Finally, the significant differences between responders and non-responders could also present a selection bias so that it can be assumed that persons participated in the study may belong to more vulnerable individuals with a higher tendency for the manifestation of depressive symptoms and a lower LS than the non-responders. --- CONCLUSION The results of this study indicate that the acculturative style separation is associated with higher levels of depressive symptoms in migrants of Turkish origin living in Germany as compared with integration. Therefore, gender-sensitive health promotion programs should target separated migrants to improve their integration into the German society. In addition, our findings suggest genderspecific protective and risk factors for depression. Culturespecific factors should be appropriately taken into consideration in prevention and intervention programs. Special attention should be given to identify depressive symptoms in female migrants to improve the LS . Further research should investigate the role and interrelationships of further migration-and acculturation-related stressors and non-migration-related factors as well as culture-specific aspects in the context of depressive symptoms and LS in Turkish migrants and other migrant collectives. It remains an important issue for public health research and policies to investigate mental health of migrants in representative surveys with special attention to gender-specific differences. --- DATA AVAILABILITY STATEMENT The datasets generated for this study are available on request to the corresponding author. --- ETHICS STATEMENT The studies involving human participants were reviewed and approved by Ethics Committee of the Medical Faculty of the University of Duisburg-Essen. The patients/participants provided their written informed consent to participate in this study. --- --- Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The aim of the present study was to investigate the degree of depressive symptoms and life satisfaction (LS) as well as the association between acculturation and depressive symptoms among a large sample of persons with Turkish migration background in Germany, taking into account gender-and migration-related differences. Methods: This study was part of a pretest for a large national epidemiological cohort study in Germany. Acculturation was measured using the Frankfurt Acculturation Scale (FRACC). Based on the median split of the two subscales, four acculturative styles according to Berry (integration, assimilation, separation and marginalization) were determined. Depressive symptoms were assessed with the depression module (PHQ-9) from the Patient Health Questionnaire. LS was assessed with a single item on a scale from 1 = bad to 5 = excellent. Differences in levels of depressive symptoms and LS in relation to gender and generation of migration were tested with analysis of covariance, controlling for age. Gender-stratified multiple linear regression analyses were also conducted for depressive symptoms as criterion variable. Results: 328 Turkish migrants participated (61.3% women). The cut-off-value of ≥10 for the PHQ-9 was achieved by 33.2% of the women and 26.4% of the men (p=0.209, j=0.071). In female migrants, the age-adjusted mean score for depressive symptoms was 7.81 (SD=6.42), in males 6.70 (SD=6.41) (p=0.137, h 2 =0.007). After controlling for unemployment status, women showed a trend for being more frequently depressed than men (p=0.055, j=0.117) and also demonstrated a trend for higher levels of depressive symptoms (p=0.072, h 2 =0.012). No significant gender-specific difference was found concerning age-adjusted mean score for LS (p=0.547, h 2 =0.001), also when controlled for unemployment status (p=0.322, h 2 =0.004). In both sexes, the second generation

Introduction Pastoral care and counseling are an integral and inseparable part of church ministry. This ministry is not only a task that pastors must carry out but also a ministry task for ministers and church members. Pastoral care and counseling must concern all elements in the church because all age categories experience life struggles . Small children with school struggles , teenagers with friendship struggles , youth with identity search struggles , and parents with family . It is wise and right to pay attention to all the categories in the church. The global and global Covid-19 pandemic has created a new era known as the era of disruption. Many people say that 2020 is an era of disruption, where many new things happen and emerge. The definition of disruption is an era in which innovation and change occur on a large scale and fundamentally change all existing systems and orders in new ways . The same thing was conveyed by that the era of disruption that changed the entire human order caused chaos in thinking and acting. Innovation and change on a large scale force society to adapt quickly. This certainly increases the pressures and struggles of life because humans do not only compete with humans but also with machines and technology. Another impact of the Covid-19 pandemic has also caused many problems in the socio-economic sector. UNICEF, United Nations Development Program , Australia Indonesia Partnership for Economic Development , and the SMERU Research Institute conducted a series of assessments of the impact of COVID-19 in Indonesia on the socioeconomic dimension. The study found that crises and problems occurred in various ways, including employment, income, resilience to shocks, schools, social protection coverage, and access to immunization and health services . The crisis has an impact on physical, mental, and spiritual health. The term today is called "burnout." The burnout that is not handled correctly will impact the loss of enthusiasm for life, depression, prolonged stress, and loss of enthusiasm to achieve goals or plans that have been set before. In responding to the crisis and burnout, the church expects to take a solution. The most appropriate solution is to implement and prioritize pastoral principles. The need for pastoral assistance and counseling today is increasingly felt in various sectors of society, both Christian and non-Christian. The economic, social, and political crisis, which resulted in crises in other fields, including health, education, and morals, made a crisis that had a widespread impact in Indonesia. Consciously or not, this has triggered the community's need for pastoral care and counseling. Pastoral care and counseling are services by finding and visiting congregations experiencing struggles in their lives . The purpose helps them through an interactive, reciprocal, and in-depth conversation to find the core of the problem and lead to a solution. However, for this interaction to occur, hospitality practices are undoubtedly needed so that the counselee is willing to be involved and open up. The church should be open and have hospitality towards its members. The Lord Jesus first showed this attitude: "Come to me, all who are weary and heavy laden, and I will give you rest." . Suppose the church does not open up and does not have hospitality to its members who need guidance. In that case, the church is like closing the "entrance" to the Christian community. They will choose to look for "non-Christian counselors" such as psychiatrists, psychologists, and other professional experts. This reality is a challenge for the church in pastoral care and counseling. Hospitality is the entrance to interaction in fellowship and community so that fellowship with God, who is the head of the church, can be enjoyed by all people. Therefore, the author considers hospitality a critical entry point in pastoral care and counseling. In the following article, the author will discuss Christian hospitality as an entry point for pastoral care and counseling from various perspectives. Research related to Christian hospitality has been reviewed by , which examines from the perspective of Christian religious education. The primary purpose of his research is how Christians can place themselves in social relations by using a qualitative interpretative method, the results of the study state that the value that must be possessed to achieve Christian hospitality in a pluralistic society is the law of love. --- Method The method used in this research is descriptive qualitative. The research procedure starts from the study of the Bible about the idea of hospitality that God initiated through Christ by using several texts related to the theme. Then the analysis is relevant to implementing pastoral care and counseling in the church for all categories. The literature approach also examines some expert views on hospitality concerning pastoral care and counseling. By analyzing and considering some of these views and reflections from the Bible. Conclusions are obtained to be used as input for counselors in the church-related to pastoral assistance and counseling in churches during times of crisis. Researchers, in this case, are limited to hospitality design from ISSN : Print 2598-3199 -Online 2598-3210 a Christian perspective, which is relevant to pastoral assistance and counseling. The design and construction can later become a benchmark for the success of pastoral assistance and counseling in the church. --- Results and Discussion --- Pastoral Counseling Pastoral counseling is a reciprocal relationship between God's servants as counselors and counselees. The counselor tries to guide the counselee into an ideal counseling conversation atmosphere that allows the counselee to truly recognize and understand what is happening to himself, his problems, his living conditions, where he is, and so on. So that he can see the purpose of his life in his relationship and responsibility to God and try to achieve that goal with the measure, strength, and ability God has given him . Christian counseling is different from secular counseling. At least according to , what distinguishes Christian pastoral counseling from secular counseling is that Christian pastoral counseling is a ministry entrusted by God Himself, absolutely dependent on the power of the Holy Spirit, based on the truth of God's word, and based on theology. In its integration with other sciences . Although the purpose of pastoral counseling is not the same as evangelism, both have the same primary goal: to become the human being that the Lord Jesus wants or to have an abundant life in the Lord Jesus . In contrast to secular counseling, pastoral care and counseling by Christian counselors can succeed if the path is open and the counselee has the will, determination, and courage to achieve the fullness of Christian life. The recovery in pastoral counseling can be achieved when the counselee gives himself to be visited by the Holy Spirit through the truth of God's word and accepts God's will for his life. To achieve the goals of pastoral care and counseling mentioned above, an element of openness and freedom is needed to express the feelings and struggles of life experienced by the counselee. According to , the diagnosis and analysis of the subject matter cannot be correct if there is no openness and freedom on the part of the counselee to express his problems and feelings. To be open and free to express, the counselee must feel comfortable in his relationship with the counselor. An effective counselor certainly has good qualifications as a Christian counselor in pastoral counseling. According to , the qualifications of a Christian counselor are as follows: have good spirituality, know God's will in his life, have love and are willing to help ease the burden, have knowledge of God's word and use it wisely. Righteous, gentle, patient, non-judgmental, empathetic, recognizes abilities and limitations, and relies on and relies on God's wisdom and strength. From the qualifications mentioned above, it can be seen that knowledge is not the only thing that is a requirement to become a Christian counselor. A Christian counselor must also have good character and a good relationship with God. In addition to these qualifications, the counselor must also be ready to deal with the counselee. Even though a counselor has good knowledge, character, and spirituality but does not have physical, time, or other readiness, pastoral counseling not carries out properly. --- The Nature of Christian Hospitality One of the things that can affect the course of the pastoral counseling process is communication. If communication does not run smoothly, pastoral counseling will not work well. Communication will be more complicated if the counselor and counselee do not know each other beforehand. Difficulty in communication can also occur if the counselee is reluctant to cooperate with the counselor in pastoral counseling if the first impression is not soothing or a tense/strange situation occurs . Seeing this, one thing that needs to be implemented is hospitality , although, on the other hand, there is undoubtedly the ability of the counselor to approach the counselee. Hospitality is etymologically derived from the Greek philoxenia, which consists of philos and xenos , which means love for strangers. Hospitality is more than an act . Hospitality is a way to make friends with people who are the responsibility of every human being . Hospitality is not just a concept; there needs to be action or practice in everyday life to establish relationships with others. Hospitality is an open attitude towards different. Hospitality embodies a sense of warmth in accepting, respecting, and friendly and fraternal attitudes toward others . Christian Hospitality is built on God's love, namely agape and unconditional love. God's love is love and not 'what is. If hospitality builds on 'what is up' love, then hospitality only happens to people who are friendly, kind, and love us, whereas when they do not do that, we are no longer friendly to them . This is the hallmark of Christian hospitality, which is built on God's unconditional love, does not discriminate, forgives, and accepts all people. Offering hospitality to another means creating a space where he or she feels welcome, like "at home" . A guest feels at home when allowed to be who he is. Apart from following the generally accepted rules of good behavior, he does not have to act in a certain way at the host's place to be accepted by him. He was given the freedom to come as-is. The Hospitality of Jesus is an example. We can see the hospitality Jesus shows as a manifestation of God's hospitality. In Luke's perspective, Jesus is the anointed Messiah, whose whole life and ministry were filled with the Holy Spirit to "preach the good news to the poor; and he has sent me to proclaim release to the captives, and a sight to the blind, to set free the oppressed, to proclaim the year of the Lord's mercy has come" . Therefore, in this case, according to , we can understand that the whole life and ministry of Jesus was shaped pneumatically or pneumatological. The Holy Spirit helped Jesus to serve effectively and touch today's society. Luke describes it more explicitly in the Acts of the Apostles how the Holy Spirit helped the disciples to be able to break through the walls of discrimination at that time . Jesus first demonstrated God's hospitality as one who received hospitality from others. The life of Jesus, from his conception in Mary's womb by the power of the Holy Spirit to his birth in a manger to his burial , Jesus depended on the reception of others. As the "son of man" who had no place to lay His head , Jesus relied on the goodness of many such as staying in their homes and receiving whatever was given . Jesus' hospitality can also be seen in the story of Zacchaeus climbing a tree to see Jesus , giving us a perfect model of this dynamic of Jesus' visit as an opportunity to experience God's hospitality . Zacchaeus was excommunicated for two reasons. In the first place, he was of short stature and could not see the crowds. In the second, he was a tax collector and, therefore, part of a class shunned by those who hated collusion with the colonial rulers of Rome . However, Jesus went beyond this exclusive attitude and invited Himself to Zacchaeus' house . He spoke the words of great healing and deliverance: "Today salvation has come to this house, for this man also is a son of Abraham" . In this case, God invites and accepts Zacchaeus into the kingdom of God, and this is God's hospitality which is shown through Jesus' role as a guest who comes to Zacchaeus' house. God sent Jesus into the world to call people into the community of salvation. The encounter between Jesus and Zacchaeus is the perfect paradigm of God's hospitality. Zacchaeus' healing visits and divine love show that lost humans can find welcome and new life in God's 'friendly hand.' Here we have a model for our communal life in terms of inviting and welcoming. Jesus' hospitality was shown by serving a banquet for his two disciples: "When he sat down to eat with them, he took bread, gave a blessing, and broke it and gave it to them." . In his guest role, Jesus also showed hospitality when he was invited to stay with two disciples on their way to Emmaus . As they experienced Jesus' hospitality in serving the banquet, "their eyes were opened, and they knew him" . The hospitality that Jesus displayed brought them to the knowledge of God. Likewise, when we show Christian hospitality, based on divine love, in pastoral care and counseling, the counselee will begin to know God, enter into the religious community, and fellowship with Him. Jesus showed that Christian hospitality departs from loving others, namely all people, including strangers and even people who are hostile to us. Furthermore, Jesus taught the disciples and the crowd about hospitality through the good Samaritan story . The exciting thing is that Jesus told this story sometime before the visit of Jesus, and the disciples were rejected in the area of Samaria . Despite experiencing the attitude of rejection from the Samaritans, Jesus remained friendly and brought up the story of the good Samaritan. In the story, Jesus taught that the Samaritan fulfilled the law of loving one's neighbor. The love that the good Samaritan did was to help his fellow human beings who were even his enemies because the Jews and Samaritans were enemies. Following the example of Jesus, in pastoral care and counseling, we must not choose people as counselees. The church is open to others, namely, everyone who comes asking for assistance and direction ISSN : Print 2598-3199 -Online 2598-3210 in God's love. The church should not be exclusive. The church is called to show the hospitality of Christ even if there are the poor, big sinners, criminals, adulterers, thieves, and others. It often happens when the counselee initially rejects, belittles, or even insults the church. However, like the example Jesus set, when they came to ask for assistance, the church should not repay evil for evil but show love through divine hospitality by still welcoming and inviting them. According to Pembroke, Christ came in the power of the Spirit to invite people into fellowship with God. To be indeed human means to share in fellowship with the Living God. Jesus came to call people to become fully human through the offer of a saving relationship. He is the model and bearer of this healing and liberating fellowship . He is the link between God and man. When people conform to Jesus, they are united with God. Pastoral care and counseling that brings the counselee into the community, led by the Holy Spirit and living like Christ, is shaped by radical openness and acceptance. When pastoral care and counseling function optimally, all who come are welcomed and confirmed as children of God. Christian Hospitality in Pastoral Assistance and Counseling. Christian hospitality is essential in achieving good communication in pastoral care and counseling. Suppose the encounter is not initiated and built through Christian hospitality. In that case, the communication and implementation of pastoral and counseling assistance feel 'tasteless.' Without Christian hospitality, the counselor may still be able to establish communication with the counselee. Mentoring can still carry out, but the grace of God who blesses cannot be seen and felt, so the fundamental goal of pastoral counseling is to invite the counselee to enter and stay in fellowship with God. Therefore, Christian hospitality is essential in pastoral care and counseling because it aims to present God's love to the counselee so that they accept the invitation, enter, and stay in fellowship with God. Although counselors already know about pastoral counseling, good character, and the ability to convince their counselees, this is not enough. In pastoral care and counseling, counselees need a sense of acceptance and empathy that can realize through Christian hospitality. To help others feel welcome, the counselor not only gives her undivided attention but also creates a space where guests feel free to be themselves. According to Pembroke, there is a paradox in hospitality: counselors want to create emptiness, not scary emptiness, but a friendly emptiness where strangers can enter and find themselves freely; free to sing their song, speak their language, dance their dance . Hospitality is not a subtle invitation to adopt a counselor's lifestyle but a gift of opportunity for guests to discover their style. The Lord Jesus accepts us as we are, but He does not force us to dress like Him, have nothing, be rabbis, have long hair, or so on. Jesus accepts us as we are, so we invite to feel God's love and respond to it in our way. Likewise, the counselor needs to sincerely accept the counselee to channel God's love so that when the counselee experiences God's love, the counselee can respond in his way. In providing acceptance, hospitality creates a friendly 'space' so that clients can feel free to be themselves. However, it is essential to note that this freedom has its limits. Values of love, respect, togetherness, and justice form the boundaries of Christian hospitality. In acceptance, the counselee is still invited to respect the values of love, respect, togetherness, and justice while remaining free to be himself. Clients must be challenged and guided when they go beyond these boundaries. Acceptance of the person in the practice of pastoral care and counseling is not limited but limited to their behavior. Their behavior should limit according to the character of Christ. Furthermore, apart from feeling accepted, the counselee receives attention from the counselor as a form of Christian hospitality, namely a sense of empathy. Empathy is a person's ability to feel and understand what is experienced by others which in pastoral and counseling contexts is related to what is experienced by the counselee. Etymologically empathy comes from the Greek Empatheia, which means 'to feel,' which is to have a deep feeling to understand the world of others. Empathy is different from sympathy; sympathy can be a feeling of caring for the feelings of others, but sympathy is not as deep as empathy. We have not said to be able to feel something others feel. Empathy will be more vital if we have experienced the same incident or the people closest to us have experienced the same incident. In empathy, there are also aspects: a) Warmth, a feeling that a person has to be warm towards others. b) Tenderness is a feeling that a person has to act or speak gently towards others. c) Caring is a person's attitude to giving attention to others and the surrounding environment . So the counselor needs to have a sense of empathy to explore the counselee's problems and build sensitivity to the needs and struggles of the counselee. Counselors with good empathy qualities will make the counselee feel understood and Bisma The Journal Of Counseling, Open Access, https://ejournal.undiksha.ac.id/index.php/bisma understood. The counselee will become comfortable and more open to the counselor in expressing his struggles because he feels that the counselor understands the problem. In conducting Christian hospitality, the counselor must realize the two approaches, acceptance, and empathy for the counselee. Without genuine acceptance, counselees can feel inferior or unable to express themselves and cannot even find themselves in pastoral care and counseling. However, without empathy, the counselor does not feel comfortable, and it is difficult to open up because he suspects that the counselor does not understand the problem. God first gave an example of hospitality to humans through Jesus Christ. Humans receive the feeling of acceptance through the work of Christ, namely the redemption of humans from sin that makes believers not perish but have eternal life . God accepts and saves people even while they are still sinners . Humans who are unclean because of sin are accepted into the holiness of God to be sanctified and sanctified . God also shows empathy to humans through the work of the incarnation. Namely, God emptied Himself, took the form of a servant, and became in the likeness of humans . God is not only sensitive to human problems and suffering, but He also feels all that humans feel through Jesus, who became human . The Lord Jesus feels our pain, poverty, and all our weaknesses. Therefore, Christian hospitality in pastoral care and counseling is not just a church program, a profession/position requirement, or periodic action for specific purposes and benefits. Christian Hospitality must continuously manifest as a lifestyle for church counselors and everyone involved in pastoral care and counseling. In this case, hospitality can be an entry point for everyone to attend church. Just as Jesus manifested hospitality to all people so that He became a way for many people, counselors must manifest hospitality to counselees. Hence, they have access to knowledge of God. So what is the focus of Christian counselors is not only solving problems but bringing the counselee to know Christ through the experience of divine hospitality. Christian counselors are not only required to answer the counselee's questions but are also required to make disciples of his counselees. Not everyone calls to be an evangelist, but everyone calls to be a witness who proclaims the good news . Jesus' command in Matthew 28:18-20 is the command to make disciples of all people. This is the real difference between secular counseling and Christian counseling. Everyone can find solutions to their problems anywhere and anyone skilled, but the kingdom of God and his righteousness and eternal life is only through the church whose head is Jesus Christ . --- Conclusion The practice of Christian hospitality in pastoral care and counseling follows the example of Jesus, who is the embodiment of God's hospitality to humans. Hospitality carried out by counselors needs to be based on agape love, namely God's love, unconditional love, and love as it is. This means that the counselor is not picky about people, serving all counselees who come to him for assistance, regardless of their status and actions. In realizing hospitality, counselors need to show a sincere sense of acceptance and have a sense of empathy for their counselees. The acceptance given by the counselor gives freedom to the counselee to express and feel comfortable, but the limitation is the noble values in behavior. Therefore, although accepting a person's personality in the practice of pastoral care and counseling is not limited, it is limited to their behavior. Counselors with good empathy qualities will make the counselee feel understood and understood. The counselee will become comfortable and more open to the counselor in expressing his struggles because he feels that the counselor understands the problem. Christian Hospitality in pastoral care and counseling is not just a church program, a profession/position requirement, or just a periodic action for specific purposes and benefits. Christian Hospitality must continuously manifest as a lifestyle for church counselors and everyone involved in pastoral care and counseling. In this case, hospitality can be an entry point for everyone. Christian counselors are not only required to answer the counselee's questions but are also required to make disciples of his counselees. So what is the focus of Christian counselors is not only solving problems but bringing the counselee to know Christ through the experience of divine hospitality. --- Conflict of Interest Disclosures: The authors declare that they have no significant competing financial, professional or personal interests that might have influenced the performance or presentation of the work described in this manuscript.

The need for pastoral assistance and counseling is increasingly being felt in various sectors of society, both Christian and non-Christian. The economic, social, and political crisis, which resulted in crises in other fields, including health, education, and morals, made a crisis that had a widespread impact in Indonesia. This has triggered the community's need for pastoral care and counseling. This works to help them through an interactive, reciprocal, and in-depth conversation to find the core of the problem and lead to a solution. However, for this interaction to occur, hospitality practices are undoubtedly needed so that the counselee is willing to be involved and open up. The purpose of this study was to examine the concept of hospitality as the first step in the presence of pastoral assistance and counseling in times of crisis for Christians. The research method used is descriptive qualitative. The results showed that Christian hospitality requires counselors to have empathy and openness to counselees while sharpening their ability to carry out congregational counseling. Without it, the congregation will remain in a depressed and stressed existence due to the crisis they are facing.

democracy in the United States, the concept of social capital has entered popular consciousness as a beneficial force to be promoted in communities. Yet this popular perspective overlooks that the social networks and connections in which this capital resides also have a dark side. Today, a small but growing body of research evidences the vicious potential as well as the virtuous value of these networks. They can promote coordination, cooperation, and cohesion to solve dilemmas of socially undesirable collective action. Participation in criminal gangs, terrorist cells, deadly riots, and ethnic violence has been linked to the interpersonal ties that bring and bind participants together. These social bonds then can be harnessed for ill as well as for good. This article extends the exploration of social capital's dark side to a particularly heinous and uniquely human form of collective behavior: genocidal collective violence. It examines the extraordinary events observed in Rwanda in 1994 in which an estimated one in five ethnic Hutu men mobilized to commit intimate acts of violence, overwhelmingly in groups, and primarily against their ethnic Tutsi neighbors. It asks the simple question of what determined the participation of one-fifth and the nonparticipation of four-fifths of this segment of the population. It finds that the answer lies in part in the particular interpersonal networks in which individuals were embedded, which facilitated or constrained their participation. Participants had social networks and social connections that differed to those of nonparticipants. It was the particular features of these participants' networks and connections that facilitated their selection into the violence, while it was the characteristics of nonparticipants' networks and connections that constrained their recruitment. Specifically, the article makes three central claims. First, the size of an individual's network is an important determinant. Compared with nonparticipants, participants in genocidal collective violence possess more social connections within their communities in general and more connections to other participants in particular. Second, in addition to the overall network characteristics, the characteristics of individual connections matter. Not all connections are equal and some connections matter more than others. In particular, kinship and neighborly ties-when compared with economic, social, political, and religious ties-are the strongest predictors of participation. Related to this, the strength of a connection also matters. The stronger the connection an individual has to a participant, the more likely he is to be drawn into the violence. Third, countervailing ties to nonparticipants in an individual's network do not lower the likelihood of participation. Individuals may have connections to both nonparticipants and participants, but it is the connections to other participants that prevail. I suggest several mechanisms may be at work in explaining the significance of social connections. Networks fulfill the functions of information diffusion, social influence, and behavioral regulation. Overall, the evidence suggests that social interaction is a strong predictor of differential participation. Whether you participate depends in part on whom you know. The article begins with the theoretical framework for the relational approach and sets out several hypotheses based on this. It then describes the research design, data, and methods before presenting the results and alternative interpretations of the data. It concludes with a discussion of the theoretical implications of these findings. --- Theoretical Framework The article's theoretical point of departure is two simple but fundamental observations from sociobiology and sociology, respectively. First, humans are a social species and naturally seek connections to each other . Second, through social organization humans are able to accomplish more collectively than as individuals . --- Social Capital The contemporary notion of social capital flows naturally from these twin axioms. The concept, however, has proved susceptible to diverse formulations. Its classical conceptualizations, originating with Coleman and Bourdieu, confounded the sources and resources of social capital. Whereas its sources lay in the connections, networks, and structures linking individuals and families, its resources encompassed trust, norms, sanctions, information access, obligations, and expectations among other things . Thus, Bourdieu defined social capital not only as ''what ordinary language calls connections'' but also as ''the aggregate of the actual or potential resources which are linked to possession of a durable network'' . Similarly, Coleman defined social capital not only as ''some aspect of social structures'' but also as ''a particular kind of resource available to an actor'' or the substance that ''inheres in the structure of relations between actors and among actors.'' The more contemporary notion, introduced by Putnam, did not resolve the conceptual confusion between the sources and resources of social capital, but did imbue it with a distinctly positive connotation and extended its application from individuals and families to whole communities and nations. For Putnam , who originally focused on the civic-mindedness created by participation in voluntary associations, social capital might ''overcome the poverty and violence of South Central Los Angeles . . . or nurture the fledgling democracies of the former Soviet empire.' ' Fukuyama made similarly ambitious claims, stating social capital is ''important to the efficient functioning of modern economies, and is the sine qua non of stable liberal democracy.'' The concept also influenced thinking within the development policy community. The World Bank launched a ''Social Capital Initiative,'' and one vice president stated ''social capital contributes significantly to sustainable development . . . . Without social capital, society at large will collapse, and today's world presents some very sad examples of this'' . Such thinking has since advanced and theory now more explicitly acknowledges the potentially dark side of social capital. Putnam refined the concept to distinguish between bridging capital that links heterogeneous groups and bonding capital that links individuals within homogenous groups. The latter is exclusionary of outsiders. Portes has explained how social capital can also constrain individual freedom and lead to excessive claims by group members as well as downward-leveling norms. Ostrom and Ahn point out that the mafia, cartels, and criminal gangs all depend on the cohesive power of social capital. Yet, notwithstanding the scholarly rebalancing of the concept, its positive perception persists in popular consciousness and in certain policy communities. --- Social Networks Although often conflated with social capital, social networks represent a distinct, expanding field of research. They merit analytical differentiation from social capital for two reasons. First, the distinction avoids the conceptual confusion between the sources and resources of social capital. Social networks are a source of social capital but are not the actual resources that comprise social capital . Second, a social network is a normatively more neutral term than social capital and can more readily be thought of having positive and negative consequences. As social organisms, almost without exception we each belong to some community and are embedded in some network structure. Social network theories acknowledge the empirical reality that our choices and actions are often interdependent . We do not inhabit a theoretical universe of atomized individuals pursuing choices and actions independently of each other. What others choose to do influences what we choose to do. Atomistic approaches that emphasize individual attributes in accounting for individual actions overlook the importance of the social context within which individuals operate. Given the interdependence of many individual decisions, social networks feature in explanations of a wide variety of human phenomena, but their role is particularly prominent in theories of episodic collective behaviors. They have, for example, been identified as a determinant of participation in social movements , electoral mobilization campaigns , and political revolutions . Certain scholars have argued more formally that social networks represent another solution to the collective action dilemma. Ostrom and Ahn argued that the trust generated through repeated interactions between networked individuals facilitates collective action. Others pointed to structural characteristics of networks. Marwell, Oliver, and Prahl first showed through computer simulation that the prospects of collective action increase as the density and centralization of ties increase and the cost of communication decreases in a network. Gould demonstrated that the effects of network density and size on collective action were in fact nonlinear and contingent on how centrally located volunteers are in the network. Siegel argued participation in collective action is a function of network structure type-small worlds, villages, opinion leaders, and hierarchical-and the distribution of individual motivation levels in the network. A separate and growing area of research has recognized the sociological foundations to participation in various violent phenomena. Social networks demonstrably have a dark side. The study of terrorism has made important advances using social network analysis techniques . In civil wars, voluntary recruitment in rebel movements is partly determined by an individual's preexisting friendship and family ties to other rebel group members . The concentration of social ties in a community also mediates who is denounced and targeted at the local level in civil war violence . In genocide, preexisting social ties facilitate recruitment into the violence, as do group ties forged through participation . In rebellions, participants are drawn in through preexisting interpersonal ties and rebel solidarity is maintained through trust among connected participants . In communal violence, the weakness of interethnic ties and strength of intraethnic ties predispose communities to violent confrontation . In riots, participants are better connected and more involved in community life than nonparticipants . I draw together these two separate lines of research-on the role of social networks in collective action and in violent phenomena-to test hypotheses on differential participation in collective violence. By collective violence, I mean episodic behavior that inflicts physical harm on people or property; is the product of individuals acting in groups, not individuals acting alone; and involves ordinary civilians as participants, not security professionals such as soldiers, policemen, or others already authorized and organized to use force. Collective violence then encompasses a diverse set of behaviors that includes riots, communal violence, ethnic massacres, pogroms, and lynchings among others . I examine a particular form of collective violence-genocidal collective violence-where the violence is additionally either passively or actively supported by the state. Collective violence may become genocidal in character if the state deploys its symbolic authority and material capabilities to support it . The state's involvement then affects the scale of the violence. It does not, however, determine differential participation of individuals into the violence. --- Mechanisms and Resources: Linking Social Networks and Social Capital Although distinct research areas, there is striking overlap between social network's ''mechanisms'' and social capital's ''resources.'' Social network studies have generated a diverse array of mechanisms to explain how connections work: information transmission , norms of fairness and efficacy , incentives and sanctions , coordination peer pressures , social approval , social influence , identity , and socialization . In social capital research, although ''trust'' is now the most commonly cited resource , in its original conceptualization it additionally comprised norms and sanctions, obligations and expectations, informationflow capability, and the absence of ''structural holes'' . For coherence, I propose a unifying typology to capture the similarity in functions of social network mechanisms and social capital resources. Table 1 identifies four primary functions: diffusion: networks circulate resources, tangible and intangible, among connected individuals; influence: they alter the thoughts, feelings, and behavior of connected individuals; regulation: they encourage or constrain the choices and actions of connected individuals; cohesion: they build solidarity among connected individuals and reinforce differences with unconnected individuals. Importantly, these functions are not mutually exclusive. The mechanisms they underlie may operate simultaneously in a given network. --- Hypotheses Network-Level Characteristics. Social network researchers have considered the significance of various structural characteristics of networks: size, density, centralization, reachability, inclusiveness, transitivity, and connectedness among others. In predicting participation in collective violence, I focus on the size of an individual's network. The theoretical logic is that the more connections an individual has, the more opportunities he will be presented with to participate. Information on opportunities diffuses through the network so that better connected individuals will acquire more information and more quickly. Granovetter in his seminal study of job hunters found that a high number of ties increased an individual's chances of finding a job opportunity. Weak ties in particular mattered for diffusing information as an individual could sustain more of them, given the relatively low investment required and thus be exposed to more information. Theoretical simulations have also identified the effect of tie frequency on participation in collective action . Hypothesis 1A: The greater the number of social connections an individual has, the more likely he or she is to participate in collective violence. It is not only the number of ties to people generally but also the number of ties to participants specifically that may matter. The theoretical logic here is that individuals with preexisting connections to other participants are likely to be influenced to participate themselves. The more participant ties they possess, the more influence they may experience. Research on social movements , riots , civil wars , and terrorism has pointed to existing ties to other participants as determinants of participation. Snow, Zurcher, and Ekland-Olson in their pioneering social movement study explicitly identified recruitment appeals as one particular form of influence other movement participants may bring to bear. --- Hypothesis 1B: The greater the number of ties to other participants an individual has, the more likely he or she is to participate in collective violence him-or herself. Connection-Level Characteristics. In addition to the quantity of interpersonal connections, the quality of these dyadic ties may matter. Existing research has distinguished connections in several ways: type, strength, thickness , symmetry, stability, and direction, among others. Intuitively, different connection characteristics may produce different effects. I examine connection type and hypothesize that certain connections may be more influential than others in predicting participation. Varshney distinguished between informal ''quotidian'' and formal ''associational'' ties within an ethnic group and finds the latter better at reducing communal conflict in India. Yet the panoply of potential connection types is richer than this dichotomous categorization. Kinship, economic, social , political, religious, and proximity ties each represent distinct forms of social interaction and may also involve distinct mechanisms. For example, whereas kinship ties may influence through love or affection, friendship ties may do so through peer pressure or social approval. Hypothesis 2A: An individual's likelihood of participation in collective violence will differ with the type of connection between him-or herself and the participant to whom he or she is connected. Tie strength is another theoretically important connection characteristic. Granovetter defined tie strength as a function of ''the amount of time, the emotional intensity, the intimacy , and the reciprocal services which characterize the tie.'' I hypothesize that stronger ties will be better at drawing individuals into collective violence, given the high risk involved in such an activity compared with the low-risk activity of job hunting. Stronger ties exert stronger influence. The importance of strong ties in high-risk, costly behavior has already been recognized in recruitment into activism . Hypothesis 2B: An individual's likelihood of participation in collective violence will be higher, the stronger the connection between him-or herself and the participant to whom he or she is connected. Countervailing Connections. Ties are multivalent. They may serve to inhibit as well as to promote participation . If ties to other participants represent a positive influence on participation, then ties to nonparticipants may exert a negative influence. Snow, Zurcher, and Ekland-Olson found that an individual's extra-movement ties reduced his or her structural availability to join the movement. McAdam and Paulsen tested the relative importance of competing ties on participation in a social movement and found that parents and movement peers exerted a stronger positive influence on participation than did friends, religious figures, and civil rights organizations. Alternate ties then may work to prevent participation. Hypothesis 3: The more countervailing ties to nonparticipants an individual possesses, the less likely he or she is to participate in collective violence. --- Research Design I present evidence of participation in Rwanda's genocide of 1994 to illustrate the relational theory of collective violence. The project employs a multimethod approach and relies on three principal data sources: a survey mapping the social networks of 116 residents from one Rwandan community; semistructured, in-depth interviews of participants and nonparticipants from two communities on events during the genocide; and a broader survey of 294 participants and nonparticipants across the country on events during the genocide. I begin with a synopsis of the violence at the national level and in the chosen research site before describing the data, research techniques, and robustness checks undertaken. --- Brief History of Rwanda's Genocide Although certain historical aspects of Rwanda's genocide remain deeply contested, the following account is based on a set of facts on which there is a reasonable scholarly consensus. In April 1994, a small group of Hutu extremists seized power in Rwanda and initiated a genocidal campaign that targeted the country's Tutsi minority for extermination. All told, an estimated 507,000 to 850,000 Tutsi would be killed, along with several tens of thousands of the Hutu majority, by the time the violence ended some 100 days later . The killers were soldiers, police, militia, and ordinary Rwandans. The genocide was the culmination of a civil war, begun in October 1990 and fought between a mainly Tutsi rebel army, the Rwandan Patriotic Front , and Rwanda's Hutu-dominated government. The war's roots lie in a revolution, shortly before Rwanda's independence from Belgium in 1962, which toppled the long-standing Tutsi monarchy and installed Rwanda's first Hutu Republic that would exclude Tutsi from political power for the next three decades. The revolution had sent hundreds of thousands of Tutsi into exile, and these exiles would make several unsuccessful armed attempts in the 1960s to return to Rwanda. However, it was not until the exiles' descendants initiated the civil war in 1990 and the reintroduction of multiparty politics in 1991 that the Hutu monopoly on power would weaken. In August 1993, the international community brokered a peace deal that envisaged power sharing between the incumbent regime, the newly formed opposition parties, and the rebel RPF. Hutu hardliners opposed the deal, however, and when Rwanda's Hutu president was assassinated on April 6, 1994 , these hardliners seized the opportunity to take control, reignite the civil war, and initiate the genocide. The international community failed to intervene, and it was not until the RPF finally defeated the extremist government in July 1994 that the killing would end. Explanations of the genocide abound. The cornucopia of contributing factors suggested include Rwanda's move to democratize, the ongoing civil war, intense demographic pressure on land, ethnic fears, ethnic hatred, a racist ideology, abject poverty, a strong state, an obedient population, the colonial legacy, and indifference from the international community, among others. While scholars contest the relative importance of these factors, there is little disagreement on the genocide's extraordinary characteristics. Rwanda experienced collective violence on a massive scale. In just over 100 days, an estimated three quarters of its Tutsi minority were eliminated. Furthermore, Rwandans mobilized as soon as one day following the president's assassination, and by one estimate, one in five Hutu men between fifteen and fifty-four years old committed an act of violence . 1 Overwhelmingly, in rural communities the violence was conducted in groups, by ordinary civilians, and sometimes with, sometimes without the support of local state authorities . Tilly recognizes Rwanda's massacres as archetypal collective violence. Yet, notwithstanding the extraordinary speed and scale of civilian mobilization, it is important to remember four in five Hutu men did not commit violence. It is this statistic that motivates the central research question of this project. --- Profile of Research Site I surveyed sector Tare in Butare prefecture in southwestern Rwanda. Tare was a demographically representative community for Rwanda. In 1994, it comprised 647 households and 6.3 percent of its population was ethnic Tutsi, in line with national averages. 2 Its experience of the genocide was also not atypical. In all, McDoom 63 percent of its Tutsi population were eliminated and 24 percent of its adult Hutu male population were implicated. Tare was also one of the pilot sectors for gacaca, Rwanda's experimental system of local courts created to foster truth, justice, and reconciliation following the genocide. As a pilot, Tare was one of the first sectors to complete the gacaca process. Comprehensive microdata on participants and victims then were available at the time I was in the field in 2009. Tare experienced three principal episodes of violence on April 19, 21, and 22, 1994. It did not erupt into violence immediately after the assassination of the president on April 6, 1994. The delay was due to the local prefect of Butare prefecture, Jean-Baptiste Habyarimana, himself a Tutsi. Habyarimana had resisted directives from the new Hutu extremist government in the capital, Kigali, and had managed to keep the peace in the prefecture for nearly two weeks. In response, Rwanda's new Hutu hard-liner president, Sindikubwabo, visited Butare on April 18 to replace and assassinate the rebellious prefect. Having done so, later that same day President Sindikubwabo traveled to Maraba commune, where sector Tare was located, and incited the Hutu population to attack the Simbi Parish church where a large number of Tutsi had gathered. His speech signaled the start of the genocide in the prefecture. Tare's Tutsi left their homes that same evening and many sought refuge in a local church, Rugango. The following morning, April 19, a group of Hutu men from Tare launched the first attack against a group of Tutsi gathered in the nearby Sovu health center. This first attack, however, was repelled. The next day, April 20, a security meeting was held in Tare and a new plan of action made to attack the Tutsi in the Rugango church. On the morning of April 21, a large group of Hutu men confronted and killed the Tutsi at Rugango and in the afternoon went on to kill those gathered at the Gihindamuyaga monastery close by. The following day, a renewed attack was made against the Sovu health center that was successful. The majority of Tare's Tutsi were killed in these three attacks. The hunt for the remainder, however, continued until the rebel army reached the area on July 2, 1994. --- Data and Method A total of 116 individuals from among Tare's 647 households were surveyed to obtain data on their individual egocentric networks. The sample was stratified by participant status and comprised seventy-nine nonparticipants and thirty-seven participants in the genocide. Participants were selected randomly from a list of convicted perpetrators. Nonparticipant respondents were selected from a census list of all Tare inhabitants resident in the sector in April 1994. The census list was compiled from a series of smaller lists drawn up by Tare's nyumbakumi, representatives of blocs of about ten households, as part of the national gacaca process. The nonparticipant stratum purposively sampled male Hutu residents aged fourteen and older to ensure the most realistic comparison with participants. I personally conducted the interviews with prison inmates using a Rwandan interpreter and trained a team of ten Rwandan enumerators to administer the questionnaire to those at liberty within the community. Given the sensitive subject matter, self-reports of personal connections to genocide perpetrators may be biased toward underestimating the true number. I purposively chose a community where the gacaca process had been completed to reduce the anxiety of self-incrimination. I additionally relied on two distinct techniques to collect data on individual's social networks: a roster and a name generator. Neither technique revealed to the respondent that I knew who the killers and nonkillers were in the community or that the purpose was to identify their connections to these two groups specifically. The roster technique involved reading a list of thirty named individuals from Tare to each respondent and asking him which of the thirty individuals he personally knew. The thirty individuals comprised fifteen participants and fifteen nonparticipants from Tare selected at random and whose names were also listed randomly on the survey questionnaire. If a respondent knew one of the thirty individuals, he was then asked to specify his connection to the individual. Respondents could specify more than one connection to a particular individual to allow for multiplexity. Importantly, the questionnaire asked for connections before the genocide. While clear for kinship relations, for other types of connections respondents were asked to specify when they first met an individual. If uncertain, enumerators were trained to prompt respondents using key life events as temporal markers. Respondents, for example, were asked whether the relationship was first established at school, before marriage, after children, or while their parents were still alive. The second name generator technique reversed the roster technique. It involved first specifying a connection type and then asking respondents to name individuals connected to them in this way. To establish the range of most common connections between individuals in the community, I conducted a focused group interview with eight Tare residents. Altogether thirty-four different types of connections were identified that I recoded as kinship, economic, social, neighborly, religious, and political ties. To corroborate and contextualize the quantitative network data, the project also included in-depth interviews with residents from Tare and a second nearby community, Zivu. I purposively interviewed participants and nonparticipants from each community identified as reliable and willing informants. For nonparticipants, a preliminary indicator of reliability was membership of the gacaca committee, and for participants, their status as self-confessed convicts. The interviews were semistructured and covered, inter alia, events in respondents' communities during the genocide. I conducted interviews using an interpreter and changed interviewees' names reported in this article to safeguard their anonymity. To improve the generalizability of the interviews, I also conducted a stratified two-stage cluster survey of 294 Rwandans from across the country. The stratification was by region and by participant status and comprised 151 northern respondents and 143 southern respondents. The survey instrument posed questions to 273 Hutu and twenty-one Tutsi respondents relating to events in their communities during the genocide. --- Dependent Variable The dependent variable took a binary form of participant/nonparticipant. Participation was defined as the commission of at least one act of violence against the person during the genocide. While a superior dependent variable would have been continuous in nature, no reliable data on the quantum of violence committed by individuals were available. Participation in the genocide involved many acts other than violence: looting, manning roadblocks, mounting night patrols, denouncing individuals, and encouraging participants among others. Nonetheless, an act of violence against the person represented an important threshold in antisocial behavior to cross, and this category of participant was distinguished to be consistent with other studies measuring participation in Rwanda's genocide . To distinguish participants from nonparticipants, I compared the official list of perpetrators convicted through the gacaca process with an unofficial list of perpetrators identified by prison inmates already convicted of genocide crimes in Tare. To minimize measurement error, an individual was classified as a participant only if his name appeared on both lists. To compare the networks of participants against nonparticipants, I conducted t-tests on the difference in mean number of ties and multivariate logistic regressions to measure the relative importance of the quantity and quality of ties using the binary dependent variable of participant/nonparticipant. --- Robustness Checks and Control Variables Four principal robustness checks were undertaken. First, as described, network data were generated using both roster and name generator techniques. Reliance on two techniques minimized the risk that respondents, either intentionally or unintentionally, misreported their connections to others in the community. Second, two alternate dependent variables of participation were tested: convicts and suspects. As perpetrators may have subverted the gacaca process and distorted trial outcomes through the co-optation of judges or intimidation of witnesses, I looked at both those initially accused and those finally convicted. Results from the name generator technique and from specifying suspects as the dependent variable may be found in an online appendix on the author's personal website. Third, I tested and reported separately the multivariate models using backward stepwise elimination to exclude unnecessary variables and to confirm the significance of included variables. Fourth, I tested for multicollinearity and determined variables exhibited no problematic levels of correlation. Finally, there were no missing data for the 116 respondents, no individual was an isolate with connections to no one else, and only one individual reported a homogenous network of nonparticipants only. The identification of suitable controls was not obvious as the extant literature is divided between an emerging consensus that genocide perpetrators are ''ordinary'' with few distinguishing attributes and studies that examine individual-level characteristics to predict participation. I favored caution and included controls for age, marital status, education, ideological commitment , occupation, and wealth measured objectively and subjectively. There were no controls for gender or ethnicity as the sample purposively comprised Hutu males only. --- Results --- Network-Level Characteristics The size of an individual's network is a robust predictor of participation. Participants possessed significantly larger social networks than nonparticipants. The descriptive statistics in Table 2 indicate that the mean number of connections to other residents listed on the roster reported by convicted participants was twenty compared with only thirteen connections reported by nonparticipants. The difference in means is statistically significant for both convicted and suspected participants. The multivariate logistic model corroborates this finding, controlling for individual-level attributes, and shows that for each additional resident to whom a resident was connected, the odds of their participation increased by 9 percent . This finding holds for convicted participants and suspected participants. Hypothesis 1A was confirmed. Participants were significantly better connected in their community than were nonparticipants. It was not merely having a larger social network that mattered, however. It was also the particular individuals in this network who were important. The more participants, as opposed to nonparticipants, in a resident's network, the more likely the resident was to participate in the violence. Convicts, on average, were connected to eleven other participants compared to only six for nonparticipants. Suspects had ten participant and six nonparticipant ties on average. The multivariate models confirm this finding and indicate that each additional participant to whom a resident was connected increased the odds of the resident also participating by a remarkable 24 percent for convicts and 15 percent for suspects. The finding is also robust to the use of the name generator technique instead of the roster method. Hypothesis 1B was also confirmed. The more connections a resident has to other participants, the more likely he is to participate himself. --- Connection Characteristics Connection Type. The descriptive statistics prima facie suggest that kinship, proximity , social , and economic connections were more important for participation than political or religious ties. However, when all six types of connections are specified in the multivariate model, the only significant connections that persist are those based on kinship and proximity. For each additional family member who participated in the violence, the odds of the related individual also participating increased by a noteworthy 75 percent for convicts and 51 percent for suspects. The ''name generator'' model corroborated the significance of family ties identified in the roster model. The importance of family ties makes intuitive sense if we see connections as channels for social influence. Given the violent, antisocial nature of the collective action involved, the emotional power of kinship bonds could exert a particularly strong influence and overcome prosocial, peaceful norms that may otherwise govern individual behavior. I discuss possible mechanisms in more depth in the qualitative analysis. The significance of proximity ties is somewhat different. Ties to neighbors increased the likelihood of participation regardless of whether the neighbor was himself a participant or not in the violence . This may make intuitive sense if we also see ties as channels for the diffusion of information. As social interaction likely increases with spatial proximity, the more neighbors one has the more information one would likely receive. I present the evidence of such a diffusion mechanism later. Although tests for multicollinearity did not indicate problematic levels, it is worth noting the potential correlation between kinship and proximity ties in the Rwandan rural sociocultural context. Rwandan families often lived close together, sometimes occupying the same hill, in part because Rwandan fathers customarily gifted land to their sons upon marriage to establish their own households. The significance of proximity ties then may also reflect some of the significance of the kinship ties. Connection Strength. In addition to the type of tie, the strength of the tie mattered. Ties based on the first degree of canonical consanguinity-parents, siblings, and children-are a substantively stronger predictor of participation than ties based on weaker, more distant degrees of consanguinity-uncles, grandparents, cousins, and so on. In probabilistic terms, using convicts as the dependent variable, the odds of an individual joining the violence more than doubled when a close family member also participated in the violence . If a more distant family member participated, the odds of participation also increased but only half as much as if it were a close family member . Some caution, however, is needed in drawing an inference about tie strength as a statistically significant result is not McDoom 15 replicated when suspects are specified as an alternate dependent variable. Nonetheless, the finding for convicts would be consistent with the interpretation suggested for the more general finding that family ties exert a stronger influence on individuals than social, economic, political, or religious ties. Stronger ties may be needed to draw individuals into high-risk activities such as violence and to break basic prosocial bonds of humanity between individuals. Tie ''strength'' then may be closely related to connection ''type.'' --- Countervailing Connections While a high number of countervailing ties to nonparticipants appears substantively to lower the likelihood of participation, the effect is not statistically significant. This is true for both convicts and suspects. It is also true for kinship, social, economic, political, religious, and neighborly ties . Having a family member who does not participate, for example, does not appear to counteract the influence of a family member who does participate in the violence. One possible interpretation of this finding is that inaction has a less influential effect than action. If a nonparticipant did not act to stop another individual from participating, then the actions of a participant would likely prevail. To borrow a popular adage, ''all it takes for evil to triumph is for good men to do nothing.'' --- Control Variables Notwithstanding the theoretical disagreement over whether perpetrators possess distinguishing characteristics, the most robust individual attribute-beyond being McDoom male-was occupational status. Individuals who had income other than from farming-mechanics, guards, carpenters in my sample-were more likely to participate. One possible interpretation is that those dependent on farming could not as readily afford the opportunity cost of participating in a risky and time-consuming activity as collective violence. The data also provide more contingent support for age and marital status. Middle-aged, married men appear more likely to participate. In sum, these findings suggest that both relational and attribute variables should be considered in predicting participation in collective behaviors. --- Qualitative Evidence on Mechanisms While the quantitative data predict that well-connected individuals are more likely to be mobilized, the interview data provide better analytical traction on the mechanisms behind the observed correlation. They suggest the operation of diffusion, influence, and regulatory mechanisms. To illustrate how networks fulfilled these three functions, I present survey data and interview testimony from Rwandan communities on information diffusion: how information on the target and rationale for violence first entered and circulated within communities; recruitment appeals: how individuals were selected and organized for violence; and incentives and sanctions: how material opportunities and the perceived costs of nonparticipation regulated individual behavior. --- Information Diffusion Information on the genocide entered and circulated within communities through preexisting networks. Immediately following the president's assassination on April 6, 1994, it was initially unclear to many Rwandans in rural communities what had happened or what would happen next. The sudden end to the twenty-one years of autocratic rule represented a major macropolitical shock and created acute uncertainty about the civil war's outcome. Information quickly arrived in communities, however, that attributed the assassination to the Tutsi and identified them as the enemy and called upon Hutu to defend themselves and mobilize against the Tutsi enemy. The majority of respondents reported they acquired this information locally. Only 6 percent indicated the radio was the source . The survey data illustrate the initial uncertainty created by the president's assassination and the subsequent arrival of information from outside of communities. Sixty-four percent of Hutu respondents reported initial cooperation across ethnic lines following Habyarimana's death . Tutsi participated alongside them in night patrols to look for the enemy, a fact corroborated by 80 percent of Tutsi respondents. Yet more than 70 percent of Hutu respondents also report that the enemy would eventually come to be identified as all Tutsi in their communities. The implication is that information reached communities that altered interethnic relations and defined Tutsi as the enemy out-group. Interview testimony corroborates the change in interethnic relations following the assassination and suggests the information responsible for the shift often entered communities through preexisting local networks. The following interview illustrates how the arrival of this information divided the community into Hutu in-group and Tutsi out-group: What happened in your community after the president was killed? When Habyarimana died the politicians said they [the Tutsi] had killed our leader. Then people's hearts changed. They said the enemy had killed our father. That extremist Rekeraho said McDoom 19 we should wake up and fight for our survival. Who was Rekeraho? He was an MDR [an opposition party] politician who lived in Mwendo. After a few days we started to see smoke from Gigonkoro [the neighboring prefecture] from burning houses. Soon everyone was afraid. Both Hutu and Tutsi. Who was burning the houses? They said those who were doing the burning had covered themselves in banana leaves so you could not see who they were. When people found out what was happening, the fear of the Hutu decreased while the fear of the Tutsi increased as they now knew who was the enemy. After a few days it was clear there were two groups: those being hunted and those who hunted. Leopold told us then that a local politician supplied both the information that Tutsi were responsible for the assassination and that they were deemed the enemy. The same politician also provided a rationale for mobilization: self-defense. In a joint interview, a perpetrator and his wife described how this local politician obtained the information through his political networks that contributed to his own attitude change: What was Rekeraho's attitude toward Tutsi like? [Husband] On the 10th [of April 1994], there was a baptism of a Tutsi baby and Rekeraho was to conduct the ceremony. The evening before, he went to Kambanda's [the Tutsi father] house and there he even offered a cow to a young Tutsi boy, who was single still, and said he would ask for a girl's hand in marriage. He had good relations with the Tutsi. When did he change? [Wife] He attended meetings in Butare town and that is when he changed. What kind of meetings? They were secret meetings involving other politicians. And when did it change here in Tare? Rekeraho then began to hold secret security meetings at his house with members of the MDR here and his old soldier friends. That is how the genocide started here. Initially, Rekeraho had outwardly good relations with Tutsi in his community, but later learned through his party political connections outside of the community that the Tutsi were to be targeted. He then acted as a ''bridging'' connection between a central and a peripheral network. He brought information from the main town to the rural community. Rekeraho then circulated this information within the community through his network of party supporters and friends who had also served in the military. --- Influence: Recruitment Appeals Once information on the target and the rationale for action reached communities, residents were then organized into attack or hunting groups . Two-thirds of survey respondents indicate that the violence in their communities was not spontaneous, but in fact organized . The recruitment process was often face to face and recruiters targeted individuals in their social networks as the following interview illustrates: Who told you to man the roadblock? It was the responsable of the cellule, Sibomana. How did Sibomana summon you to the roadblock? It was before the umuganda [compulsory labor] that he came to my house and told me to work on the roadblock. He said no-one must stay at home. Do you live near Sibomana? Yes, I live nearby. How far? It is on a hill. I live at the bottom, and Sibomana lives near the top, but you can call from one to the other. Do you have family ties to Sibomana? No. We had an alliance founded on religion. Sibomana's mother was the godmother of my sister. Do you have any economic links to Sibomana? Did you work for him? I was not poor. I did not work for Sibomana. Were you friends? Yes. . . . How did you first know Sibomana? We grew up together on the same hill. And we kept cows together and we went to each other's house because Sibomana's mother was the godmother of my sister. In a community of more than 1,000 individuals, it would be impossible for the cell responsable to recruit everyone in this manner in such a short span of time. We learn, however, that the two also share a personal, multiplex connection. They had been neighbors since childhood, and the responsable's mother was the godmother of his sister. These ties may then explain why Sibomana selected Deogratias to work at the roadblock. Kinship ties were also important in recruitment. In the following interview, a participant told us that the core group of killers in his community comprised three related individuals. The senior relative, Karimunda, recruited his junior kin: How did the genocide start in your community? Karimunda started it. People from another cellule, Musekera, kept coming and asking why have you not started yet? At first they burned some houses. Then, on Sunday the killing began. Karimunda found his younger brother Philip and his nephew standing by the roadside with some other young men and he told us we have work to do. We followed him and we burned houses. . . . The next day the ibitero started at 7.00 a.m. People gathered at the bar as Karimunda had told them to and they went hunting again. Then at night they went on patrols. Why did you go? They forced us to do it because if you missed it, they would fine you. Who forced you? It was Karimunda and his relatives. Even the responsable of the cellule was forced to do it. In the next interview, a nonparticipant from the same community told us that the same nephew was a ''good man'' before the genocide and implied that it was his uncle who influenced him into committing ''bad things'': What did people think of the organizers of the genocide? [read list of organizers] Karimunda. He was a butcher and knew how to use a machete. He was the kind of man who liked conflicts. When you met him you could see he was different from anyone else. Even the councilor was afraid of him. Jean-Claude N.: He was Karimunda's nephew. Before the genocide he was a good man. But afterwards, because he was McDoom 21 related to Karimunda, he went everywhere with him and did bad things together with him. Philip S. [Emmanuel's brother]: He worked together with Karimunda. They were both very bad people. There was a time when he went to Buhoro with Karimunda and they took spears with them. They came back smelling of blood. --- Regulation: Incentives and Sanctions Once the violence began, residents rapidly learned that there were material benefits to joining attack groups as well as costs for refusing to participate when recruited. More than half of survey respondents cited the opportunity to gain materially or financially as the reason why other residents joined attack groups . It was by far the most common motivation reported. Participation in attack groups then presented an opportunity to loot. To benefit, however, you would need to know of the opportunity and the plan of action. In the following interview, Leopold told us that this information was acquired through membership of the attack group and that while coerced at first, participants went voluntarily once they learned of the material advantages: How did they decide whom to attack? When they attacked a home, somebody in the group would say so and so had cows and then people would plan to meet there. In the first few days people went by force. But later they went willingly because they saw they could get property like roofing tiles and mattresses. Once the group reported that so and so's house had been completed, the extremists and the politicians would tell them where to go next once. A refusal to participate also incurred costs. Interviews revealed three widely perceived sanctions for being a refusenik: social stigma, evidenced in the accusation of ibyitso ; monetary or in-kind material fines; or physical harm: What happened to those people did not go with the attack group? They would fine them money and if they could not pay, then they would take their livestock. These people were called ibyitso. Was anyone actually harmed for refusing? No one was killed in my community. But I heard others were beaten up elsewhere. The data, however, are conflicting on the consequences of nonparticipation. Almost two-thirds of survey respondents also indicate nothing happened to refuseniks in their communities . The discrepancy may reflect the evolution in individual motivations to participate to which Leopold alluded. Individuals may initially have been coerced. Subsequently, however, many may have participated willingly once the material advantages became known, thereby obviating the need to enforce the participation of more residents. --- Alternative Interpretations of the Data The research design permits descriptive rather than causal inferences. It predicts wellconnected individuals are more likely to participate, but it does not prove that their connections explain why they participated. Unobserved heterogeneity may account for the observed correlation between connectedness and participation. One alternative interpretation is that the findings are attributable to homophily, the well-established observation that ''similarity breeds connection'' or that ''birds of a feather flock together'' . Participants may share some unobserved characteristic that makes them prone to committing genocidal violence, and it is this similarity that makes them seek out connections to each other. Homophily is one of the well-known challenges to demonstrating causality using observational network data in which the dependent and independent variables are measured simultaneously . It raises the question of endogeneity. We cannot readily know whether it is the connection that is causing the similar behavior or whether it is the similar behavior that is causing the connection. One method to control for homophily would be to observe changes in connections and changes in the behavior over time. Such an empirical strategy, however, is practically infeasible for rare and difficult-to-observe behaviors such as participation in genocidal violence. 3 Nonetheless, in the absence of conclusive longitudinal data, there remain five suggestive reasons for believing that participation in genocidal violence is attributable to the influence of social connections rather than to some unobserved characteristic that participants share. First, I control for many of the characteristics that extant theory and evidence suggest may make individuals prone to such violence: age, marital status, education, wealth, employment status, and ideological commitment. The statistical significance of connections persists in the quantitative analysis. Second, the qualitative data are corroborative and strongly suggest the importance of connections. They suggest networks diffused information on the rationale for mobilization and the targets of violence, enabled the use of interpersonal influence to recruit individuals into attack groups, and regulated participation in the violence through incentives and sanctions. Third, if the findings were due to some unobserved similarity between participants, we would expect participants to have not only more connections to other participants than do nonparticipants but also fewer connections to nonparticipants. Homophily predicts participants would prefer connections to other participants because they possess similar characteristics. Yet the descriptive statistics show that perpetrators have more connections to both participants and nonparticipants. Moreover, these differences are statistically significant. Fourth, it is widely accepted that homophily does not account for all social connections between individuals and that some relationships are formed through simple availability . The availability effect would be particularly acute in the context of rural Rwanda where individuals are tied to the land on which they subsist and where social and geographic mobility are consequently highly limited. Finally, and particularly suggestively, I test for the relative importance of voluntary and involuntary connections and find involuntary connections to be a stronger predictor of participation. Homophily assumes individuals voluntarily seek out connections to each other because they share similar characteristics. To model the potential for homophily, I coded as voluntary connections any relationships that two individuals were freely able to enter and to end. In the context of rural Rwanda, these included friendship, marriage, membership of a political party, and membership of a voluntary organization. Conversely, I defined involuntary connections as relationships that either one or both individuals could not freely enter or end. In Rwanda, these most commonly included kinship blood ties and clientelist ties based on land, jobs, or other important economic benefits where clients may be unwillingly tied to their patron. In both the name generator and roster models, involuntary ties proved statistically and substantively stronger than voluntary ties in predicting participation. As many involuntary ties are also kinship ties, and the power of kinship ties has already been noted, this result is unsurprising. In interpreting this finding, it should be noted I assume that preferences are socially learned rather than biologically determined. Family members are not born with a shared proclivity for violence but instead develop it over time through repeated and prolonged interaction with each other. It is the kinship connection then that leads to the similar proclivity. --- Discussion and Conclusion The evidence presented here reinforces the growing consensus that social networks and social ties, the sources of social capital, have a dark side . The better connected an individual is to his community generally and to other participants specifically, the more likely he is to be drawn into extreme antisocial group behavior such as genocidal collective violence. These connections may supply individuals with pernicious information, expose them to undesirable influences, and create nefarious incentives and sanctions. Family members and neighbors appear to play particularly prominent roles in this process. The perhaps counterintuitive implication is that social misanthropes may be less likely to engage in such antisocial activities than the socially gregarious, given the collective character of the violence. The interdependence created by strong social bonds structurally constrains the freedom to act independently. The evidence also draws sociological theory and political science research a little closer together to reinforce the widening consensus that social structure is a key predictor of participation in various forms of violence. It suggests a relational approach should complement the attribute approach that presently predominates in extant research on participation. The atomistic quest for distinguishing demographic, socioeconomic, attitudinal, or psychological attributes that may indicate a dispositional susceptibility for such violence has largely proved elusive. It has led to the consensus in genocide research that perpetrators are simply ''ordinary '' men . Yet I submit the inquiry should not be limited to the individual but should extend to the social context in which he is embedded. The evidence here supports the position in modern sociology, wherein structure and agency are complementary rather than competing forces in explanations of social action . Both pull and push factors matter, and we should guard against excessive determinism of either approach. While individuals are distinct atoms, they are bound together in interdependent molecular structures. Who you know matters as well as who you are. Network structures may also provide an important link between micro and macrolevels of analyses of human phenomena . The growing focus on the microfoundations of violence has for example revealed a ''macromicro disjunction'' . Macrovariables do not consistently predict important microvariations in violence within complex and aggregated events such as civil wars, ethnic conflicts, and genocides. In identity-based explanations of violence, for example, it is unclear how a macrovariable such as ethnicity affects micro-level outcomes such as why certain individuals commit violence and others not and why particular individuals are targeted but others unharmed. The answer may lie in the complex set of social relations at the local level that mediate participation and victimization . More generally, local factors may matter for observed microvariations. The micropolitical, microeconomic, and microsocial forces that shape patterns of violence at the local level may themselves be shaped by the structure of local networks and ties through which they flow. Lastly, the evidence also provides some insight into the microdynamics of Rwanda's genocide. As often observed, Rwanda had the highest population density of all African states, a remarkable 305 persons/km 2 in 1993. Rwanda's genocide, as previously noted, was also remarkable for the speed of the violence and the scale of popular participation. These facts have led to speculation that the violence was the product of some sort of neo-Malthusian resource crunch: too many people, too little land . This article suggests, however, that the extraordinary speed and scale of the violence may have sociological rather than ecological origins. In highly densely populated societies likely exist highly dense social networks. Highly dense networks often signify numerous connections exist between individuals and that frequent face-to-face interaction occurs, particularly in technologically simple societies where the means of communication and transportation are basic. It is these many ties and frequent contact that enmesh individuals, providing both opportunities for and constraints on actions. As already noted, high network density improves the efficiency of collective action more broadly . Mass mobilization and rapid violent contagion, the remarkable features of Rwanda's genocide, may then have their roots in Rwanda's remarkable population density. --- Notes 1. The estimate is for 367,000 Hutu men to have committed category two crimes under Rwanda's 2004 genocide law. Broadly, category one crimes covered organizers and sexual offenders; category two covered violent crimes against the person; and category three covered property crimes. The estimate is calculated by extrapolating data on conviction rates personally collected at the local level from the pilot sample of gacaca communities. An unofficial tally, based on the presently unreleased official gacaca report, puts the final number of category two convictions at 361,590. See the ruling party's website: http://rpfinkotanyi.org/wp/?p¼1958. Accessed January 16, 2013. 2. In April 1994, Rwanda comprised 11 prefectures, 145 communes, 1,545 sectors, and 9,000þ cells. A sector was home on average to about 800 households. In addition, in some places there was a fifth layer. The nyumbakumi were unpaid individuals representing a collective of ten households. Rwanda's 1991 Population Census reported a Tutsi population of 8.4 percent. 3. Other strategies to improve causal inferences from network data include fixed effect modeling, the use of instrumental variables, cross-lagged structural equation modeling, and quasi-experimental research designs . --- Declaration of Conflicting Interests The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. --- Funding The author acknowledges the award of a postdoctoral fellowship by the UK's Economic and Social Science Research Council during which the fieldwork for this paper was conducted.

Although popularly perceived as a positive force important for objectives such as economic development and democracy, social capital may also be linked to less desirable outcomes. This article highlights a dark side to social capital by pointing to its role in a particularly pernicious phenomenon: genocidal violence. Drawing on a survey of residents from one community that experienced violence during Rwanda's 1994 genocide, I show that individual participation in the violence was partly determined by the features of residents' social networks. Perpetrators possessed larger networks in general and more connections to other perpetrators in particular. The quality as well as quantity of connections also mattered. Strong ties generally, and kinship and neighborly ties specifically, were strong predictors of participation. In contrast, possession of countervailing ties to nonparticipants did not reduce a resident's likelihood of participation. Drawing on in-depth interviews to explore the possible mechanisms behind these findings, I suggest participants' networks fulfilled functions of information diffusion, social influence, and behavioral regulation. More broadly, the findings suggest the importance of social structure and social interaction for participation in collective violence. Relational data should complement individual attribute data in predicting participation. The findings also suggest, contrary to the neo-Malthusian interpretation, that the role played by Rwanda's extraordinarily high population density in the violence may have been

immunity disorder, haematological deficiency , stress, local trauma, infections and infections systemic is considered as predisposing factor. 5,7 Histology lesion show epithelium ulceration, with exudate on its surface, tissue necrotic, infiltrated of cell inflammation, lamina propria edema with various degrees neutrophils and infiltration cell mononuclear, as well degeneration hyaline. There are a number cell surrounding vessels blood inflammation. It can be seen an expansion and congestion capillaries, cell endothelium vessels blood enlargement and narrowing of the lumen of the vessels. Inflammatory process play role important in emergence of apthous stomatitis. 8 Pain can originate from excessive inflammation and irritation, chemical afferents that are not once ends on connection layer epithelium and subepithelium. This pain can hinder action general like chewing, speaking, swallowing, besides influence quality patient of life. 6,9 Management of stomatitis is still focused for relieve the symptoms. The goal treatment is to reduce inflammation, reduce pain, lengthen period between plague disease and accelerate healing. There are many approaches treatment, however topical corticosteroids still become gold standard. They have showed a profitable effect in reduce pain and ulcer duration. However thus, its continued use continuous and not appropriate can cause side effect, mainly related with possibility of drug absorption systemically. 5 Previous findings revealed that the prevalence of stomatitis is slightly higher among children whose parents have a low level of education, those who live in rural areas, as well as those that are lower economic status, but the difference is not significant. Likewise, Patil et al. put forward that stomatitis is more common in the samples with low socioeconomic status. 2 In conversely, some studies find it higher prevalence in families with better socioeconomic status. Like Suryanata et al. mentioned, stress can become one of the risk factors for developing RAS and occurs in high economic status. 4 There is differences in health status of oral and oral health quality between ethnicity. Malay reported bad quality of oral health when compared with the Chinese and Indians. Difference quality of life between ethnicity associated with disturbance in food and talk with obviously, with the Malays feel little impact more Lots compared to with the Chinese. 10 Based on findings from previous studies, then need done study with subject ethnicity Malays in Indonesia. Study using IFLS data in which the author ever before analyze risk factors of toothache. 11 The aim of the study is to examine the relationship between a number of factor consists of age, gender, education , socioeconomic status and general health conditions with incidence of stomatitis in ethnicity Malays in Indonesia use IFLS data of 2014. Variable main are related data with experience ulcer during last four weeks, which is obtained from question IFLS survey, " Have you experienced oral stomatitis in last four week?". Besides that, the covariate hypothesized potential related with ulcer including age, gender, place residence , economic status , level education level, and general health conditions. --- METHOD Data analysis using SPSS version 25 for analyze effect independent from covariates on the outcome question canker sores . We also categorize variable independent become two group and do labelling in accordance with the proposed hypothesis. The predictors were: age , gender , area housing , economic status , level parental education . All procedure reviewed in an accurate way and approved by institutional review boards in the US and Indonesia at Gadjah Mada University. Informed consent was obtained from respondents and parents or legitimate representative in a manner law. Informed consent was given before starting work. Statement of anonymity and confidentiality has finished before beginning survey. All procedure studies done in accordance with principles Declaration of Helsinki. respondent with prevalence the highest followed with respondents who did not school and elementary school and junior high school , meanwhile respondent with education last diploma/S1/S2/S3 is respondent with least prevalence . --- RESULTS Respondents' ethnicity Malays living in rural areas own prevalence oral stomatitis compared with respondents living in urban areas . Respondents with health status unhealthy and very unhealthy was higher in prevalence 20%. Whereas healthy group was 11% and that is very healthy amounted to 5%. Respondents with economic status quintile currently occupy order highest case oral stomatitis followed with quintile low and very low . As for the economic status high 6% and very high 10%. The results of the analysis in Tables 2, 3, 4, 5, and 7 show no significant relationship between type gender, age, grade of education, residence, and economic status with incident thrush . Table 6 shows significant relationship between general health with incident of oral stomatitis . --- DISCUSSION Oral stomatitis is inflammation that occurs in some location in the mouth such as on the tongue, palate, cheek mucosa, lips mucosa, gums, floor of the mouth and other locations. Oral stomatitis can lower quality life somebody like difficulty in speaking, chewing, and swallowing. The circumstances can too reduce the quality sufferer in studying and working. 3,12 Prevalence of oral stomatitis by ethnicity Malay in Indonesia was 11%. It accordance with study previously suggested its prevalence range between 5-25%. The study was conducted by Mathew, reported that the prevalence of RAS is 2.1% in south India. 13 In other study conducted by Patil et al. , reported that prevalence of RAS in northern India found by 21.7%. 2 A study conducted in Maharashtra reported that only 72 patients out of 71,851 patients were diagnosed in a manner clinical with aphthous stomatitis with prevalence 0.1%. 3 On this study, the prevalence of oral stomatitis was slighly more among the respondent's female compared to with male. In a study conducted by Patil et al. have reported that there are higher prevalence among female than among male . 2 It has been suggested by several studies that hormonal factors may be responsible for the higher prevalence of stomatitis among girls. 14 In contrast, few studies have reported this higher prevalence among men. 15 The results show no significant relationship between age, gender, residency, education, and economic status with incidence of oral stomatitis. In contrary with research conducted before demonstrated there is difference significant between male and female. Likewise, previous findings disclose that more oral stomatitis often happened to the respondent with low economy. 2,4 Besides that, study previously put forward that prevalence more oral stomatitis higher among respondents who live in rural areas from urban. 13 The study also finds significant relationship between condition health general with incident of oral stomatitis. Where thing the confirmed with exists frequent cases of stomatitis occurs in patients with disease systemic certain such as diabetes mellitus, anaemia, infection systemic and immunity disruptive. Systemic disease usually become predisposition of stomatitis. [16][17][18] This study had limitations so that need study more carry on in development method. Limitations from study were; 1) measuring study tools used questionnaire "Is respondent experience ulcer in last four week"; 2) characteristics respondent with category rural and very unhealthy were few compared to others; 3) no followed longitudinally; and 4) no question related what factor trigger happening thrush/oral stomatitis. --- CONCLUSION Prevalence of oral stomatitis by ethnicity Malay in Indonesia is 11%. No significant relationship between age, gender, residency, education and economic status with incidence of oral stomatitis. There is significant relationship between health condition with incidence of oral stomatitis. Study results recommend necessity of oral hygiene in patients who have unhealthy condition. Disease oral stomatitis is necessary get attention from the dentist using education assembled prevention The conclusions.

Background: Indonesia Family Life Survey (IFLS) was used to formulate various policy government. In that survey, respondents were also asked about experience thrush (oral stomatitis) during four last week. Objective of this study is to analyze relationship of several risk factors with incident of oral stomatitis of Malays in Indonesia. Method: This cross-sectional study done based on the data obtained from IFLS in 2014. Data analysis used excel software (univariate data) and IBM SPSS version 25. Prevalence of oral stomatitis in ethnic Malays in Indonesia are 11% (26 cases of 236 respondents). Analysis results SPSS shows not significant relationship between age, gender, place residence, education, and economic status with incidence of oral stomatitis. Result: There is significant relationship between condition health general with incidence of oral stomatitis (p < 0.05), Conclusion: Study results recommend necessity look after and care cleanliness cavity mouth in patients who have unhealthy condition.Oral stomatitis/thrush is general terms of inflamed and painful mouth and may happens anywhere inside the mouth ,including buccal mucous, gums, tongue, labial mucous, and palatum 1 . This can cause painful lesions that cause difficult to eat, drink and swallow. Besides pain, multifactorial and generally repeated. This characterized by its periodicity and for limit himself alone. Prevalence range between 5 and 25% of the population. [2][3][4] Stomatitis manifests with appearance one or multiple painful ulceration, covered by pseudo membrane and white or greyish and surrounded by a clear erythematous halo. Lesion usually located on the mucosa non-keratinized mouth and can show recurrence after period variable remission. 5 Stomatitis can be caused by a viral infection or bacteria, disease, or bad oral hygiene and gums. Other factors including smoking reaction, wound burn consequence of hot food or drink, or allergic reaction. 6 Stomatitis is also caused by local infection, systemic disease, irritation, physique or chemical or allergy reaction. This lesion was unknown etiology, however several factors like heredity,

Introduction Methadone maintenance treatment is seen globally as one of the most effective methods, both in terms of cost and patient outcome, to help people withdraw from using opioid drugs [1,2]. Many countries have introduced programs to assist in reducing the number of people using these drugs [3] so as to prevent morbidity and mortality directly in relation to opioid drugs, but also due to co-morbidities, such as infection [4,5]. MMT has shown efficacy in improving the quality of life of drug users and reducing their need to use other drugs in several studies worldwide, such as in China [6,7] and the US [8,9]. In the US in 2007, intravenous drug use was estimated to cause an economic burden of $193 billion, mainly due to loss of productivity and absenteeism [10]. Reducing this economic burden of drug use through methadone maintenance is seen as highly advantageous, in and can lead to greater numbers of drug users in the workforce and therefore greater productivity [11]. In Vietnam, there are roughly 260,000 HIV patients [12] and 180,000 people who inject drugs [13]. Both groups are strongly interlinked as around 30% of the HIV positive population are PWID [13,14]. HIV in Vietnam is largely concentrated among the drug using population so targeting of this group through MMT services to reduce HIV is of benefit. Services for both have been rapidly increased in recent years to now include 364 HIV outpatient centers [15] and 251 MMT centers [16]. This improvement has been mainly fueled by investment from international AID organizations, such as PEPFAR and the Global AIDS fund; however, more recently, funding has declined. Maximizing the public benefit from both services is therefore very important. Using MMT services to target HIV patients or those at risk of HIV has already been studied in Vietnam. While drug use and HIV transmission have already been linked [17], alterations to drug usage through MMT programs have been shown to improve HIV/AIDS patient quality of life within the country [18]. MMT services could, therefore, be used to assess HIV risks and prevention among this vulnerable group of patients. Currently, most data in Vietnam regarding MMT services in relation to HIV knowledge, attitudes and practices focuses mainly on urban areas and cities [19]. There is substantially less information regarding MMT and HIV in mountainous provinces. People here have been identified as having increased rates of health problems, including drug use and HIV remain among others [20]. Access to services in these areas is also limited giving rise to a larger hidden drug using population and more people who do not know their HIV status [21]. As more patients utilize MMT services in these regions, the hope is that information regarding HIV will also become more prevalent. To maximize health services resources, examining the effectiveness of MMT programs in improving the KAP of HIV/AIDS is important. Ideally, as patients utilize the services, they should improve their KAP towards HIV and take this into the community to reduce the transmission and risk of HIV/AIDS among the hidden drug using population as well. This study aims to evaluate the changes in HIV knowledge, perceived risk and HIV testing uptake among MMT patients follow a year of treatment. This could help to identify areas where improvements are needed. --- Materials and Methods --- Study Setting and Population Pre-and post-assessments with no control group from December 2014 to December 2015 in six clinics in three mountainous provinces namely Dien Bien, Lai Chau, and Yen Bai province. Six clinics were selected, including Provincial AIDS Centers in Dien Bien, Lai Chau, and Yen Bai City; and District Health Centers in Thanh Xuong, Phong Tho and Nghia Lo. We invited patients who met following inclusion criteria: Initiating MMT service at one of the selected clinics at the time of the baseline assessment in 2014; volunteering and agreeing to participate in research by giving written consent; and having good physical and mental health to be able to answer the questionnaire. MMT patients were interviewed in the first month and 12th-month of MMT program. In this study, the sample size was calculated using the formula for estimating the difference between two population means. With absolute precision ∆ = 0.05 and standardized mean difference = 0.18 [22], the sample size required in each province was 100. The total sample size was 300 patients. The study first screened 300 patients participating in the MMT program; after 12 months of treatment, 56 patients withdrew, and 244 patients remained. We compared results in the first month and 12th-month of 244 remaining patients. The percentage of male participants in our sample was 100%, since only male patients registered at the clinics we studied at the time of investigation. --- Measures and Instruments Face-to-face interviews were conducted by well-trained researchers at the Hanoi Medical University using a structured questionnaire in Vietnamese. The questionnaire was piloted in the small group with 10 patients. After receiving feedback from patients regarding logical, language and text perspectives, we revised the questionnaire to make sure that the content was appropriate to the context of study, as well as the quality of data. Socio-demographic characteristics were collected consisting of age, gender, ethnicity, education, marital status, and occupation. --- Knowledge of HIV Treatment and Prevention The knowledge included: HIV/AIDS transmission; susceptibility to HIV infection; and HIV prevention according to five questions set by the Vietnam Ministry of Health in 2007. The respondents were asked to answer five True/False questions. According to the guideline of the MOH, MMT patients had good knowledge if they correctly answered all of five questions, while others did not have good knowledge if any one of five questions was answered incorrectly. --- Attitudes of HIV Treatment and Prevention The attitudes included: Self-assessed HIV status; reasons for risk of HIV/AIDS infection; and reasons for no risk of HIV/AIDS infection. --- Practices of HIV Prevention and Treatment The practices included: Using HIV Testing and Counselling Services before beginning MMT; provider-initiated testing and counseling; HIV Testing; and HIV Testing result. --- Statistical Analysis We described frequency and prevalence of variables at the baseline and in follow-up data. Chi-squared test and Fisher's exact test were used to assess the difference between before and after MMT intervention. Statistical significance was set at p-value < 0.05. The effect size was performed to determine the effect of MMT program. Among effect size, Phi was used to estimates the extent of the relationship between two binary variables and Cramér's V was used with variables having more than two levels. In the context of effect size level classification, very small, small, medium and large effect sizes equating to the values 0.01, 0.2, 0.5 and 0.8, respectively were used [23]. Data analysis was performed by using STATA software version 12.0. --- Ethics Approval This study was approved by the Vietnam Authority of HIV/AIDS Control's Scientific Research Committee. --- Results Among 244 patients in this study, majority belonged to the Kinh ethnic background , had less than high school education , were living with spouses or partners and being employed as freelancers . Table 1 depicts the change in knowledge of HIV/AIDS prevention and treatment among MMT patients over 12 months. Most participants thought a healthy-looking person cannot be HIV infected, both at the start of the study and after 12 months . Significantly more participants at 12 months knew that sharing needles and syringes was a method of HIV transmission, the effect size of change for this indicator was 0.022. There were 94.7% and 96.7%, respectively, of participants at baseline and 12 month time point identified injecting drug users as being susceptible to HIV. Majority of participants at both time points identified faithfulness and condom use as being protective against HIV, however, they also incorrectly identified mosquitoes and sharing food as modes of transmission. Fewer participants at 12 months said their knowledge was not good . The changes to attitudes of HIV risk among MMT patients at 12 months are described in Table 2. Most patients at both times point perceived that they were not at risk of HIV . Drug injecting was the most common reason identified for increased risk of HIV while 15.2% and 17.4% at 12 months also identified sex without using condoms as a reason for risk. 72.7% and 70.7% at 12 months identified not sharing needles as the main reason for no risk of HIV. Being faithful, using condoms and not having sex with female sex workers were other common reasons for no risk. Table 3 depicts the changes to practices of MMT patients for HIV treatment and prevention. Slightly more patients had used HIV testing and counseling services at 12 months than at the start . District level health bureaus were used by 91.8% of patients at the start and 94.1% of patients at 12 months for provider-initiated testing and counseling. --- Discussion In this study, we aimed to assess the level of the knowledge, attitudes, and practices regarding HIV/AIDS of MMT patients in mountainous areas and possible improvement after one year of following MMT program. In general, the majority of our participants displayed fairly sufficient knowledge regarding HIV transmission and prevention, appropriate HIV risk perception and adequate HIV testing update. However, very limited changes were found regarding the KAP at 12 month follow up compared to program initiation. This highlighted the difficulties of enhancing HIV/AIDS KAP in less developed areas and called for greater educational and counseling efforts. Over the course of 12 months, the only significant KAP improvement in our study was seen in the knowledge of MMT patients about sharing needles as HIV transmission mode, with a small effect size of the change. Such finding was in line with a study in China examining the long-term effects of methadone maintenance treatment that also demonstrated improvements in knowledge with regards to needles sharing [24]. Another study in Nepal concerning safer drug use interventions, such as sterile injecting equipment and education, also showed that knowledge of HIV of participants increased while unsafe practices decreased as they progressed with the intervention program [25]. Nonetheless, although MMT services have been expected to be a potentially beneficial route for targeting HIV risk and prevention as HIV prevalence among PWID was substantial [26], literature looking specifically into how MMT impacts the KAP of HIV have been limited. The lack of improvement in KAP found in our study suggested that there were probably insufficient educational efforts at MMT clinics to enhance HIV-related knowledge of MMT patients. To our knowledge, in mountainous MMT clinics, patients gained information mostly through health staffs-provided informal advice and routine counselling when coming for their daily MMT dose. Organized, formal educational campaigns on HIV-related issues were scarce. In addition, while most participants correctly identified main HIV modes of transmission and methods of protection-transmission via needles sharing, protected by protected sex etc., a substantial proportion of respondents reportedly considering mosquitoes and food sharing as HIV transmission route. This suggests potentially serious shortcomings in HIV knowledge of those in mountainous areas, which could also be a result of the shortage of educational and counseling activities/programs available that otherwise should be included in MMT clinics. While MMT services may have been adequate at providing the actual methadone treatment, they seem to lack a holistic service and complete healthcare intervention for the patients. As the HIV epidemic in Vietnam is concentrated among PWID, to improve HIV testing uptake among PWID and to maximize the expenditure on services, it would be beneficial for the patients to have access to both MMT and HIV-related services simultaneously. MMT patients have also indicated their desire for integrated and decentralized services [27] and one study has demonstrated that when MMT services are integrated with HIV centers, HIV testing and counseling rates increase [28]. Such integrated approach to tackle HIV-related issues in drug-using patients would be of even greater help for a mountainous population in Vietnam, as previous studies have shown that geographical barriers, such as distance to services, which likely to undermine adherence to MMT program, would also hinder HIV service uptake [29]. Several implications can be drawn from this study. Greater efforts should be spent on improving the knowledge of MMT patients with regards to HIV risks, probably through educational and counseling programs provided by the staffs of the MMT clinics. Potentially more beneficial is the integration of MMT and HIV services within an MMT clinic, so that an MMT clinic would offer a comprehensive package of services covering MMT, HIV testing and HIV counseling. In the Vietnam context, the benefits of such comprehensive services-offering clinics have been appraised by different parties involving in these services, including clinic directors, providers, staffs, and patients [30]. Policy and clinical-level movements are recommended to be conducted, including issuing legal framework and policies facilitating integration, staff training and improving collaboration across different services providers to tackle documented barriers to such integration [30]. In addition, our findings suggested some directions for further researches on the subject of HIV-related KAP-in particular the sources from which patients obtain HIV-related information, the basis of their beliefs with regard to this knowledge, or whether there were any other influences during the time between points of data collection. --- Limitations Of 300 patients, only 244 participants completed the survey as 56 withdrew before the 12 month period was over. The withdrawal of patients, typically those with lower adherence to MMT and worse KAP toward HIV risks and prevention, may lead to bias in the changes seen. The patients that completed the study and adhered to treatments may already have better KAP about HIV risks and prevention even in absence of MMT enrolment, implying that there would be greater positive changes if the 56 patients were included. Although the sample was quite small, six centers across provincial and district level services were chosen hopefully making the data representative of MMT users generally in mountainous areas. There is a risk that the sample is slightly skewed towards people who know more about HIV than the generic drug using population as these patients are actively seeking medical care and therefore may care more about their health than the remaining hidden drug using population. --- Conclusions In conclusion, this study found a limited improvement in HIV knowledge and testing uptake among MMT patients following a 12-month period. It also highlighted some shortcomings in the KAP of these patients, in particular, incorrect identification of HIV transmission routes, among patients both at program initiation and follow-up. The findings lent support to the argument for enhancing education and counseling efforts at MMT clinics regarding HIV, as well as for improving access to preventive and health care services through the integration of MMT/HIV services. ---

Methadone Maintenance Treatment (MMT) program has been considered a medium through which human immunodeficiency virus (HIV) risks assessment and prevention on drug

Introduction Background to the study Condoms are barrier contraceptive devices for dual protection available in two types, the male and the female condoms World Health Organization, [14] . When condoms are used correctly and consistently they are most effective in prevention of unwanted pregnancy and prevention of spread of sexually transmitted infections including HIV/ AIDs. The female condom prevents conception of unwanted or unplanned pregnancy in similar proportion to that of male condom that is 95 per cent. The female condom is a female empowerment tool to increase use of condoms for family planning and prevention of HIV/ Aids, compared to male condom; a female partner has more control to negotiate for the female condom than a male partner. [14] . In a research done in the United States on availability of messages related to female condom use, it was found out that majority of the websites had little details in promotion of female condom use. This posed a barrier in use of the female condom. Those websites which had the message on female condom were not detailed but just mentioned the female condom as a barrier contraceptive method for family planning. Women from low income urban U.S, reported that when the female condom was introduced for the first time many women tried to use it increasing the incidence of using the female condom. However, there wasn't sustainable or continued use of the female condom after the first use due to unavailability of the condom and challenges faced during the first trial. It was reported that female condom were available during the time of campaigns and promotion of the female condom but were not available in the nearby shops for continued use [5] . In Sub-Saharan Africa, women accept and are ready to use the female condom when they are given access to the device and exposed to interventions which support the use of the female condom. This was found out after analyzing various researches on acceptability of female condom in Sub-Saharan Africa [1] . In another study in Zimbabwe, and China it was revealed that unavailability of the female condoms in the shops and other sales outlets led to low utilization of the female condom. The respondents reported that whenever they sought the female condom they never accessed them [3] . Low utilization of female condom was related to low level of awareness among the youths, acceptance and availability in shops and chemists determined use of female condom in Ghana. The youths who readily sought and found the female condom used it while those who didn't find the female condom in the nearest shops and chemists reported not to have used the female condom [1] . On accessibility, the distance to the nearest shop and availability of the female condom can promote use of the female condom or deter one from having safe sex. Male condom dispensers are located strategically near point of use to reduce the distance to pick them; however when one travels long distance to access female condom and incase of unavailability of the female condom this reduces chances of using the female condom [12] . --- Methodology A cross-sectional descriptive study design was used to generate both quantitative and qualitative data through selfadministered questionnaires, focus group discussions and indepth interviews respectively. This design enabled the researcher to collect data assessing both the use of the female condom before and after at a short specific period of time. The study population in this study was youths between 18 and 35 years of age in selected tertiary training institutions in Migori County. Youths in tertiary colleges come from diverse backgrounds and regions, are accessible and most of them have the knowledge about HIV/ AIDs and its prevention mechanisms. The study employed two sampling techniques: Purposeful sampling of the tertiary institutes in Migori County and proportionate random sampling of research participants. Purposive sampling was used to identify the target study population on the basis of specific considerations by the researcher such as, the health of the youths is of paramount importance to the building our nation since they represent majority work force and professionals. Reliability of the instrument was estimated using the split half method. Odd-even split method was used to obtain the two halves. The following Spearman Brown prophecy formula is used to calculate the reliability coefficient. The Cronbach's alpha was calculated and was found to be at which showed a high degree of reliability of the variables. --- Data collection and analysis The questionnaire comprised of sections that looked into the demographic characteristics of the youths, accessibility of the female condom by the youths, the cost of the female condom, and socio-cultural barriers to use of the female condoms. Qualitative data was collected from the youths using open ended questions, focused group discussions and in depth interviews with the health care workers at the tertiary institute health clinics. The study was done in two phases; phase one involved data collection using selfadministered questionnaires administered to youths and the second phase involved conducting two Focus Group Discussions and in-depth interviews with women of 18 -35 years of age in each of the selected institution. The data collected was then grouped into categories, tested for homogeneity and normality of distribution, Chi Square was used to check for the significant associations between the variables at 95 percent confidence interval, a p value of p ˂ 0.05 was used. The qualitative information collected was coded into the respective categories illustrating the various themes, then was entered into computer for thematic analysis using N-Vivo version 11. Quantitative data was analyzed using Statistical Package for Social Sciences . Bivariate analysis was used to measure the strength of associations while the multivariate analysis was used to adjust the confounders. Descriptive statistics specifically tables and bar graphs were used to present the findings. --- Results The ages of the respondents were ranging from 18 to 35 complete years. More than half of the participants were aged between 18 and 25 years of age, while the remaining were aged between 25 and 35 years. Most of the youths were aware of the topic under study; however increase in age was associated with minimal or low level of awareness of the female condom. Expectedly majority of the respondents were youths who were single , while the remaining were married and others separated or cohabiting, and respectively. Most of the respondents were Christians by denomination 95.3% while the remaining participants belonged to Muslim denomination. On source of income, most of the respondents reported to receive their main financial support from their parents/ guardians, from relatives and their partners 80.5%, 13.4% and 6.1% respectively. Table 1 shows a summary of demographic characteristics of the youths under study. --- Awareness and use of the female condom Majority of the respondents reported that they were aware of the female condoms and that they can access them from the chemists/ pharmacies 50%, health facilities and shops/ super markets 38.2% and 11.8 respectively. The source of the female condom was not significant, the youths never minded where to access them from so long as they get the device for use when there is need to use the female condom . However, the youths who were aware of the female condom use were more likely to use the device than the ones who not aware of it benefits and where to access it from. This was significantly associated with use of the female condom . There were more posters for male condom use than for female condom use. With the availability of the posters and other information and communication materials for male condoms being easily available, more youths were conversant with ease use of the male condom than the female condom. Respondents were able to explain the correct procedure for male condom use than for female condom use. However, availability of the female condom posters was significantly associated with the use of the female condom . In a research done on knowledge, acceptance and utilization of female condom it was found out that increased awareness of the female contributed to increased utilization of the female condom in Ghana [1] . In India, incorporation of counseling on female condom in provision of family planning services was found to increase utilization of the female condom . However, in this study it was found out that the youths are aware of the female condom and its benefits but rarely use it due to others factors. --- Availability and use of the female condom Most respondents 98.2% reported that the female condom were not easily available in the college's reproductive health clinic while 1.3% said sometimes the stocks are there and the remaining 0.5% agreed that the stocks were available in specific points of sale. Availability of the female condom was significantly associated with use of the device at a p value . Logistic regression was done on where to access the female condom from and availability of the female condom at that point, it was found out that most of the time the female condom was out of stock and this affected its use among the youths. The youths reported that it took time before replenishing the stocks and also the working hours for the reproductive health clinic were limited as they spent the opening hours of the clinic in class. However, there was no statistical significance between where to access the female condom from and availability of the condom . In a study by Obembe et al., accessibility of the female condom, however, the study didn't indicate significance of where to access the female condom from [13] . When the researcher probed more on female condom distribution, the respondents reported that mostly they are supplied with male condom dispensers latter than the female condom dispensers within their place of residence. Availability of the female condom dispensers in accessible points was significantly associated with use of the female condom in most of the places of residence the female condom dispensers were not there 93.3%, while the remaining portion of participants reported that though the dispensers were not there, they can access the reproductive health clinic which is proximal to their place of residence 6.1%. During an interview with the key informant, it was evident that the institutions had procured and had in place male condom dispensers within the place of residence. This promoted use of male condom compared to female condom, however even the male condom was not consistently used. During focused group discussion, members reported that whenever they sought contraceptives in the reproductive health clinics, they are given other options and the female condom is rarely championed for as reported by one of the respondent that; Whenever I go for family planning services am given the option of injectable method, a jadelle or sure , the nurse has never bothered to demonstrate to me how to use a female condom, may be they also don't know how to use them . On the place where the point of distribution for female condoms was situated, majority of the respondents reported that the place was not accessible, while 15%%, 17.4% reported that they were not sure and some can access the point of distribution respectively. It was found out that whenever there was need for the female condom use, most of the supermarkets and shops around the students' hostels had no stocks. In the reproductive health clinics, there was no specific place or point to pick the female condom from; the respondents reported that in case one needed to get one, she had to be physically given the condom by a health provider. This barred many youths since they need a place where they can walk in freely any time of the day and night to pick the female condom and use. Therefore the places where the female condoms were stored or distributed from weren't accessible at all times, especially at night when the youths would like to use the device. This was significantly associated with use of the female condom . Most youths during focus group discussion reported that the male condom was easily accessible at all times hence likely to be used more often than female condom. During an in-depth interview with the nurse in charge of the reproductive health clinics in the selected institutions, it was evident that they too had no stocks for the female condom but enough stocks of the male condom. It was found out that whenever they sought the device, many shops, supermarkets and chemist never stocked the female condom. The place of residence had male condom dispensers and no female condom dispenser; this made use of the male condom preferred in comparison with the female condom. Availability of female condoms within the place of residence was a determinant of female condom use, most youths reported that whenever they opted to use the device, little did they know where to get them from, this contributed to low utilization of the female condom among the youths. Acceptability and availability of the female condom were associated with low utilization of female condom in Tanzania [5] . These were replicated in the current study and also in china in a study by Ananga et al., which reported that limited access to the female condom led to low utilization of the female condom [1] . The female condom dispensers and female condom posters and other information, education and communication materials about female condom significantly affected use of the female condom. Availability of male condom dispensers within the colleges and place of residence made a male condom a common good in the market; however lack of female condom dispensers in critical places like place of residence, hostels and social places within the tertiary institutions made it difficult for the youths to access and use the female condom. When key informants from each tertiary institution were interviewed it was evident that most health care facilities do not stock the female condom and none of the institutions had a female condom dispenser, others even lacked the demonstration models for female condom insertion. These findings concur with those revealed by [10,7] . --- Cost and use of the female condom The cost of the female condom was considered relatively expensive compared to a male condom, 37.9% of the youths were not sure if the female condom was affordable or not while 35% disagreed, 6.8% strongly disagreed, 20% agreed and 0.3% strongly agreed that the device was affordable , this meant that the female condom was 15.4 times more likely to be used if it was made affordable. However, during focused group discussion it was evident that even for those who can afford weren't able to use the device due to unavailability of the stock. In the selected tertiary institutions, the device was not stocked frequently due procurement issues surrounding the cost of the female condom. During sessions of group discussion, it came out clearly that those who had used the female condom previously they had gotten it from health workers during campaigns for HIV/Aids prevention where they were taught on how to use and were given some for use. Later after using the few devices given, they haven't used the device again because they scarcely know where to get the female condom from and cost implications. When I went to buy a female condom from the supermarket, imagine one packet goes for two hundred Kenyan shilins while that of a male condom was sold at fifty Kenyan shillings. And it is used only once. The packet had only one condom while a packet of male condom had three. My partner opted for the male condom . The researcher probed the respondents to establish if the female condom was available and affordable will that increase the utilization, majority of the respondents 77.6% were willing to use the device while 16.8% said they won't use and 5.6% proportion of the respondents was not sure since other factors will determine the use or disuse of the device. The attitude of the user was discussed during focus group discussion to determine the level of utilization of the female condom if it was available and affordable . When discussing on issues of financing in focus group discussion, it came out from the participants that the youths were not fully satisfied with the support they receive from the principal supporter therefore seek more money from both relatives and sexual partners. Majority of the respondents reported to have gotten main financial support from the parents, some who received support from partners and relatives were likely to have casual relations and this increased use of the device among those who engaged in casual sexual relationships, . The main source of income was also considered significant in this study, those who received enough money for use reported low use or disuse of the female condom compared to those who do not get adequate funding from home sought extra money from casual sex partners as this was revealed in focused group discussion. On further probing and analysis, it was revealed that the cost of purchasing the female condom was a barrier to the users; this is in line with the findings of [11,13] which indicated that the high cost of the female condom lowered its availability in shops and other places of sale. These results were replicated in a study by Mashanda-Tafaune in Botswana [8,6] . The significant variables that were identified in this study are shown in table 2. --- Conclusion Awareness on where to get the female condom from was significant in relation to use of the female condom. The youths were not aware of the female condom and its benefits had not used it. However, utilization of the female condom was low. Unavailability of the female condom significantly affected and limited use of the condom. Those youth who felt the need to use the female condom were unable to access it when needed. Where to access the female condom from was not significant so long as it was available. Affordability of the female condom was associated with disuse of the female condom. Whenever the youths were able to afford the female condom, it was not stocked in the shops and kiosks. The male condom was found more stocked than female condom in the supermarkets. The cost of the female condom also posed a barrier to use among the youths. Male condoms were reported to be distributed for free and there were male condom dispensers near the place of residence unlike female condom dispensers. On accessibility, factors related to availability of female condom and the cost significantly affected use of the female condom. Reproductive health clinics lacked model for demonstrations for female condom insertion process and female condom dispensers and rarely promoted use of the female condom. However, awareness on where to access the female condom from was not significant compared to awareness on benefits of the female condom use. --- Assumptions The study assumed that the respondents gave honest information. --- Study limitations The research was limited to youths aged 18 -35 years, who gave personal report of being sexually active. The respondent's feelings, attitude, social interaction and failure to respond to certain items in the questionnaire due individual's culture and values were not easily controlled but large sample size was taken to minimize the errors in data collection.

The female condom is a dual protection tool for both prevention of HIV/ Aids transmission and unwanted or unplanned pregnancies. The study was done in Migori County among the youths attending selected tertiary training institutes, with specific objectives of identifying the extent in which access to the female condom and other social cultural factors affecting use of the device in Migori County. A cross-sectional descriptive study was used to collect both quantitative and qualitative data. A pre-test study was conducted in Msomi teachers training college to test the validity and reliability of the data collecting tools. The tertiary institutions were purposively selected while the research participants in the selected institutes were randomly and systematically selected and a sample size of n= 385 was chosen proportionately. Data was collected using structured questionnaires, in-depth interviews and focus group discussions. It was evident from the research findings that majority of the youths 83.9% were single with 61.3% of the youths getting information about female condom use from media, 73.9% used it to prevent pregnancy than to prevent transmission or contracting HIV/Aids. Availability and cost of the female condom was significant at p≤0.001, attitude of the user and partner was also significant at p value of 0.024.

Introduction Today's youth have almost universal access to the internet and frequently engage in social networking activities using various social media platforms and devices [1]. This is a phenomenon that hate groups are exploiting when disseminating their propaganda [2,3]. Data gathered from a demographically balanced sample of over a thousand youth in the United States showed that approximately half of them, within a study period of three months, experienced exposure to hateful material while online [4]. Analyses of interviews with right-wing extremists demonstrated that communication over the internet provides an effective networking method amongst their supporters. These groups use the internet to convey their racist messages and adapt communication strategies that are appealing to youth, with the scope of recruiting new members, including the use of images, videos, music, and online games [5,6]. The online space also provides hate groups with a unique opportunity to portray an image of unity and identity [6,7]. This online collective identity self-perpetuates as a welcoming space of like-minded individuals, providing sought-after validation for potential members' societal grievances and attracting socially isolated individuals to become members of a community that accepts them [8]. Furthermore, the anonymity of the internet creates an environment where hate groups speak or act out more radically compared to what they would do in person [9]. This is in part attributed to a psychological process referred to as online disinhibition, a process that most individuals experience when online compared to real-life leading to a lack of restraint and increased openness of expression [10,11]. The literature describes two types of online disinhibition: toxic and benign disinhibition. Toxic disinhibition manifests as a propensity towards a variety of negative attitudes and behaviors such as anger, vicious criticism, outgroup hatred, cyberbullying, racism, and aggression [12][13][14][15][16][17][18][19]. On the contrary, benign disinhibition is mostly described as a positive process by which individuals feel increased comfort in manifesting acts of kindness when online compared to in-person which, however, in some cases can lead to undesirable situations [13,[20][21][22][23]. The investigation into online disinhibition and exposure to hate remains relatively nascent-to our knowledge, there has only been one study that has demonstrated an association between online disinhibition and exposure to hate online [12]. The presence, form, and function of hateful material available on the internet has been extensively discussed by academics and practitioners [24][25][26][27][28]. Past research has focused on long-term exposure to hateful content online showing that it might reinforce discriminatory views and could lead to developing defensive and hyper-vigilant attitudes [29][30][31]. Some have theorized that access to such material influences the likelihood that an individual will, eventually, engage in hateful or violent behavior [31][32][33]. However, there is limited knowledge on the risk factors that lead an individual to be exposed to online hate in the first place and on the consequences of such exposure. In recent years, the investigation into the risk factors for exposure to hate in the online space has burgeoned. In 2011, a US-based national study demonstrated that amount of general technology use and age are predictive factors for almost all technology-based violent experiences and exposures [32]. Recent research in the US has identified various socio-demographic, psychological, and behavioral factors that may put an individual at risk of exposure to hate online, such as young age, white race, male gender, level of education, online victimization, low trust in government, and time spent on the internet [34,35]. Such risk factors provide invaluable information on a pernicious phenomenon that has persisted at high levels over the past decade in the US. Indeed, data from the US have consistently indicated high levels of exposure to hate messages online among nationally sampled populations, with 53%, 65%, and 87% of respondents indicating exposure to such messages in 2013, 2015, and 2016, respectively [34][35][36]. The international knowledge base of the correlates for exposure to hate online has similarly grown over the past decade. Research from Finland has demonstrated that exposure to hate is associated with young age, social trust, weak family attachment, time spent online, social media use, visiting risky websites, offline victimization, and the production of hate messages [29,37]. Data in this context have also indicated high levels of online exposure to hate messages, with 38% and 48% of sampled respondents indicating exposure in 2013 and 2016, respectively [29,37]. Additional research from four countries , has also supported the notion that high levels of exposure to online hate persist cross-nationally [36]. As informative as this research base is, to our knowledge, no studies in the US have attempted to identify predictors of exposure to hate messages using data from after 2016. With such high levels of exposure both in the US and internationally, it is necessary to keep the information on risk factors up to date. Despite the growing body of literature surrounding the prevalence and risk factors for exposure to online hate internationally, to date, most studies have examined witnessing hate as either a victim-whereby the messages are targeted towards one's own group; or as a bystander-whereby, the messages are targeted towards a different group [29,38]. Interestingly, there exists a dearth of literature that focuses on the interactions between those espousing hateful messages and those who are exposed to them. While there is extant research on the efficacy of persuasive racist narratives on youth, to our knowledge no studies have identified the risk factors associated with encountering those who attempt to persuade youth by utilizing such narratives-referred to in the literature as persuasive storytellers [31,39]. Our study aims to address this gap in the literature by identifying factors that exist upstream of the interactions between youth and persuasive storytellers, which can inform educational initiatives that focus on preventing the interaction with them in the first place, rather than discrediting their narratives after youth have been exposed to them. There is concern that the internet may provide youth with a gateway to online hate communities and expose them to a dizzying array of sites containing hateful material and individuals espousing hatred [40]. As such, it is important to understand who are most likely to experience such exposure, to equip them with the requisite knowledge to critically assess the material they may come across while online. Careful attention must also be given to identifying various factors that may place an individual at greater risk for encountering someone who attempts to persuade them of hateful views, to provide educational initiatives opportunities to prevent this encounter from occurring. To this end, the present study aims to contribute to the growing body of literature on online hatred by attempting to elucidate the various risk factors that expose youth to hateful messages and those who espouse them in the online space. The research objectives for this exploratory study were as follows: 1. Identify youth demographic characteristics, behaviors, and psychological processes associated with exposure to hate messages online. --- 2. Identify youth demographic characteristics, behaviors, and psychological processes associated with encountering someone trying to convince the respondent of racist views. --- Materials and Methods We implemented a cross-sectional study design, using a quantitative questionnaire, in two high schools in Massachusetts . The questionnaire was implemented in December 2018 in one high school and in April 2019 in the other. At both schools, the questionnaire was implemented on paper during school hours as part of Health Class. Prior to implementation, the questionnaire was pilot tested in Utah and Massachusetts, and questions were revised based on youth feedback. A copy of the questionnaire can be found as a Supplementary Attachment. For the analyses presented in this study, we explored youth demographic characteristics, online behaviors, psychological processes, and their association with online exposure to hateful content-no other outcome variables were considered for this study. Below we present the independent and dependent variables considered in this study, as well as the methods for data analysis. Parents were provided with information on the study and opt-out forms one month prior to data collection. Consent to participate in the study was obtained from the students prior to responding to the questionnaire. The Helsinki ethics protocol was followed throughout the course of this study [41]. The study protocol and the survey questions were deemed exempt by the Harvard T.H. Chan Institutional Review Board as well as by the ethical committees of the school districts where the study was implemented. Below, we describe the socio-demographic characteristics of the participants. Following, we provide descriptive statistics of the sample and results from simple and multiple regression models. --- --- Independent Variables Respondents were asked about the amount of time they spend using technology each day, which social media tool they use, and how frequently they use it, with answer options ranging from 1 to 6 . As social media use across platforms is correlated, an overall social media use variable was created by a summative score created by converting the distribution of frequency of use for each social media platform into a normal standardized distribution and adding the resultant scores across all social media platforms [42,43]. To assess which social media platforms were related to the outcomes, a Spearman's correlation matrix was created. Respondents were asked how many of their social media followers they knew in person, and if they had recently removed any strangers from such followers; similar questions were asked about respondents' friends' social media behaviors, under the assumption that they would be less likely to misreport friends' habits compared to their own. Respondents were asked to report how frequently they played online video games, and if they chatted with strangers while doing so. Online disinhibition was measured using the online disinhibition scale after adapting the questions to the young age of the study population [13]. The factor structure of the questions was assessed with a factor analysis using principal component analysis for factor extraction, and as a result, a scale with a score ranging from 7 to 28 was formed, with higher values indicating a more disinhibited behavior. Participants' perception of online risk was measured by asking the respondent to rate the risk of seven online scenarios. In this case as well, a factor analysis was performed to assess the structure of the questions, and as a result, a scale was created with scores ranging from 7 to 35, with lower values indicating lower risk perception. For both scales, Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett's test of sphericity were used to test for the suitability of the data for factor analysis. Cronbach alpha was computed to assess the scales' reliability. Questions about engagement in risky online behaviors were also asked, such as chatting with strangers or sharing personal information with online contacts. Finally, respondents were asked about parents' supervision of their online activities and if they had a trusted adult to ask for help if they encountered an uncomfortable situation online. --- Dependent Variables Two dependent variables were created to measure exposure to online hate during the two months prior to the survey. Respondents were asked to report how frequently they had come across insulting verbal or written expressions against a specific group because of their race, religion, disability, sexual orientation, ethnicity, gender, or gender identity . Respondents were also asked if, in the same period, they encountered someone trying to convince them of racist views. Both variables were dichotomized as yes/no. --- Statistical Analyses Simple and multiple logistic regression models were used to study the association between the independent and dependent variables. A Box-Tidwell procedure was used to confirm that the benign disinhibition score, the only continuous variable in the model, had a linear relationship to the log odds of each dependent variable. Independent variables were included in the multiple models when a statistically significant association was found in the simple models . Gender, race, and school year were included regardless of their significance in the final models because of theoretical relevance. Hosmer-Lemeshow tests were used to assess the goodness of fit of the models. Race and gender were also tested as interaction terms [44]. Data analysis was performed using Stata Statistical Software: Release 15.1. College Station, TX: StataCorp LLC. --- Results We provide below the sample characteristics and descriptive statistics for social media use and online behaviors, adult supervision of online activities, and exposure to hate online. Following, we present the results of the factor analyses for the online disinhibition and risk perception scales, and the results of the simple and multiple logistic regression models for exposure to online hate. --- Sample Characteristics and Descriptive Statistics The majority of the respondents reported spending over three hours a day interacting with technology, excluding schoolwork. YouTube, Snapchat, and Instagram were the most frequently used social media tools with over 90% of respondents reporting their use. Only 7% of the respondents knew all of their social media followers, the majority of them reported that in the two months prior to the survey, they chatted with someone on social media that they had not met in person and 45% reported they believe their friends did so as well. During the same timeframe, 67% removed a follower from their social media account they had not met in person, and 48% reported to have shared personal information, such as their school or town name when posting on social media. Among those playing video games online, 52% reported that in the two months prior to the survey they chatted with someone they did not know while gaming. Regarding parents' supervision, the majority of students had parents that occasionally asked about their online activities, but only 25% reported that their parents had rules for what they did online and checked on them to make sure they followed the rules. The majority of students reported they had a trusted adult they could ask for help if they experienced an online situation that made them feel uncomfortable. More details on the students' online activities are provided in Table 1. Fifty-seven percent of students reported to have come across hate messages on social media or on a website in the two months prior to the survey and twelve percent reported to have encountered someone online that tried to convince them of racist views during the same time period. Details on students' exposure to hate online is provided in Table 2. In Table 3, we report the correlation between social media platforms and hate online. Exposure to hate messages online was significantly correlated with Twitter use and Houseparty use . Encountering someone trying to convince the respondent of racist views was only significantly associated with YouTube use . --- Factor Analyses and Descriptive Statistics of Risk Perception and Online Disinhibition --- Risk Perception Scale Bartlett's Test of Sphericity and Kaiser-Meyer-Olkin measure of sampling adequacy indicated that the data on the seven questions designed to measure risk perception were suitable for factor analysis. A factor analysis was computed resulting in one factor with an eigenvalue greater than one, all items had a factor loading greater than 0.4, and 52% of the variance in the data was explained by the model. Cronbach's alpha was 0.84. Risk perception scores were negatively skewed with a mean of 29.3 and a median of 31 . In the simple and multiple models, risk perception was examined as a dichotomous variable. Individuals with scores less than or equal to 28 were defined as having "low risk perception" . Individuals with scores greater than 28 were defined as having "high risk perception" . --- Benign Disinhibition Scale Bartlett's Test of Sphericity and Kaiser-Meyer-Olkin measure of sampling adequacy indicated that the data on the online disinhibition questions were suitable for factor analysis. A factor analysis was computed on the 11 questions resulting in two factors with eigenvalues greater than one which explained 45% of the variance in the data. After oblique promax rotation, one question with a factor loading below 0.4 was discarded. The resulting two subscales were similarly structured to the scale developed by Udris et. al. [13] The benign disinhibition subscale contained seven questions and had a Cronbach's alpha of 0.72. The toxic disinhibition subscale contained three questions. Due to its low internal consistency , toxic disinhibition was not included in the simple and multiple regression analyses. Benign online disinhibition scores had a mean of 18.4 and a median of 18 . Box-Tidwell test results from the simple regression analyses supported the assumption that the distribution of benign online disinhibition scores was linear to the log odds of respondents' exposure to hate messages online and to the experience of encountering someone online that tried to convince them of racist views . Descriptive statistics of the risk perception and benign disinhibition scale can be found in Table 4. In the simple regression models, the following variables were significantly associated with the dependent variable -exposure to hate messages online: time spent online , removing a follower from a social media account -whom the respondent had not met in person , communicating with someone on social media that the respondent had not met in person , presence of parental rules for online activities , benign online disinhibition , and good academic performance . Detailed results for the simple models of exposure to hate messages online can be found in Table 5. --- Encountering Someone Trying to Convince the Respondent of Racist Views In the simple regression models, the following variables were significantly associated with the dependent variable-encountering someone online that tried to convince the respondent of racist views: school year , playing video games online , and benign online disinhibition . The only significant categorical predictor with more than two categories was chatting-while gaming online-with someone the respondent had never met in person . Detailed results for the simple models of exposure to hate messages online can be found in Table 5. --- Multiple Models --- Exposure to Hate Messages Online The overall Likelihood Ratio chi-square test statistic for the multiple logistic regression model exploring the association between the independent variables and exposure to hate messages online was significant . Hosmer-Lemeshow Goodness of Fit test results confirmed that the model was a good fit for the data . Time spent online was associated with increased odds of exposure to online hate messages-youth that spent three or more hours a day online had 2.4 times the odds of reporting exposure to hate messages , compared to those who spent less than 3 h a day online. The odds of reporting exposure to hate messages among those who communicated with someone on social media that they had not met in person were 1.7 times that of those who had not done so . Benign online disinhibition was associated with reporting exposure to hate messages online-each one-unit increase in score on the benign disinhibition scale resulted in a 6% increase in the odds of reporting exposure to such messages. Good academic performance was also associated with exposure to online hate messages, students who reported receiving grades greater than a C had 3.4 times the odds of reporting exposure to such messages compared to students who were receiving Cs or lower grades . Gender and race were used to study their interaction with the following variables: time spent online, benign disinhibition, and communicating with a person not met in person while online. None of these interactions resulted to be significant. Detailed results for the multiple models of exposure to online hate messages can be found in Table 6. --- Encountering Someone Trying to Convince the Respondent of Racist Views The overall LR chi-square test statistic for the model investigating the association between independent variables and encountering someone trying to convince the respondent of racist views was significant . Hosmer-Lemeshow Goodness of Fit test results confirmed that the model was an appropriate fit for the data . Only benign online disinhibition was associated with students' risk of encountering someone trying to convince the respondent of racist views. Each one-unit increase in benign disinhibition score resulted in a 19% increase in the odds of experiencing this situation. There were no significant interaction terms observed. Detailed results can be found in Table 6. --- Discussion and Conclusions Technology has become increasingly important in the lives of adolescents who are heavy users of various forms of electronic communication such as instant messaging, e-mail, social media, and sites where they share opinions, photos, and videos [45]. Although teens usually find valuable educational support and information on the internet, they can also be exposed to negative influences, such as online propaganda, hate messages, and racism. The work presented in this manuscript aimed to explore some of the risk factors for such negative online influences. Specifically, our goals were to identify youth demographic characteristics, behaviors, and psychological processes associated with online exposure to hate messages and encountering someone trying to convince the respondent of racist views. In our study, we found that the more time youth spent online the more likely they were to be exposed to hate in the online space. This result is consistent with the previous literature [29,34,35]. Not surprisingly, communicating with strangers online was associated with an increased risk of being exposed to hate. Interestingly, good academic performance was also associated with increased risk of exposure, this may be due to increased awareness and ability to recognize the online material as hateful or by some interest in the topic expressed by higher educated youth, as found in previous research [35]. Finally, our data indicate that the more individuals felt disinhibited online, "loosened up, felt less restrained, and expressed themselves more openly", the more likely they were to be exposed to hateful propaganda and to encounter individuals attempting to convince them of racist views. Our results raise interesting questions about the nature of disinhibition and the underlying processes that moderate the relationship between disinhibition and exposure to hate. What is of specific interest here, however, is that benign disinhibition was significantly associated with exposure to hateful material. We believe that while toxic disinhibition may be associated with active engagement in hate and harmful activities, benign disinhibition as a whole may be associated with passive exposure to hate [12]. These preliminary results pose intriguing and critical questions for future research. First, does someone exhibiting only benign disinhibited behaviors stray free from the innate risks of hateful content? Second, what role does mere exposure play on one's own future behavior, feelings, and psychology? The results from this exploratory study can play a useful role in the development of educational interventions and programs that aim to prevent youth from encountering individuals espousing hateful views and equip youth with the knowledge they need to appropriately react to such material. In many prevention spheres, prevention efforts have focused on educating about what risks are present online, and indeed the prevalence and nature of nefarious online actors and groups. Our results pose important implications for efforts to increase online safety, in that educational initiatives need to prioritize self-reflection and self-awareness as much as content-based knowledge. Preliminary studies into online safety education programming show that initiatives focused only on enhancing knowledge might be missing the spot for long-term education [46]. In an evaluation of existing internet safety resources and programs, results have shown that experiencing a risky online situation is not about "lack of knowledge," but rather omission of the necessary skills, referred to as digital literacy, needed when navigating the internet [47]. Furthermore, existing literature in the European context, has demonstrated that there is a gap in the digital literacy of adolescents, especially concerning the ability to critically assess the credibility of information they encounter online [48]. This is particularly relevant to the current study, as the use of misinformation and the propagation of racialized myths-referred to in the literature as persuasive storytelling-is central to the internal communication and recruitment strategies of hate groups [24,31,39,49]. Past research has recommended that school-based education programs aimed at improving digital literacy and online safety should include the promotion of "soft competencies," including the development of knowledge, attitudes, and skills to critically and safely experience the online space [50]. Based on the results of the current study, such competencies may include learning how to limit one's own time online, self-awareness of disinhibited behavior, and avoiding risky situations, such as engaging with strangers. Interestingly, these requisite competencies for safe online activity may not be limited to adolescents. Previous research on a sample of parents of high school students indicated that on average, parents were only able to correctly answer half of the questions related to the ability to assess the credibility of online situations and have safe interactions with strangers [51]. Other studies focused on high school teachers show similar weaknesses in their ability to assess the credibility of information found online, which is part of an overall lack of digital literacy [52,53]. Informed adults are a critical component of improving the digital literacy of students [52]. As such, when attempting to improve the digital literacy and online safety of students, it is necessary to focus on not only the students' competencies but also that of the influential adults in their lives-their parents and teachers. Future research should explore the relationship between the digital literacy competency of students and influential adults in their lives, including their parents and teachers. When reflecting on the findings of this study, it is imperative to recognize its limitations. The primary limitation is its cross-sectional design. As such, the observed relationship between independent and dependent variables, albeit plausible, should not be assumed as causal. Additionally, due to the cross-sectional nature of these data, the direction of the observed relationship between exposure and outcome can only be speculated. As some of the exposures and outcomes could be perceived as negative, there is potential for social desirability bias to influence students' responses. As a non-random convenience sample, it is important to acknowledge that these results are not necessarily generalizable outside of the sample, and that selection bias may have also influenced the results. The reported study results are also limited in their generalizability due to the relatively small size of the sample from which they were obtained. Regarding social media use, we believe there are limitations to measuring the effect of individual platforms. As the use of individual social media platforms was correlated, we did not feel it was appropriate to include them for investigation in the regression models we generated. While our results indicated that Twitter and Houseparty were associated with exposure to hate online , and YouTube was associated with exposure to an individual trying to convince the respondent of racist views , we decided against including social media platforms as individual predictors in our models due to concerns of collinearity . Future research should more systematically investigate the influence of individual platforms on exposure to hate online. Finally, there may be limitations with the measurement of online disinhibition within our sample due to the lack of variance explained by the scale and the fact that toxic disinhibition could not be reliably measured. Similarly to previous research, our study of a population of high school students found an association between exposure to hate messages in the online space and time spent online, academic performance, communicating with a stranger on social media, and benign online disinhibition. In our sample, benign online disinhibition was also associated with students' risk of encountering someone online that tried to convince them of racist views. This study represents an important contribution to understanding youth's risk factors of exposure to hateful material online. These findings can lead to important preliminary recommendations for the development of educational activities aimed at improving online safety that should focus on teaching youth how to limit the time they spend online, helping them recognize the disinhibition effect of the internet and how passive online behaviors may also generate risk and influence decisions they make when online. Based on the existing literature, we recommend that future initiatives that focus on preventing youth exposure to online hate also incorporate a component for parents and teachers, as these adults play a critical role in how youths behave online. Finally, the results from this exploratory study can contribute to an actionable knowledge base of the determinants of exposure to hateful online material by providing the basis for future longitudinal investigation and hypothesis testing of the impact of online behaviors and novel psychological processes, including online disinhibition, on the frequency of exposure to hate messages and those who espouse them. --- Supplementary Materials: The following are available online at http://www.mdpi.com/1660-4601/17/22/8531/s1, The questionnaire used for data collection is included as a supplementary file. The data gathered during the course of the project are available upon request, due to research subjects' privacy considerations, via the national public repository of data Criminal Justice Data , hosted by the Inter-university Consortium for Political and Social Research , and can be found at the following link: https://www.icpsr.umich.edu/web/NACJD/studies/37338. Author Contributions: N.H. Performed the statistical analyses and drafted the introduction, methods, results, and tables. N.S. and T.C. drafted the discussion and conclusions. M.S. Supervised the data analysis and provided feedback for the results and methods sections. M.A.T. Developed the study design and supervised the data analysis E.S. Developed the study design and provided feedback on all sections of the manuscript. All authors have read and agreed to the published version of the manuscript. Funding: Research reported in this publication was supported by the National Institute of Justice under award number 2016-ZA-BX-K001 Evaluation of the Peer to Peer : Challenging Extremism Initiative and award number 2018-2A-CX-0002 Operation250: An Evaluation of a Primary Prevention Campaign focused on Online Safety and Risk Assessment. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Justice. ---

Today's youth have extensive access to the internet and frequently engage in social networking activities using various social media platforms and devices. This is a phenomenon that hate groups are exploiting when disseminating their propaganda. This study seeks to better understand youth exposure to hateful material in the online space by exploring predictors of such exposure including demographic characteristics (age, gender, and race), academic performance, online behaviors, online disinhibition, risk perception, and parents/guardians' supervision of online activities. We implemented a cross-sectional study design, using a paper questionnaire, in two high schools in Massachusetts (USA), focusing on students 14 to 19 years old. Logistic regression models were used to study the association between independent variables (demographics, online behaviors, risk perception, parental supervision) and exposure to hate online. Results revealed an association between exposure to hate messages in the online space and time spent online, academic performance, communicating with a stranger on social media, and benign online disinhibition. In our sample, benign online disinhibition was also associated with students' risk of encountering someone online that tried to convince them of racist views. This study represents an important contribution to understanding youth's risk factors of exposure to hateful material online.

THE PREVALENCE OF AUTISTIC TRAITS IN A HOMELESS POPULATION Autism is a neurodevelopmental condition characterised by difficulties with social relating, social communication, flexibility and sensory processing 1 . It occurs in approximately one percent of the population, and persists across the lifespan . Autism is a dimensional condition, representing one end of a continuum of traits that extends throughout the general population . A realistic understanding of autism should not only focus on the difficulties arising from the condition, but must also include consideration of the strengths of autistic people. For example, the majority of autistic people diagnosed according to current conventions have fluent language and an IQ in the normal range . Furthermore, many have capacities that stem directly from their autism . Despite their autism-related strengths, and the fact that many autistic people lead satisfying adult lives, under current systems of care people on the autism spectrum are at high risk of poor adult outcomes . These include social isolation, educational and occupational under-attainment, difficulty establishing independent living, poor quality of life and increased risk of an early death . In the current study we seek to investigate the link between autism and a different adult outcome, namely homelessness. This has received very limited attention in the empirical literature to date, and may well represent an important part of the picture of adult outcomes of autistic people. 1 In this paper, we use the term 'autism' as a direct synonym for the DSM-5 diagnostic entity of 'autism spectrum disorder' . We have chosen not to use the term ASD, as we do not accept the assumption it conveys, that autism is inherently a state of mental disorder. Homelessness is an umbrella term which covers a range of different situations. It refers to rough sleepers, that is people who sleep or bed down in the open air, or in buildings or other places not intended for human habitation . It also includes people who do sleep in a place designed for habitation, but who do not have any legal title to their accommodation or access to any private spaces for their social relations . Homelessness both arises from and contributes to vulnerability: it has severe negative effects on physical and mental health . If autistic people are more likely to become homeless, it is important to document this so that a potential subset of the homeless population can be identified and appropriate resources extended to them. This would also help with the development of targeted measures to help prevent autistic people becoming homeless in the first place. The current study aims to explore the relationship between autism and homelessness, by making an initial estimate of the prevalence of autistic traits in a homeless population. This work was initially motivated by anecdotal reports from autism clinicians and keyworkers in a homeless support service that rates of autism may be elevated in this population . In line with this, there is indirect empirical evidence to support the idea that autism is a risk factor for homelessness. Autistic adults, compared to those without autism, experience elevated rates of mental health problems, greater difficulties attaining independent living conditions, lower educational and occupational attainment, and a higher risk of social isolation . All of these characteristics are known risk factors for homelessness ). We know of no studies published in peer-reviewed journals testing directly whether autism predisposes people to homelessness. However, there are two studies in the 'grey literature' that support this idea. In one small-scale internal audit in a UK National Health Service setting, a psychiatrist investigated the presence of autism in a group of 14 homeless men with social difficulties . Seven members of this preselected group were judged to have shown strong signs of autism, based on a non-standardized but thorough assessment involving interviews with 12 of the 14 homeless individuals, interviews with workers, and also reviewing case notes. In another study, the National Autistic Society in Wales surveyed 415 autistic adults and family members of people with autism. Twelve percent of these autistic adults reported having been homeless at least once since leaving school . These findings are based on suboptimal methods of sampling and measurement and have not been subjected to peer review, and so must be treated cautiously. Nevertheless, they suggest the value of a more systematic investigation of the link between autism and homelessness. The task of assessing rates of autism in a homeless population is difficult. Diagnosing autism in adults is in itself challenging , and homelessness complicates assessment further. The ideal process of assessing autism in adults involves combining the results from standardised self-report, direct observation and informant report measures to gain a picture of current behaviour and developmental history . This intensive process requires a high degree of engagement from the person being assessed, and from someone who knew them as a child. Difficulties with engagement are ubiquitous in work with homeless people and gaining any kind of history can be extremely difficult in this group, as they have often lost contact with family and friends . Other factors such as high rates of substance misuse, mental health problems, and a disjointed social environment all also complicate the process of assessment . Furthermore, there are no autism measures that have been validated for use with homeless people . Reflecting the challenges of assessing autism in homeless people, our work has the following features. First, we directly acknowledge that we are unable to offer definitive diagnoses of autism in the current study. Instead we seek to derive an initial estimate of autistic traits, including those of sufficient quality and quantity to be suggestive of a DSM-5 diagnosis. Second, instead of using self-report and/or direct observation, we chose to measure autistic traits by informant report, with the informants being keyworkers in a homeless support service. These are staff members who work directly with homeless adults to help them make positive changes, and also coordinate their contact with diverse services. In this role keyworkers work with their homeless clients over a sustained period of time and generally know them well. Our decision to use informant report was in response to the likelihood that a large proportion of the homeless population we sampled would not engage with research. Those with autistic social communication difficulties would likely be among the least likely to participate, which would introduce a bias into any estimate of prevalence. A similar informant-report approach was adopted by Fraser and colleagues with another hard-to-engage population, when they estimated autism prevalence amongst patients in youth mental health services by interviewing their key clinicians. Whilst we acknowledge that this approach does not offer a gold-standard autism assessment, it does allow us to investigate the full caseload of a homeless service, thereby limiting sampling bias. A third key feature of this study is that, given the lack of relevant measures validated for homeless people, we collected data using an in-depth interview, structured according to the DSM-5 description of autism spectrum disorder. Any measure should possess reliability and validity . We assessed the inter-rater reliability by blind double-coding a random selection of interviews. Criterion validity is the extent to which a measure co-varies with another measure of the same construct. In this study, we administered the most appropriate extant autism measure for keyworker assessment of autistic traits, the Autism Spectrum Disorder in Adults Screening Questionnaire ; and checked how this correlated with outcomes from our semi-structured DSM-5 interview. Construct validity is the extent to which an instrument shows the pattern of association with other measures that would be predicted based on what we know about the construct being measured . We made three a priori construct validity hypotheses. First, since autism is associated with higher risk of social isolation , we predicted that if the interview possesses validity, those with high autism trait scores should show elevated levels of social isolation. Second, we predicted that high autism trait scores would be associated with lower levels of substance abuse in the homeless population. This was based on reports from homelessness experts we consulted that, compared to the non-autistic homeless population, their autistic homeless clients are less likely to abuse substances. In support of this is the meta-analytic finding that rates of drug dependence and alcohol dependence are very high in the general homeless population . By contrast, such problems are much less common amongst autistic adults, as shown by a recent whole-population study that found 3.4% of their autistic participants had substance-use related problems . Our third construct validity hypothesis concerned non-autistic psychopathology. Mental health problems are very common amongst homeless people, and such difficulties could artificially inflate scores on any measure of autistic traits. For example, if a person is socially withdrawn due to psychosis or social anxiety, this could mistakenly be taken as a symptom of autistic social impairment. Therefore, we reasoned that if our instrument has construct validity, it will not be strongly associated with non-autistic mental health difficulties in the homeless population. In summary, our overall aim is to derive an initial estimate of the prevalence of autistic traits in a homeless population using informant reports. To this end we sought to address the following questions: 1. Does the DSM-5 keyworker interview that we used to evaluate autistic traits show interrater reliability? 2. Does the DSM-5 interview show criterion validity, as indicated by agreement with another professional-report measure of autistic traits? 3. Does the DSM-5 keyworker interview show construct validity, as indicated by those with higher autistic trait scores, compared to those with lower autistic trait scores, being more socially isolated, less likely to abuse substances, and having equivalent levels of mental health difficulties? 4. What proportion of the caseload of a long-term homeless service have elevated levels of informant-reported autistic traits, consistent with a DSM-5 diagnosis of autism? --- METHODS --- Procedure The study was based in a homelessness outreach team in an urban area in the UK. In this team, each homeless person has a keyworker, a member of staff who coordinates their contact with services and works directly with them over a sustained period of time. At the start of this project, the research team provided an autism training workshop for the keyworkers to improve the quality of reporting; to reduce biases that could arise from variations in keyworkers' autism knowledge; and to engage the keyworkers in the project. This training workshop included a presentation and structured case discussion facilitated by the research team, and lasted two and a half hours. Subsequently, all keyworkers agreed to participate. This involved them completing a separate interview for each of their homeless clients. The only homeless clients who were not the focus of an interview were those born outside of the UK and the Republic of Ireland. We suspect that homelessness for those born outside the UK and Ireland, including refugees, is a distinct phenomenon, in terms of its causes and characteristics . This study received ethical approval from the University College London Research Ethics Committee, reference 8359/001. All keyworkers were provided with an information sheet and consent form. We followed procedures to protect the privacy and confidentiality of the homeless people who were the focus of the research interviews. The homeless people were not identifiable to the research team: we were not told names or any other identifying information such as date of birth. The joint first authors conducted the interviews. At the time of the research they were clinical psychology trainees working as psychologists in the UK National Health Service and studying for their doctorate in clinical psychology. This role involves extensive general training on assessment, and they also received specialist training in autism from the third and fourth authors , which included feedback on pilot interviews they had carried out. --- --- Measures --- DSM-5 based semi-structured interview We used keyworkers' knowledge of their clients to identify those homeless individuals with observable traits of autism. To do this we created a 'DSM-5 Autistic Traits in the Homeless Interview', which we call the 'DATHI'. This allowed us to gather in-depth information about the individual's presentation. The DATHI was developed through the following sequential process: 1) consultation with experts on homelessness, including those with experience of working with autistic homeless people; 2) going through the DSM-5 criteria in detail and creating a draft interview; 3) consultation on this draft with the local Adult Autism Special Interest Group, which comprises clinicians from several local adult autism assessment services; 4) Piloting the measure with keyworkers from the homeless outreach team. At each stage the interview was adapted based on feedback received. The DATHI, which is presented in the online supplementary materials for this article, was based on DSM-5 criteria for autism spectrum disorder. It has separate sections for each of the seven criteria, with general questions followed by specific prompts. For example, a question about eye contact was followed by prompts about whether eye contact was absent, or whether the individual had a fixed gaze. Some questions were adapted to the homelessness context, based on the information gained from experts in the development phase. An example of this was that one prompt in the section based on DSM-5 criterion B2 asks about ritualised behaviour in relation to sleep sites. The focus here was on whether there were especially fixed patterns of sleep site selection, or if the person set up their sleep site in a ritualistic fashion. The DATHI was scored by rating whether autism symptoms were present for each of the seven DSM-5 criteria. A range of scoring options were used to ensure that a criterion was only classified as 'Present' if there was good evidence that this was the case, as we expected that there would be a wide variety in presentations and it would be difficult in some cases to decide whether or not a particular behaviour was a characteristic of autism. The other potential scoring options were: 'Possibly present', 'Not present', 'Present but attributable to cause other than autism', and 'Insufficient information to classify'. Scores on individual criteria were combined to make an estimated overall classification for each homeless person. There were four possible summary outcomes: screened positive -high likelihood of DSM-5 autism; marginal -medium likelihood of DSM-5 autism; screened negative -low likelihood of DSM-5 autism; unclassified -insufficient information. The rules to assign each of these summary outcomes are shown in Table 1. [Table 1 here] Autism Spectrum Disorder in Adults Screening Questionnaire The ASDASQ is an informant-report autism screening measure, developed for mental health clinicians to rate autistic symptoms of their patients . This asks questions about the person's current presentation, with answers in a yes/no format. Potential scores range from 0-9, with higher scores indicating a higher probability of being autistic. Given the prevalence of mental health conditions among homeless people and the fact that it is designed to be completed by professionals, we considered that the ASDASQ was the most suitable instrument to use in the current study, in order to explore the criterion validity of the DATHI. --- Additional information gathered to test construct validity Information on mental health and substance use was gathered via a structured questionnaire completed by keyworkers drawing upon client's notes. A score of 1 was given for mental health diagnoses if clients had one of more formal diagnoses. Substance use was coded separately where a score of 1 was given for the use of alcohol or any illegal drug of any amount taken weekly to monthly. An additional semi-structured interview was used with keyworkers to gather observable information about their clients' social contacts. These qualitative data were then quantitatively coded using content analysis , with respect to four main categories of relationships . A code of 1 was given for each category if there was evidence of a current partner, any peer relationship including 'drinking buddies' or acquaintances known through drug taking and any contact with any family including by text or phone. If a person scored 0 on all 3 categories they were given a score of 1 in the totally isolated category. For this coding process, inter-rater reliability was calculated based on a second rater blind coding twenty interviews, which had originally been coded by MR. This showed high level of agreement across the categories; partner , peer relationships and family contact: . We also gathered information about whether any individuals had pre-existing diagnoses of either autism or intellectual disability. --- Data analysis --- Reliability checking and assigning final classification After classifications were made by the primary researchers the reliability of the DATHI was investigated. This was done by selecting all the 'screened positive' and 'marginal' cases and a random selection of cases that had 'screened negative' . The decision to over-select positive and marginal cases, rather than take a random selection from all cases screened, was made to provide a more rigorous test of the reliability of the measure. All written information collected in the assessment was shared with the raters who were blind to scores assigned in the DATHI, and to the final classification. Reliability was checked for each of the seven DSM-5 criteria and for the overall classification made, using Fleiss's kappa . By convention kappa values below .20 indicate limited reliability, .41 to .60 'moderate' agreement, .61 to .80 'substantial' agreement and .80 to 1 'very strong' agreement . The reliability raters are experienced in assessing autism in adults in both clinical practice and research. After reliability checking was complete, a consensus decision was made by the whole research team about classification for those cases where there was a disagreement between the original classification and that made by the reliability rater. --- Examining criterion validity In addition to generating an overall classification for each individual, classifications for individual DSM-5 criteria were converted into numerical scores. If an item on the DATHI screened positive it was given a score of 2; if it screened marginal it was given a score of 1; if it screened negative it was given a score of 0. These scores were then summed to provide an overall DATHI score, as well as subscale scores for DSM-5 Criterion A and Criterion B . Correlations between these DATHI scores and the ASDASQ were calculated to examine criterion validity. --- Examining construct validity The construct validity of the DATHI was examined by comparing those identified as having elevated autistic traits with those without elevated autistic traits on the following variables reported by keyworkers: substance misuse; mental health diagnoses; social connectedness. Group differences for these categorical outcomes were expressed as odds ratios with 95% confidence intervals. All analyses were conducted using IBM SPSS Statistics version 24. Fleiss's kappa was calculated using a plug-in for SPSS downloaded from the IBM developerWorks website . --- RESULTS Figure 1 shows the process of carrying out interviews and classification. It has details of total numbers screened, how many did not meet inclusion criteria, and when reliability checks were carried out and final classifications made. Interviews with keyworkers took 20-60 minutes per case, as more complex cases required more follow up questions on the DATHI. [Figure 1 here] --- Reliability of the DATHI Inter-rater reliability coefficients are shown for each DSM-5 criterion on the DATHI in Table 2. According to widely used guidelines for interpreting kappa , inter-rater reliability for criterion A1 is in the 'moderate' range, whilst for the other six DATHI items it is 'substantial'. We also looked at inter-rater agreement for overall classification, in terms of whether or not the DATHI identified an individual as screening positive. Fleiss's kappa was 0.69, 95% CI [0.37, 1.0], p < .001, indicating a substantial level of agreement between raters on this outcome . [Table 2 here] Criterion validity of the DATHI The overall score from the DATHI was significantly and substantially correlated with the ASDASQ . The ASDASQ was also significantly correlated with the DATHI scores for DSM-5 Criterion A and for Criterion B . --- Informant reported autistic traits in a homeless population Table 2 shows the proportion of homeless people reported to show difficulties consistent with each DSM-5 autism criterion. The full range of answer codes for DATHI items were used, but the 'Attributable to other causes' code was applied for only two homeless individuals where either use of alcohol or a psychotic episode led to a very brief and obvious shift in the individual's presentation. Table 3 gives examples of keyworker observations that led to individuals scoring positive for specific DATHI items. In some instances superficial details in this table have been altered to maintain the confidentiality of clients. [Table 3 here] Item-level DATHI scores were used to make overall classifications according to the a priori algorithm described in Table 1. After the final classification 13 of the 106 cases screened positive, showing sufficient keyworker-reported autistic symptoms to meet DSM-5 criteria. This equates to a prevalence in this population of 12.3%, 95% CI [7.0%, 20.4%]. Nine cases were identified as showing marginal evidence of DSM-5 autism, 72 as not showing any evidence of autism, and 12 as being insufficiently well known to services to be given a classification. Table 4 gives basic demographic details and length of homelessness for each classification. Supplementary Table S1 shows the profile of autistic symptoms on the DATHI for each individual who screened positive or marginal for autism traits. [Table 4 here] Characteristics of high and low autism traits scorers -construct validity of the DATHI As is shown in Table 5, in line with our construct validity predictions, homeless people who were identified by the DATHI as having elevated autistic traits were more socially isolated than low trait scorers. They were less likely to have a reported substance abuse problem. Autistic traits on the DATHI were not associated with having a nonautistic mental health diagnosis. With regards to intellectual disability , four people out of the total sample were known to have ID diagnoses. One of these screened positive on the DATHI, three screened negative. One individual was reported as having a pre-existing diagnosis of autism, and this person screened positive on the DATHI. [Table 5 here] DISCUSSION We sought to investigate a possible link between autism and homelessness, by gathering initial evidence as to the prevalence of autistic traits in homeless people. We developed an interview to be administered to keyworkers, based on DSM-5 diagnostic criteria, and used this to screen the entire caseload of a homelessness service in a large English city. There was evidence in support of the reliability and validity of the DSM-5 interview we developed. According to reports of keyworkers, 12.3% of homeless people had a range of autistic traits consistent with meeting DSM-5 diagnostic criteria. This is substantially higher than the general population autism prevalence of 1% . Rates of autism may therefore be raised in this homeless population, and further investigation is warranted to understand links between autism and homelessness. --- Reliability and validity of DATHI To our knowledge, there is no prior research in peer-reviewed journals on autism and homelessness. This likely, in part, reflects the considerable challenges of assessing autism in homeless adults. Many homeless people are reluctant to engage with professionals, reports from relatives are often impossible to attain, presentations are complicated by co-occurring difficulties , and no homeless-specific measurement instruments have been validated . To address some of these challenges we collected data using a keyworker interview, which we call the 'DSM-5 Autistic Traits in the Homeless Interview' . We chose to use an interview, rather than a questionnaire or direct observation tool, to allow for in-depth consideration of whether particular behaviours are indicative of autism. For example, the back-and-forth discussion between interviewer and interviewee can help both parties reach a decision about whether an individual's lack of eye contact is pervasive across situations, or only occurs when that person is under the influence of substances. Our adoption of a keyworker interview also allowed us to screen an entire caseload of homeless people in one service, thus minimising sampling bias and increasing the generalizability of our findings. This approach was made possible by the fact that in the service we based this study, it was the norm for keyworkers to have longstanding relationships with their homeless clients. Given that is was designed for the current study, a crucial question is whether the DATHI is reliable and valid. To investigate reliability, we measured agreement between blinded raters on a subsample of interviews. This was a stringent test of inter-rater reliability as we deliberately over-sampled 'marginal' cases. For all but one of the DATHI's items , inter-rater agreement was 'substantial', with the other item showing a 'moderate' level of agreement. Further, when we considered the instrument's ability to distinguish between those who screened positive for autism and those who did not, inter-rater agreement was 'substantial' . These findings suggest that the DATHI has adequate reliability. The case for the DATHI's content validity is supported by the fact that it was closely based on DSM-5 diagnostic criteria for autism spectrum disorder. Its face validity was checked by receiving comments on drafts from homelessness and autism experts, and by piloting the interview with keyworkers. We have included a copy of the DATHI in supplementary materials for this paper, and also present examples of behaviours coded in this study in Table 4, to allow readers to make up their own minds about content and face validity. Initial evidence for criterion validity comes from the high levels of association with another professional-report measure of autism traits, the ASDASQ . We acknowledge that this is not an especially strong test of criterion validity, as the ASDASQ is a screening measure, and was designed for psychiatric populations, not homeless people. However, as our aim was to begin to develop an evidence base in this area we considered that results from the ASDASQ would at least provide some information regarding the criterion validity of the DATHI. One challenge to the DATHI, and any assessment of autism in homeless people, is the high level of mental health difficulties and substance abuse problems in this population. The risk is that behaviours, such as social withdrawal or atypical non-verbal behaviour, could be mistakenly labelled as autistic in nature, when really they reflect a mental health problem or the effects of substance misuse. Therefore it is reassuring that in this study higher DATHI scores were not associated with higher rates of diagnosed mental health problems. Also there was an inverse relationship between reported substance abuse and autistic traits. These findings support the construct validity of the DATHI. Also, our prediction that homeless people with higher DATHI scores would be especially socially isolated was supported by the data, and this provides further evidence for the interview's construct validity . --- Autistic traits and homelessness people Given the above initial evidence that the DATHI is an adequate keyworker-report measure of autistic symptoms in the homeless, our findings that autistic difficulties are over represented amongst homeless adults should be taken seriously. We believe they raise the following possibilities that are worthy of future investigation. First, autism is likely a risk factor for becoming homeless. Our findings hint at one mechanism that could underpin this, since we found that those homeless with autistic traits were more socially isolated. Perhaps a lack of social capital makes people more vulnerable to becoming homeless in the face of other risk factors such as poverty and unemployment . Second, autistic homeless people may have a distinct profile of needs that impact on their daily life and chances of exiting homelessness. For example, sensory difficulties could make it hard for someone to live in a noisy hostel; and executive problems could make a transition to independent accommodation especially difficult. --- Limitations We have already discussed at length the challenges of assessing autism in homeless people and acknowledged that our findings are preliminary. Further validation of the DATHI will be valuable so that the instrument can be used in future investigations of autism and homelessness. This should involve testing the DATHI in the homeless population against the criterion of clinically diagnosed autism, based on a multi-disciplinary assessment . Such work will lay the ground for a more precise estimate of the true prevalence of autism amongst homeless people, and for studies that seek to identify the characteristics and needs of autistic homeless people. There was a sizeable group of homeless people in this study who were so poorly known to services that no data could be gathered about the presence of ASC symptoms. They received the classification 'insufficient information to classify'. These individuals in general actively avoided contact with keyworkers, and while this could have a wide variety of causes it seems plausible that this would be the type of behaviour an autistic person might display. This may mean that our estimate of prevalence is too low. With regards to the generalisability of our findings, we avoided sampling bias with respect to our target population, which was all the UK and Republic of Ireland-born clients of a specific English homelessness service. Nevertheless, this target population is not perfectly representative of the general homeless population, since they come from a service for the longterm homeless, who tend to have more complex presentations . Future work should investigate autism in more diverse homeless populations. --- Clinical implications and future directions This study has provided initial evidence that rates autistic traits are raised in homeless populations. While this cannot be more than a tentative conclusion, this would be consistent with the well-evidenced poor outcomes for adults with ASC . Lai and Baron-Cohen refer to a 'lost generation' of adults with ASC who did not receive a diagnosis because of lack of knowledge about the condition, and the individuals we have identified may be part of this group. If autistic difficulties are common among homeless populations this has important implications. Many people are homeless in the UK; the most recent estimate is that there are almost 5000 rough sleepers at any one point , and there is a much larger group of people with no stable accommodation who are termed the 'hidden homeless' . There may therefore be a considerable number of homeless autistic adults who are not having their needs met, and who are in an extremely vulnerable position. Some organisations have recently developed ways of supporting homeless autistic adults, and they have provided anecdotal evidence of success . These interventions have used expertise from the autism field to inform keyworking, and relatively straightforward adaptations have reportedly allowed the engagement of adults who had previously refused support. It will be valuable to manualise and empirically test such interventions, to begin to build an evidence base for supporting autistic homeless people. Also, it will be important to research pathways into homelessness for autistic people, to understand the mechanisms of risk. This can then be used to design preventative strategies to help autistic adults avoid homelessness. van Heijst, B. F., & Geurts, H. M. . Quality of life in autism across the lifespan: A metaanalysis. Autism, 19, 158-167. Client is so poorly known to services that any attempt to match their behaviour to criteria would be a guess . No eye contact, blank facial expression, has to be told when interactions are finished as would not pick up on this otherwise. Difficult to tell how they are feeling from their facial expression, does not use or pick up on body language. A3: Deficits in developing, maintaining, and understanding relationships Acts the same in different contexts. Difficulties in social interaction led to leaving accommodation. No friendships or interest displayed in making friends, acts the same to everyone they meet. No interest shown in making friends, small talk. Often rude and aggressive in interactions. B1: Stereotyped or repetitive motor movements, use of objects, or speech Plays with cuffs and rubs legs all the time, moves papers repeatedly in and out of envelope. Seems like there is a rhythm to these behaviours. Repeatedly moves arms in a very particular way. Speaks in an oldfashioned way, stereotyped oldfashioned way of saying goodbye. Described as talking like a character from a nineteenth century novel. Consistently uses unusual name for people in authority. B2: Insistence on sameness, inflexible adherence to routines, or ritualized patterns of verbal or nonverbal behavior Room precisely ordered with similar items placed in rows, but extremely dirty. Items organised in rows in room. Day follows precise routine of when they sleep, watch TV. Always take same route to shop and has rules about where shopping can be placed in flat. Possessions organised very precisely on shopping trolley, would take this on same route each day. When trolley was stolen was devastated. B3: Highly restricted, fixated interests that are abnormal in intensity or focus Talks a lot about food, very picky about foods, when shopping will stare at one product for a long time reading all ingredients. Everything they buy has a particular animal on it. Likes one colour a great deal, painted their room an intense shade of this colour. Makes lists of obscure musicians, has a large collection of broken electronics. --- B4: Hyper-or hyporeactivity to When fire alarm went off Sensitive to texture of clothing, Oversensitive to sound and light, sensory input or unusual interests in sensory aspects of the environment unexpectedly seemed like they would scream, always has curtains closed, burnt hand badly but seemed under-reactive to pain . checks before purchasing anything. Attracted by flashing light. Has TV on very loud, while in hospital would have TV screen very close to face. does not like TV being left on when is in office with keyworker, refused own TV. Complains about noises others cannot hear. --- SUPPLEMENTARY MATERIALS AND ACCESSING RESEARCH MATERIALS The main interview used in this research has been uploaded in supplementary materials. Other questionnaires used and data relating to the study can be accessed by contacting Alasdair Churchard.

Anecdotal evidence suggests that autistic people experience an elevated risk of homelessness, but systematic empirical research on this topic is lacking. As a step towards filling this gap in knowledge, we conducted a preliminary investigation of the prevalence of DSM-5 autism symptoms in a group of long-term homeless people. The entire caseload (N=106) of a UK homeless outreach team was screened (excluding individuals born outside of the UK or Republic of Ireland) using an in-depth, semi-structured interview with keyworkers, based on DSM-5 diagnostic criteria. This showed adequate inter-rater reliability, as well as evidence of criterion and construct validity. Of the sample, 13 people (12.3%, 95% CI [7.0, 20.4]) screened positive, meeting DSM-5 autism criteria by keyworker report. A further nine people (8.5%, 95% CI [4.5, 15.3]) were 'marginal', having autistic traits that were not quite sufficient to meet DSM-5 criteria. Those with elevated autistic traits, compared to those without, tended to be more socially isolated, and less likely to use substances. This study has provided initial evidence that autistic traits are over-represented among homeless people; and that autistic homeless people may show a distinct pattern of characteristics and needs. Further investigation is required to build upon these provisional findings.

Tis article has been retracted by Hindawi following an investigation undertaken by the publisher [1]. Tis investigation has uncovered evidence of one or more of the following indicators of systematic manipulation of the publication process: Discrepancies in scope Discrepancies in the description of the research reported Discrepancies between the availability of data and the research described Inappropriate citations Incoherent, meaningless and/or irrelevant content included in the article Peer-review manipulation Te presence of these indicators undermines our confdence in the integrity of the article's content and we cannot, therefore, vouch for its reliability. Please note that this notice is intended solely to alert readers that the content of this article is unreliable. We have not investigated whether authors were aware of or involved in the systematic manipulation of the publication process. In addition, our investigation has also shown that one or more of the following human-subject reporting requirements has not been met in this article: ethical approval by an Institutional Review Board committee or equivalent, patient/participant consent to participate, and/or agreement to publish patient/participant details . Wiley and Hindawi regrets that the usual quality checks did not identify these issues before publication and have since put additional measures in place to safeguard research integrity. We wish to credit our own Research Integrity and Research Publishing teams and anonymous and named external researchers and research integrity experts for contributing to this investigation. Te corresponding author, as the representative of all authors, has been given the opportunity to register their agreement or disagreement to this retraction. We have kept a record of any response received. --- Introduction In recent years, the national policy on family planning has been continuously relaxed. It is an interesting domain which got importance due to mass increase in the population of the world and family planning is one of the most important methodologies for controlling it. It is defined as "the exercise to develop a proper management or control policy to ensure how the number of children has to reduce, preferably one, and the intervals between their births , especially through the contraception or voluntary sterilization." Although the current laws in some areas stipulate that social maintenance fees shall still be levied for those who have more than two children, there have been significant changes compared with the previous policies of "double only two children" and "single two children". After the implementation of the "comprehensive two child" policy ①, the increase of China's population fertility did not meet expectations [1]. Other studies ④ show that there is a close relationship between fertility and the allocation pattern of educational resources, and infant education plays a very important role in educational resources. ,e lack of educational resources and differential allocation will inevitably bring competition for educational resources and also bring great pressure to children's parents. Under this condition, people put more energy and financial resources into limited educational resources by reducing fertility, in order to get better educational resources and less pressure. "No one takes children" has also become one of the important factors affecting the fertility rate due to the lagging development of educational resources. According to the survey ⑤ of fertility willingness released by the National Health and Family Planning Commission in 2015, 74.5%, 61.1%, and 60.5% of the factors are unwilling to have a second child because of economic burden, too much energy, and no one to look after them, respectively. A survey conducted by the NLD shows that 68.1% of mothers who have given birth to one child but do not intend to have a second child are because their children are unattended ⑥. In 2016, the China Women's Federation and Beijing Normal University issued a survey report on the impact of the implementation of comprehensive two child education on family education. ,e report showed that 53.3% of one child families had no intention of having two children. ,e survey found that the four public service resources of education, medical treatment, health, and living environment were the most important factors affecting the decision-making of two children. With the development of society, infant care has gradually become a more significant social problem. Infant care is not only the responsibility of the family, but also needs the support of the government and the supply of public services [2]. ,erefore, in recent years, infant public care services have attracted more and more attention. ,e employment rate of women in cities is higher than before, and under the influence of the new social culture, they pursue a higher quality of life. At the same time, they also bear greater life pressure, and the energy and time of childcare are obviously affected ①. At the same time, these women are also more expected to return to the workplace. For ordinary families, the average marriage age has gradually increased in recent decades, resulting in grandparents' increasing age when looking after infants and young children, a relative decline in their ability to take care of children, and a relative increase in the pressure on grandparents to take care of young children [3]. ,ese reasons may lead to a decrease in grandparents' willingness to take care of young children ②. Children's early education and care is the basis for learning cognition and various skills. Children's later development largely depends on the quality of children's care and education ③. ,erefore, parents have higher and higher requirements for the quality of their children's care. ,ey hope that their children can learn more scientific knowledge and skills and develop their abilities appropriately while being taken care of. ,is is obviously not available in traditional care, and public childcare services have these advantages and can meet the needs of parents [4]. In October 2017, in the report to the 19th National Congress of the Communist Party of China, it was mentioned to implement "education for children, learning and education," and put "education for children" in the first place in the progress of people's livelihood. However, at present, China's infant care services are still seriously insufficient [5]. According to the data released by the National Health and Family Planning Commission, the enrollment rate of 0-3-year-old infants in various kindergartens in China is only 4%, and the enrollment rate of urban infants under the age of 3 is less than 10%. Recent surveys in Changsha, Yongzhou and other places also found that 90% of parents believe it is necessary to let children participate in childcare and early education, but nearly half of them did not participate because of economic and social reasons ⑤. ,erefore, facing the huge gap, the establishment of a complete public childcare service system is imminent. As for the demand for childcare services, a large number of scholars have studied the macro environment: first, women's market participation has increased. Since the marketization, women have gradually been liberated from the family and participated in the labor market. As a result, women are facing the conflict between childcare and job choice, which also increases the pressure on them to take care of children and forces them to find other ways of infant care. Second, population aging and care deficit. Facing the problem of population aging, the state has issued policies to delay the retirement age [6]. At the same time, with the improvement of grandparents' requirements for the quality of life of the elderly, they prefer to enjoy the life after retirement rather than help take care of their grandchildren, which makes there less infant care resources from grandparents. ,ird, population mobility and miniaturization of family structure. With the reform and development of the economic system, the scale of population flow is unprecedented, and the family tends to be smaller. ,e parents need to go out to work, and the grandparents and core families live in different places, which increases the childcare cost within the family, leads to the lack of "helpers" and the obvious weakening of the family's care function [7]. With the strong support of government policies and the strong demand of the people, the establishment of the childcare service system should receive more attention and research, however, this research filed is still in its infancy and careful attention is required from the academia and research organization. ,erefore, it is necessary to explore the influencing factors of the people's demand for childcare services. To address the issue mentioned above, we have developed and reported a big-data-enabled system for the accessibility of urban non-profit public nursery services especially those which are required to be ensured by the concerned government. Additionally, we have devised a methodology that a government will ensure that these services and measures are needed to be taken on timely basis. ,e rest of the paper is organized as given below. In the subsequent section, an extensive review of the available techniques, preferably those which are closely linked to the problem at hand, is presented along with various problems in the existing approaches are highlighted. In the next section, the proposed methodology is well explained in various sections and subsections along with how these issues could possibly be resolved. Results and discussion are provided in a separate section. Finally, concluding remarks are provided in the last section of the paper. --- R E T R A C T E D --- Related Work In recent years, with the relaxation of the national fertility policy, more and more families have entered the ranks of "two child families". With the emergence of various parenting problems, the Chinese government is also actively introducing various parenting policies. Experts and scholars in many fields such as education, sociology, and economy are also actively exploring, and the "public childcare service" has gradually become a hot spot in the academic community. Different scholars have different views on the definition of childcare service. ③ Childcare is an alternative service. When families cannot take care of infants and young children for some reason, childcare services can take care of infants and young children instead of families. At the same time, childcare services can significantly reduce the rearing pressure of some families, liberate family productivity, and promote social development. [4] It is considered that nurseries are services to assist families in providing care, health care, and infant development experience. [5] Childcare is a variety of day care and arrangements chosen by the parents for their children. ,e purpose is to enable infants and young children to be cared for, protected, and fully educated. [6] ,e right to care is an important part of civil rights. ,e essence of childcare is the guarantee of the right to care. As for the function of childcare services, Peng Shuhua believes that childcare services can not only promote women's employment, but also enhance family stability. ② infants aged 0-3 are in the early stage of life cycle. ,eir life experience and experience during this period have an important impact on personality and intelligence. Good public care services are conducive to the development of infants and young children and the formation of good habits. ③ the "elderly floating" group has difficulties and pressure in infant care. A perfect social infant care service system is conducive to liberate the elderly who care for their grandchildren and reduce the burden of the elderly group. ④ the survey found that in the case of a shortage of public childcare services, urban women who have two children and take care of children aged 0-3 have a higher proportion of family impact on work, a larger gender gap, and some women are forced to interrupt their work. ,e perfect public childcare service system is conducive to liberating women's productivity, narrowing the gender gap, and reducing the impact of work on families. For the discussion of the mode of nursing service, ⑤ the ways and methods of providing services are different for the nursing institutions established by different subjects. According to the different supply subjects and scale, nurseries are mainly divided into family nurseries, enterprise nurseries, community nurseries, institutional nurseries, and PPP nurseries. ⑥ taking the new public service theory as the framework, this paper puts forward seven elements: concept, mode, project, specification, mechanism, effect, and policy to identify and classify the practice mode of infant care service in China. Infant care services originated in western countries. In the 1950s and 1960s, the birth rate in western countries increased significantly after the war, which led to the increasing inability of traditional families to take care of infants. At this time, a large number of socialized childcare institutions appeared, and childcare services came into being. At present, the research literature on international childcare services mainly includes the introduction of childcare service model, development experience, and evaluation of treatment in childcare institutions in foreign developed countries, but there are few related in-depth discussions [8]. Some social democratic countries in northern Europe pursue the value orientation of inclusive development, as well as the principle of highly non-commodity delimitation and universalism. In these countries, the government provides infant care services, advocates the socialization of the family cost of raising, creates a social environment suitable for taking care of children, and the government provides various needs in the process of infant growth. ,is welfare policy can not only ensure the family's ability to raise infants and young children, but also reduce the burden of family care for infants and young children through free choice ①. In countries like Norway and Sweden, each and every woman is allowed and encouraged, i.e., through various means, either to take part in the working environment where men are supposed to work. As we know that various hurdles have been reported but participation of the women in the working environment especially under the concerned policy is very encouraging. Additionally, these type of services removes the barriers for women which are in different societies throughout the world. Apart from this, these countries have specifically contributed to the education sector where it is ensured that education provided by these countries is generally integrated education, that is, the combination of trusteeship and early education. Some other countries in Europe pursue the value orientation of filling vacancies, survival, and re-family. ,ese countries attach great importance to the maintenance of traditional family relations and regard taking care of children as the responsibility of mothers. ,e government will take care of children only when the family is unable to take care of infants and young children. ,e service models of these countries emphasize the family's ability to support. In this case, the development of public childcare services is greatly limited. It is difficult for general infants to obtain childcare opportunities, and women's rights and interests have also been hurt to some extent. ,is category includes Germany, Italy, and other countries. In such countries, the way of childcare is generally phased service. For example, in Germany, infants aged 0-2 only receive childcare services, and infants aged over 3 receive early education at the same time. In addition, some developed countries pursue the value orientation of filling vacancies and survival, but from the perspective of family, they pursue the concept of de-familialization. ,e type of social welfare in these countries is generally liberal welfare. ,ese countries emphasize the regulatory role of market mechanism. ,e state only provides necessary help to poor families and only guarantees the minimum level of welfare. Few infants and young children --- R E T R A C T E D finally receive the childcare services provided by the state ③. Like the Nordic countries, these countries generally advocate the principle of educational integration, and children can receive early education while receiving custody services [9]. Foreign childcare services developed earlier, and both the development of policy system and the large-scale development of childcare institutions are relatively perfect. Compared with domestic, foreign public childcare services started earlier. ,e American Association of Preschool Education issued a certification standard for high-quality childcare institutions in 1984. Its core idea-the development of suitability education has had a wide impact all over the world. In view of the differences in the development of childcare services at home and abroad, Qiu Baili ④ through the comparison between the kindergarten evaluation standards formulated in Zhejiang, Jiangsu, and Shanghai and the certification standards of high-quality childcare institutions formulated by the American Preschool Education Association believes that there is still a certain gap between China and foreign advanced educational ideas [10]. Reference [11] believes that foreign values are more advanced, the system is more perfect than that is in China, there are laws with high feasibility to provide protection, and a variety of allowances to provide meticulous care for infants and young children and economic support for parents. Humanized welfare policies and all-round childcare services can meet the different needs of all kinds of children and families to the greatest extent. Reference [12] through the comparison with EYFS nursery institutions in the UK in terms of examination and registration, teacher-child ratio, qualification, and training, it is considered that there is still room for improvement in the supervision and implementation, teacherchild ratio, employee education, and training mechanism in China's preschool education. Generally speaking, the foreign public childcare service system is not perfect, and there are also problems such as the distribution of rights and responsibilities, financial investment, the qualification of institutions, and educational security personnel [13]. However, many of its systems and concepts are more advanced and worthy of our study. --- Childcare Services Childcare service is a part of children's welfare. It is transformed from the word "childcare." At first, it means "entrusted parenting," that is, to entrust infants and young children to people or institutions other than family caregivers for parenting and care. With the changes of the times, people's demand for childcare services is not only satisfied with the initial upbringing and care, but also organically combines "education" and "maintenance," ,e United Nations defines childcare services as the care and education received by children outside the family from groups or institutions. In OECD countries [14], however, it is important to note that childcare is not only concerned with the outside environment as various authors have perceived it differently such as childcare services start from the home, this is actually the thought of certain group of authors. Additionally, a certain class of authors believe that childcare service starts exactly when a child has to interact outside his/ her home. It is considered that all services including formal and informal care and education for preschool children are childcare services. Public childcare service is a kind of childcare service, which refers to the childcare service provided or participated by the government or public forces. In China, the service content mainly includes early education and nursing. For the definition of infant age, infants aged 0-3 are generally regarded as the research object of public childcare services in China. In fact, in more developed big cities, the strict sense of nursery institutions are mainly kindergartens, community childcare centers built by the government, and internal custody centers established by some enterprises. Some enterprises participate in the provision of nursery services in the form of internal employee welfare. However, the market-oriented force that really takes the supply of childcare services as the core operation and management business is still in the blank stage, and the real market-oriented supply of childcare services is relatively lacking ①. It is to be noted that generally the public childcare service is how education is provided to these children. However, in this paper, childcare service has been referred to as environment where education along with other necessary services are provided especially by the concerned government. Additionally, it is to be noted that it is the responsibility of the concerned government departments to ensure supply of subjects directly or to the childcare institutions directly or indirectly. However, this entire process is desperately needed to be managed by the government for 0-3-year-old infants, including daily life care, nutrition matching, habit formation, and early education, in order to achieve their development, learning, and growth goals. Finally, if some of these services are missed or not provided on a timely basis then it may affect the life of the infants. --- Data Description and Model Setting Generally, we know that for providing a big-data-enabled model, data is needed. As we have described below, a dedicated survey was conducted to collect the desired data needed for the verification of the proposed study and to be in a better position to enable or at least develop a system where childcare services are ensured by the concerned government. --- 4. 1. Data Sources. ,e data used in this article comes from the data obtained from the extensive survey conducted by the research group of Jiangsu Institute of social security from January 2019 to March 2019. ,e research group took married young people aged 20 ∼ 40 as the survey object, and collected the basic information of the respondents and their families, the survey of the respondents' fertility willingness and fertility, and the survey of infant rearing and childcare services to study the residents' infant rearing and childcare services. A total of 683 valid questionnaires were collected, involving 93 cities in 17 provinces such as Jiangsu, Anhui, and Guizhou. As the research objects are concentrated in the respondents and their families with infants aged 0 ∼ 3 years, --- R E T R A C T E D the survey data corresponding to the families with the youngest child aged 3 years and below were selected. At the same time, after deleting the missing data and abnormal data, a total of 190 valid sample data were obtained [15]. --- Variable Description. In this study, proper description of the various variables which are being utilized such as explanatory variables. ,ese variables were used to explain whether there is a demand for childcare services in a particular region; likewise, the explanatory variables are used to represent various possible factors that affect people's demand for childcare services. Various effects of these factors are family income, the current care mode of children, the age difference of children in multi-child families, whether there is an only child in the parents' generation, whether they live with their grandparents now, the working hours of parents, the distance between childcare institutions and home. ,e characteristic variables of nursing institutions, such as the charging level of nursing institutions and the charging level of institutions they can bear. --- Model Design. In order to calculate the regression model, it is assumed that: f ty � α + 􏽘 n i�1 β i x i + e i . ( )1 Where, if ty is the basis of decision-making, if ty � 1, representing the need for childcare services; if ty � 0, which means no nursery service is required. x i is the variable for each sample, β i is the parameter corresponding to each variable, α is the intercept term, e i is the random error term obeying the standard normal distribution. x i impact on the demand for childcare services is determined by formula : P if ty � 1|x i 􏼁 � P if ty � α + β i x i + e i > 0 􏼁 � P e i ≤ α + β i x i 􏼁 � F e i � α + β i x i 􏼁. Logistic distribution: P if ty � 1|x i 􏼁 � 1 1 + exp -α+β i x i 􏼒 􏼓 When α + β i x i tends to infinity, P is infinitely close to 1, whereas, when α + β i x i is infinitely small, P is close to 0. --- Analysis of Empirical Results Analysis is one of the methods which is available in literature and are used to ensure applicability and feasibility of a newly developed system in the expected environment where the proposed system is supposed to be deployed. --- Empirical Analysis. Logistic regression was used to analyze the impact of different factors on whether people choose childcare services. As shown in Table 1, the impact of parents being the only child on the choice of childcare services shows that the impact of respondents being the only child on the choice of childcare services is negative, and the probability ratio of respondents choosing childcare services for the only child is 0.486 times that of the respondents not being the only child, ,e probability ratio of respondents choosing childcare services for their only child is 0.456 times that of respondents whose spouse is not the only child, and the probability ratio of families with both husband and wife being the only child is 0.358 times that of families with both husband and wife not being the only child. It fully shows that families with only one child are less likely to choose childcare services. At the same time, when controlling the variable that the respondents are the only child, the regression results show that the longer the respondents work at home, the less inclined they are to choose childcare services. ,e probability ratio of choosing childcare services relative to not choosing childcare services decreases by 2% for each additional hour of family working time. Generally, in our society, we have observed that for those couples, whose working schedule is hard and comparatively longer than usual, it is high likely that they will surely require services, i.e., help in most of the case, to ensure proper caring of their infants. ,is could possibly be ensured through either a servant or if there are other family members like parents of the husband or wife or both, however, this is very rare in different societies and thus other alternatives are needed to be provided. In the situation that the concept of childcare has not been formed and the development of the childcare market is not perfect, the first choice is the help of grandparents, so that people will not choose to buy childcare services first. However, this does not mean that with grandparents to help take care of children, there is no need for childcare services. ,e probability of families with parents and grandparents taking care of their children choosing childcare services is 1.279 times higher than that with only parents. Table 2 shows that when controlling the variables such as the age of respondents, the age of the first child in the family, whether the parents are the only child, and whether the grandparents live together, it is found that the greater the age difference between the second youngest child and the youngest child, the less likely they will choose childcare services. ,e data show that the age difference between the second youngest child and the youngest child has an impact on whether the family chooses childcare services at a significant level of 5%. When the age difference increases by one year, the probability of families choosing childcare services decreases by 251% compared with not choosing childcare. For families with more than two children, the greater the age difference between the second child and the youngest child, the less pressure the family has to take care of. When the age of the respondents increases by one year, the probability of requiring childcare services increases by 95% compared with not requiring childcare services. Table 3 shows the impact of the situation of childcare institutions on people's demand for childcare services. ,e data in the table shows that people who care about the distance between the nursery institution and their home need more nursery services. People who care about the --- R E T R A C T E D facilities, environment, and food safety of care institutions do not need care services compared with those who do not care much. People who care about the fees charged by nursery institutions have more demand for nursery services. People who have a higher bearing on the charge level of childcare services are more likely to choose childcare services. --- Result Analysis . ,e discussion from the micro-personal aspect mainly focuses on the following aspects: on the one hand, the requirements for the improvement of education quality are in conflict with the educational ideas of grandparents. After the introduction of the two-child policy, the grandparents are old, and their enthusiasm for the care of their grandchildren is not high, and people's professional requirements for the level of infant care are gradually improving. Even if grandparents are willing to take care of their children, in the face of the conflict of parenting concepts of grandparents when taking care of their grandchildren, more and more parents refuse the care of their grandparents. On the other hand, the parents work intensively and lack professional parenting knowledge. Parents face problems such as high work intensity and insufficient parenting experience and are unable to give scientific, full, and all-time care to infants and young children. ,rough the above analysis, it is not difficult to draw a conclusion: people's demand for childcare services exists in reality, and the demand is large. According to the data released by the National Health Commission, the current shortage of childcare services in China is very serious. ,e enrollment rate of 0-3-year-old infants in various childcare institutions in China is only 4%, far lower than the proportion of 50% in some developed countries ①. According to Wang Hui's research results, the enrollment rate of children under the age of 3 in the Chinese cities is less than 10%, and there are differences among provinces. ,ere is a great difference between the expectation of local people and the actual proportion of local children. Among them, 90% of parents in Beijing believe that "childcare is necessary for all development," and the proportion in Shanghai is as high as 95.3%. Liu Zhongyi's research results also show that at present, the parents of infants under the age of 3 in China have a strong demand for institutional care services. ,e demand for 0-3-year-old infant care services in urban areas is more prominent. It is important to note that, packet loss ratio is defined as the ratio of the number of packets which are lost during the transmission to the total number of packets generated in the system, networks in this case. When the cycle length is 104 network packets, the false positive test results of malicious packet loss are shown in Figure 1. In the malicious packet loss false positive test, the horizontal axis of the experimental results in Figure 1 is the detection threshold. When the detection threshold of the system designed in this paper is about 0.25%, the occurrence probability of false positive approaches 0. ,e detection threshold of the original system needs to reach about 0.37%, and the occurrence probability approaches 0. Under the same experimental environment, the false negative test results of malicious packet loss are shown in Figure 1. In the false negative test, the random packet loss rate of abnormal nodes varies from 0.2% to 0.8%, and the cycle length is 105 network packets. As can be seen from Figure 2, while the packet loss rate of the two systems increases, the false negative rate decreases. When the packet loss rate of the original system is greater than 0.8%, the false negative probability is less than 2.0%. When the packet loss rate of the system in this paper is greater than 0.52%, the false negative probability is less than 2.0%. --- R E T R A C T E D ,e false positive test results of malicious packet modification of the two systems are shown in Figure 3. In the test process, in the absence of malicious nodes, the probability of abnormal self-contracting caused by the network itself is 0.2%, and the cycle length is 104. When the sampling rate of the two systems increases, the probability of false positive decreases. In this system, when the sampling rate is 12.3%, the probability of false positive is 5%, and when the sampling rate in the traditional system is 50%, the probability of false positives decreased to 5%. It shows that the system designed in this paper can achieve a high safety level when the sampling rate is small. ,e test results of malicious packet modification of the two systems are shown in Figure 4. Finally, the false negative probability of malicious packet modification in the presence of abnormal nodes is tested. ,e threshold setting of the two systems is 0.4%. As can be seen from Figure 4, when the packet change rate is 0.4%, the probability of false negative is 1%, and gradually approaches 0 with the increase of packet change rate. When the packet change rate of the original system is 0.5%, the false negative rate gradually approaches 16%, on the other hand, it also verifies that the system in this paper has high accuracy in the intelligent location of network abnormal nodes. --- Conclusion In this paper, we have presented big-data-enabled methodology to ensure that proper and timely public childcare services are provided to every individual especially children both in the cities and the urban areas. ,ese services are defined as the service required to be provided to a child when he/she interact with outside environment such as education. Moreover, these services must be ensured by the government especially through a dedication department. ,e statistical results show that people's demand for childcare services does exist. What also exists is the contradiction between people's demand for childcare and the insufficient construction of the number of childcare institutions and the single supply of the provider. ,is contradiction is the problem we urgently need to solve. If both husband and wife are the only children, compared with families with non-only children, the demand for childcare is lower, and the occurrence ratio of choosing childcare is 35.8%; the longer the family's working hours, the smaller the demand for childcare services. ,e above data show that the work and health status of grandparents in the family and the parenting concept of the two generations have an impact on the demand for childcare services. In addition, in families with more than two children, the greater the age difference between the second child and the youngest child, the less the demand for childcare services. For each additional year of age, the probability ratio of requiring childcare services to not requiring childcare services increases by 95%; the higher the respondents' psychological acceptance of childcare services, the greater the demand for childcare services; people who care about the facilities, environment, and food safety of nurseries have less demand for nurseries. [16][17][18]. --- R E T R A C T E D and Xiangeng Zhang. All authors participated in the review of the paper. --- Data Availability ,e datasets used and analyzed during the current study are available from the corresponding author upon reasonable request. --- Conflicts of Interest ,e authors declare that they have no conflicts of interest. --- Authors' Contributions ,e conception of the paper was completed by Yi Li, and the data processing was completed by Xiaoli Deng, Shan Liao, Xiaoli Liang, Hongyan Wang, Hua Fang, Zhen AI, Pei Chen,

In today's society, the pressure of family care is increasing and people's demand for childcare services is increasing. At the same time, the government also pays more attention to the problem of 0 ∼ 3-year-old infant care services which are desperately needed in every society throughout the world. General studies have been carried out in literature where results show that in families with only one child, long working hours and with more than two children, the three factors of children's age gap, attention to nursery facilities, attention to environment and food safety have a negative impact on people's demand for nursery services. From these studies, we have extracted three vital factors that have a positive impact on people's demand for childcare services: grandparents' help to take care of children, parents' age, and psychological tolerance of childcare service fees. In this paper, we have utilized one of the most commonly used methodology, i.e., big data, to smoothly resolve the accessibility issue of urban non-profit public nursery services. In order to verify and evaluate our claim, we have implemented the proposed scheme and compared the results which shows that exceptional performance of the propose scheme.

Background Fertility levels and trends in the world generally and in a country particularly vary across time. The casual factors are diverse, but they may be broadly grouped into direct or proximate determinants and indirect or background factors [1]. The proximate determinants are bio-behavioural mechanisms that act to reduce fertility and are popularly labelled as intermediate variables. Bongaarts defined eight proximate variables impacting the fertility of a population: the proportion of women of reproductive age married in the exposure group; contraception and induced abortion fall under the second group; the third group included lactational infecundity, frequency of intercourse, sterility, spontaneous intrauterine mortality, and duration of the fertile period [2]. And, from these eight, Bongaarts identified four variables as the most important proximate variables because of their higher and more direct inhibitory effect on fertility, as together they account for nearly 96% of the variation of fertility in a population [3]. Further, Bongaarts' model has quantified the contribution of these four factors to the observed fertility level. This resulted in a simple but powerful model for analyzing how fertility changes overtime or differs from one group to another; therefore, it is the most extensively used models in the world to measure the extent to factors reduce fertility. The proximate determinants model is mostly used to identify the main factors behind fertility changes, the changes associated with fertility through time, and to compare changes in fertility among countries or regions of a country [4]. The National Population Policy of Ethiopia was adopted in 1993 with an overall goal of harmonizing the rate of population growth and economic development primarily addressing the high fertility. The policy in particular, aims to achieve a total fertility rate of 4.0 children per woman in 2015. The policy envisages meeting this target by employing different strategies such as expanding clinical and community-based contraceptive distribution services; promoting breastfeeding as a means of increasing the time span between pregnancies; and raising the minimum age at marriage for girls from the current lower age limit of 15 to, at least, 18 years [5]. In addition, the Health Policy emphasizes the need to improve the coverage and quality of family planning services in the country [6] and the Women Policy also acknowledges the need to ensure women's access to family planning and other reproductive health services as one of the strategies to empower women [7]. The National Reproductive Health Strategy stipulates as one of its goals to reduce unwanted pregnancies and to enable individuals to achieve their desired family size [8]. The Health Extension Program involves the delivery of family planning information and services to the community through home visits [9]. The Family Law, revised in 2000, put 18 years of age as the legal marriage age for both men and women [10]. With a population of nearly 98.1 million in 2015, Ethiopia is the second most populous country in Africa next to Nigeria and is projected to be 165.1 million in 2050 [11]. The population grows at a rate of 2.6% per annum [12]. The vast majority of the population in the country lives in rural areas and is heavily dependent on rain-fed agriculture. Women of reproductive age constitute one-quarter of the total population of the country and about 48% of the female population [12]. The main feature of fertility in Ethiopia is that it was at its highest levels at the end of the twentieth century, and that the rural-urban differentials are the highest in Africa. The total fertility rate of the country was declined from 6.6 children per woman in 1990 to 5.5 children per woman in 2000 and 5.4 children per woman in 2005 [13][14][15]. This shows that in the fifteen-year period, since 1990, the total fertility rate declined by only one child per woman, and stagnated between 2000 and 2005. In particular, there was a slight difference between the 2005 and 2011 total fertility rate [15,16]. Fertility declined in all regions of the country except Afar, Somali, and Benishangul-Gumuz National Regional States during the last decade. It remained constant in Afar National Regional State and Benishangul-Gumuz National Regional State between 2000 and 2011. However, in Somali National Regional State the total fertility rate increased from 5.7 children per woman in 2000 to 7.1 children per woman in 2011 [14][15][16]. In Amhara National Regional State, however, had the greatest decline in the total fertility rate during the last decade as compared to Oromiya National Regional State , Southern Nation Nationalities and Peoples National Regional State , and Tigray National Regional State [14][15][16]. The onset of fertility decline in Amhara National Regional State has engendered much interest since it has been suggested that the fertility transition in the region is different in certain important aspects from that experienced in the past elsewhere in the country. The available studies on proximate determinants of fertility in Ethiopia were carried out at a national level of aggregation [17,18]; the results might have been affected by problems of aggregation bias. Since the contribution of components of proximate determinants of fertility vary from area to area, macro-level analysis rarely take into consideration of region-specific factors. In addition, these studies provide only estimation of predicted fertility levels and do not attempt to investigate socioeconomic differentials and trends in the fertility change overtime. Moreover, they do not explain the decomposition of fertility change into components of proximal determinants. As these researche gaps, there are two dimensions in applying the proximate determinants model in this study. First, keeping the regional difference and heterogeneous society in mind, the region-specific proximate determinants are the focused of this study because of the lack of a comprehensive analysis of trends in proximate determinants of fertility at the regional level in the country. Secondly, previous research lacks decomposition analysis; the present study emphasizes analysis of the trends in the fertility change between the two time points to determine contributions of the four proximal determinants. Therefore, the aim of this study was to determine the relative contribution of each of the proximal determinants in inhibiting fertility during the last decade in Amhara National Regional State, Ethiopia using the Bongaarts' Proximate Determinants of Fertility Model. Detailed study on the trends in proximate determinants of fertility change is important for assessing the implementation of population/reproductive health policies and family planning programs and the efforts achieving the overall goal of changing fertility behaviour, that is, increasing contraception, reducing unmet need for family planning, and rising age at first marriage. In addition, the decomposition of proximate determinants is also a basis for future projections of changes in the overall level of fertility which may be expected as a result of changing socioeconomic, demographic, cultural, institutional, psychological, health, and environmental conditions. --- Methods --- Data sources The 2000, 2005, and 2011 Ethiopia Demographic and Health Surveys were the data sources for this analysis. They were designed to provide population and health indicators at the national and sub-national levels. The sample was selected using a stratified two-stage cluster design. Enumeration areas were the sampling units for the first stage and households comprised the second stage of sampling. A complete listing of households was carried out in each of the selected enumeration areas and a representative sample of households was selected proportionally. All women of reproductive age found in the selected households were interviewed. A representative sample of 2059, 2158, and 2163 households were selected for the 2000, 2005, and 2011 survey, respectively, in Amhara National Regional State. The 2000, 2005, and 2011 Ethiopia Demographic and Health Surveys successfully interviewed 3820, 3482, and 4433 women of reproductive age, respectively, in the region. --- Analysis Bongaarts proposed the proximate determinants of fertility model at a time when it was assumed there was very little intercourse outside of marriage. Hence, all of the calculations he proposed were based only on currently married women of reproductive age. Bongaarts initially proposed the proximate determinants framework; social and cultural mores have been shifting to greater or lesser degrees across the world and girls and women are now having sexual intercourse before and outside marriage. Stover proposed updating the framework to account for the fact women are having sex before and outside marriage and these same women may use contraceptives and seek induced abortions [4]. This study was based on Bongaarts' model because in most societies of the region births out of wedlock were not accepted and virginity was a prerequisite for marriage. In addition, the median age at first marriage of women of reproductive age in 2000, 2005, and 2011 was about 14.5, 14.4, and 15.1 years, respectively, while the median age at first sexual intercourse of the 2000, 2005, and 2011 was about 14.5, 14.7, and 15.3 years, respectively. That is, age at first marriage is synonymous with age of entry into sexual relations in the region [14][15][16]. For these reasons, we opted to conduct the proximate determinants analysis using currently married women of reproductive age. In addition, when contraceptive use is concentrated at later ages, the Bongaarts' model is a poor estimator of the fertility-inhibiting effect of contraception because women use contraception to stop rather than to space births [19]. These problems, however, seem to be minimal in Amhara National State where women use contraception at all ages and mainly for spacing purposes. In the region, the age pattern of modern contraceptive use was similar across the three surveys. It increased largely up to the age group of 30-34 and then declined rapidly towards the ends of the age cohort [14][15][16]. --- A) Role of proximate determinants of fertility The model indices were calculated for each survey separately. The model expresses the total fertility rate as a product of four indices and total fecundity rate [2,3,20]. Mathematically, it can be expressed as: TFR ¼ TF à C m à C c à C a à C ið1Þ Where, TFR is the total fertility rate, TF is the total fecundity, C m is the index of marriage, C c is the index of contraception, C a is the index of induced abortion, and C i is the index of postpartum insusceptibility. Each index has a value between 0 and 1. When the index is close to 1, the proximate determinant has a negligible inhibiting effect on fertility, whereas, when it takes a value very close to 0, it has a large inhibiting effect [2,3]. That is, the complement value of index represents the proportionate reduction in fertility attributable to the fertility-inhibiting determinant . The effect of abortion on fertility could not be assessed in this analysis because of a non-availability of data from the Ethiopia Demographic and Health surveys. Data on induced abortion are very rare in Africa due to illegality of induced abortion unless performed to save the mother's life [21]. Before 2005, abortion in Ethiopia was permitted only in the case where the life of the woman was physically at risk. With the 2005 revised abortion law, abortion is allowable due to fetal impairment or deformity, rape or incest , minors , and physically and mentally unable to care for the would-be born child [22]. --- i) Total fecundity The total fecundity is a hypothetical or potential value that the total fertility rate would take if all the indices of the proximate determinants were 1. That means, a fertility level that would prevail if all women age 15-49 were married, there was no contraception use in the population concerned, no postpartum insusceptibility , and no induced abortion [2,3]. According to Bongaarts and Potter, the total fecundity of most populations falls within the range of 13 to 17 births per woman with an average of approximately 15.3 [2]. Therefore, multiplying all of the indices with 15.3 as the maximum number of births produces a predicated total fertility rate of the population. --- ii) Index of marriage The index of marriage represents the proportion by which total fertility rate is smaller than the total marital fertility rate as a result of marital pattern . The higher the level of marriage in the population, the less the inhibiting effect on fertility and vice versa. The index of marriage is calculated as the proportion of total fertility rate over the sum of age-specific marital fertility for a given population at a given time, that is, C m ¼ Σm a ð Þg a ð Þ=Σg a ð Þð2Þ Where, C m is the index of marriage, m is the agespecific proportions of women currently married and is produced by dividing the number of married women of a particular age group by the number of women in the same age group and g is the age-specific marital fertility rate and computed by dividing the age-specific fertility rate by the proportion of women currently married in each age group. iii) Index of contraception The index of contraception describes the effect of contraception on reducing the risk of conception . In terms of the marital fertility rate, the index of contraception gives the proportion by which the total marital fertility rate is smaller than total natural fertility . However, the contraception index value depends on the current contraceptive prevalence rate and the average use effectiveness of the contraception. The index of contraception varies inversely with the contraception prevalence and use effectiveness of modern contraception practiced by couples. The index of contraception is calculated as the proportion of currently married women of reproductive age currently using specific methods of modern contraception, that is, C c ¼ 1:00-1:08à uà e ð Þ ð3Þ Where, C c is the index of contraception, u is the contraceptive prevalence among married women age 15-49, and e is the average effectiveness of contraception. The constant 1.08 given in Eq. is an adjustment factor designed to remove infecund women from the equation so the contraception index would become zero if effective prevalence reached 92.5%, and the remaining women are assumed to be infecund [2]. The value of e is estimated as a weighted average of the method specific use effectiveness ) level, the weights being the proportion of women employing a specific method ), that is, e ¼ Σe i ð Þà u i ð Þ=uð4Þ The indices of use effectiveness proposed for particular contraceptives are pill = 0.90, Intrauterine Contraceptive Device = 0.95, sterilization = 1.00, condom = 0.62, and others = 0.70 [2]. --- iv) Index of postpartum insusceptibility The index of postpartum insusceptibility is intended to describe the effects of breastfeeding or abstinence on fertility in the population. According to Bongaarts and Potter, in the presence of breastfeeding and postpartum abstinence, the average birth interval equals, approximately 18.5 months plus the duration of postpartum insusceptibility. The constant 20 in Eq. represents the average birth interval length if neither breastfeeding nor postpartum abstinence is practised [2]. Therefore, the index of postpartum insusceptibility is estimated as the ratio of the mean birth interval without postpartum insusceptibility to that with postpartum insusceptibility, that is, C i ¼ 20= 18:5 þ i ð Þ ð5Þ Where, C i is the index of postpartum insusceptibility and i is the mean duration of postpartum infecundability measured in months. --- B) Decomposition of Change in Proximate Determinants of Fertility The decomposition of proportional change in the total fertility rate into the components of the proximate determinants was also carried out. The purpose of decomposition analysis of proximate determinants was to identify the sources of changes in the predicted total fertility in the last decade. A decomposition method developed by Kitagawa and later elaborated by Das Gupta was used to analyze the data [23,24]. The decomposition was computed for three different scenarios: for the proportional change between 2005 and 2000, 2011 and 2005, and 2011 and 2000. The negative signs on the percentages indicate those proximate determinants having inhibiting effects on fertility during the period. The decomposition procedure is described below. Step 1: Decomposition of change in total fertility rate over the time frame can be calculated as: TFR t2 =TFR t1 ¼ C m t 2 ð Þ=C m t 1 ð Þ Ã C i t 2 ð Þ=C i t 1 ð Þ Ã C c t 2 ð Þ=C c t 1 ð Þà R t 2 ð Þ=R t 1 ð Þð6Þ Step 2: Proportional change for each proximate determinants and the residual between two periods a) Proportional change in total fertility rate TFR t2 =TFR t1 -1ð7Þ b) Proportional change in total fertility rate due to a change in the index of marriage C m t 2 ð Þ=C m t 1 ð Þ-1ð8Þ c) Proportional change in total fertility rate due to a change in the index of postpartum insusceptibility C i t 2 ð Þ=C i t 1 ð Þ-1ð9Þ d) Proportional change in total fertility rate due to a change in the index of contraception C c t 2 ð Þ=C c t 1 ð Þ-1ð10Þ e) Proportional change in total fertility rate due to a change in the remaining proximate variables -fecundability, intrauterine mortality, and sterility R t 2 ð Þ=R t 1 ð Þ-1ð11Þ Step 3: Total proportional change in total fertility rate overtime due to proportional change in the proximate determinants, residual, and the interaction factor P f ¼ P m þ P i þ P c þ P r þ Ið12Þ --- Results Biological and Behavioural characteristics of respondents A skewed method-mix had been apparent in the region with only three methods, namely injectables, implants, and pills accounting for about 97% of all contraceptive use in 2011. Injectables were by far the most predominant method in all surveys. The share of injectables rose over the last decade, from 52% in 2000 to 75% in 2005 and 80% in 2011. Between 2000 and 2005 the proportion of implants to the total contraceptive use remained constant at 2% but there was an increase to 12% in 2011. In 2011 only 5% of total contraceptive use was the pill and it was 45% and 23%, respectively, in 2000 and 2005. Other methods taken together contributed to less than 3% of overall contraceptive use in 2011. In Amhara National Regional State, the median age at first marriage among currently married women increased slightly from 14.5 years in 2000 to 15.1 years in 2011. The proportion of women who married at age 15, 18 and 20 declined from 56.5%, 87.0%, and 92.1% in 2000 to 48.3%, 74.9%, and 83.0% in 2011, respectively. The early marriage was highest in the country though it showed only a 12% decline in the last decade from 87% in 2000. In Amhara National Regional State, the mean duration of amenorrhea declined from 22.4 months in 2000 to 18.4 months in 2011 in the total sample of currently married women. The mean duration of abstinence in the total sample of currently married women increased from 2 months in 2000 to 2.7 months in 2011. The mean duration of insusceptibility declined from 22.6 months in 2000 to 19.0 months in 2011. --- Trends and levels of predicted fertility Table 2 depicts the model total fertility rates for each survey among currently married women age 15-49 by residence and education. Accordingly, the predicted total fertility rate of the region declined from 5.5 children per woman in 2000 to 5.1 in 2005 and further to 4.1 children per woman in 2011. Although there was an increase in the predicted fertility level in urban areas from 2.6 children per woman in 2000 to 2.8 children per woman in 2011, in rural areas the predicted fertility declined from 5.9 children per woman in 2000 to 4.5 in 2011. Women with secondary and above education had the lowest predicted fertility rates in 2011 compared with women with primary education and those who had no formal education. The model total fertility rate of women with secondary and above education was about three children lower than those with no education and about two children lower than those with primary level education. --- Changes in components of proximate determinants of fertility The fertility-constraining effect of contraception increased as evidenced by a decrease in the index value overtime. The estimated contraception index declined from 0.93 to 0.65 during the period. The fertility-constraining effect of postpartum infecundability decreased as evidenced by an increase in its index value overtime. The estimated index of postpartum infecundability increased from 0.49 in 2000 to 0.54 in 2011. The index of marriage overtime remained unchanged with an index value of 0.71 in 2000 and 0.70 in 2011. Table 2 shows the fertility-inhibiting effect of various proximate determinants overtime by educational status and place of residence of women. The magnitude of the influence of proximate determinants varied across time and different segments of the society in regulating fertility of a population. Of the three indices considered, postpartum insusceptibility and marriage showed the while postpartum insusceptibility remained the leading factor in rural areas of the region. In addition, over the period considered, the influence of contraception in controlling the fertility of the rural women increased more than the other proximate determinants. The fertilityregulating impact of marriage in rural areas was stable overtime ; on the contrary, even though the effect of marriage declined from 0.42 in 2000 to 0.65 in 2011, its impact on fertility-inhibition was higher than rural areas. In 2000, postpartum insusceptibility and marriage were seen the primary fertility-limiting factors for illiterate women and women with secondary and above level of education , respectively. Contraception also played a role in limiting fertility among women with secondary and above level of education while its impact was minimal among illiterate women . A similar pattern was observed in 2005 and 2011 except the fact that contraception had gained prominence in limiting fertility among women with a level of education above primary school . Moreover, contraception showed a large shift of impact in controlling fertility among illiterate women and women with primary level of education. --- Decomposition of changes in proximate determinants of fertility Once the indices of the three major proximate determinants had been calculated for each survey, the next step is to decompose the change in the total fertility rate observed between two time points to determine contributions of the three proximate determinants . A decline in the total fertility rate observed in Amhara National Regional State during the period 2000 to 2011 was much stronger towards the end of the period. In the period 2000 to 2005, a 6.9% decline in the total fertility rate was observed as opposed to a 19.9% decline in the total fertility rate witnessed during the period 2005 to 2011. In the first-half of the decade, 2000 to 2005, it was only the use of contraception had played a role in the decline of fertility while marriage and postpartum insusceptibility were seen to have increased fertility in the region. In the second-half of the decade, an 8.0% increase in the total fertility rate was due to a change in postpartum insusceptibility while marriage and contraception were responsible for a 4.0% and 21.5% decline in the total fertility rate, respectively. The changes in the total fertility rate by education and residence were also decomposed by the proximate determinants. There was a consistent increase in the total fertility rate in urban areas whereas fertility in rural areas of the region had declined. The largest portion of the increase in fertility of urban areas between 2000 and 2011 happened as a result of a change in the proportion of married women of reproductive age over the period despite the fact that postpartum insusceptibility and contraception were associated with a decline in fertility. The percent change in the total fertility rate of rural women during 2005-2011 was twice that observed during the first-half of the decade . Change in the proportion of rural women using modern contraception was responsible for a significant proportion of the decline observed in fertility over the period 2000 to 2011. From 2000 to 2005, the decline in fertility observed among women with no formal education was larger than women with above primary education. The responsible factor for this decline was a change in the prevalence of contraception over the period among illiterate women, however, the decline in fertility among women with secondary and above level of education was associated with factors that were not included in the model . On the contrary, the increase in fertility among women with primary level of education was largely attributed to the changes in the proportion of married women during the period and reduction in the duration of postpartum insusceptibility. In the second-half of the decade, 2005 to 2011, a decline in fertility in all three categories of women's education was observed. The decline in fertility observed among women with primary level of education was the largest and all the proximate determinants played a role albeit, in varying degrees. The overall decline in fertility witnessed during the 2000-2011 period by level of education of women was highest for women with secondary and above level of education and least for women with primary level of education . --- Discussion The predicted total fertility rate decline of 25.5% between 2000 and 2011 was decomposed into a 29.3% decline due to an increase in contraceptive practice and a 1.9% decline due to a decrease in the proportion of women in union, and a 9.0% increase due to shortening of the duration of postpartum infecundibility. The change in postpartum insusceptibility contributed the increase in the total fertility rate cancelling out a substantial part of the potential impact of contraception and contributing to a stall in fertility decline. Consistent with this finding, Nasir and Sinha reported the fertility-inhibiting effect of natural fertility control being gradually replaced by contraceptive use [25][26][27]. The index of postpartum insusceptibility declined in the values of its indices in the survey periods signifying a decrease in traditional cultural practices of postpartum abstinence. Contrarily, Ngalinda and Teklu et al. found an increase in the effect of postpartum infecundability in controlling fertility overtime and a primary factor for reducing the prevailing levels of fertility compared to its biological maximum [18,28]. The model also confirmed that postpartum insusceptibility due to breastfeeding had the greatest inhibiting effect on fertility among illiterate women and women residing in rural areas. In societies where breastfeeding is generally prolonged and universal and contraceptive use is rare, the primary determinant of birth interval is the duration of breastfeeding [29][30][31]. However, this was not true in the urban areas where contraception played the leading role in inhibiting fertility. The general observation is that the duration of breastfeeding declines with development. In particular, breastfeeding declines with urbanisation and education [32]. The fertility-constraining power of contraception showed remarkable increase overtime. Fertility decline in Africa initially began in Kenya, Botswana, and Zimbabwe due to high levels of contraceptive practice, ranging between 27 and 44% among currently married women of reproductive age [33]. Slowing of fertility decline and contraceptive prevalence in sub-Saharan Africa have impacted on the pace of fertility transition [34][35][36][37], that is, a weakened family planning service environment is contributing to the stall in fertility decline [38,39]. An increase in contraceptive prevalence has emerged as a dominant explanation for the observed fertility declines [40][41][42][43][44][45]. However, Ngalinda and Teklu et al. found the use of contraception had only a very minor fertility-inhibiting effect compared to its biological maximum [18,28]. Contraception led to a remarkable change in inhibiting fertility among illiterate and rural women between 2000 and 2011. This may be due to the deployment of health extension workers since 2003 in each Kebele . The accompanying diffusion effects brought by such new endeavours are contributing factors in shaping the contraceptive use and fertility behaviour of the society, especially in rural areas of the region [9]. The index of marriage contributed a minor proportion of the decline in the total fertility rate in the region due to the small increase in age at first marriage, as marriage is a customary and socially established practice that has been carried on for generations in most communities of the region. However, the proportion of marriage is one of the major fertility-inhibiting factors and played increasing role overtime [18,25,[45][46][47][48][49]. The predicted total fertility rate of the urban areas in the region consistently increased during the period due to a reduction in age at first marriage. In Amhara National Regional State, there has been a decline of more than two years in the median age at first marriage from 22.9 to 20.6 years in urban areas especially amongst the youngest women [14][15][16]. Consistent with the finding reported by Westoff where a decline of one year of age at first marriage lead to an increase of 5-6% in the total fertility rate [50], the finding of the present study also suggests a change in age at marriage in urban areas to some degree brought about the slight increase in fertility during the period in the region. This study has a number of strengths. First, the study utilized a large population-based sample, a randomly selected population, a high response rate , the use of standardized surveys, the well-tested validated questionnaires, and the high quality of data due to extensive training of data collectors and support during the fieldwork, with concurrent data entry and editing with feedback during fieldwork. Second, calculations were done after the data were weighted for the sampling probabilities and non-responses. Third, decomposition of proximate determinants was applied to understand the sources of change in fertility overtime. Finally, the study used a five-year interval surveys and can provide important data for monitoring population/reproductive health policies and family planning programs. Our study also has some limitations. During decomposition analysis, the index of induced abortion was taken as 1 in all populations, that is, induced abortion was assumed to have no inhibitory effect on fertility of a woman because of a lack of data in each survey. The exclusion of abortion in the model could have an effect on the estimates of total fecundity especially in areas where it is high. It is documented that induced abortion rates in sub-Saharan Africa are growing very fast, particularly in urban areas and amongst the youngest [51][52][53][54][55][56][57]. Another limitation of Ethiopia Demographic and Health surveys was that they employed a cross-sectional retrospective design to collect information about women's birth. The validity, reliability and completeness of crosssectional retrospective data depend on the respondent's recall of past events. Differences in recall periods had been shown to influence the quality of data in relation to fertility, since individuals tend to recall the most recent period. Fertility estimations in each survey were dependent on the accuracy of the respondents' current age and the quality of birth history data. A common feature of age distributions in each survey was heaping on ages ending with 0 and 5. The birth history collected from each woman age 15-49 is essentially a listing of all the live births the woman has had. The most important attributes of each birth, apart from the fact of the birth, are the month and year of birth; whether the child has survived to the date of interview; and, if the child has died, the age at death. The reliability of the total fertility rate estimates depends on the accuracy and completeness of reporting and recording of births and deaths. Typically, the most serious source of non-sampling errors in each survey that collects retrospective information on births and deaths arises from the omissions of children and the displacements of birthdates. --- Conclusions An increase in the level of contraceptive use and effectiveness overtime among fertile women was the single most important contributing proximate determinant to the recent fertility decline in the region. Postpartum insusceptibility contributed to an increase in fertility due to a shortening of the duration of postpartum infecundibility during the last decade. Regardless, this increase was compensated by the fertility-inhibition effect of contraceptive use in Amhara National Regional State. Marriage contributed a little in constraining fertility in the 11-year span due to the unchanging of age at first marriage at early age in the region. Although fertility declined overtime in the region, a close look at fertility experience of women disaggregated by various characteristics revealed that those who live in rural areas and had primary or no education were the ones with high fertility. Fertility transition in the region could be accelerated through significantly increasing girls' education beyond the primary level of education and strengthening education and communication about the status of women. On the other hand, strengthening of existing reproductive health and family planning programs is required in order to increase the quality and quantity of contraceptive use and achieve higher use effectiveness which will lead to a greater contribution to fertility decline. In this regard, strengthening the Health Extension Workers, Health Development Armies and other volunteers' community health workers could have a powerful role in reaching women with family planning messages, and impacts the ability to maintain the current impetus in contraceptive use. There is the high level of unsatisfied demand for fertility regulation and the increasing contraceptive use in the future particularly in rural areas would have a major impact. In addition, the implementation of the family law which asserts the minimum age of marriage as 18 years should be ensured by creating community awareness using social mobilizations, strengthening local enforcements, and establishing legal age assessment committees in each rural Kebeles. --- --- Authors' contributions NMA, GNT and TDD conceived and designed the study and performed the analysis of the data. All authors read and approved the final manuscript. --- --- --- Competing interests The authors declare that they have no competing interests. ---

Background: The overall decline of fertility in Amhara National Regional State between 2000 and 2011 was the highest in Ethiopia. The aim of the present study was to determine the most significant proximate determinant of fertility change during the last decade in the region using Bongaarts' model.The sources of data were the 2000, 2005, and 2011 Ethiopia Demographic and Health Surveys. The model indices were calculated for each survey. Decomposition of fertility change into components of proximal determinants was also carried out. An index value close to 1 is a negligible inhibiting effect while a large inhibiting effect when the value very closes to 0. Results: The fertility-constraining effect of contraception increased from 0.93 in 2000 to 0.65 in 2011; however, it was lower than the effect of postpartum insusceptibility at all given times. The index of marriage remained unchanged in constraining fertility over the period (0.71 in 2000 and 0.70 in 2011) while the influence of postpartum insusceptibility slightly declined from 0.49 in 2000 to 0.54 in 2011 but was stronger than contraception and marriage. The contribution of contraception was most important in urban areas (0.46 in 2011 from 0.52 in 2005 and 0.64 in 2000); however, in rural areas, it became an important determinant over the period (0.95 in 2000 and 0. 69 in 2011). The effect of postpartum insusceptibility in rural areas showed a decreasing trend (0.48 in 2000 and 0. 53 in 2011). The index of marriage in rural areas was stable overtime (0.75 in 2000 and 0.73 in 2011) while in urban areas the effect declined from 0.42 in 2000 to 0.65 in 2011. Marriage was the most important proximate determinant of fertility among women with secondary and above education but the impact declined during the period (0.41 in 2000 and 0.61 in 2011). The importance of postpartum insusceptibility in limiting fertility among women with secondary and above education declined overtime (0.77 in 2000 and 0.87 in 2011) whereas the contribution of contraception became more important (0.44 in 2000 and 0.35 in 2011). Conclusions: An increase in the level of contraceptive use and effectiveness overtime was the single most important contributing factor for the recent fertility decline in the region.

Introduction The study of environmental issues has taken two approaches to the relationship between humans and nature: environmentalism and ecologism. The difference between both is that environmentalism has an anthropocentric perspective on the environment, i.e., concern for nature revolves around the needs and interests of humankind . In contrast, ecologism has an ecocentric perspective, since it postulates that nature has an intrinsic value, positioning humans as one more element of nature . One of the major discussions in environmentalism and ecologism takes place in the context of the environmental crisis, characterized by an interrelated set of specific problems such as depletion of natural resources, different forms of environmental pollution, decrease of biodiversity, etc., . This environmental crisis is related to the effect that human behavior has on the environment at different scales, like changes in production and consumption patterns and other processes exacerbated by globalization . These effects, in turn, have resulted in risks for humans and ecosystems , 2014). The marine ecosystem, among others, fulfills economic and psychological functions . At the economic level, it is a supply of food and other natural resources obtained from extractive activities , 2014; Saavedra and Mardones, 2021). At the psychological level, it is a source of wellbeing that allows the expression of cultural and aesthetic manifestations, provides spiritual wellbeing and stimulates recreation , 2014; Thushari and Senevirathna, 2020). These functions indicate the predominance of humanity's instrumental view of the sea. From the perspective of environmental psychology, the more dominant this instrumental approach is, the greater the potential for endangering the ecosystem . This instrumental view of the sea might be related to the predominance of neoliberalism in Peru and Chile during the last decades . Neoliberalism is an economic and political system that reduces government intervention in favor of the free market, placing producers and consumers in charge of regulating the system . Under this system, the government ensures the right conditions to maintain macroeconomic growth, even at the expense of the people and the environment . Thus, some of the social consequences of neoliberalism are economic growth at the expense of increased inequalities and a society with individualistic and depoliticized values . These, in turn, justify the system with attitudes, values and practices that prioritize individual and interpersonal interests over the public one . Hence, the environmental crisis facing the marine ecosystem goes unnoticed at the political, economic and social levels due to a system that makes it invisible or delegitimize it . The influence of social and economic systems, such as neoliberalism, on environmental concern might be understood through, what the New Ecological Paradigm identifies as, the two dimensions at the basis of environmental attitudes and behaviors: ecocentrism and anthropocentrism . From an ecocentric perspective, environmental protection is important for its intrinsic value, transcending the capacity of natural resources to satisfy functions in favor of humanity . In that sense, people with an ecocentric perspective, following biospheric values, will choose to perform pro-environmental actions based on the cost and benefit for the environment . In contrast, the anthropocentric perspective argues for the protection of nature based on the value of natural resources and ecosystems in maintaining and improving quality of life . This anthropocentric dimension is aligned with selfish and altruistic values. Those who follow selfish values will decide to perform pro-environmental behaviors considering costs and benefits for themselves, while those who prioritize altruistic values will take into account the repercussions that marine pollution has on other individuals or society . In relation to this, environmental attitude incorporates cognitive, affective and dispositional components, characterizing an opinion or predisposition to act in favor of the environment . In this sense, people's opinion about environmental protection and natural resources precede environmental behavior . The cognitive component refers to the knowledge and information about environmental problems, their causes and the mechanisms to avoid and correct them. The affective components are psychological factors such as feelings or attitudes of concern for environmental conservation. The dispositional component includes personal attitudes towards pro-environmental action and the cost of implementing environmental policies . Associated with environmental attitudes, the connection with nature is a cognitive and emotional connection based on affective experiences associated with staying in nature . While being in connection with nature, love and concern for it involve a sense of responsibility and commitment to environmental protection . In that sense, the greater the connection with nature, the more likely it is that pro-environmental behaviors will be performed . Pro-environmental behavior comprises deliberate and effective actions aimed at protecting the environment . However, despite the wide acceptance of the research of environmental attitudes, several studies report inconsistent results regarding a statistically significant relationship between attitude and pro-environmental behavior . This might be linked to the weighting of values, such as competition and individualism, within society that influences selfish attitudes . However, from a more ecological perspective of pro-environmental behavior, four levels of analysis are proposed to explain the different range of involvement, from individual action to cultural representation . It is proposed that the levels of pro-environmental behavior could be understood as expressions of a model of socialization of environmental attitudes and behaviors. This is based on Bronfenbrenner's ecological theory, which states that human socialization occurs in the interaction between the different systems humans engage with . Based on what Hoffman and Graham proposed, the first level of pro-environmental behavior involves individual actions of people committed to the environment, such as cleaning up litter, segregating waste, using reusable products, etc. These pro-environmental actions occur within systems people belong and interact with, such as family, work peers, friends, or volunteer groups; this is known as microsystem . These systems, although independent of the individual, are influenced by them and vice versa . For example, if the individual lives with a family that promotes pro-environmental behaviors, they are more likely to develop environmental protection habits. Similarly, individuals who volunteer to promote the use of non-plastics are more likely to reduce their plastic consumption, and their actions could influence social groups close to them. According to Bronfenbrenner , these systems in which the individual participates within the microsystem can be connected to each other through a network that constitutes the mesosystem. For example, a plastic bottle recycling campaign at work encourages collaborators and their families to participate in the activity. The second level of pro-environmental behavior is constituted by the particular policies implemented at the local level in favor of the environment . These policies are implemented in settings where the person does not actively participate but is affected by occurring events; this is defined as exosystem . Particular policies are issued by authorities, policymakers and people in power who have an impact at the local level. These institutions form systems that influence people's pro-environmental behavior although they might not actively participate in. For example, municipal ordinances, although established by local authorities, regulate the environmental behavior of individuals even if they have not been directly involved with it. The third level refers to a set of public policies promoted by an interest group . This level, like the previous one, also belongs to the exosystem. The difference is that these policies have a wider scope such as national, regional or international. For example, the consumption tax on plastic bags has led stores to apply a fee to customers for plastic bags , promoting sustainable alternatives. The fourth level refers to the society's cultural proposal of prioritized values , such as pro-environmental values and environmental protection. These cultural or ideological elements are human-related and have influence on lower systems, hence, as a group, they form the macrosystem . Consequently, a society that promotes ecological attitudes and behaviors would build a citizenship culture that fosters environmental protection, expressed in the demand for environmental policies and the promotion of sustainable habits. In the present study, we explored the relationship of these levels of analysis and socialization of pro-environmental behaviors in marine environments, whose physical, chemical and biological properties have been altered by global warming and pollution , 2014). The main problems affecting this ecosystem are the warming and acidification of its waters, destructive fishing practices, petrochemical exploitation of the seabed and plastic pollution . Plastic pollution in the sea is an alarming global problem caused by human behavior at the individual, collective and societal levels . Some sources of plastic pollution are industrial, recreational or economic activities . Another source of plastic comes from solid waste generated on land and offshore, that decomposes into small particles: microplastic, which can generate the death of marine species and alter the marine ecosystem . Furthermore, marine plastic pollution has an impact at the social and economic level. At the social level, there are negative consequences on human health related to food safety and exposure to chemical substances due to microplastics in the food chain , on people's lifestyles, mental health, identity and cultural heritage . At the economic level, there is a loss of income associated with marine and coastal activities such as fishing, aquaculture and marine tourism, due to the high costs of remediation activities such as beach cleanup . In this context, Peru and Chile might be two key countries involved in marine plastic pollution since, in 2017a new plastic island near the coasts of both countries was found . In those lines, the most common products found on these coasts are single-use plastics such as bottles and plastic bags . Studies in Peru and Chile have reported that part of the plastic pollution comes from economic activities such as fishing, since the greatest polluting source is abandoned, lost or discarded fishing gear . Lost gear traps and attracts different animals, which can lead to their death and the continuation of a cycle of entanglement , and disperse invasive species altering the marine ecosystem . Plastic pollution might have a terrestrial origin by sewage spills, polluted river basins, etc., , and a marine origin by the consumption of food and beverages on boats, the loss of equipment, products for boat maintenance, among other sources of microplastics . Thus, marine plastic pollution might be one of the most environmental problems these countries face. A global poll identified that Peru and Chile were two countries that recognize the importance of a plastic global treaty the most in the region. While Chile supported more the prohibition of single-use plastic than Peru, the latter was involved in a proposal to the United Nations to consider the role of the life cycle of plastics in environmental pollution . Nonetheless, in Peru, plastic remains one of the most common sources of marine pollution , despite the existence of public policies such as Law No. 30884 which, as of 2018, prohibits the consumption of such materials in protected natural areas. Despite its approval, Congress claimed that part of the law, specifically the ban on the manufacture of expanded polystyrene , affected small and medium-sized companies and should therefore not be implemented . This institutional discrepancy might demonstrate the authorities' lack of interest in environmental issues. In the case of Chile, in 2021, Law No. N.° 21,368 was passed, banning the free service of single-use plastic items in stores throughout the country, limiting the supply of plastic items, and promoting environmental education about the ecological impact of single-use plastic engaging companies, municipalities and the Ministry of Environment . Years before, the establishment of fines for littering in beaches and riverbanks were approved . These actions were possible due to the growing concern that governmental institutions and non-governmental organizations were having about marine litter, which influenced the involvement of multiple municipalities in reducing the consumption of single-use plastics . The success of such efforts also depends on the actions of citizens, which may be affected by a given context. In this regard, a study conducted in coastal regions of Chile found that those who are more willing to engage in pro-environmental behaviors valued their cultural history, felt a connection with nature, and their regions had an economy based on sustainable tourism. This was not the case in regions whose economy was based on mining or aquaculture, and with few spaces to enjoy nature . Despite these advances, there is discontent among Chileans, manifested in massive protests, towards their political institutions, as environmental issues have a direct impact on the residents of regions where extractive activities predominate . This is consistent with several studies that report that when people feel connected to an ecosystem, they tend to engage in pro-environmental behaviors . However, as identified in the contexts of Peru and Chile, when such preservation measures affect the interests of large companies, the protection of ecosystems is given low priority . Based on all the above, this study aims to contribute to the knowledge about representations of the sea and the understanding of pro-environmental behavior as a product of a socialization process, taking Bronfenbrenner's ecological model as a reference. In addition, it seeks to highlight the structural dimension of pro-environmental behavior while identifying its individual and collective levels of action. Therefore, the general objective of this study is to explore the attitudes and reported-pro-environmental behavior of Peruvian and Chilean citizens regarding marine plastic pollution. To this end, three research objectives were proposed: explore the representations that citizens of both countries have about the sea and the problems associated with this ecosystem; identify the causes and consequences that citizens of both countries recognize of marine plastic pollution; and learn about the actions that citizens of both countries propose to take to reduce marine plastic pollution. --- Methods --- --- Data collection A semi-structured interview guide was prepared to facilitate dialogue with the participants, a focused analysis of the details of the discourse, and delving into new and relevant information . Its structure was based on three thematic axes, with a total of ten main questions with their respective sub-questions . The interview guide was elaborated based on the literature review and the study's objectives. --- Procedure After recruiting the participants, the interview's time and date were coordinated through the Zoom platform. Before each interview, the informed consent protocol was implemented. The interviews lasted approximately 45 min and were conducted in two phases. Fieldwork was conducted in Peru from January to August of 2022, while in Chile from April to October of the same year. Once data saturation was reached, the recruitment was closed and the interviews were transcribed. Finally, the information from both countries was analyzed. The overall procedure, including the data collection and the information analysis was carried out in Spanish. For the purpose of this article, the results and final report of the study were translated to English. --- Information analysis The analysis of the information consisted of a thematic analysis, the aim of which was to identify the most relevant and frequent patterns of meaning in the participant's discourse. This procedure was carried out in three stages using Atlas.ti 9 software. The first stage was open coding, this served to identify the content of the interviews systematically and objectively. Thus, analysis codes were obtained and systematized in a codebook . These codes were established since an inductive logic, that is the identification of themes from the analysis of the participant's discourse that accounted for their feelings and experiences . Once some families of codes were identified, axial coding was carried out in parallel, which sought to link the codes according to their dimensions and properties around categories representing thematic axes . During the coding process, the saturation of the information was considered, which marks the moment when, after analyzing multiple interviews, new information cannot be found and, thus, the coding analysis concludes . Finally, the categories were analyzed through the revised literature and reviewed by the research team and other experts in the field. --- Results --- Representations of the sea: influences of recreational and economic use The Peruvian and Chilean citizens' representation of the sea is characterized mainly, at the cognitive and affective level, by a positive appreciation of this environment . This might be related to the fact that most of them had frequent contact with this ecosystem . According to the participants, the sea is the origin of life, a space that allows human life and where a wide biodiversity resides. Thus, the sea is represented as "a great source of oxygen, food and recreation for humans" . Based on this, the participants attributed two main purposes to the sea: recreational and economic use. Most of the participants highlighted the recreational use, the perception of the sea as a space for sports, games and social interaction. According to the Peruvian participants, who reported this purpose most often, "most people go to the sea to swim or drink beer. They see it as a recreational space" . This brings them happiness, which is described as a feeling of exaltation, wellbeing and fun: "a feeling of pleasure, one of the greatest pleasures I feel is when I run, jump into the ocean and swim" . Furthermore, this recreational use is often associated with the feeling of connection with nature, understood as the feeling of being connected with the sea, the life that converges in it and with oneself. For some participants, this connection is experienced as a mimicry with the sea, and for others it is "a connection with yourself, you discover yourself in the water" . In this way, the participants recognize the sea as a refuge: a safe and appropriate place to relax and express emotions. In this sense, having access to the sea produces tranquility in most of the participants from both countries, understood as the feeling of not having problems or worries: "… it is something very comforting, if I am stressed or worried about something, the marine environment relaxes me" . The association between this representation of the sea with the feelings of connection with oneself, security and relaxation generate in them respect for this environment: "I think it is a safe space for people, the feeling of being isolated when you are in the sea is soothing and relaxing. This can give you a great respect for the sea, to understand that it is the only place where you are alone, where you can be safe and you can do what you like" . In this way, for some participants the sea represents a space for disconnection from routine, to get away from work and the city: "those of us who study, work, feel stressed in the day-to-day feel that the sea is like a balance that takes you out of the routine and recharges you with energy" . This energy renewal is described by the participants as a return to a sense of well-being. In addition, the disconnection from routine produces freedom in most participants from Peru and Chile, the transitory state of having no restrictions or responsibilities of one's own free will. Moreover, the sea acquires value for what it offers to people: "[the sea] is an environment for anyone who wants to take a space to renew themselves, to recharge their energy, to de-stress, to free themselves. Also, I think it is an invaluable space for society" . To some Peruvian and Chilean participants, the importance of the sea is associated with aesthetics, the perception of the sea as something beautiful and pleasant to observe: "for me it is significant and beautiful, the colors, the sun, the clouds, the wind, the movement is different, it is wonderful" . This appreciation of the sea comes from their own definition of magnificence, the feeling of witnessing an immense element that surpasses their own capacities: "I recognize the strength and power it has, and that means that I recognize my human weakness. It is about respecting that there is something bigger" . The respect for the sea is also inspired by its representation as an unknown element, which cannot be controlled, and as a potentially dangerous place for people, which cannot always be trusted: "I believe that nature cannot be challenged, because when it wants to show itself, it does. I am not going to face a rough swell, I do not go near if there is a high tide, because it is said that the sea is treacherous" . However, in opposition to respect, recreational use is also linked to the exploitation of the sea, understood as the low appreciation and low prioritization of marine protection by the general population. For the participants this is because "in the minds of people the sea is a place where you can go, enjoy and be happy. There is no sense of marine protection" . There is a perception that the lack of concern about marine protection might be related to a limited representation of the sea as only a source of wellbeing and/or economic profit: "I think that, first, as humans, as a population, we are mainly responsible for not having been able to protect our sea, first because we think that it is only a source of economic benefit or aesthetic-visual attraction, but we were never educated in its importance" . In this way, the exploitation of the sea could also be associated with its economic use which consists of economic and extractive activities. In this sense, the sea "is a work device…, thanks to this, we can have a development of economic progress" . For some participants, mostly from Chile, this representation of the sea could be influenced by the neoliberal model, an economic, political and social model that prioritizes the exploitation of resources and consumption. This model justifies "that agreements can be generated between families and political conglomerates that have created laws in favor of extractivism" . Therefore, the participants identified that at the regulatory level, extraction is allowed over marine protection. This would promote overexploitation, which entails an excessive and indiscriminate exploitation of marine resources by the industries. According to some participants, especially Chileans, companies engage in this "to generate more income, to make fishing more profitable. This is a problem because in reality they do not allow the maintenance of the sea's life cycle because they are interfering more than they should" . It's been identified that, regarding the purpose associated with the sea, Peruvian and Chilean participants tend to focus on the recreational and economic use, respectively. While Peruvians link the sea to social activities and aesthetics more often, Chileans highlight that the sea is mainly seen by their society as a source for economic benefit, even above its protection. In those lines, Peruvian participants attribute the lack of concern about marine protection to a representation of the sea that solely focuses on the positive emotions generated by its recreational purpose, whereas Chilean participants relate the diminished marine protection to an economic and social model that allows harming extractive practices in search of economic gain. --- Marine plastic pollution: socio-individual and structural causes and consequences The prioritization of the recreational and economic use of the sea above its protection might be linked to marine pollution, which for the participants represents the main marine environmental problem. Marine pollution is seen as a set of various elements: "solid waste dumped by boats or people, or perhaps because there is a sewer or a drainage system nearby. A person can be swimming or enjoying themselves, and suddenly find garbage, excrement, dead birds, plastic" . In this sense, several types of marine pollution were identified by Peruvian and Chilean participants: marine plastic pollution, the presence of plastic and microplastic in this environment, pollution by other materials, such as glass, debris, animal remains and drainage, oil spill pollution or other types of fuels, and pollution from construction near the sea that causes its deterioration and produces waste that ends up in the beach. Among all these types, marine plastic pollution is recognized by the participants of both countries as the most visible and relevant problem that the sea faces: "the first thing that comes to my mind is plastic pollution, and how it damages life, the life of the animals that live there. For me, it is the most serious problem since I have seen it the most" . Similarly, some participants report that the main pollution comes from the disintegration of plastic waste into microplastics in the sea or on the beach: "Whenever I walk along the shore I see plastic, bottles, small bottle caps… beach toys and, especially, microplastics, which are those tiny plastics that are almost the size of a grain of sand. I can see them and you can spot a lot of them on the shore" . It has been identified, based on the participant's perspective, that marine plastic pollution has socio-individual and structural causes. Within the socio-individual causes, the one most identified by the participants from both countries is the lack of knowledge, described as the limited information people have about marine plastic pollution and its consequences: "it has to do with the individual agency of people or small groups who are not aware of the effect that throwing plastic waste into the sea can have. Every little thing can increase the problem. They do not know the consequences" . Although, for some participants from Peru and Chile, scarcity of information is not the main cause of the problem, but the lack of responsibility towards marine protection. This means not taking responsibility for the behavior that pollutes the sea, even if they are aware of the damage it generates: "it's not that you do not know you have to clean up litter or that there are not enough garbage cans, and even if there are no garbage cans you can take it with you and then throw it away, but people are irresponsible and lazy" . This lack of responsibility, according to some participants from both countries, might stem from not associating their actions with the effects of marine plastic pollution: "there is no awareness that your action has repercussions, and if there is no awareness, you cannot grasp the logical connection; and since you do not see it and everyone else pollutes, you do not feel responsible" . It also, based on the opinions of mostly Peruvian participants, stems from selfishness as it leads people to prioritize themselves and their needs over marine protection: "people think: 'since I'm going to get out and I'm not going to come back, I can throw things here' [the sea], so they pollute. This is what we see happening the most. And because they think that it is not their habitat or that they are not going to come back, they pollute it and do not worry about it" . These individual causes might be related to one of the most reported, mainly by Chileans, structural causes that involve societal Frontiers in Psychology 08 frontiersin.org and systemic factors that allow marine plastic pollution: the lack of environmental education, which refers to the lack of visibilization and broadcasting of information about environmental protection and the scarce active participation in environmental initiatives. According to some participants, the lack of environmental education hinders the perception of the sea as a crucial ecosystem for humanity that needs protection: "if they told us in school that the sea is important. Nobody talks about these things; it is very invisible. If they would talk about it, it would be different, it would be part of us" . Another structural cause of marine plastic pollution, identified by some participants from both countries, is the massive consumption of plastic, the widespread use of plastic due to its easy access and production that ends up reaching the sea: "it is easy to acquire it and, therefore, there is also mass consumption. People tend to pack and carry their food in plastic when they go to public spaces such as the beach" . It should be noted that the lack of environmental education and the massive consumption of plastic are mainly attributed to users of the sea such as bathers and athletes, as well as coastal communities. In this regard, some Peruvian and Chilean participants recognize that another structural cause of the problem is poor waste management, meaning the deficient management of waste systems in both countries, which begins with the way people dispose of it and ends with how it's processed. The poor waste management, according to participants, can be attributed to a deficient handling of the problem by the authorities and companies: "at a systemic level, recycling is wrongly planned. It has been proven that, even if there are waste segregation containers, litter ends up all together in a sanitary landfill" . Based on the above, it has been identified that both socioindividual and structural causes are perpetuated over time due to a lack of regulation of plastic fabrication and consumption by the authorities, that consists of poor control and lack of sanctions established for individual, groups or companies' polluting behavior. In that sense, some participants hold their governments responsible for allowing the perpetuation of marine pollution: "in large industries, if control measures or rules are not set, they will never be aware of the issue. It also applies to people. If the issue is not taken seriously by authorities by setting rules against marine pollution, people will continue doing it. This also replicates with the fisherman, if the authorities do not tell them anything and allow them to do their fishing activities without conscience or care, the problems will continue" . The perpetuation of marine plastic pollution, based on the opinions of Peruvian and Chilean participants, has an ecological, social and individual impact. At the ecological level, it is recognized that plastic causes severe damage to nature, manifested in the ecosystem disruption and the impact on climate change. In one hand, the disruption of the ecosystem includes the negative modification of the interaction between marine species and their environment, due to the presence of plastic waste: "the alteration of the habitat of the beach, of the water, of the species that live there… it has been seen that where there used to be nesting birds, today due to plastic pollution, they no longer remain in that area" . On the other hand, plastic pollution is thought to be a catalyst for an environmental crisis that might have an impact on climate change: "I imagine that it is associated with the melting of the poles, the deterioration of the ozone layer, I think that marine pollution increases the problem" . At the societal level, the reported consequences of marine plastic pollution are the impact on the food chain, the scarcity of resources, and the difficulty in cleaning up the sea. According to the participants, the most relevant social impact would be the effects on the food chain, such as the ingestion of plastic waste by marine species that end up reaching humans through the food chain: "there are several studies that show that a great number of fish that we consume in the capital, that come from the sea, consume microplastics and therefore, when we eat them, we also ingest these microplastics" . This effect is related to resource scarcity, the risk that the resources provided by the sea will be depleted due to the presence of plastic in large quantities: "yes, because it affects all of us. The animals begin to die, we run out of fishery resources" . In this way, for some participants, this permanent presence of plastic waste through microplastics generates a difficulty in cleaning up the sea, and thus maintains the risks to the health and the future of societies. At the individual level, most Peruvian participants report that marine plastic pollution results in a loss of enjoyment of the sea, understood as the low satisfaction in its recreational use due to the presence of plastic waste: "for athletes, it is not pleasant to enjoy a sea full of plastic" . Since the recreational use of the sea is seen as one of its main purposes, the lack of it might trigger negative emotions in some of the participants, making them more aware of other negative effects of marine plastic pollution. In fact, most participants have reported experiencing emotions such as anger, sadness and hopelessness. Anger refers to the feeling of being upset about marine plastic pollution: "sometimes it makes me angry, because I see it as something unfair, as something that should not be" . On the other hand, sadness refers to the feeling of sorrow generated by seeing how pollution affects marine life: "sadness because due to us, animals that do not know how to defend themselves are directly affected" . Participants also experience hopelessness because they perceive that other people, institutions and their authorities do not and will not take responsibility for dealing with the problem: "Sometimes the problem is huge for one person or a specific group to take charge. And one feels impotence to see how different companies that are mainly responsible for this pollution, due to plastic production, do nothing. I think: what am I going to do? what can I do?" . This hopelessness is associated with the participants' assessment of the actions they take in response to marine pollution. Most of them perceive that they have limited self-efficacy in response to the problem: "more than anger, this situation makes me feel sorrow, and I try to do something about it. Yes, we try to educate as much as we can, but we cannot control everything, we cannot solve everything" . In this sense, especially for Peruvian participants, there is a concern for future generations, understood as the concern produced by the belief that other generations will not be able to benefit from the sea due to pollution: "we can survive thanks to the sea and future generations also will. One time I heard: "I hope climate change… will be in 100 years, " but in 100 years other people will be alive. There will be people who will not be able to have these moments of enjoying the sea" . However, there is some hope left as a group of participants report feeling a need to protect the sea, which refers to the communities' will to take actions to address marine plastic pollution and take care of the sea: "more than emotions, it's the need to protect. That feeling that made me feel annoyed with the fact that it was dirty, untidy, we turned it into something positive. In the end it was like a manifestation of unification, we started to protect" . The willingness to turn negative emotions towards marine plastic pollution into collective efforts for marine protection, although reported by a reduced number of participants, might be a useful path to promote pro-environmental behaviors. As presented above, Peruvian and Chilean participants have a general knowledge of the causes and consequences of the problem that, although is superficial, reflects a general scope of individual and societal factors and stakeholders at play in perpetuating marine plastic pollution. This might be influenced by their college education; however no causal association can be made. It is relevant to highlight that Peruvian participants tend to focus more on individual causes and consequences rather than structural ones, in comparison to their Chilean counterparts. This pattern seems to repeat in their proposals for solution alternatives. --- Individual and structural solution initiatives for marine plastic pollution Finally, the participants suggested different individual and structural initiatives that they have or intend to have carried out or think that might be implemented by various stakeholders to reduce marine plastic pollution . Regarding individual initiatives, the majority of participants from both countries recognize the importance of reducing plastic consumption, such as avoiding the use of plastic in daily life and switching to biodegradable products. Thus, they believe that "we can go shopping with our reusable cloth bags, so that people could appreciate its importance, I mean, there are people doing it, and it is not that complicated" . This initiative could be useful to reduce the massive consumption of plastics, but it is stressed that to achieve this, the authorities and companies need to get involved in offering sustainable alternatives. Another alternative that could be implemented with the collaboration of these stakeholders is beach cleanup, understood as picking up litter from the beach. Participants from both countries consider this as a useful initiative since it contributes, slowly, to the reduction of marine plastic pollution: "yes, it is slow, but effective. Although, with the support of some organization or the government, it could be more effective. People do it purely for the environment" . Despite the perceived contribution of this action, some people, specifically Peruvians, consider it ineffective in the grand scheme of the problem, as it does not address the causes of marine plastic pollution: "it is not targeting the root cause of the problem itself, which is how plastic enters the sea not the cleaning of it" . However, the general consensus is that beach cleanups might increase awareness about the relevance of marine protection. Although most of the participants expressed not having been engaged in these initiatives, some of them, mostly Peruvian participants, report doing the following pro-environmental behaviors: picking up litter, understood as cleaning up their own or others' waste when being on the beach, and calling out other people for disposing their waste on it by disapproving their behavior and sharing information about the consequences of the problem: "we keep a close eye on the beach goers to see if they litter… If they try to hide the litter, we intervene to guarantee the correct waste disposal, requesting them to 'take their litter home, as they came here without it, and found the beach litter-free. Let us leave it as we found it'" . Nonetheless, most participants recognize that their reportedproenvironmental behaviors are few and far between, thus they manifest having intentions to participate in more meaningful pro-environmental behaviors, such as recycling. Recycling involves collecting, sorting, and classifying plastics so it can be reused or taken to recycling centers to be managed or transformed. For those rare participants that do it, recycling might be a manifestation of a society's increased concern for the environment: "at my house, we all recycle. Once my car is full, I take it to the recycling center. It's satisfying to see the center's containers filled with recyclable materials. This indicates the occurrence of a cultural shift" . In that sense, for particularly Chileans, the purpose of the individual initiatives is to make marine plastic pollution more visible. This involves educating themselves about the issue, but also sharing it with others, raising awareness about it: "we must take action, protect our beaches and seas, educate those with little awareness, speak out when necessary, and actively participate in this process" . These individual level solutions must be supported by structural proposals with political impact. In this regard, for participants from both countries, a key initiative to take is environmental education, focused on promoting and broadcasting information about environmental protection with the involvement of the government and the citizens, in order to mitigate the lack of knowledge and the lack of responsibility towards the problem: "If in my community, I learn that plastic is harmful to the planet, I will refuse to use it, and if I see it I will try to recycle or reduce it. I will try to make a change, but if I have not been educated about plastic pollution, it would be hard for me not to do it" . In addition, participants propose structural initiatives that should be implemented as part of a collective involvement, understood as the joint participation of private companies, government institutions and citizens in generating pro-environmental significant change. In the case of companies, some participants, especially Chileans, consider that businesses should focus on reducing plastic manufacturing and replacing it with sustainable alternatives: "here in Chile there was also a bill in which supermarkets stopped giving out shopping plastic bags, so every time we went to the grocery store, we had to know that we should bring our reusable bag" . Furthermore, some participants from both countries believe that companies should take actions and implement policies to reduce their environmental footprint. Regarding the structural initiatives that should be taken by the government, Peruvian and Chilean participants mainly suggest the establishment of laws and sanctions. This implies the elaboration of norms that regulate and punish people's polluting behavior. In other words, the authorities need to "enter a little more into politics, create new laws, establish regulation, and take strict measures to deal with cases of pollution" . It is important to highlight that, for a small group of Chilean participants, the Constitutional Convention, the body in charge of drafting the new constitution in Chile, should support this initiative. Also, most of the Peruvian participants consider significant the involvement of the authorities in improving waste management. They believe that "those in the government have to make a good recycling plan, manage the waste, 10.3389/fpsyg.2023.1308796 Frontiers in Psychology 10 frontiersin.org collect it from the whole district and city, gather it in a place where it will be buried, burned, and segregated between plastic and organic waste" . According to mostly Chilean participants, these pro-environmental initiatives have been undertaken by civil society rather than the government or private companies, thus they believe that this stakeholders need to finance and promote pro-environmental behaviors since its their responsibility to do so: "initiatives are achieved more at the local level than at the governmental level, there should be support for self-managed organizations from municipalities or the government because it is something they are supposed to do, but we do it for them" . Faced with a limited engagement of the authorities in addressing marine plastic pollution, some participants propose that citizens increase their political involvement, which entails a political participation in favor of environmental protection by selecting authorities who support it, and organizing the community for protest: "just as they call to protest for certain social issues, they should protest for environmental issues. We and those to come are the ones who will be the counties' leaders" . Along with this political involvement, some participants, specifically Chileans, believe that academic research should be encouraged, which implies carrying out research with the purpose of learning and informing citizens and other relevant social stakeholders about the causes and consequences of the problem in order to develop possible solutions. In those lines, some participants from Peru and Chile recognize that the main purpose of pro-environmental initiatives should be the revalorization of the sea, which entails the acknowledgement of its intrinsic value and the prioritization of marine protection at the societal and governmental level. For this, a joint effort is needed: "on one hand, individuals need to stop polluting marine areas, on the other hand, raising awareness within large companies is necessary to make them understand the significance of the sea; generating a collective environmental concern that can transcend in time" . Therefore, although Chileans participants tend to focus more on the role that the government and private companies have in establishing structural initiatives than the Peruvian participants who mainly highlight the individuals' responsibility, there is a general consensus that to effectively address marine plastic pollution various stakeholders must work integrally in generating, institutionalizing and participating in individual and structural initiatives that pave the way for a cultural change towards marine protection. --- Discussion Peruvian and Chilean citizens represent the sea as a space that provides refuge, disconnection from routine and, overall, connection with nature. This is based on affective experiences associated with being close to the sea , such as tranquility and happiness. In this way, the sea acquires positive value from the meaning that the participants build around it . --- An anthropocentric and neoliberal view of the sea According to some studies, having a connection with nature might generate a greater environmental concern and a commitment to protect the environment . However, in line with studies that found a gap between pro-environmental attitudes and pro-environmental behavior , our results suggest that although the positive perceptions of the sea might encourage intentions of engaging in pro-environmental behaviors, in the case of most participants these intentions do not translate into pro-environmental behaviors, especially if it involves greater effort. Participants acknowledge that even if they feel connected to the sea, they do not engage in collective actions aimed at reducing marine plastic pollution in a systematic way. Even those who report having engaged in pro-environmental behaviors refer to an individual level of action, such as picking up litter from the beach. This distance between the participant's environmental concern and their behavior could be due to the fact that their positive perception of the sea is more linked to the recreational benefits it brings them rather than to a commitment to its protection regardless of the costs or benefits. The recreational use, in addition to fulfilling the cultural function of providing a space for aesthetic manifestations and spiritual wellbeing , 2014; Thushari and Senevirathna, 2020), allows enjoyment through leisure and social interaction. The prioritization of this recreational use of the sea, according to the participants, leads people to not care about it beyond the experiences it provides them. In this case, the concern about plastic pollution could be linked to selfish values, which influence their decision to engage in pro-environmental behaviors according to the cost or benefit related to it . Therefore, people tend to engage in pro-environmental behaviors that represent less costs from them, such as occasionally picking up litter and calling out pollution behavior rather than reducing the use of plastic or recycling, while still benefiting from their connection to the sea. The economic use attributed to the sea fulfills an extractive function, as this ecosystem is seen as a supply of food and other natural resources , 2014; Saavedra and Mardones, 2021). Participants from both countries report that, as a result of this attribution, overexploitation represents one of the main problems that damages the sea and has repercussions on its pollution. This perception of the sea as a vehicle for extractive and economic profit is associated with the neoliberal model, defined by the participants, especially Chileans, as the economic, political and social model that prioritizes the exploitation of resources and consumption. The neoliberal model would be evidenced in the decision of the authorities of both countries to apply laws on the manufacture and use of single-use plastics to specific locations, such as tourist areas, so that they do not generate changes to business sectors of greater economic interest for the country, such as small and medium companies in Peru , and aquaculture industries in Chile . The passing of the law in tourist locations might respond to the fact that tourism in coastal regions is an economic activity that relies on the recreational use of the sea, which is harmed by plastic pollution in it; thus, in the authority's perspective, marine protection in coastal regions would be aligned with the economic interests. Conversely, laws that prioritize marine protection over profit in regions whose industries focus on the exploitation of marine resources might be perceived by the authorities as threats to the economic interests of the population, even though different civil society organizations and, in the case of Chile, local authorities have been making efforts to bring to their attention the consequences of plastic pollution . A similar situation has been found in contexts whose economic model is neoliberal: the proposed solutions to the environmental crisis are rejected, as they challenge the neoliberal system that responds to social sectors that seek the greatest possible profit, even if this results in the government's inability to address environmental issues . Hence, the decision of authorities to apply pro-environmental policies only in sectors where this would be beneficial to their economy could be influenced more by altruistic values than by biospheric values, given that the main interest is in protecting the economic welfare of the population and not in marine conservation . In that line, we propose that the altruistic values that might be behind the authority's decisions correspond to the preservation of the economic interests of today's pollution rather than concern for future generations. The lack of addressing marine plastic pollution consequences on a larger scale by the authorities might be the reason why a few of the participant's report being concerned for future generations, and why these reports have come from the Peruvian sample, since Peru, in comparison to Chile, has been less in favor of prohibiting single-use plastics . Therefore, the environmental concern about marine plastic pollution corresponds more to, what the New Ecological Paradigm calls, the anthropocentric perspective, which is focused on the superiority of human beings over nature. This perspective directs environmental concern towards the consequences that the environmental damage has for the individual or for society . Thus, it was found that the concern of the participants falls more on the social and individual consequences of the problem, such as the loss of enjoyment of the sea or the impact on the food chain, than on the ecological ones, like the disruption of the ecosystems. Under that anthropocentric and neoliberal perception of the sea, the environmental attitude is characterized in this study, at a cognitive level, by the lack of information about the problem, its consequences and its relationship with human behavior. At the affective level, marine plastic pollution causes negative emotions, such as sadness and annoyance, next to hopelessness, which is triggered by the acknowledgment of the magnitude of the problem and the negative environmental attitude of other citizens. At the dispositional level, participants recognize that their societies have a low predisposition to marine protection, manifested in the high consumption of plastic, due to its easy manufacture, use and disposal , in the poor waste management, and in the lack of oversight and establishment of plastic legislation to regulate the industries' bad practices . In this context, both Peruvians and Chileans present limited individual efficacy of response, i.e., they perceive that their pro-environmental behaviors are insufficient to solve the problem. The latter would explain why the participants, despite having a connection with the sea, do not consistently maintain a commitment to protect this environment, limiting their actions to an individual, loosely organized and low-impact level . This limited individual efficacy of response could be understood by the proposed model of socialization of pro-environmental behavior , which depends on the interaction and interdependence between the different systems in which a person develops. Thus, the individual pro-environmental behaviors that the person carries out in the different systems to which they belong must be integrated with municipal or national actions and must be supported by values of marine protection and conservation from the society to which they belong in order to achieve the desired impact. This environmental attitude is linked by the participants to two causes of the problem: selfishness and overexploitation. These causes are part of the macrosystem, that is, the system of beliefs and values of the societies to which the participants belong and which influence individual and institutional behavior. In the Peruvian case, marine plastic pollution is attributed to individual characteristics such as selfishness, which is usually related to individualism. The latter is seen negatively in collectivist societies such as the Peruvian one, as it prioritizes individual development and well-being over the collective ones. The Peruvian participants identify this prioritization of individual well-being over collective well-being as selfish acts of people in the face of the problem. In the Chilean case, the participants recognize a link between the overexploitation of the sea and the neoliberal model, as they report that industrial fishing generates large quantities of plastic waste into the sea. They also point out the prioritization of the consumption of plastic products over marine protection. It is important to note that both causes are interrelated and are part of the neoliberal model to which both societies belong. Peruvian participants tend to identify selfishness as a trait of their society, while Chilean participants tend to link overexploitation to the country's economic system. This more critical and structural view of the problem in the Chilean participants could be related to the political landscape that Chile is going through and to the more active political involvement of its citizens . Neoliberalism as a social and economic system has disarticulated the relationship between government and civil society . In Peru, over the last decades neoliberalism brought economic growth at the expense of increasing informality, inequality and institutional distrust, which caused civil society to disengage from its political role and to not question the economic and political system . In Chilean, although the neoliberal economic system has been questioned, there is still no commitment to change due to the fact that Chilean society maintains its tendency to legitimize social hierarchies . This social dominance orientation has been found in other studies to be negatively related to pro-environmental attitudes and environmental policies . --- What it's being done: individual actions and local policies Based on the above, two of the four levels of pro-environmental behavior are mainly identified : individual actions and local level policies . Individual actions are reduced to beach clean ups, picking up litter and reprimanding polluting behaviors, which reflects a general and limited view at possible solutions to address the problem, since, instead of addressing the causes, they are aimed at mitigating the effects. Despite this, for the participants, mainly Peruvians, such actions are feasible to implement in practice and promote the visibility of the problem. It should be noted that this type of initiative is usually developed in immediate contexts where the individual belongs and participates actively, such as volunteer, family or friend groups, which correspond to the personal microsystem . In this sense, the attitudes towards marine plastic pollution held by systems closest to the individual influence their attitudes towards the problem, and vice versa. Reducing plastic consumption and recycling are other individual initiatives aimed at mitigating plastic pollution and informing about its consequences, which, according to participants, require the collective involvement of citizens, businesses and government institutions. While participants report having the intention to engage in these individual actions, they recognize that the involvement of social stakeholders and institutions involved in policy making and with the power to take pro-environmental behaviors to a more institutional level is needed. This, according to Bronfenbrenner's model, would correspond to the interaction between the behaviors involved in the mesosystem with the decisions, actions, policies or collaborations that can be made among various social stakeholders at the exosystem level. It is proposed that for individual actions to have an impact outside the individual, family or community sphere, it is necessary for various stakeholders and institutions to support, promote and replicate these actions through the implementation of local-level policies. In this way, the reduction of plastic consumption and recycling, articulated with an improvement in waste management, would provide greater institutional support to the objective of reducing marine plastic pollution. Moreover, academic research, which can take place in the exosystem, would allow the elaboration of public policies based on scientific evidence, under which authorities could establish laws and sanctions regarding the problem. Although local policies would drive concrete actions, these represent a necessary step to achieve a profound change in society characterized by the establishment of a system of beliefs, policies and values focused on environmental protection . Even though, in Chile some initial steps have been taken to involve different stakeholders in addressing plastic pollution , this would be difficult to develop and/or improve in both countries, since, according to the participants, there is a low commitment from the authorities to protect the environment at a more structural level. --- What needs to be done: public policies and environmental belief system As can be seen in Figure 3, from the point of view of the participants, the proposed solution initiatives at the individual level are inefficient for reducing marine plastic pollution, since they only influence the microsystem. In this sense, participants believe that ideally these actions should be complemented by structural initiatives that involve various stakeholders. Structural initiatives, seen in Figure 4, belong to the exosystem and macrosystem levels, in which a set of local and public policies together with a social belief system would reduce marine plastic pollution. It should be noted that the next two levels of pro-environmental behavior, a set of public policies promoted by interest groups and a culture of environmental protection , are less recognized to be executed in the participants' perception. These structural levels of pro-environmental behavior play the role of sustaining individual actions and local policies . An initiative known by the participants, especially Chileans, that corresponds to the public policy level is the support of pro-environmental initiatives. Although it is identified as a possible solution, most participants have little knowledge of initiatives focused on marine protection, such as "Hazla Por Tu Playa" or "Científicos de la Basura" . Even though, specifically in Chile an informal alliance has been established between different stakeholders to take action against plastic pollution , in the perspective of the participants of both countries there is a weak national presence of environmental activism articulated with governmental and non-governmental institutions. In fact, these initiatives tend to arise from the concern of citizens and not necessarily from the governments, which would demonstrate why the participants do not acknowledge the practical implementation of structural measures, even though they are considered important to solve marine plastic pollution. According to Bronfenbrenner's ecological model, this type of public initiatives would correspond to the exosystem, since they are spaces in which the individual does not necessarily have a direct participation, but indirectly influence their behavior. Thus, despite the fact that most of the participants are not actively involved in an environmental activist organization, the activities that these types of entities carry out in favor of marine conservation have repercussions, in some way, on the environmental actions of the individual. However, the latter would not be possible if there is not, in turn, a system of beliefs or values that promote environmental protection, which corresponds to what is known as the macrosystem . That is, the cultural elements of the macrosystem have influence on the lower order systems, so, as other studies have suggested , if this system does not promote a cultural proposal that encompasses positive values and attitudes in favor of the environment, the development of pro-environmental behaviors at the other levels will hardly take place. This is because the macrosystem influences the way in which individuals behave within their social context . However, it can be observed that there is not an acknowledgement of more wide-ranging levels of pro-environmental behavior, beyond individual actions and, to a lesser extent, policies at the local level. In that sense, there is also no acknowledgement of a belief system rooted in society that promotes environmental protection . On the contrary, it dominates a culture of exploitation, supported by a neoliberal system, which serves the economic interests of the authorities, private companies and plastic industries. In this context, a measure that, according to the participants, can confront this extractive predisposition and promote a culture of protection is environmental education, supported by the respective authorities and internalized by the general public. Environmental education programs applied in Chile, with high school students, had a positive impact on the development of greater environmental awareness and environmental knowledge about the effects of debris in marine environments . These interventions would not only increase positive environmental attitudes towards the sea, but also promote pro-environmental behaviors, such as improved plastic consumption practices . In Peru, although there is no evidence of successful environmental education programs, awareness campaigns led by environmental activist organizations, such as "HAZla por tu Playa, " have promoted significant pro-environmental behaviors. This organization has managed to convene more than 10,000 volunteers to carry out more than 1,235 beach cleanup campaigns nationwide and more than 52 awareness-raising campaigns, increasing environmental awareness in the participants and encouraging them to improve their plastic consumption habits and veer into a sustainable one . Therefore, this study proposes that environmental education should include cognitive, affective and behavioral components that express a predisposition to act in favor of the environment. In this sense, it is necessary to generate greater knowledge in the population through the visualization of the causes and consequences of marine plastic pollution and the various strategies to address it . Participants' perception of structural solution initiatives for marine plastic pollution. Elaborated by the research team. Saavedra et al. 10.3389/fpsyg.2023.1308796 Frontiers in Psychology 14 frontiersin.org Furthermore, to encourage environmental concern for the effects of plastic pollution on humans and nature, a connection with nature and a sense of responsibility towards the marine ecosystem must be nurtured . Finally, it is crucial to supply sustainable alternatives to replace plastic products and promote effective waste management, such as recycling . Integrally, this type of education should be transversal throughout the individual's development, in order to promote values, beliefs, attitudes and practices focused on environmental protection. Thereby, environmental sustainability and social welfare might be achieved. --- Conclusion In synthesis, the environmental representation of the sea, characterized mainly by the attribution of recreational and economic purposes, orients environmental concern towards the consequences of marine plastic pollution on individuals and society. This would reflect, according to the New Environmental Paradigm , an anthropocentric perspective of the problem, since environmental concern about the sea would be based on the value it has for the maintenance and improvement of the quality of human life . In this sense, instead of engaging in pro-environmental behaviors based on biospheric values that assign an intrinsic worth to nature, the various social stakeholders will behave according to selfish values, which emphasize the cost and benefit for themselves, or altruistic values, which prioritize, in this case, the economic interests of society. Furthermore, it is found that the participants' pro-environmental behaviors correspond to individual actions, developed in their personal microsystems, whose objective is to generate visibility and awareness of the problem in the closest social groups. It is recognized that for these actions to have an impact that transcends family, work or friendship contexts, a collective involvement with companies, institutions and academia is needed to take these pro-environmental behaviors to a structural level. In this sense, the pro-environmental behaviors carried out individually or collectively, corresponding to the microsystem and mesosystem, should be articulated with local, national and international policies established, promoted and executed by various social stakeholders such as companies, governmental and non-governmental organizations present in the exosystem. However, this articulation between individual actions and public policies might be hindered by the neoliberal system which promotes a culture of exploitation of the sea, attributable to selfishness, in the Peruvian case, and overexploitation, in the Chilean case. Thus neoliberalism, as a social and economic system that prioritizes economic gain over the environment and obstructs pro-environmental behaviors , might be a core factor in the perpetuation of marine plastic pollution. Although both countries have adopted this system and currently deal with its social consequences, it's relevant to highlight that only the Chilean participants verbally acknowledge its possible influence on the causes of marine plastic pollution and the authorities' lack of commitment towards resolving it. Therefore, there is a need to build and promote a culture of marine protection that encourages citizens, academia, industries and authorities to adopt more environmentally sustainable forms of consumption and production, in order to reduce the use of plastics and, ultimately, plastic pollution. To achieve this, starting from the anthropocentrism that prevails in society, it is suggested to use selfish values, linked to leisure and resource extraction, to promote marine protection in order to preserve its recreational and economic use. Upon this, it is proposed to promote altruistic values that re-orient marine protection in favor of the maintenance and improvement of the quality of life of the individual and society. This to, finally, develop biospheric values under which harmony is sought between the welfare of society and environmental protection, so that environmental concern is oriented towards the intrinsic value of nature, beyond any anthropocentric interest. In oversight, this study contributes to literature by integrating two theoretical models, Hofmann and Graham's levels of analysis of pro-environmental behavior and Brofrenbrenner's model of socialization and assessing the role of each level of analysis into the systems of socialization to analyze the environmental attitudes and behaviors of Peruvian and Chilean participants, in light of possible influences of political, social and economic factors, like the neoliberalism system. Regarding the limitations of the study, it should be noted that participants were mostly urban residents, which hinders the opportunity to assess if the proximity to the sea has any influence on the connection with nature or positive perceptions of the sea. Likewise, another limiting factor is that during the analysis of the results no differences were found based on demographics, since most of the participants shared similar social, economic and education backgrounds. In addition, the qualitative study design did not allow further generalizations of the findings. Lastly, it is recommended to carry out research with strategic social stakeholders involved in decision making, such as authorities and entrepreneurs in the plastic industry. In addition, research should be expanded to explore the impact that plastic consumption habits have on marine pollution. --- Data availability statement The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author. --- Ethics statement Ethical review and approval was not required for this study involving human participants in accordance with the national legislation and institutional requirements of Pontificia Universidad Católica del Perú, institution responsible for this project. The studies were conducted in accordance with the corresponding local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants' legal guardians/next of kin in accordance with the corresponding national legislation and institutional requirements. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. --- Supplementary material The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1308796/ full#supplementary-material

Marine plastic pollution remains one of the greatest problems worldwide. Hence, this study explores the attitudes and pro-environmental behaviors of Peruvian and Chilean citizens regarding marine pollution, with an emphasis on plastic pollution. For this, forty-four semi-structured interviews were conducted with Peruvian (n = 24) and Chilean (n = 20) citizens, of which, through thematic analysis, semantic patterns were identified. Results show that the participants' representation of the sea is positive and related to the connection they report having with this environment. Additionally, it was found that the sea fulfills recreational and economic purposes, reflecting an anthropocentric perspective, since it is associated with leisure and resource extraction, respectively. Both purposes are related to the causes of plastic pollution, although with differentiated effects. Anthropocentrism is also reflected in the direction that environmental concern takes towards the impact of this type of pollution predominately on individuals and society. Regarding pro-environmental behaviors, most of the initiatives proposed by the participants in response to marine plastic pollution correspond to individual actions, which could be due to the fact that they perceive a low commitment level from authorities to address the problem. In particular, the Chilean participants attributed a greater role to their authorities in dealing with plastic pollution, which would indicate a more institutional perspective of the problem. Thus, it is proposed that to address marine plastic pollution it is necessary to articulate individual actions with public policies carried out by social stakeholders such as governments, companies and non-governmental organizations, in order to build a more efficient culture of marine protection. KEYWORDS marine plastic pollution, representations of the sea, pro-environmental attitudes, pro-environmental behaviors, anthropocentrism

Introduction Smoking during pregnancy is recognized as one of the most important preventable risk factors for adverse pregnancy and birth outcomes. Smoking is associated with fetal growth restriction, stillbirth, premature delivery, premature rupture of membranes, and sudden infant death [1][2][3][4]. Besides increasing the risk of miscarriage, ectopic pregnancy, and placenta previa, smoking during pregnancy can also predispose the offspring to behavioral and cognitive disorders, overweight, obesity, respiratory diseases, reduced lung function, and tobacco addiction [2,3,[5][6][7]. In addition to its direct effect on perinatal outcomes, smoking during pregnancy has a role as a potential mediator of socioeconomic differences in these outcomes [8][9][10]. It is strongly associated with socioeconomic condition [11], being more prevalent among less educated women [12][13][14][15][16][17] and those in the lowest income group [13,18]. Most studies analyzing time trends in smoking during pregnancy have reported a decrease in prevalence in any period studied [15,[19][20][21][22][23][24][25][26][27][28]. Yet, the downward trend of smoking prevalence has varied by levels of socioeconomic condition among pregnant women, with smaller decreases among those in the lowest level [26,28]. Smoking cessation during pregnancy is as expected higher among high-income pregnant women [20,22,25,29]. However, studies of time trends in smoking cessation are scarce with inconsistent findings showing increase [20], decrease [25,30], and stability of prevalence [22]. Time trends in prevalence of smoking cessation during pregnancy have been little investigated, and no studies could be found of trends in socioeconomic differences in smoking cessation during pregnancy. In addition, there is a scarcity of population-based studies of smoking and smoking cessation during pregnancy in Brazil. This study aimed to examine the time trends in prevalence of smoking before and during pregnancy and in prevalence of smoking cessation during pregnancy in 2007, 2010, and 2013 by family income quintile, maternal level of education, maternal skin color, and maternal age. --- Methods A cross-sectional study is carried out in Rio Grande every three years to assess pregnancy and childbirth care among newly delivered women living in this municipality. Rio Grande is located in southern Brazil, about 300 km from Rio Grande do Sul state's capital Porto Alegre, and it has a population of about 210,000. The present study was based on data collected in three surveys conducted in Rio Grande in 2007, 2010, and 2013. The target population was all women living in the municipality who gave birth in all maternity units in the city between January 1 and December 31 in each year studied. Given that over 99% of deliveries take place in the municipality's hospitals, these are census surveys; that is, they include all births taking place in the municipality during the study period. Newly delivered women of newborns born before the 20th week of gestation and/or weighing less than 500 g were excluded from the study. All information here presented was collected using a standard questionnaire administered by trained interviewers. Interviewers attended a 40-hour training about a month before they began working on data collection each year of survey. The training sessions included questionnaire and instruction manual reading and simulation interviews. We carried out a pilot study in the maternity hospitals to be included in the study and tested the questionnaire and data collection logistics. Female interviewers paid daily visits early in the morning to the municipality's two maternity hospitals. They checked admission records for any birth delivery that took place the day before and then double-checked this information against the hospital's Medical Statistics Service Database . When there was a record of a birth the day before, they made a note of the mother's name and proceeded to the postnatal ward. The interviewers approached the new mother and confirmed she lived either in the urban or in the rural area of Rio Grande. They then described the study and read a free and informed consent form and invited the mother to participate. Those who agreed signed two copies of the consent form and they were given one copy for their records. After the interview, questionnaires were coded and sent along with a copy of the signed consent form to the study site. All consent forms were filed at the study site and all questionnaire answers were reviewed and coded for data entry. The data were double-entered into Epidata 3.1 by different staff; they were entered in the second time in the reverse order. We conducted data entry checks for every set of 100 questionnaires and corrected any discrepancies. After completing data entry, a single file was created containing the data of all mothers and their newborns. All information about smoking before and during pregnancy was collected in standard manner in all three surveys. "Smoking before pregnancy" was defined as having smoked at least one cigarette a day in the six months prior to the current pregnancy. "Smoking during pregnancy" was defined as smoking at least one cigarette a day every day during at least one trimester of pregnancy. "Smoking cessation during pregnancy" was defined as having smoked at least one cigarette a day in the six months prior to the current pregnancy and not smoking throughout the current pregnancy. The exposure variables included maternal age ; maternal self-referred skin color , or black); maternal level of education ; and monthly family income . Fieldwork supervisors carried out quality control by performing phone interviews with a reduced sample comprising 10% of the participants. The kappa coefficient varied according to the characteristics studied and survey year. We calculated the coefficient of agreement for 24 questions and it ranged from 0.61 for "reason for cesarean section" to 0.92 for "type of delivery. " For most other questions, it was greater than 0.70, which is considered satisfactory. In addition, birth records at the maternity hospitals were examined by the interviewers on a daily basis and by the supervisors on a monthly basis. We used descriptive statistics to calculate proportions and related 95% confidence intervals for the variables studied. The chi-square test for linear trend in proportions was performed to examine time trends for each survey and for the entire study period. The chi-square test of heterogeneity was conducted to assess differences across categories within each survey as well as in percent decreases during the study period. All the statistical tests were two-sided, and the significance level was set at 5%. The analyses were performed with Stata 12 software . The survey studies were approved by the Universidade Federal do Rio Grande Research Ethics Committee and the A.C. Santa Casa do Rio Grande Heath Research Ethics Committee. All participants signed a free and informed consent form before the interviews. They were also informed about their right to decline participation and their right of confidentiality respect to any information obtained in the course of the research project. --- Results 7,572 women were interviewed in the three surveys: 2,540 in 2007, 2,379 in 2010, and 2,653 in 2013. The rates of losses and refusals were 1.3% in 2007, 2.8% in 2010, and 2.3% in 2013 . To calculate sample loss and refusals we used data from the National Live Birth Database as gold standard. SINASC gathers birth information by mother's place of residence from all Brazilian municipalities. Most mothers were 25 years or older and self-referred as white and had nine full years of schooling or more . Table 2 shows that overall prevalence of smoking before pregnancy fell from 27.9% in 2007 to 22% in 2013, a 19.7% decrease . A downward trend was seen in smoking prevalence among women with high-income level , 9-11 years of schooling , and white skin color and in almost all age groups except for 25-29 years old. In all three surveys, smoking before pregnancy was generally more prevalent among women in the lowest income quintile , less educated and self-referred as black. Table 3 shows that overall prevalence of smoking during pregnancy fell from 22% in 2007 to 18% in 2013, an 18% decrease . In all three surveys, smoking was generally more prevalent among women in the lowest income quintile . When we examined smoking prevalence by family income between 2007 and 2013, we found that this decrease was not uniform across income quintiles: there was a 17% decrease in the lowest quintile while it was twice as much in the highest quintile . The largest percent decreases were observed among women in the highest quintile , of white skin color and younger . Table 4 presents the prevalence of smoking cessation during pregnancy. There was a slight reduction during the study period, though not significant . In all three surveys, smoking cessation was generally more prevalent among more educated women and those who were in the highest income quintile . Regarding level of education, in all three surveys, there was an increasing trend in smoking cessation as the level of education increased . The same trend was observed in all three surveys by income quintiles as smoking cessation increased with family income. --- Discussion This study found a significant reduction in the prevalence of smoking before and during pregnancy between 2007 and 2013, which was more pronounced among younger women, of white skin color, in the highest income quintile. Smoking cessation during pregnancy was more prevalent among more educated women and those who were in the highest income quintile. One of the limitations of this study lies in the fact that smoking was self-reported by the mothers. Despite the potential underestimation of smoking prevalence and the potential overestimation of smoking cessation during pregnancy, there is no reason to believe they occurred differently in the three surveys studied. Nevertheless, self-reported smoking status is the most widely used method for assessing smoking in population-based studies. While self-reported smoking is recognized as an appropriate measure to assess exposure to tobacco during pregnancy [31,32], some authors have argued it may underestimate the prevalence of smoking [33][34][35] and affect estimates of the effects of smoking [36]. Researchers have suggested biochemical markers for measuring tobacco exposure in pregnant women, but the prevalence of concealment of smoking status may vary depending on the choice of cutoff points [37]. The use of biochemical markers is costly and requires adequate cutoff points for pregnant women. Variations in nicotine metabolism need to be taken into account when setting up cutoff points for this specific population [38] and these cutoff points have to reliably differentiate intermittent from passive smokers [39]. Recent populationbased studies conducted in Norway [31] and the US [40] showed that, among women who reported not smoking during pregnancy, only 2% showed cotinine levels consistent with active smoking. Another limitation of this study is its cross-sectional design that may give rise to recall bias. In order to minimize recall bias, specific questions about smoking and smoking cessation were asked at different times between the six months before pregnancy until the last trimester. The objective was to attain a higher consistency in the response, as the interviewee answered several questions on the subject. We believe that this brought more reliability to the information collected, because all of the answers should have been consistent. This differs from asking a single isolated question about the issue, in which case it would not be possible to evaluate inconsistencies. The operational definition of what constitutes smoking is a major concern in both public health and clinical practice settings. This issue is of particular importance in population-based studies because smoking and smoking cessation are characterized using low-cost straightforward approaches including few questions. However, there is a lack of a consistently applied definition of smoking and smoking cessation during pregnancy in epidemiological studies. In this study, we applied the definition of smoking during pregnancy as proposed by Santos et al., 2008 [28], that is, smoking at least one cigarette a day every day during at least one trimester of pregnancy. As for smoking cessation, we did not find any other study in the literature that used the same information we collected in our study in the operational definition of "smoking cessation. " Therefore, we developed the criteria presented in our manuscript. Yet, the different definitions of what constitutes smoking and smoking cessation during pregnancy must be taken into consideration when comparing studies. In contrast to the reduction in smoking before pregnancy observed in our study, previous studies reported that the prevalence of smoking 3 months before pregnancy was stable at around 26% [20] and 22% [22]. The downward trend in prevalence observed among women with higher income , 9-11 years of schooling, and white skin color and in almost all age groups except for 25-29 years old suggests some women of reproductive age are quitting smoking to be prepared for pregnancy or even a reduction in the smoking initiation among those women. Similarly to that reported in previous studies, we found higher smoking prevalence before pregnancy among less educated women [20] and those who were in the lowest income quintile [18]. A reduction in the prevalence of smoking during pregnancy is consistent with that reported in most other studies [15,[19][20][21][22][23][25][26][27][28] despite socioeconomic differences between countries. The 18% decrease in the prevalence during the study period corresponded to a 3% annual decrease and was similar to that observed in New South Wales, Australia, between 1994 and 2007 [26]. However, a study conducted in Maine, US, from 2000 to 2010, reported a 0.02% annual increase in the prevalence of women who smoked in the last trimester of pregnancy [41]. We found in our study that the downward trend in prevalence of smoking during pregnancy was significant only among women in the upper and lower income quintiles. However, similarly to that reported in previous studies [26,28], it was not uniform across the categories studied, with marked reductions among high-income women . This finding conforms with the predictions of a model of the cigarette epidemic in developed countries that was proposed by Lopez et al. in 1994 [42]. According to this model, individuals of higher socioeconomic condition are more susceptible to the influence of mass media campaigns about the health risks of smoking. These results are confirmed by Santos et al. [28] who reported a decrease in the prevalence of smoking from 24.9% to 8.7% among high-income women compared to 43.7% to 33.6% among low-income women. Another study conducted in New South Wales found a 67.9% reduction in smoking prevalence among pregnant women of higher socioeconomic condition compared to just 25.9% among those of lower condition. In contrast to our results, these authors found further prevalence reductions among older women [26], but smoking during pregnancy was defined as "ever smoked during the current pregnancy. " The same was reported by Tong et al. [22] who found a significant decrease in smoking during the last trimester of pregnancy only among women of age 35 years or more. Increasing trend in smoking prevalence as family income decreased is in agreement with some earlier studies conducted in Canada [13] and the United States [18], both based on secondary data. Although there was no significant trend in level of education, we found higher smoking prevalence among less educated women , as previously reported in other studies based on secondary [12][13][14][15][16] and primary data [17,43]. Considering the harmful effects of smoking during pregnancy, pregnancy can be a favorable time to encourage smoking cessation since the possibility of harming their offspring may motivate smoking abstinence among pregnant women. Many women quit smoking as soon as they learn that they are pregnant, but those who continue to smoke are likely a subgroup that would be more resistant to change and thus require further investigation [44]. Pregnant women who quit smoking during pregnancy often have a profile that contrasts to that of those who continue to smoke. Those who quit are more likely to be primiparous women, of higher income, more educated, married, have a planned pregnancy and early prenatal care, and smoke fewer cigarettes per day [20,22,25,45]. Our study found, in all three surveys, an increasing trend in the prevalence of smoking cessation with income and education increase. US studies conducted based on the Pregnancy Risk Assessment Monitoring System surveillance data that defined smoking cessation as when women who smoked three months before pregnancy quit smoking by the last trimester reported a direct association between the prevalence of smoking cessation and education [20,22]. In view of appeals and effort to raise awareness of the harmful health effects of smoking during pregnancy, we expected an increase in prevalence of smoking cessation during the study period. However, we found no other studies in Brazil that examined smoking cessation trends among pregnant women. Studies conducted in other countries have reported inconsistent results: upward [20], downward [25,30], or even stable trends [22]. The results of this study may be due purely to chance because of very high 𝑃 values . However, it may also suggest that governmental antismoking programs have been effective at preventing initiation of tobacco use but not at promoting smoking cessation. Brazil has established the National Tobacco Control Program comprising actions that promote smoke-free environments and smoking cessation projects. This program recommends that all pregnant women and nursing mothers have access to a cognitive-behavioral approach for smoking cessation. However, there are no interventions or programs specifically targeted to pregnant women. It seems that current antismoking programs in Brazil have had varied effectiveness for preventing initiation of smoking and for smoking cessation during pregnancy. There is a need for different approaches. However, this should be monitored continually to acquire better evidence and, if necessary, to understand the causes of the decrease in smoking cessation in order to develop new actions for promoting smoking cessation during pregnancy. The finding of an association between prevalence of smoking and smoking cessation with socioeconomic status over time could be explained by the Fundamental Cause theory. According to this theory, the SES is related to several risk factors, influences multiple outcomes, and involves access to resources that increase survival. Through new knowledge, this may lead to the establishment of preventive measures against risk factors, to the control of diseases, and to the prevention of diseases and associated complications [46]. However, this theory has some limitations including the fact that the resources for interventions are not always available at the time and in the required amount. Thus, the mechanisms that cause particular diseases change over time, the relationship between the SES and the occurrence of unfavorable outcomes can be weakened or even disappear, exposure can become irregular or unstable, and there may not be enough knowledge to propose new interventions. Finally, considerable challenges remain to be tackled. In addition to those stated in the Fundamental Cause theory, the challenges for governments are to develop interventions, to reduce risk factors and prevent unfavorable health outcomes, to eliminate socioeconomic inequalities in health and access to health interventions, and to develop interventions of acceptable costs that are at its core more equally available among different population groups. Otherwise, this is likely to perpetuate the fact that the most affected are those from the lowest socioeconomic groups. --- Conclusions Despite the observed reduction, smoking before and during pregnancy is still highly prevalent among some women groups and prevalence of smoking cessation is low in almost all groups studied. There is a need to identify and provide treatment to female smokers before pregnancy or early in pregnancy during prenatal care. Moreover, it is necessary to strengthen ongoing actions and target low-income, less educated, black pregnant women. --- Conflict of Interests The authors declare that there is no conflict of interests regarding the publication of this paper. Submit your manuscripts at http://www.hindawi.com --- Stem Cells International Hindawi

Objective. To examine time trends in prevalence of smoking and smoking cessation during pregnancy by family income, maternal level of education, skin color, and age. Methods. We conducted three population-based surveys in 2007, 2010, and 2013 with newly delivered mothers living in the municipality of Rio Grande, Southern Brazil. Data were collected using questionnaires administered after delivery in all (two) maternity units in the city, at Dr. Miguel Riet Corrêa Júnior Hospital and at Santa Casa de Misericórdia. Time trends were analyzed using chi-square test for linear trend. Results. Data of 7,572 women showed that the prevalence of smoking before pregnancy decreased from 28% (26.2-29.7) in 2007 to 22% (20.8-24.0) in 2013 (𝑃 < 0.001). Prevalence of smoking during pregnancy decreased from 22% (20.4-23.7) in 2007 to 18% (16.6-19.5) in 2013 (𝑃 < 0.001). This reduction varied across income ranging from 17% (poorest) to 35% (richest) (𝑃 < 0.001). The lower the income, the higher the smoking prevalence during pregnancy. Smoking cessation was more prevalent among women of higher level of education and income. Conclusions. Smoking before and during pregnancy is still highly prevalent and the prevalence of cessation is low pointing to a need to strengthen actions targeting low-income, less educated, black pregnant women.

Introduction People in minority groups of all kinds, particularly black, Asian and minority ethnic groups, continue to face daily adversity. In the UK, the Equality Act 2010 was introduced to deter prejudice and discrimination on the basis of nine protected characteristics -age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race, religion or belief, sex and sexual orientation. However, while the act has been beneficial for some families in terms of providing reasonable adjustments for children and young people , disparities remain in a broad range of areas, including healthcare. Disparities in healthcare have been highlighted by the coronavirus disease 2019 pandemic which has affected disproportionately people in various minority groups, including people with learning disabilities and those from black, Asian and minority ethnic communities . According to Khunti et al , these disparities may be --- Why you should read this article: • To understand the concept of cultural safety linked to employment in front-line services, social living circumstances and vitamin D deficiency compromising immune responses. Further, there have been higher levels of COVID-19 vaccine hesitancy in certain groups, such as black African, Pakistani, Bangladeshi, white non-British and Gypsy and Traveller communities . It has been suggested that cultural factors and misunderstandings can influence people's willingness to take up vaccination opportunities . However, further research is required to understand the reasons why certain groups have been affected disproportionately during the COVID-19 pandemic and in healthcare generally. If nurses fail to understand the reasons why such disparities in healthcare exist for minority groups, then they may miss opportunities to protect and improve child and family health. It is important for nurses to be aware of these groups' perceptions of threats, security and fairness to attempt to understand the lived reality of individuals and their varying circumstances and what health, illness and care means to them. Open dialogue about the structural injustices and unconscious biases that affect minority groups can provide an opportunity to address these issues in a constructive and appropriate manner. It is essential that healthcare professionals assure all patients, including those from minority groups, that they have regard for their experiences and needs. The Nursing and Midwifery Council Code in Action: Caring with Confidence advocates recognising diversity and includes a section designed to support nurses to be inclusive and to challenge discriminatory behaviour. Similarly, the Royal College of Nursing Equality and Inclusion Strategy 2017-2020 encourages nurses to be 'bolder and more courageous in advocating for the centrality of equality and inclusion'. The authors of this article suggest that to treat people impartially and with respect for their individual needs, conditions and living circumstances is the ethos of a cohesive society as well as the foundation of effective, compassionate care. --- Defining culture Defining the precise meaning of culture can be challenging, in part because the term and its context are highly complex and nebulous, alluding to refinement, linguistics, values and beliefs. Helman defined culture in a broad sense as 'a set of guidelines that individuals inherit as members of a particular society, and that tell them how to view the world, how to experience it emotionally, and how to behave'. Each person emerges into a culture at birth and throughout their life they continue to accrue and be shaped by it. Culture helps each person to manage their expectations of daily living, which may guide their aspirations and affirm who they are . It also links an individual to others who seem like them -by whatever criteria -and serves to distinguish them from others who perhaps live differently. Culture can help a person to feel secure, define progress and success and enable them to relate to wider society . The authors of this article suggest that culture: » Is an accumulated way of understanding the world, living, perceiving needs and experiencing fears. » Supports people to identify their core beliefs and fundamental values. » Influences people's traditional customs and habits and provides them with patterns of living. » Assists people to negotiate help and offer appropriate support to others. Purnell listed several factors that may contribute to a person's 'cultural being', including family, community, aspirations, spirituality, foods, religion, rituals, illness, education and positive and negative life experiences. Culture is particularly important for those who belong to groups that are a statistical minority. Although people in such minority groups may come from different backgrounds and cultures, they often share common ground in experiencing a lack of enquiry about their needs and expectations in relation to healthcare . --- TIME OUT 1 Imagine that you are supporting a family from a different culture to yours. Consider how to adjust the way in which you respond to them. Enquire about their needs and preferences. What part might culture play in relation to the following: » Providing advice on dietary modifications for a patient with diabetes mellitus » Exploring optimal approaches to alleviate pain » Understanding how to engage with a patient's family carers effectively --- » Arranging rehabilitation after an injury --- Role of culture in healthcare The experience of health and illness is cultural since patients, families, nurses and other healthcare professionals perceive notions of 2018). Giger and Davidhizar explained that culture shapes the behaviours people use to protect, maintain and improve their health and that of their children. Therefore, culture affects people's responses to illness and their expectations of how it should be prevented or treated. Customs and rituals can be important when prescribing treatment products, as these may be codified in religious and/or other beliefs. Patients' expectations of nursing care are also often cultural, as are nurses' expectations of patients. It is important that all healthcare professionals recognise the effect that an individual's social conditions and living circumstances may have on their responses to questions about their health. In addition to the standard care they provide, therefore, it is crucial that children's nurses pay attention to the cultural experiences, needs and expectations of children, young people and families. Many people who access healthcare services may believe that their culture is ignored or adversely affects how they are treated . In stressful circumstances such as illness, it is crucial to respect a person's culture and enable them to manage issues in accordance with their values and beliefs. People who feel oppressed need reassurance that they are valued and that their concerns are heard. If children and young people's voices are not listened to or prioritised, services become inequitable. Nurses need to consider that if health is experienced in cultural terms and diverse ways, then so too are healthcare and services. --- Key points Clinical policies and guidelines, such as the RCN's equality and inclusion strategy , The Code: Professional Standards of Practice and Behaviour for Nurses, Midwives and Nursing Associates , and Standards of Proficiency for Registered Nurses , acknowledge the role of cultural factors in nursing. Furthermore, discrimination based on perceived difference is widely condemned and regarded increasingly as intolerable by professional bodies such as the NMC . Incorporating cultural aspects into clinical management can improve the efficiency of interventions . This affects not only the overall quality of care received by the child and their family, but also their lived experience which will influence whether they access healthcare services in the future. Therefore, there must be a focus on inclusive care that is based on professional curiosity, with consideration given to values and priorities, culturally sensitive communication and appropriate decision-making, actions and goal setting. --- TIME OUT 2 Consider an interaction you have had in your practice where there was some form of miscommunication. What assumptions might each person have made that could have contributed to the misunderstanding? What do you think might have prevented each person from conveying their concerns or needs clearly? Cultural differences and assumptions Each person's cultural self is internalised and usually so ingrained that they do not think about it consciously. Therefore, there is a risk that people will unwittingly bring expectations to healthcare encounters that are then used to judge whether or not their experiences were satisfactory . Further, it is possible that nurses' approaches towards patients, families and other healthcare staff may be culturally limited which can lead to incorrect assumptions and unconscious bias. Cultural differences can represent a threat, opportunity or resource; people can be either adversarial or welcoming of diversity. For example, children's nurses may demonstrate displeasure in subtle ways when families choose not to act on their advice which, the authors suggest, could be due to a lack of cultural awareness, listening or understanding. Nurses should therefore reflect on whether they may subconsciously expect parents to respect their 'authority' and use this realisation to consider acting differently. Attitudes can be interpreted through facial expressions, posture, body position or the way an adult relates to a child. However, this carries a risk of misinterpretation or mistaken assumptions about what others feel or need. As nurses have a duty of care to empathise with and understand the experiences of children and their families, it is important that they check that their interpretations are accurate. Problematic intercultural communication in nurse-patient relationships, for example those not based on mutual respect, can be a contributory factor to negative health outcomes . The way that individuals respond to challenging, sensitive or highly charged situations, such as grief or the loss of a young person, are often shaped by culture. However, it is often 'gut reactions' that can lead people to judge others and events . Children's nurses must recognise R C N i nursingchildrenandyoungpeople.co.uk | PEER-REVIEWED | © RCN Publishing Company Limited 2022 that they have the power to influence others significantly and should reflect on their own cultural premises. For example, a nurse may assume that a young person's mother is their primary caregiver or make assumptions about how a child's family manages their child's pain in terms of administering medicines. In reality, family roles are diverse, while sometimes pain may be managed according to custom and traditions such as homeopathy or homemade tinctures passed down through generations . In nursing, there are various approaches to considering culture and encouraging inclusivity in healthcare. For example, Prosen emphasised the importance of mutual understanding of one another's cultural values, beliefs and practices, which is known as cultural congruence. When using such approaches, nurses support healthcare choices that are personally and culturally meaningful to the service user. However, they may also unwittingly prescribe ways to consider and resolve issues that might disadvantage the family and/or lead to feelings of resentment. --- Cultural safety The concept of cultural safety was developed by Irihapeti Ramsden and other Maori nurse leaders in New Zealand to meet the diverse needs of Maori healthcare users and has since become established in global policy . Cultural safety incorporates ideas such as cultural awareness, sensitivity and competence. It is based on the notion that the cultural assumptions and expectations of a majority group can unthinkingly lead to the oppression and disadvantage of minority groups . In this respect, it may be worth considering that everyone could be vulnerable to forms of unconscious bias or prejudice. Using a cultural safety approach is about respecting diversity and individualising care and it serves to celebrate uniqueness and neurodiversity rather than dismissing or oppressing others due to differences and diverse thinking . Box 1 details the characteristics of culturally safe thinking. Cultural safety builds on the principle of professional curiosity and supports a familycentred approach to care, with which children's nurses are familiar. This can help to better inform children by signposting them to appropriate support and empowers them to take responsibility for their own health . There is also evidence that promoting community engagement, for example through Sure Start or Flying Start programmes, can increase people's sense of belonging, improve risk assessment and reduce health inequalities for those who are most vulnerable . Cultural safety can be applied to children and young people with any acute or chronic condition. For example, when caring for a child or young person with type 1 diabetes, standard dietary advice would typically relate to Western mealtimes and food groups and their effects on blood glucose levels. However, in addition to this advice healthcare professionals need to explore dietary restrictions in accordance with a child or young person's cultural or religious beliefs and consider food that the family can realistically afford or access and that they usually consume. To facilitate cultural safety, efforts would have to extend to the food industry and advertisements and the messages which promote people's food and drink choices. Ball described five principles to promote cultural safety . Practising these principles can increase the likelihood that a child or young person seeking care will feel respected, included and protected in terms of their cultural identity, and can support care planning by promoting mutual goal setting. --- TIME OUT 3 Children's nurses may be perceived by some children, young people and their families as authority figures. How might differences in perceived power affect the quality of interactions with others and how could this issue be addressed? Box 1. Characteristics of culturally safe thinking » Promotes healthcare professionals' self-analysis and awareness of the effect of their actions during any therapeutic encounter » Recognises that a power gradient exists between the healthcare professional, the child or young person and their immediate family » Encourages the use of fundamental interpersonal skills that can be applied sensitively by all healthcare professionals. An example is active listening to gain mutual understanding » Rejects ideas about what constitutes a traditional family and the method of rote learning the cultural characteristics of a particular ethnic group or applying stereotypical assumptions. Do not make assumptions about an individual's needs but instead enquire about these » Enables healthcare professionals to appraise and reflect on their role during social interactions, learn from positive cultural encounters and use negative experiences to make improvements in the future » Recognises that all interactions must be based on dignity and respect for every individual and --- Importance of culturally safe communication Communication can be considered one of the most important means of achieving cultural safety, since people who feel disempowered can find it challenging to communicate with healthcare professionals who they may perceive as uninterested or having power and authority over them. Social interactions between children, young people, families and healthcare professionals require the use of interpersonal skills which are defined by cultural orientation . This incorporates not only the language a person speaks most fluently and comfortably, but also health literacy and the use of appropriate colloquial expressions in place of professional jargon and medical acronyms. It also includes non-verbal forms of communication such as body language, gestures and facial expressions. All these aspects of communication must be acknowledged and an environment that favours mutual dialogue needs to be established . Effective communication is particularly important when caring for children and young people because culturally mediated nuances are learned throughout childhood and may not only be used imperfectly but may also be susceptible to disrespect from people outside the cultural group. Early on in their development children have an incomplete sense of who they are as cultural beings. They are also vulnerable to confusion if they see the cultural norms of their family treated with disrespect or prejudice or ignored. It is a vital part of anti-discriminatory practice and diversity awareness that a child or young person's developing cultural self is treated with respect and sensitivity. This requires insight, reflection and self-awareness on the part of children's nurses. --- TIME OUT 4 Reflect on a time you spent working with a colleague who was an influential role model, for example a mentor, practice supervisor or practice assessor. Did this colleague exhibit awareness of people's cultural needs and did they facilitate cultural safety? --- Advantages of cultural safety One advantage of cultural safety is that it aims to ensure that aspects of culture are considered in every encounter between nurses and children and their families. Culture should become the reference point for care, progress, how the family feels about accessing healthcare services and the support that children's nurses provide. An important characteristic of cultural safety is that it is the person receiving the care who determines whether it feels culturally appropriate and safe and whether they feel adequately listened to and informed . Therefore, cultural safety can supplement everyday practice to enhance the quality of healthcare encounters through sensitive communication which promotes empathy, improved access to care and participation. This encourages children's nurses to develop their enquiry-based skills and to work autonomously and with others to mutually influence decisions about children's healthcare. Previous approaches to cultural care have tended to focus on establishing the cultural orientation of the group to which the family belongs. However, such approaches fail to recognise diversity and the degree to which individuals vary within cultural groups -not every child behaves or experiences illness, pain or bereavement in the same way. In addition, autistic children may express pain in a way that is different from those who are neurotypical , so such approaches also fail to consider neurodiversity. There can be a risk of inadvertent stereotyping of people by making assumptions about what children or families should do or believe based on generalisations or myths about the cultural groups to which they belong. However, there are patterns in communication and behaviours that could be useful to consider. For example, there may be an expectation that fathers will be more forthright and assertive in their communication patterns and that mothers will be more tentative in presenting their views and needs . Nurses may need to consider such biases or factors --- Table 1. Five principles to promote cultural safety --- Principle Actions that nurses can take Protocols Find out about cultural forms of engagement and respect these. An example of this would be using the preferred form of a child's name --- Personal knowledge Become mindful of your own cultural identity, socio-historical location in relation to service recipients, commitments to certain beliefs and ways of conceptualising health and wellness. Be prepared to share information about your own background if this will help to establish equity and trust with the child and their family Partnerships Promote collaborative practice, whereby a child is welcomed into a joint problemsolving approach as an expert patient who has important information and knowledge --- Process Engage in mutual goal setting with frequent checking in to ensure that proposed action plans align with the values, preferences and lifestyle of the child and their family --- Positive purpose Ensure that positive steps can be taken to achieve a child's goals or resolve issues and that these are likely to be beneficial to their health and well-being R C N i Teal CR, Gibb AC, Green AR et al Helping medical learners to recognise and manage unconscious bias towards certain patient groups. Medical Education. 46, 1, 80-88. doi: 10.1111Education. 46, 1, 80-88. doi: 10. /j.1365Education. 46, 1, 80-88. doi: 10. -2923Education. 46, 1, 80-88. doi: 10. .2011.04101.x .04101.x © RCN Publishing Company Limited 2022 in their therapeutic encounters and cultural safety, which recommends enquiring about people's roles , may assist in deconstructing gender-based assumptions. This is an important advantage because working in this way places the responsibility for cultural appropriateness on the children's nurse, while giving families the opportunity to report when care does not feel culturally safe . --- TIME OUT 5 Contemplate an encounter with parents who are refusing to allow their child to receive vaccinations. What is your role in such a scenario? Would you: a) Discuss with them the evidence for, and risks associated with, vaccination? b) Explore their beliefs, values and aspirations in the context of the evidence on the benefits and risks of vaccination? c) Simply challenge them to think again? --- Conclusion Cultural safety can be an effective way for children's nurses to work with people from within diverse and complex communities to empower them and promote their needs and autonomy. At a time when healthcare services and staff are stretched to their limits, it may be reassuring to be aware that cultural safety does not require significant resources, training or documentation. Instead, it requires healthcare professionals, including nurses, to reflect continually on their attitudes and to address the potential effects of these on the people they care for. This will enable healthcare services to become increasingly accessible to children and families from minority groups. When considering the optimal approach to care for various patient groups, listening to children, young people and families can provide a useful starting point. --- TIME OUT 6 Identify how cultural safety applies to your practice and the requirements of your regulatory body --- TIME OUT 7 Now that you have completed the article, reflect on your practice in this area and consider writing a reflective account: rcni.

The aim of this article is to enhance children's nurses' understanding of culture and the idea of cultural safety, and to support them to recognise these as crucial components of the care they provide to children and young people and their families from a range of backgrounds. After reading this article and completing the time out activities you should be able to: » Recognise the ways in which people from minority groups experience disparities in healthcare. » Explore the meaning and components of culture. » Outline the principles of cultural safety and how these can be applied in practice. » Describe the potential advantages of implementing cultural safety in the care of children and young people.De D, Richardson J (2022) Using cultural safety to enhance nursing care for children and young people. Nursing Children and Young People.

Introduction In 1995, 45.5 million abortions that took place worldwide were considered illegally induced. Concerned by this trend, many countries liberalized their laws to increase the women's access to safe abortion. By 2003, illegally induced abortions were reduced to 42 million in the world. This declining trend of induced abortions was more or less similar in the developed countries, where abortion was legal and easily accessible, and developing countries, where it was predominantly illegal and restricted. However, there was a striking difference noted in relation to their safety, with 55% of all induced abortions being 'unsafe' in the developing countries, compared to a much less proportion in the developed countries [1,2]. An unsafe abortion is defined as 'a procedure for terminating an unintended pregnancy either by individuals without the necessary skills or in an environment that does not conform to minimum medical standards or both' [3]. It leads to acute life-threatening as well as long-term disabling morbidity. Approximately 68,000 die from complications following unsafe abortions each year, giving a median mortality ratio of 34 per 100,000 live births in countries where abortion is illegal [4]. Consequences of such abortions are crucial in Asia, where it is concentrated among the poor, with a higher tendency towards severe complications [1]. It emphasises the need to prevent unsafe abortions among women facing an unintended pregnancy. In Sri Lankaa country in South Asia, with the expansion of its national family planning programme, a steady increase in the use of contraceptives has been noted from 32% in 1975 to 68% by 2006-7 among women in the reproductive age [5]. Despite this success, induced abortion remains a commonly practised fertility control method at the rate of 45 per 1000 women in the reproductive age [6]. This rate is an indication of its growing demand although abortion is illegal in Sri Lanka. Despite being the most easily preventable cause [7], unsafe abortions is contributing considerably to the maternal morbidity and mortality in Sri Lanka, without showing a declining trend over the past few years [5]. In this backdrop, providing targeted interventions to women vulnerable to unsafe abortions would be of immense value. Women perceive that many factors related to their socio-economic circumstances, fertility intentions, knowledge and attitudes on abortion, partner preferences, and availability and access to abortion services 'push' them towards an abortion [8][9][10][11]. These factors are populationspecific and vary based on the legal, cultural and socioeconomic background of a country. In Sri Lanka, past studies have implied poverty, poor knowledge and varying attitudes on abortion among women who seek abortion [12][13][14]. However, these findings may not be applicable in today's context, since the fertility aspirations of modern Asian women have drastically changed over time. Worldwide, 25% of unintended pregnancies end up as 'unwanted or mistimed child births' [15], highlighting that not every woman would invariably opt to undergo an abortion. More importantly, it further suggests that in addition to many 'push' factors that favour abortion, women also have 'pull' factors that work against abortion. All these factors during the decision making would determine a woman's vulnerability to abortion. Although many factors favouring induced abortion have been identified among abortion seekers [16], their risk compared to women who decide against an induced abortion has not been evaluated. Identifying the risk of these factors would empower women to make the right choice through modification of the circumstances under which they make their decisions. This would be most crucial especially for women who would decide to undergo an unsafe abortion. Thus, this study intended to assess the risk of circumstances that surrounded women who decided to undergo an unsafe abortion, compared to women who decided to carry an unintended pregnancy to term. --- Methods We carried out an unmatched case-control study including a qualitative component in nine hospitals in eight out of the 24 districts in Sri Lanka. Five of these hospitals were selected based on the highest frequency of all types of abortions reported in the Indoor Morbidity and Mortality Registers for each district [Medical Statistics Unit, unpublished]. Two hospitals were intentionally selected to ensure the representation of Muslim and estate-sector-Tamil populations. In the district of Colombo, both apexreferral tertiary hospitals for women were included. Cases were women in the selected hospitals with complications following an unsafe abortion. Controls were mothers in postnatal wards following the delivery of an unintended pregnancy carried to term. The required minimum sample was based on 80% power to detect potential associations between the cases and controls at 5% alpha error; 20% minimum probability of exposure in the controls; odds ratio of 2; and 1:4 unmatched casecontrol ratio. During recruitment, all women admitted to the gynaecology and medical/surgical casualty wards were screened consecutively over a period of six months for signs and symptoms suggestive of an abortion . Of them, the potential 'cases' were identified by a confirmed diagnosis of 'induced abortion' based on the World Health Organization criteria [17] under three categories: 'certainly induced' , 'probably induced' and 'possibly induced' abortion. Of these women, all 'certainly induced' abortion cases and only the women who showed definitive clinical signs of infection plus received intravenous antibiotic treatment in the 'probably induced' abortion category were recruited as 'cases' for the study. This restriction ensured that the group of cases represented only the women who underwent an unsafe abortion. For comparison, controls were selected during the same study period using a systematic sampling method among mothers admitted to postnatal wards of the same hospitals following the delivery of an unintended pregnancy carried to term. 'Unintended pregnancy' was identified using an interviewer-administered-questionnaire based on the WHO definition of 'any pregnancy of a woman contracepting during the cycle of conception or not contracepting due to reasons other than desired pregnancy' [17]. Core questionnaire developed by the WHO for multicentre hospital-based descriptive studies on abortions [17] was used for data collection. It was modified to suit the local conditions and its judgmental validity was assessed by a panel of experts in maternal health care, public health and clinical psychology. It was administered by pre-intern medical officers who were not involved in providing care in the ward. Prior to data collection, they were trained by a group of psychologists and experts in qualitative research. Using the validated questionnaire, data were collected on demographic, socio-economic and reproductive characteristics, and on decision making . Circumstantial details on pregnancy termination were obtained only from the 'certainly induced' abortion cases . Furthermore, in-depth interviews were conducted by the principal investigator in a sub-sample of women to explore their decision making process. This sample size was decided by the point of saturation at which no new information was generated. Ethics clearance was obtained for the study from the Ethics Review Committee of the Faculty of Medicine, University of Colombo. Data were analysed using Statistical Package for Social Sciences -Version 20. Descriptive statistics included proportions calculated for categorical data and mean and standard deviation for quantitative data. The risk of abortion associated with women's socio-economic characteristics and their decision making was assessed by comparing the cases and controls using odds ratio and 95% confidence intervals . --- Results Of the 822 potentially induced abortions that were initially identified, 171 were recruited as cases for the study. The controls consisted of 600 post-partum mothers. There were two controls who had previously undergone an induced abortion and two cases who had done it for the second time. The majority of women in both groups belonged to 25-29 age group, were married and of Sinhala-Buddhist origin. Compared to the controls, the significant characteristics of cases were being unmarried, primigravid, not having an informed decision on family size, less-educated, employed, having children and having longer last birth intervals . Of the employed, more cases were in elementary occupations as manual labourers or as plant/machine operators . The non-primi gravid women were at significant risk for unsafe abortion, if they had longer average birth intervals and longer last birth intervals . Nearly 50% of the partners of both cases and controls had completed their upper secondary education, while nearly all were in employment. Cases differed significantly from their partners who were more educated and employed in better-skilled occupations than them . No such difference was seen between the controls and their partners. Prior to their unintended pregnancy, all women in both groups had heard about induced abortion 'as a method of pregnancy termination' and its 'illegal' status. It was mainly through television and women newspapers and to a lesser extent from immediate associates such as friends , neighbours and family/relations . When arriving at a decision about their pregnancy , compared to the controls, a significant proportion of cases sought advice from others , with more reliance placed on non-medical sources such as their spouse/partner, friend, neighbour and family/relation, and also from women who had previously undergone induced abortion . In the qualitative enquiry, cases revealed that their unwillingness to approach qualified persons was due to the illegal status of abortion, while some admitted to have relied on abortion experiences of others since they had not received any formal education on abortion before. Although women relied on partners during decision making, a higher proportion of cases made their final decision on abortion all by themselves, compared to the controls . Remarkably, 12.3% of the cases initially considered continuing with their pregnancy, while 42.7% of the controls considered undergoing an abortion. The commonest reason given by cases for terminating their pregnancy was not being economically stable . Other reasons included poor support from the partner, too old for carrying a pregnancy and too short last birth interval. The commonest underlying reason given by controls for continuing with the pregnancy was due to religious beliefs that portrayed abortion as a 'sin'. In 12% of controls, it was due to cultural constraints that portrayed abortion as a socially unacceptable event. Good support from the partner, being economically stable and not having a short last birth interval also helped them in their decision. After making their decision to abort, the majority of cases approached their partners and/or immediate associates to obtain more information about the persons/places available for pregnancy termination and to accompany them to abortionists . It was revealed in the qualitative enquiry that in their absence, women depended on unknown sources such as taxi drivers for this purpose. Only 11% of cases attempted methods of self-induction. 46% of cases visited the abortionist more than twice and 14% more than five times. In the majority of cases, septic procedures with no pain relief were performed during termination by non-qualified abortionists, for a wide range of payment of Rs. 1,000-30,000. In the qualitative inquiry, women reported that most places lacked sterile or proper equipment and were run without assistance in the back room of a boutique, own home or in the house of a relative or abortionist. None took place in government hospitals or clinics run by non-governmental organizations. The most commonly used methods were trans-vaginal insertion of rods and injections. The worst experience reported in the qualitative inquiry was one woman collapsing following the insertion of a castor oil plant stem into the vagina for a fee of Rs. 30,000. In response to future intentions, 14% of cases alleged that they would resort to abortion in the event of another unwanted pregnancy, while 47.5% were not sure of their decision. --- Discussion This study sheds light on the decision making process and circumstances leading to an unsafe abortion, in comparison with that leading to a mistimed birth, in a setting where abortion is illegal. It should be noted that these findings are not applicable to women who succeeded a safely induced abortion, but to those who developed complications following an unsafe abortion. Thus, this study is unique, providing evidence on the women who are most vulnerable to maternal morbidity and mortality. Health status of Sri Lanka is sustained by the Government policy of 'free health for all' that has been instrumental in achieving a high health literacy and health seeking behaviour, especially among the females [5]. Our findings however are contrary to the expected, highlighting deficiencies in the provision of information to women on the health implications of unsafe abortion and women's utilisation of unsafe abortion services. In comparison, previous studies have shown that women face difficulties in obtaining both information and services on abortion in low-resource and low-literacy settings, irrespective of the legal status of abortion [11,14,18]. --- The role of access to information in the decision making Our study shows that a decision to abort using unsafe techniques is significantly associated with the risk of women seeking access to information about its safety during the decision making process . Furthermore, in the absence of any formal channel providing awareness, cases seemed to rely more than their controls on non-medical sources for knowledge. Despite the free health service that ensures access to state-owned health facilities throughout the country [5], the reluctance of women to approach a healthcare worker in the event of an unintended pregnancy, was probably due to the illegal status of abortion in Sri Lanka. Stigma attached to abortion could also be a major deterrent [11,18]. With the wide accessibility of mobile telecommunication services even in the most remote areas in Sri Lanka, exploring the possibility of establishing 24-hour help lines for reliable information at the community level would be of much value [19]. Abortion education through media did not play a significant role in the decision making in our study. Nevertheless, close monitoring of any sensationalising abortion news by media is highly recommended. Furthermore, to prevent the vulnerable groups being misinformed, women need to be empowered with knowledge on the safety of abortion from an early age. Adolescence would be the ideal time to approach them through educational opportunities such as school-based internet inquiry [20,21]. Though sexual health is already a vital component in the school curriculum, there may be cultural barriers in discussing these topics at school. It is thus important to provide skill development for teachers. Previous research has implied varying attitudes on unsafe abortion among women who seek abortion [1,14,18,22]. Our study implies that the risk of unsafe abortion associated with access to unreliable non-medical sources of information is not only because it leads to poor knowledge on its safety but also because it formulates positive attitudes towards induction. Close associates seemed to play a crucial role in changing women's perceptions on abortion as a 'safe' procedure. In particular, women showed a tendency to confide more in other women's experience of successful inductions. This type of 'herd behaviour' was probably facilitated by their low socio-economic status characterised by low level of education and employment in unskilled/semi-skilled occupations. The equally low social status of their co-workers seems to reinforce the women's poor access to correct information as well as to safer abortion services. It is shown that unlike home-bound women, employed women are more at risk for unsafe abortions, with many opportunities for clandestine sexual relationships and peer influences [23]. Therefore, multi-pronged interventions in the form of counselling services should take priority in these high-risk settings. With high literacy among women in Sri Lanka, work settings could be equipped with educational material and contraceptives. Evidence recommends behaviour change communication campaigns through inter-personal approaches that engage community leaders and influencers [18], dialogue-based groups that negotiate the social support they need when making decisions [24] and community intermediaries that create an enabling environment [25] as feasible methods. In previous studies in Asia, a significant decision maker has been the partner with their direct involvement through "orders" to abort or indirectly through denial of Women who admitted to have undergone an unsafe abortion were only included in the analysis. 2 Includes women who were reluctant to reveal the information. 3 Also includes respondents who were not sure of the method used. 4 Also includes 13 women who attempted termination by themselves. responsibility for the pregnancy [11,26]. Our findings are in contrast to this, highlighting the changing women status in Asia. Although partners of cases were of better socio-economic status than of their controls, partners' influence seemed to be minimal when making the final decision to abort a pregnancy. However, once the decision was made, men were utilised more by cases to accompany them to abortionists. --- The role of safe abortion services in the decision making Increased availability as well as accessibility have shown to improve the quality of induced abortions in countries where abortion is legalised [22,27,28]. Our study verifies that the vulnerability of Sri Lankan women to unsafe abortion is mainly due to poor access to affordable abortion services. This was explicit, with an increased risk of unsafe abortion seen among women of low socio-economic status. It is alleged that safe abortion services are not uncommon in the country, but almost always offered for a substantial fee. This is in contrast to the health seeking behaviour in the past, during which 'menstrual regulation methods' were offered to clients for an affordable fee by a few non-governmental organizations [29]. --- Circumstances leading to unsafe abortion Worldwide, the commonly reported reasons for unsafe abortion are postponement or completion of child-bearing and socio-economic concerns such as poverty and unemployment [10,29,30]. However, a recent study concluded that a decision to undergo abortion is typically motivated by multiple, diverse and inter-related reasons on financial constraints and lack of partner support [31]. These factors coincide well with our findings, which confirm that a decision favouring unsafe abortion was predominantly based on their economic instability and poor support given by partners. Other than the 'push' factors, several 'pull' factors that prevent a decision of abortion were also identified in our study. Sri Lanka is a multi-ethnic country with strong religious convictions. In a local study among women, 94% felt that abortion was a 'sin' while 66% felt that it should be available on demand for married women [12]. It is interesting to note that women who decided against abortion considered more about the ethical aspect than its legal status. This implies that even if abortion is legalised, safe abortion services would remain under-utilised, unless the stigma attached to abortions that leads to the reluctance of both women accessing the services and the health care staff providing it, is addressed. Under-utilisation of safe abortion services due to poor knowledge [18], religious beliefs [32,33] and stigma [14,34] has been observed elsewhere. Holmgren & Uddenberg [32] have reported that women's main moral dilemma was not a conflict between the woman and her foetus, but a conflict between several close relationships, also concerning the prospective father [32]. It is imperative that health managers ensure that clients' information against unauthorised disclosure is protected by creating a respectful environment, with physical space for assuring privacy, a wide-range of skills for building rapport with women in a culturally-attuned empathic manner and attitudinal changes at all levels of health care providers to treat them with dignity, so that women are comfortable in discussing their decision making with health staff. Sri Lanka Government health policy aims to facilitate equity through increased access to health services and quality of care. --- Methods of abortion Circumstances under which most of the unsafe abortions took place illustrate the extremely poor quality of the services provided by Abortionists. Most were done by unqualified persons for a wide range of payment in unsterile environments using septic procedures. It is evident from our study that the procedures used by abortionists had been more or less similar, compared to the methods used on abortion cases in hospitals a few decades ago [29,35]. Legal enforcement of severe punishment to these unqualified abortionists would be effective. This would prevent the less resourced districts to have access to unsafe abortion services. Decisions to refer or to keep in their care by abortionists are strongly associated with the patients' financial resources [36]. In our study too, women's health seeking behaviour following an unsafe abortion was determined by the advice provided by the abortionist; poorer women to visit the hospital immediately and others to be treated in a private hospital. --- Limitations Our sample has not captured the women who sought care outside Government hospitals following postabortion complications. This could have led to an overrepresentation of cases of poor socio-economic status, since the state owned health services are free of charge and thus, predominantly accessed by people of relatively low socio-economic status. However, this selection bias was minimised by having controls accessing the same hospital services s as the cases. --- Conclusions A decision favouring unsafe abortion was predominantly based on their economic instability and poor support given by partners, whereas a decision against it was based on ethical considerations over its legal implications. Reliance on non-medical sources of information such as immediate associates leading to poor knowledge as well as positive attitudes on its safety played a crucial role in the decision making process towards an unsafe abortion. Unsafe methods of termination used have not changed over time. It is believed that the findings of this study would also be applicable to similar settings in developing countries. --- Competing interests The authors declare that they have no competing interests. ---

Background: Following an unintended pregnancy, not every woman would invariably choose to undergo an unsafe abortion. It suggests that in the decision making process, women face both 'push' factors that favour abortion and 'pull' factors that work against it. This study assessed the circumstances that surrounded a woman's decision to undergo an unsafe abortion, compared to a decision to continue, when faced with an unintended pregnancy in Sri Lanka. Methods: An unmatched case-control study was conducted among 171 women admitted to nine hospitals in eight districts following an unsafe abortion (Cases) and 600 women admitted to the same hospitals for delivery of an unintended term pregnancy (Controls). Interviewer-administered-questionnaires and in-depth interviews assessed women's characteristics, decision making process and underlying reasons for their decision. The risk of abortion related to their decision making was assessed using odds ratio (OR) and 95% confidence interval (CI). Results: Compared to controls, the cases were significantly less-educated, employed, unmarried and primi-gravid (p < 0.05). All knew the 'illegal' status of abortion, mainly through media (65.5% cases versus 80% controls). When making a decision, the risk of undergoing an unsafe abortion was significant among those who sought assistance (44% versus 32%; OR = 1.7 (95% CI = 1.2-2.4)), with more reliance placed on non-medical sources such as spouse/ partner, friend, neighbour and family/relation. Speaking to women with past experience of induced abortions (31% versus 21.5%; OR = 1.6 (1.1-2.4) and failure in making the final decision with partners also imparted a significant risk for abortion (64% versus 34%; OR = 3.4; 2.4-4.8). A decision favouring unsafe abortion was predominantly based on their economic instability (29.5%) and poor support by partners (14%), whereas a decision against it was based on ethical considerations (44% religious beliefs: 12% social stigma) over its legal implications (4%). Most abortions were performed by unqualified persons (36.1% self proclaimed abortionists; 26.2% not revealed their qualifications) for a wide range of payment in non-sterile environments (45.9% unknown place) using septic procedures (38.5% trans-vaginal insertions; 24.6% unaware of the procedure). Conclusions: Women's risk of unsafe abortion was associated with unreliable sources of information during decision making that led to poor knowledge and positive attitudes on its safety; poor access to affordable abortion services; and their economic instability.

Introduction The years 2020 to 2022 will remain a stand-out time in human history due to the outbreak of coronavirus [1][2][3] and the European energy crisis due to the Ukrainian War [4] that led to one of the fastest-phase periods of digitalization and boosted focus on sustainability [1,2,5] in different industrial, educational [6], municipal, and NGO contexts. Even the most traditional companies shifted their actions, too [7]. Historically, pandemics with similar big problems tied to them, like in the case of the COVID-19 outbreak, have happened before. These diseases have killed millions of people around the world, but none of them had as severe social distancing limitations. Examples of previous experiences with diseases include the smallpox outbreak in 430 B.C. and the spread of the Ebola virus in 2014. During the early stages of the COVID-19 outbreak, the vision toward the virus was not consistent [8]. While some governments and individuals considered it to mostly resemble a fairly common virus, similar to the common flu, others took it differently and raised the warning about serious challenges the virus could bring to the table. However, the harsh reality was quite evident in 2020, and the entire world did acknowledge the seriousness of COVID-19 in terms of its rapid global spread and impact on individuals' health, countries, continents, global societies, and economies [1,2]. Behav. Sci. 2023, 13, 5 2 of 16 According to the World Health Organization and Worldometers, on 5 December 2022, severe waves of COVID-19 hit 220 countries, territories, and international conveyances, resulting in 650,126,812 confirmed infected people and 6,646,993 death cases [9,10]. This wave affected the global ecosystem and led to a change in habits of interacting and dealing with others [11,12]. Also, governments acted to set emergency regulations to contain the outbreak of the virus, such as social distancing, travel bans, and partial or complete lockdowns. This included restrictions leading to an ability for people to move outside, limiting movements between districts, and confinement of people to their own houses, without being able to go out [13,14]. This change pushed the pre-COVID-19 digital era transformation into high gear. One of the major affected industries because of the pandemic is individuals' trends toward e-commerce, which solved several logistic issues that assist human survival [15,16]. Thus, several research studies were conducted to explore the critical factors of e-commerce resilience [15,17] and the societal behaviors and trends toward e-commerce during and after the pandemic [16] in different industries and countries, such as food and beverage [18], retailing [19], fashion [20], and real estate industries [21]. However, such studies are limited in Middle Eastern countries to explore and predict the social trends of individuals toward utilizing e-commerce. This would be a barrier to answering significant questions related to the future of e-commerce and online business models in the Middle East, according to the behavioral changes after the pandemic hit. Therefore, this study aimed to explore the likelihood of the post-pandemic utilization of e-commerce by consumers. Saudi Arabia and Egypt were selected to conduct this study. The following sections of this research paper contain a review of the global impact of COVID-19 that led to the research questions. Then, the research methodology is discussed before the analysis of the collected data. This analysis is divided into four stages: the demographical analysis, the health impact of COVID-19 on participants, the trend analysis toward utilizing e-commerce and e-services, and the regression analysis. After that, the results are discussed before the study is concluded, and limitations are highlighted to propose the future research direction and to guide researchers to fill the gap. --- Literature Review --- COVID-19 Global Impact The global reaction of individuals and governments to control the rapid spread of COVID-19 resulted in several positive, neutral, and negative consequences, affecting multiple domains. Domains like, but are not limited to, local and international government policies [22,23], the economy [24][25][26], tourism [24,27], nature and the environment [28,29], society, psychological [30] and social behavior [31][32][33], health [34,35], and impact on food security [36,37]. As we can see, the consequences of COVID-19 affected both the local and international government policies in almost every country that tended to create new policies and/or modify or suspend current policies [22,23]. For example, if COVID-19 impacts are the enforcement of social distancing, the early travel ban, border closings, airport and seaport closings, strict COVID-19 testing, emergency laws, the allocation of emergency budgets [22,23], and the shifting online of some services and the previously highly face to face activity leaned [24] education system [38], crowd participation activities [39] changed the pre-COVID-19s business model innovation models [40] to new norms. The outbreak of COVID-19 made the local and global economies less stable and demanded more trust in places [41] where natural trust between players was more evident in history and dynamics of supply chains have developed a lot lately. Several economies even ended up in a critical situation, such as loss of jobs, reduced business market values, and reduction in sales [26,42]. The educational sector was in crisis at the start of the COVID-19 pandemic. Globally, most school systems were not prepared to jump to a complete remote teaching model [43], but on the other hand, the crisis did generate a huge load of new educational innovations, too [44]. Additionally, the global stock markets and crude oil prices were severely hit [45,46]. On a positive note, some sectors were able to adjust fast, and they found significant business boost opportunities from their skills and resource utilization capability to reinvent their business models. This added to their business values and stocks. Good examples are technology business sectors, autonomous and robotized solutions providers, online services, medical, sports , and also some public services too. The international tourism, travel, and people transportation services industries were the sectors most negatively affected by COVID-19 due to the lockdown, travel ban, and border closings [49,50]. The tourism industry was affected from the first day of the lockdown, as travelers, who are the main customers of this industry, suddenly stopped booking flights, trains, and hotels, renting cars, and engaging in entertainment activities [27] as some of the individuals shifted to the virtual tourism [51]. In addition, the COVID-19 crisis had a further severe impact on the travel industry when people started asking to cancel and receive refunds for existing bookings [27]. The continuous spread of the virus led to an extension of the travel ban in entire countries and even domestically, resulting in the historic bankruptcy announcements of well-known airports and airlines [52][53][54][55]. From the natural and environmental point of view, the effect of COVID-19 surprised the globe. For instance, as most people stayed at home during the lockdown, air quality improved noticeably in several contaminated areas, such as high-density cities in south and southeastern Asian countries [29]. In addition, a variety of wild and rare animals were spotted on city roads as they were enjoying the quiet environment of the lockdown and roaming for different purposes, such as looking for food [28]. The unexpected impact of COVID-19 on several domains included society and individuals' social and psychological behavior, such as coping with social distancing [31,56]. This required distancing came with the tendency to stay at home to reduce the likelihood of being infected and virus transmission. However, some individuals decided to take the risk of interacting with others due to their panic about food availability or having less income by staying at home for a long period [57]. Reaching the required level of food security is one of the top global requirements for living [58]. Thus, COVID-19's effect on this domain left some countries facing a critical challenge. The outbreak of COVID-19 and the aforementioned consequences resulted in severe food security impacts in several countries, such as Nepal [36], Kenya, and Uganda [37]. The sudden policies and restrictions led to a greater scarcity of food; workers, laborers, handcrafters, and farmers could not work as usual, while the restriction of movement prevented individuals from buying what they needed from other areas [36,37]. The issue of obtaining food from other areas was mitigated in several advanced countries by the use of different solutions, such as food delivery applications [58]. The mentioned global situation raises e-commerce and e-services market-related short and long-term effects and impacts the connected question, which is: What are the social behavior and demands of individuals toward utilizing e-commerce and services during and after the COVID-19 pandemic in Saudi Arabia and Egypt? The next section will review the impact on e-commerce and e-service providers in Saudi Arabia and Egypt. --- E-Commerce Market The E-commerce market in the Middle East and North Africa region is continuously evolving and grew by 35% in 2020 compared to the previous year. In more detail, the market value of e-commerce in MENA was equal to $11 billion in 2018. This number increased to $15 billion and $22 Billion in 2019 and 2020, respectively. The market is expected to reach $30 billion by the end of 2021 [59] as a result of different factors, including the spread of COVID-19 that led individuals to shop online rather than visiting stores physically [59][60][61], where the global digital transformation was boosted as a result of the pandemic [5]. Moreover, it is predicted that the value of e-commerce in the MENA region will reach $104.1 Billion and $183.4 by the end of 2022 and 2026, respectively [62]. Saudi Arabia, UAE, and Egypt are among the top players in leading this positive trend. The market of e-commerce in Saudi Arabia is considered the 25th largest e-commerce market globally [63] and is one of the top e-commerce markets in the MENA region that offers a variety of e-commerce and services to consumers by having tens of online stores that compete to gain more from the market [64]. In 2020, Saudi Arabian e-commerce revenue increased by 34% compared to 2019. Namshi.com scored the highest rate in Saudi Arabia in net sales by $126 million. It is followed by jarir.com, extra.com, amazon.com , and noon.com with net sales of $91 million, $64 million, $52 million, and $51 million respectively [65]. As the market could be affected by several factors and conditions, the forced lockdown played a significant role in boosting the Saudi market of e-commerce and services that led to a significant increment in the e-commerce and e-services market [60]. The Egyptian e-commerce and e-services market has the potential to be one of the main markets in the MENA region, where the number of internet users within the border of Egypt exceeded 53.5 million in 2019 [66]. This large number led to the rank of the Egyptian e-commerce market as the 48th largest market in the world [67]. Also, the Egyptian ecommerce and e-services market took a significant advantage in 2019 and 2020, which was affected by the global market condition and by the COVID-19 lockdown [61]. Considering the net sale in 2020, Souq.com scored the top rank with total net sales of $116 million. It was followed by bthech.com, which scored second with net sales of $21 million. The next three largest e-commerce players in the Egyptian market were lcwaikiki.eg, elarabygroup.com, and hihonor.com, with net sales of $17 million, $13 million, and $12 million respectively [67]. Not only did Saudi Arabian and Egyptian markets gain from the pandemic condition, but the global market took advantage of the lockdown and social distancing. For instance, according to Yahoo Finance, the price of a single Amazon share in the stock market was $1847.84 on 31 December 2019, while the very high demand for buying products and services online boosted the stock price dramatically to $3499.12 by 1 September 2020 [68]. In addition, eBay's stock price was investigated in the same period. It was valued at $36.11 and rose to $53.65 [69]. --- The Gap and Research Questions So far, the reviewed literature has shown a considerable effect on individuals' daily lives due to the impact of COVID-19. However, as already mentioned, some industries were able to adjust quite fast, and those industries truly took advantage of the many consequences of the outbreak. Information and communication technology , warehousing [70], logistics [71], and supply chain-related companies are among the sectors that gained the opportunities to make significant changes and gain profits from their activity during the pandemic. Interestingly, Digital citizen science projects [72] or community-based monitoring activities [39] did not seem to jump to new levels, even though social distancing could have given people a lot more time to contribute and digitalization had provided tools to do so. Additionally, several industries, such as IT business values, stocks, and food delivery companies, did skyrocket, and they found new, never before realized customer sectors [68,69,73]. Remaining at this high position in the market after the crisis requires considering several factors in terms of business operations and future plans. Porter's Five Forces model emphasizes five main factors to maintain one's business position in the market. Those five factors are potential entrants, the power of the suppliers, the power of the buyers, industry competitors, and the power of the substitutes [74]. Therefore, the impact that came from the power of the buyer presented a high demand in some online commerce and e-services when several other sectors suffered severely from the crisis and many announced bankruptcies. Thus, it would be better to investigate the possible social trend that would affect enterprises' strategic and tactical decisions on the national and international scopes. Therefore, some research studies highlighted this issue from several perspectives, such as comparing consumer trends before and after, but not during, the pandemic [75,76] or focusing on some specific industries, like exporting services [77] and the healthcare industry [78]. However, there is a lack of research highlighting the possible social trend toward utilizing online services and commerce in Saudi Arabia and Egypt as two of the main online markets in the Middle East, as the current studies were limitedly explored. Also, the other side of the gap would be presented in the current studies by focusing on the periods before and after the pandemic, as mentioned before, or only during the COVID-19 period, but not on exploring the shift in consumer preferences during and after the pandemic in the Middle East. This briefly described gap led to the development of the main research question, which is as follows: • What are the social behavior and demands of individuals toward utilizing e-commerce and services during and after the COVID-19 pandemic in Saudi Arabia and Egypt? Answering this question will provide more insights and give the ability to build a hypothesis for a large-scale study. --- Methodology To answer the main research question and to understand the possible social behavior of utilizing e-commerce and services after the COVID-19 period. The methodology of this study was started by analyzing the current literature, which was followed by collecting the necessary data. Then, a statistical analysis was conducted. This stage consists of four stages. The demographic analysis comes first to understand the nature of the sample. Then, the health impact of the pandemic was explored on the sample to find the possible relationship between the sample health record and their behaviors toward e-commerce utilization. The third stage explored the status of consumers' online purchasing and their future likelihood trend in the e-commerce industry, while the fourth stage focused on the regression analysis to get more insights into the relationship between the COVID-19 health record and consumer trends in e-commerce. Collecting data was essential to obtain insights. Thus, an online survey was created to obtain the required data. The sampling method followed the snowball sampling technique, where each participant was asked to invite others to participate. The data was collected via multiple sources by utilizing a semi-structured online survey [79,80]. This survey consists of four sections, as illustrated in Figure 1. This survey was mainly in Arabic, the main language in the investigated countries. Thus, a brief description of the study and a consent statement was written at the beginning of the online survey in Arabic and English to ensure a complete understanding of the main aim and the research-based intention in using and analyzing the collected data, as well as to give participants the complete rights and freedom to accept or reject the invitation to be part of the study. scopes. Therefore, some research studies highlighted this issue from several perspectives such as comparing consumer trends before and after, but not during, the pandemic [75,76 or focusing on some specific industries, like exporting services [77] and the healthcare industry [78]. However, there is a lack of research highlighting the possible social trend toward utilizing online services and commerce in Saudi Arabia and Egypt as two of the main online markets in the Middle East, as the current studies were limitedly explored Also, the other side of the gap would be presented in the current studies by focusing on the periods before and after the pandemic, as mentioned before, or only during the COVID-19 period, but not on exploring the shift in consumer preferences during and afte the pandemic in the Middle East. This briefly described gap led to the development of the main research question, which is as follows: • What are the social behavior and demands of individuals toward utilizing e-com merce and services during and after the COVID-19 pandemic in Saudi Arabia and Egypt? Answering this question will provide more insights and give the ability to build a hypothesis for a large-scale study. --- Methodology To answer the main research question and to understand the possible social behavio of utilizing e-commerce and services after the COVID-19 period. The methodology of this study was started by analyzing the current literature, which was followed by collecting the necessary data. Then, a statistical analysis was conducted. This stage consists of fou stages. The demographic analysis comes first to understand the nature of the sample Then, the health impact of the pandemic was explored on the sample to find the possible relationship between the sample health record and their behaviors toward e-commerce utilization. The third stage explored the status of consumers' online purchasing and thei future likelihood trend in the e-commerce industry, while the fourth stage focused on the regression analysis to get more insights into the relationship between the COVID-19 health record and consumer trends in e-commerce. Collecting data was essential to obtain insights. Thus, an online survey was created to obtain the required data. The sampling method followed the snowball sampling tech nique, where each participant was asked to invite others to participate. The data was col lected via multiple sources by utilizing a semi-structured online survey [79,80]. This sur vey consists of four sections, as illustrated in Figure 1. This survey was mainly in Arabic the main language in the investigated countries. Thus, a brief description of the study and a consent statement was written at the beginning of the online survey in Arabic and Eng lish to ensure a complete understanding of the main aim and the research-based intention in using and analyzing the collected data, as well as to give participants the complete rights and freedom to accept or reject the invitation to be part of the study. The first section of the survey focuses on demographic information, i.e., independent variables such as age, gender, and marital status. The following second section highlights factors that could describe the financial situation of the participant. The second section also provides independent variables that would affect decisions about post-COVID-19 trends in utilizing e-services and e-commerce. Employment status and monthly income are the factors explored in the second section. The third section investigates the direct impact of COVID-19 on participants from two perspectives: the stay-at-home orders during the peak periods of COVID-19 and whether the participants or any of the people surrounding them were infected with COVID-19. In direct connection to the issue, the fourth section investigates the participants'the consumers-intention toward e-services and e-commerce in four timeframes. These timeframes are before the outbreak of COVID-19, during the peak of the pandemic and the highest movement restrictions , after the partial ease of the governments' restrictions, and after the COVID-19 stage, when the world gets proper and complete medicines and protections against the pandemic. To distribute the developed survey, a professional survey distribution organization was contracted to distribute, collect, and manage the process. This process included the following stages: --- • Selecting the correct sample, such as the geographical location and the age; • Preparing data for analysis at the preprocessing stage. This included removing incomplete records and incorrect entries. After that, data were prepared for analysis. The following section will analyze and present the results of the conducted study. --- Data Analysis and Results The conducted survey was distributed in two Middle Eastern countries: Saudi Arabia and Egypt. The invitation to participate in the questionnaire was sent via social media with the help of a hired marketing specialist. Thus, the convenience sampling methodology was used to collect the required data [81], which resulted in a sample size of 155 responses. The analysis of the collected data went through four main stages. The first stage was the implementation of demographic analysis to view the main statistics and obtain the first insights. This stage covers the first two sections of the conducted survey, illustrated in Figure 1. The second analysis stage highlights the direct health effects to conduct a regression analysis that investigates the relationship of several factors that could affect online consumer behavior toward e-commerce and e-services. The third stage exposed the trend of utilizing e-commerce and e-services in four timeframes, as mentioned in the methodology section. The fourth analysis phase investigated the relationship among variables. For the findings of these stages, the detailed analysis and the obtained results are discussed in detail in the discussion section at the end of this manuscript. --- Demographic Analysis To understand the nature of the participants, Table 1 shows the demographic information and the independent variables of the collected data. The data, as shown in Table 1, presents the 155 people who participated in the data collection process: 104 from Saudi Arabia and 51 from Egypt. With regard to age groups, the participants were divided into five groups: 18-30, 31-40, 41-50, 51-60, and 61+. However, the analysis revealed that there was only one participant older than 60 years. Therefore, the last two groups were merged into one group to cover participants older than 51. The table also shows the approximate similarity between the participating genders. The sample contains 82 males and 73 females, while, in terms of marital status, the sample consists of 84 single participants, 69 married participants, and only two divorced participants. With regard to employment status, the study divided participants into five groups. The majority were unemployed: 63 . The second highest group of 42 participants were employed in the private sector. The third highest group was participants employed in the government sector: 24 . This was followed by students and business owners: 14 and 12 , respectively. To gain more insights into the financial situations of the participants, Table 1 is used to reveal that the obvious majority of 107 of the participants were earning less than 10,000 every month in the local currency; those currencies are the Saudi Riyal and the Egyptian Pound . The main reason for not calculating the income using a unified currency is to consider the different costs of living in both countries. In addition, about 46% of the studied sample's monthly income was negatively affected by COVID-19, while for 52%, it was not affected. Only 3% found more opportunities to raise their income during COVID-19. Lastly, nearly half of the participants were suffering from full-day curfews that the governments ordered to contain the outbreak of COVID-19. The following subsection discusses the trends of utilizing e-commerce and services by the studied sample. --- Health Impact of COVID-19 The impact of COVID-19 affected the behavior and the future willingness to utilize e-commerce and e-services. Thus, every participant was asked if they, their families, or their relatives had been infected with, or had died from, COVID-19. It is shown in Table 2 that 88 participants were not infected by the outbreak of COVID-19. This leads to the conclusion that 43% were affected directly or indirectly by the infection of the virus. In detail, 17 participants, 33 of their families, and 42 of their relatives and friends were infected with COVID-19. Also, 5 of participants said that their family members had been killed by COVID-19, and, surprisingly, 39 reported that their relatives and friends had died of the disease. * Please note that each row would take a value from 0 to 155 , as some participants could be infected PLUS have some of their families, relatives, or friends infected as well. Note: The phrase "at least . . . " in the statistics of the above paragraph is mentioned as the participants were asked whether or not they had experienced cases of COVID-19. A "yes" answer indicates that they would have experienced one or more cases of COVID-19. --- Trend Analysis The initial step of exploring the participants' views toward utilizing e-commerce and services was dedicated to exploring the general intention of utilization . Four time periods are used: before COVID-19, during the peak of the virus , after the easing of restrictions with the enforcement of social distancing), and after the virus. The table shows a slight increase in utilizing e-commerce and services during the outbreak from the first stage. This would indicate that the majority of stage 2's customers were the same customers as those before the outbreak. Participants who did not utilize the investigated services in the first two stages were asked for an explanation for their decisions. These were a desire to see and touch the products physically before buying them so as to feel comfortable and ensure the intended quality, the preference for going outdoors, the trust level toward the offered products and service providers, and late delivery. Also, the intention toward e-commerce and e-services decreased a little with the ease of the movement restriction, to be 0.57 instead of 0.63 in stage 2. Although the difference between the stages was not significant, participants were asked about the reason for this reaction. The intention to go outdoors with care and to feel free was the main reason. The information presented in Table 3 gives general indications of utilizing e-commerce and e-services by individuals in the four explored periods. However, it was expected that some participants did not consider specific activities while surfing the web, such as utilizing e-commerce and e-services activities. Therefore, Table 4 presents detailed information as a result of asking each participant. * Please note that each row would take a value from 0 to 155 . The percentage of those details is illustrated to be readable in Figure 2. Regardless of the last column of each category , which shows participants who do not like to utilize the examined products and services, the busiest period was during the curfew. However, the figure gives a negative indication of the future behavior of utilizing e-commerce and e-services. Most of the utilized e-commerce products and services returned to normal situations as before the outbreak of COVID-19. Additionally, the results show that utilization of some products or services is expected to be even lower than the normal trend when one compares the periods before the outbreak and after the virus, such as subscribing to online TV and video streaming services. Regarding the results in Figure 2, buying food and groceries online was the trend among 43% of the participants during the curfew compared to only 13% before the outbreak. This number declined by about half, to 21%, after the easing of movement restrictions. In addition, participants are aiming to reduce their online buying to only 15% after the end of the pandemic. The trend of buying cooked and fast food online and taking advantage of delivery applications was admitted to by 29% before the outbreak. This trend jumped to 52% during the curfew and went back to 35% during the social distancing period, then to the expected 28% after COVID-19. Regarding the results in Figure 2, buying food and groceries online was the trend among 43% of the participants during the curfew compared to only 13% before the outbreak. This number declined by about half, to 21%, after the easing of movement restrictions. In addition, participants are aiming to reduce their online buying to only 15% after the end of the pandemic. The trend of buying cooked and fast food online and taking advantage of delivery applications was admitted to by 29% before the outbreak. This trend jumped to 52% during the curfew and went back to 35% during the social distancing period, then to the expected 28% after COVID-19. Buying hygiene products was also affected clearly by the hit of COVID-19. The increase in the utilization ratio from 18% before the outbreak to 39% during the curfew proves this claim. Again, this number declined to 19% during the ease of the restriction period and is expected to go down to 14% after the pandemic ends. Another example is the trend towards subscribing to online meeting services for personal purposes. The restriction of movement led to a skyrocket in the utilization ratio, from 29% before the pandemic to 66% during the curfew, which was the highest utilized category in the figure. This ratio went back to 39% during the social distancing period and is expected to decline to 21% after the pandemic. The next subsection will run a regression analysis to explore relationships between the factors that would affect the trend of utilizing e-commerce and e-services. --- Regression Analysis The main objective of a regression analysis is to explore the possible effect of COVID-19 infection on the participants' decisions to utilize e-commerce and e-services during the social distancing period and after COVID-19. For the analysis, the examined variables are Buying hygiene products was also affected clearly by the hit of COVID-19. The increase in the utilization ratio from 18% before the outbreak to 39% during the curfew proves this claim. Again, this number declined to 19% during the ease of the restriction period and is expected to go down to 14% after the pandemic ends. Another example is the trend towards subscribing to online meeting services for personal purposes. The restriction of movement led to a skyrocket in the utilization ratio, from 29% before the pandemic to 66% during the curfew, which was the highest utilized category in the figure. This ratio went back to 39% during the social distancing period and is expected to decline to 21% after the pandemic. The next subsection will run a regression analysis to explore relationships between the factors that would affect the trend of utilizing e-commerce and e-services. --- Regression Analysis The main objective of a regression analysis is to explore the possible effect of COVID-19 infection on the participants' decisions to utilize e-commerce and e-services during the social distancing period and after COVID-19. For the analysis, the examined variables are presented in Table 5. The independent variables were classified into two main categories: infection with COVID-19 and death caused by COVID-19. In the first category, participants were divided into four groups according to the closest infection case to the participant. The first group contains participants who were infected directly by the virus. The second group contains people who were not infected but whose family members were. The third group includes participants who did not suffer from COVID-19, and nor did their families, but whose relatives or friends did. The last group contains people who did not encounter COVID-19 cases in their inner circle of family, friends, or relatives. The second category explores the effect of deaths resulting from COVID-19 in terms of three groups. The first group contains participants whose family members have died of COVID-19. The second group includes participants whose relatives or friends died of COVID-19. The third group contains participants who had encountered any death cases due to COVID-19. The presented data and its analysis show that the relationship between the intentions of utilizing e-commerce and e-services during the social distancing and after-COVID-19 periods is loosely coupled, where all p-values are above 0.05. Thus, the null hypothesis is accepted for all cases in which none of the virus infections or death cases would affect participants' intentions. --- Theoretical and Practical Implications --- Principal Findings The collected sample covered individuals from Saudi Arabia and Egypt. More than half of the sample was aged between 18-30, and the sample was approximately balanced between males and females. More than half of the collected sample was not infected by COVID-19, and the other half were either infected or related to infected people. In addition, more than a quarter of the sample encountered death cases because of COVID-19 in their families or relatives. The results of the trend analysis Section 5.3 were clearly in favor of the claimed significant direction toward e-commerce and e-services by the public, which led to the skyrocketing of stock prices for companies in this sector [68,69]. However, the result raised some concerns regarding the future status of those companies. As the consumer is considered one of the five main forces that Porter highlighted as their effect on the rivalry of businesses in the market [74], the intention of the majority of participants indicates that the utilization of e-commerce and e-services at this high rate would not continue after the pandemic ends. Although it was discovered in the literature that market research predicted the rise of e-commerce in Saudi Arabia in the coming years [65], this study is distinct in highlighting some alerts conducted in considering the post-pandemic of local consumer trends, including the drivers and challenges. This could open the doors for expanded scientific and market research to consider the highlighted factors on a larger scale to facilitate the promising future of e-commerce industry in the Middle East, especially in Saudi Arabia and Egypt. --- The Effect of COVID-19 Before the regression analysis was conducted, it was expected that there would be a limited relationship between the health effect of COVID-19 and the psychological intention of utilizing e-commerce and e-services. However, the presented results in Table 5 exceed the expectation by neglecting the effect of the virus on the intention level of utilizing ecommerce and e-services completely. This would indicate that though individuals adapted to the online norm to process different daily tasks and businesses, people might be eager to return to pre-COVID-19 behavior after the pandemic ends, which unfortunately does not seems to do so, at least before 2023 [82]. However, there are several questions that should be considered in further detailed research, such as how fast individuals might return to the pre-COVID period. Are they going to return immediately or gradually? Do individuals need to keep some of their new online skills or avoid them completely? What is the proper date to consider as the starting point of the post-pandemic era? --- Conclusions COVID-19 is considered to be one of the most significant historical pandemics so far. It is the first pandemic that stopped the world as we know it, especially in travel and social entertainment, and recreational contexts, which ended up transforming individuals' ways of living, at least for the time being, but in many permanent ways, as some already have predicted. The pandemic has "forced" nations and the global population to do fastphased industrial digitalization and engage in the societal digital transformation of daily lives, the utilization of social communication channels, and change the way companies do their business. Still, some of the areas of digitalization did not jump into higher gear, like mass-customized products . Nevertheless, the advancement of information and communication technologies led to the offering of several products and services via online channels and generated great shortages of physical devices needed for online channels, like webcams, mobile internet access modems, computers/tablets, etc., in general. This push towards a robust online presence helped provide most nations with the essentials of living, and individuals relied on e-commerce and e-services heavily during the outbreak. This study explored the social trend of utilizing e-commerce and services during the easing of restrictions and social distancing as well as during the expected stage after the end of the pandemic. The results indicated that this exceptional trend toward e-commerce and e-services is expected to shrink after the normal situation returns. In addition, the results have shown that the lockdown period did change habits during the pandemic, and it is necessary to explore the facts of how this change would last after the pandemic, particularly if the post-pandemic period is like the pre-pandemic period, where people claim that they would want to return to the old norms and habits, as they are waiting for the end of the pandemic so that they can return to their "normal" lives. However, the reality is that many companies have changed their business models and operational practices, and they will never return to being exactly as they were in pre-pandemic times. This means there will not be a pre-pandemic "normal" available for people, no matter what they want. Business organizations, owners, and economists should consider the expected hope of the customers to return to pre-COVID-19 norms, which could mean a second wave of needs to adapt again with their business models to keep up with the behavior changes of their customer base and to avoid any unexpected issues, such as the availability of products and services on-premises, and the possible impact on the global stock markets. --- Limitations This research study encountered several limitations. Although the studied sample consists of 155 individuals from two countries, it is recommended to conduct similar studies on wider ranges. Also, researchers should consider conducting similar studies in different countries, especially from the advanced world, as the results would help to create a comparison of the effect of the virus on several regions globally. Additionally, as the trend analysis presented the limited effect of COVID-19 on the individuals' intention toward e-commerce and e-services, it was planned to conduct a regression analysis to examine the effect on each country individually. However, dividing the participants would limit the sample size in each country and might lead to an unreliable result. Thus, it is recommended again that similar studies be conducted to focus on other demographic variables. --- Future Work The limitations that were discussed in this study led to the proposal of a new project aiming to develop research hypotheses to forecast the future global trend of societies toward e-commerce and e-services. The proposed sample size would exceed 10,000 to synthesize the results. --- Data Availability Statement: Data will be available upon request. Acknowledgments: Authors would like to acknowledge LUT, for providing Ari Happonen financial support for internationalization, which has made this international collaboration publication possible. --- Funding: This research received no external funding. Institutional Review Board Statement: Ethical review and approval were waived for this study due to the nature of the study, that did not concern or expose any personal information. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. --- --- Abbreviations

COVID-19 is a major global crisis affecter, changing global norms and societal behavioral models. Many companies have faced existential crises, but on the other hand, businesses that were and are helping others to boost digitalization, ICT and software solutions deployment, remote communications integration, e-commerce & e-services, and so on, have boosted their businesses, as people shifted online during the global lockdown and international travel restrictions. Our work explores the trend of e-commerce and e-services utilization during the ease of restrictions and the social distancing period to forecast the trend continuation patterns after the pandemic. An online survey was conducted and targeted individuals in Saudi Arabia and Egypt, resulting in 155 participants. The data were analyzed from four perspectives: demographics, COVID-19 health impact, trend analysis, and regression analysis. The results indicate heavy utilization of e-commerce and e-services during the global movement restrictions and travel bans. This trend has, however, significantly reduced during the ease of restrictions and social distancing period. Utilizing e-commerce and e-services in Saudi Arabia and Egypt, based on the research data, is positively correlated to the outbreak conditions. On the other hand, current data still does not give clear indications, and this pattern is going to be mostly, partly, or not at all permanent now as societies are returning to mostly a free movement of people and marginally restricted social distancing times.

publicity, 46 study characteristics , 2 10 30 58 the use of selection processes, 2 10 30 57 interest in other areas of study/careers, 2 30 a lack of support 2 49 and a lack of self-confidence. 49 How these factors influence the different types of motivation and one another is unknown. Knowledge about the mechanisms and the interplay between factors can help study counsellors to stimulate a choice based on autonomous motivation. This can enhance student performance and well-being. To understand the complex mechanisms, students' background characteristics, such as gender, SES and ethnic and parental background, need to be taken into account as well. For example, female 7 8 45 47 48 59 and ethnic majority students 32 report more altruistic motives, while male 45 47 48 59 and low SES students 32 report more extrinsic motives. This disparity has consequences for study choice counselling and for assessment of motivation during selection processes. It has been argued that for purposes of widening access, interventions should target under-represented groups to encourage them to apply to medical school. 9 To be able to do this effectively, there needs to be a greater understanding about the factors that encourage or discourage students to apply to medical school. 46 Most studies addressing the motivation for studying medicine have gathered information retrospectively from medical students 5 8 42-45 49 51 52 56 or even physicians. 47 Only few investigated this among high school students 2 10 32 46 50 55 or undergraduate students 55 57 during their study choice or application process. Moreover, a qualitative study can help in understanding the complex interplay between the factors that are involved. We had the opportunity to study this in a setting where admission is changing from lottery-based to selection-based. In the lottery, all applicants had a chance to be admitted and the chances of admission increased parallel to applicants' pre-university grade point average, while in selection only the top-ranked applicants based on academic and non-academic selection criteria are admitted. This policy change has profound consequences for those who are interested in studying medicine, which makes it an interesting context for research. The following research questions guided our study: 1. What types of motivation do high school students have for applying to medical school and how can these motivations be classified according to SDT? 2. Which factors play a role in reaching these different types of motivation? and 3. Do factors playing a role differ for students with different background characteristics? --- Methods Because the mechanisms and the interplay between the factors that are involved are unknown, we chose a qualitative design. We drew upon a constructivist point of view, in which knowledge is considered socially constructed by the people involved in the research process and our reality is viewed as constructed by our social and individual contexts. 60 We took a phenomenological approach, and the aim was to understand the complex world from the point of view of those who live it, that is, the subjective experiences of high school students during their study choice process. 60 61 We conducted interviews because interaction between the researcher and the researched is crucial to come to an understanding of meanings. 62 Multiple perspectives and perceptions of a variety of people were sought to enable better interpretations of the meanings of the phenomenon under study. We used the COnsolidated criteria for REporting Qualitative research checklist for reporting qualitative research for the assurance of good quality . 63 setting In the Netherlands, high school students graduating from 6-year pre-university education with science subjects are eligible for entry into medical programmes. Students choose these subjects in the third grade. At the time this study was conducted, March-December 2015, the admissions system was in the process of changing from lottery-based and selection-based admissions to only selection-based admissions from 2017. The lottery procedure was weighted for students' pre-university grade point average, and students' likelihood of admission increased with a higher grade point average. In addition, applicants could participate in a selection process by applying halfway through their final year; for the lottery, students could apply until nearly the end of the final year. The proportion of the students that medical schools were allowed to admit through selection processes increased from 50% to 100% through the years. Annually, nearly 2800 students can enrol in medical programmes in the Netherlands. --- --- Procedure Directors of two high schools were approached and invited to participate in this research. First, in semistructured interviews with the study counsellors of the high schools that were involved in the guidance of the students, we gathered information about which elements they identified in students' study choice processes , which yielded topics to address in the subsequent interviews with students. These interviews also served to gather information about the composition of the schools' student populations and the way in which study choice counselling was organised. The study counsellors purposively recruited students they knew considered or had considered studying medicine for one-time participation in the study. All students participated and none dropped out of the study. The one-on-one interviews were conducted by AW at the two high schools and lasted between 31 and 52 min. At the beginning of the interview, AW introduced herself and explained the purpose of the study. At the end of the interview, students were asked to draw a graph showing how their motivation had increased and/or decreased over time from the moment they considered medicine as a possible study choice and to write down when and why changes occurred. This graph served as a summary and stimulated the recollection of further important aspects. Interviews were semistructured and were altered based on arising topics in earlier interviews in an iterative process and adjusted to the type of participant and to the expected study choice phase of students based on the grade they were in. See box for the interview guide used in the interviews with the students. Interviews were audiotaped and transcribed verbatim. All transcriptions were checked for accuracy. A summary of each interview was presented to the respective interviewee as a form of member checking. Students either approved their summary or provided no response. When sufficiency and data saturation were reached, no more interviews were conducted. data analysis AW read and reread and familiarised herself with the data. AW coded all interviews using ATLAS. ti. The first interview was coded together with a second researcher, UI. Further in the coding process, another four interviews were independently coded by AW and UI. Whenever there were differences in coding, these were discussed until consensus was reached. Because there were very few differences and they mainly concerned the phrasing of new codes, AW coded the remaining interviews. We analysed all data using SDT as a theoretical framework. The different motivation types formed the basis for a coding template for students' reasons to pursue a medical career. 64 The quotations were first identified as expressions of a particular motivation type and further categorised . Open coding was conducted to identify the factors that influenced students' motivation in a positive or negative manner. 65 Factors identified in quotations were first categorised as factors of influence, next as positively or negatively influencing factors, and finally according to the topic and subcategory . Quotations that contained expressions of students' basic psychological needs were also coded. The findings were discussed and agreed upon by the full research team. Reflexivity is a critical aspect in the analytical process. We acknowledge that while using SDT as a theoretical framework for our analysis allows for a deeper understanding of students' motivation for studying medicine, this particular focus may also cause us to overlook other relevant aspects. We tried to keep this to a minimum by keeping notes, 'memos', of findings outside our theoretical framework that seemed relevant for understanding the mechanisms. --- ethical considerations Participation was voluntary. We provided participants with information about the purpose of the study and the handling of personal data, after which they signed an informed consent form. The audio-recordings and interview transcripts did not contain identifying information about the participants, have been stored with anonymous descriptions and are accessible only to the principal researcher. Students received a €15 gift card for their participation. This study has been approved by the Netherlands Association for Medical Education Ethical Review Board , dossier number 408. --- results Three study counsellors and 24 students participated in this study. Tables 2 and3 provide overviews of the distribution of participants, and the identified categories and subcategories, respectively. --- Motivation for applying to medical school Both subtypes of autonomous motivation were addressed to a great extent by the students. Intrinsic motivation was seen in a interest in medicine and the working of the human body and altruistic reasons . The desire to help people was more prominent among females than among males. Students stressed the importance of choosing an area of study they find interesting. [The most important reason is] that I am interested in the human body, in combination with working with people. […] This has actually always been most important. I believe that when you choose a study, interest, wanting to learn, is most important. You can say, then this, then that, then you make a lot of money or something, but you have to persevere in your studies, which can only be done when you are interested. With regard to identified regulation, students expressed that they were interested in studying medicine because they value and acknowledge the importance of medical study and the medical profession. --- Open Access --- Main factors influencing students' motivation for studying medicine Getting acquainted with the medical profession The most prominent factor that influenced students' autonomous motivation and strength of motivation was experiencing or gaining knowledge about the medical profession. This was reached when students themselves or relatives had been patients or when students had come into contact with the medical profession due to having medical professionals in the family or other role models, which led them to realise the importance of the medical profession and elicited their interest in medicine. Well, that thing with my cousin, that made quite an impact. It really made me see how much doctors do. A general belief was that having a medical doctor as a parent made it easier to gain information about the medical profession, as well as getting access to internships in healthcare. However, while several school A students had a medical doctor parent, none of the school B students did. Internships were considered most valuable for getting acquainted with the medical profession, and gaining such experiences was the main factor that had a positive influence on autonomous motivation for studying medicine. Students experienced how doctors make diagnoses and treat diseases, which often confirmed their interest in medicine. The internship was so much fun. Getting acquainted with the medical profession through medical professionals and internships also served as a reality check, which reduced their autonomous motivation. Learning about the downsides made some female students realise that their image of the profession was inaccurate and that they had romanticised the aspects they valued and liked about the profession. Students considered whether the information they had gathered about the profession matched their own preferences and perceived capabilities. Yes, that is partly why I lost interest in medicine. I thought, when you, as a doctor, hear such things, you must be able to cope with not everyone being healthy, and that anything can happen. That makes me wonder, can I handle that?' For some female students, TV programmes also played a role. While reality shows elicited autonomous motivation, popular TV series, such as Grey's Anatomy, made students aware of the 'coolness' of the medical profession. Yes, in primary school we watched TV programmes. […] Sometimes about history, sometimes about nature, about animals. Sometimes it was about human beings, that they were ill and things like that. I think that made me realise what I liked and didn't like. selection Overall, many students appeared to have little knowledge about the admissions process. There seemed to be differences between students from the two schools. Students from school B in particular lacked knowledge about the admissions process. This was true even for those in their final year of high school. They were either unaware of the change from the lottery to selection or what selection entailed. While males mentioned that selection had no effect on their motivation, for a number of females it had a significant impact; one female student even refrained from applying due to the presence of a selection process. Their motivation decreased due to the difficulty in gaining admission to medical programmes and students lost hope of being successful in the selection process because they felt insecure about their competence, particularly compared with other applicants. Sometimes [my motivation] was higher, sometimes lower. It was always high, but sometimes I thought, I always wanted to help people, and this of course is very much helping people by means of health and by means of curing people, enhancing quality of life. But at some point, at age 12, I knew how difficult it is to become a doctor. Then I thought, I need an alternative. Building a CV appeared to be the aspect of the selection process with the most demotivating effect on students. There seemed to be differences in the experiences of the students from the two schools with regard to building their CV and their chances of success in the selection process. Some students from school A thought that attending their school would be an advantage in the selection process because the education was perceived to be superior to that of other schools. Moreover, some students from school A reported that they had people in their network who could help them prepare for the selection process because of their experience as a student, medical professional or because they were involved in the selection process at a medical school. In addition, many of those students had easy access to healthcare-related activities for CV-building purposes. One student from school B expressed that she heard that applicants with a foreign last name would be discriminated against. Open Access high school subjects Studying high school subjects related to medicine, seeing how these relate with practice and finding out how the human body works boosted students' interest and enjoyment. During the biology classes, I noticed that I found it all very interesting, especially the human body, and that I found it interesting to learn how it works. Then I figured that the field of study that addresses the human body most is the medical field. --- study choice-related activities Attending open days and actively thinking about their career choice generally increased students' strength of motivation. Open days could also have a demotivating effect when the talks were boring. Another common factor that was mentioned as having a negative influence on motivation for studying medicine was a growing interest in other courses. 'When I attended the open day, they were talking about the study of medicine. It was so tedious. And also, they didn't present it in an animated way, it was so boring. If I compare it to pharmaceutical sciences … That is not really something I aspire to study, but they presented it in a much nicer way. It came across as more interesting than the medical field, while I actually want to study medicine. study and job characteristics Some aspects of medical education and the medical profession made students doubt their study choice, such as the hierarchical culture, job perspectives, responsibility for the life and death of patients, working conditions and the difficulty and length of medical programmes. There was one factor that particularly played a role for female students from school B. They expected a clash with their religious beliefs . Yes, what I hear a lot lately is that the job opportunities after graduation are very limited. That is really something that holds me back a little. Otherwise I would definitely have chosen medicine, I'm sure about that. I would love to work in a hospital, but there are all kinds of [clothing] regulations, and I am not sure whether that suits me. I am … I wear a headscarf, and well, I comply with the dress code of the Islam, and I am not sure whether that goes together. These aspects did not seem to prevent most students from choosing medicine, which was explicitly expressed by students. --- Basic psychological needs Autonomy Students described how they wanted to feel autonomous in making their career choice. Although they mentioned that parents' and, to a lesser extent, friends' opinions were important, they stressed the need to do what they want. Many students pointed out that they expected it to be difficult to remain satisfied with their professional life when their career choice would be driven by what others want instead of what they want. Students from school A expressed this more firmly than students from school B. I didn't want to do something because of my parents. I wanted something for myself. If I do something to prove myself to someone else, it won't be fun. Parental influence was reported by the study counsellors as one of the most important factors influencing students' study choices. Parents seem to stimulate their children to make their own choice, although at the same time strongly conveying their expectations to their children. Parents may have a stronger influence than is perceived by students, and students' need for autonomy may therefore be less satisfied than they perceive it to be. So far, my father ranked medicine number one, but I don't think this had an influence on me, or that I wanted to study medicine because of him. I'm doubting a little, […], whether I am influenced by people. competence Female students expressed feelings of competence that were related to the study of medicine, the medical profession and their participation in a selection process, while both male and female students expressed the need to be challenged in their studies. Then I really thought, okay, this is me. That sounds really stupid. This, I just know, sounds very stupid and really very arrogant, but I think I will be a good doctor. I think I just have the capability to do all that. But yes, the admissions criteria did not scare me because I knew I met them. I found it a bit scary, the lottery and selection […]. And I am very smart. I don't want to be arrogant. I am intelligent, but I don't have a pre-university attitude. And yes, my grades are always about 6.5 average. And yes, I thought, this doesn't quite fit … my attitude and yes … It was a bit scary; it made me a little insecure. Feelings of competence increased students' motivation for studying medicine, for example, passing a test in a school subject related to medicine. medical profession. This gave some students the desire to belong to that professional group. Nice people in the profession, that's the impression I get. […] I like all my mother's colleagues, very much. So … they all have a very good sense of humour. They all make funny jokes, for example about gastroenterology being the diarrhoea-unit. That kind of stuff. So, that is a lot of fun. summary An overview of the factors that play a role in high school students' motivation for studying medicine is provided in figure 1. We found that students' main motivation was of the autonomous type, such as interest in science, helping others and valuing the medical profession. Controlled motivation concerned parental pressure, prestige and salary. Healthcare experiences, the presence of selection processes and having a parent in the medical field were important factors for students' study choice process and motivation, but social disparities in exposure to medicine exist. --- dIscussIon --- Principal findings Mainly autonomous motivation for studying medicine, that is, the desire to help people and a scientific interest, was reported, but controlled motivation, such as prestige, was also reported. This is in line with previously reported findings. With regard to how the motivation for studying medicine is formed and the role of selection processes in this, there appears to be a complex interplay between healthcare experiences, growing up in a medical family, the use of selection processes and motivation . We found exposure to healthcare to be a crucial factor impacting students' motivation. Such exposure fostered an interest in medical subjects and the desire to help others and made students value the medical profession, but it could also serve as a reality check and make students realise that they previously had unrealistic expectations. The beneficial effect of healthcare experiences with real patients on students' motivation has been reported previously. 32 Moreover, the variety of factors in combination with the interplay between these factors highlights how the motivation for studying medicine is shaped depending on the individual student. Growing up in a medical family enhanced students' motivation to study medicine and provided easier access to acquiring healthcare internships and assistance in the selection process. Healthcare experience is often one of the selection criteria. This highlights an inequality in access to resources relevant for both the study choice process and the selection process related to students' background characteristics. In our study, we found that the perceived inequality could be a demotivating factor for students without a medical network. This may induce self-selection among students without such a network. This might especially be true for students who grow up with the notion that they are not fit for a university and have no doctor role models to inspire them. 2 32 As a result, the intended diversity within the student population could be compromised. The change from lottery-based admissions to selection-based admissions, as is occurring in the Netherlands, allows student diversity to be monitored to Open Access determine whether the use of selection processes negatively effects student diversity. --- Implications of the study Our findings provide initial insight into which areas can be addressed to help students choose a field of study based on autonomous motivation. Although most interventions aimed at widening access are focused on facilitating the access of under-represented students by stimulating the fair treatment of applicants in selection processes, 66 67 it seems worthwhile to simultaneously direct efforts towards the recruitment of a diverse group of prospective applicants. Recruitment efforts need to start with guiding high school students towards a study choice based on the right information and autonomous motivation. SDT posits motivation along a continuum and this can change from the autonomous to the controlled type and vice versa. 33 34 The current study shows the complexity of how motivation is formed. Each individual student is motivated by a combination of different factors. Similarly, students experience a variety of demotivating factors. These demotivating factors can 'pull' or 'push' students away from their choice to study medicine. 68 A student can be pulled away from her choice because she develops interest in other subjects or because she finds that another study better suits her values . A student can also be pushed away from her choice because she wants to avoid negatively valenced aspects of her choice to study medicine . The interplay between motivating and demotivating factors will lead to differently formed choices for each individual student. Therefore, tailor-made guidance of high school students appears necessary. Based on the findings, we plan to create an interview guide for study counsellors to determine which factors play a role for each student. The interplay between the factors can then be explored by means of mind mapping. This way, the motivation for choosing a study can be determined and when the controlled type of motivation is dominant, the student can be stimulated to choose a study based on autonomous motivation. A similar approach could be useful for guiding medical students towards their specialty choice. Knowledge about how autonomous motivation for different specialties is formed may provide possibilities to motivate students to opt for specialties that are aligned with the changing healthcare structure and societal needs, 69 such as family medicine and geriatrics. Previous research has shown the negative impact of the lack of connections to physicians and the misconceptions of both students and parents about the medical field on under-represented students' pursuit of a medical career 57 58 and there seems to be a desire for more information about the medical profession and study, and the admission processes. 32 58 Exposure to healthcare seems crucial in this respect. 2 58 Both high schools and medical schools should realise that students who do not grow up in a medical family have less access to such information and activities. 58 To stimulate a choice based on autonomous motivation, high schools could encourage students to consider which subjects they like in school and offer time and possibilities for internships. In addition, a programme, such as the one used at school B, in which students who are first in their family to enrol in a university are assigned a mentor relevant to their career preferences, can provide such students with information. Despite the study choice activities offered at school B, however, many of those students still seem to be disadvantaged due to a lack of medical professionals in their network. A striking finding was that many students seemed to be poorly informed about the selection process. 58 Moreover, students felt they would be less likely to succeed than their competitors because they felt that others may have more knowledge about medicine and selection processes and more resources to prepare for the selection process. Indeed, our findings indicate an inequality in this respect. Under-represented students' concerns about the complexity of the admissions process 58 and not being able to compete with other applicants have been reported previously. 57 When medical schools wish to incorporate healthcare experiences as one of their selection criteria, they should acknowledge the unequal access to such experiences among high school students. For those who lack the opportunity of gaining healthcare experiences through their own network, medical schools could facilitate this. Furthermore, considering that some students have access to more information about the selection process through their network, selection committees should be as transparent as possible about their process. 45 48 49 52 strengths and weaknesses of the study The qualitative design allowed the exploration of the complex mechanisms underlying students' motivation for studying medicine and more specifically the role of selection processes. The use of SDT as a theoretical framework was helpful in understanding how autonomous motivation, the most desirable type, is formed. Having members with various backgrounds on the research team strengthened the validity and reliability of the data analysis. Using interviews with study counsellors, in addition to findings from previous literature, to generate topics for the student interviews, enabled us to touch upon more relevant factors to answer our research questions. Moreover, the timeline that was provided to the students at the end of the interviews yielded additional information in some cases. Our study has several limitations. We only included students from urban high schools. Other factors could play a role in the motivation of students from more rural areas. However, our findings are in line with previously reported findings. Therefore, we expect that the mechanisms can be generalised to a larger student population. Further research is required to test this hypothesis. In addition, the influence of parents on students' motivation for studying medicine was rather unclear. The study Open Access counsellors' and students' perceptions with this regard seemed to differ. Including parents in the study might have provided more evidence about their role in students' study choice process. --- strengths and weaknesses in relation with other studies This study is one of the few that investigates the motivation for studying medicine among those who are still in the process of making a study choice. In addition, the qualitative design yielded richer data on motivation and factors that influence motivation than the quantitative surveys used in most previous studies. This enabled the investigation of the interplay between the factors involved. We investigated the mechanisms while acknowledging the importance of the qualitatively different types of motivation; therefore, our study adds a new dimension to the previous findings. --- Future research The mechanisms for the interplay between healthcare experiences, growing up in a medical family, the presence of selection processes and motivation have been hypothesised, but still require quantitative testing on a larger scale. One potential next step would be to validate the model in different countries and cultures. In addition, in this explorative study we found indications for differences and inequalities between students due to their backgrounds. In future research, the findings with regard to the different types of motivation for studying medicine and the extent to which the identified factors discourage students from applying could be compared for students with different background characteristics. Research could also focus on which types of motivation mainly occur in which cultural settings to identify the optimal setting for fostering students' autonomous motivation for studying medicine. Correction notice This paper has been amended since it was published Online First. Owing to a scripting error, some of the publisher names in the references were replaced with 'BMJ Publishing Group'. This only affected the full text version, not the PDF. We have since corrected theseerrors and the correct publishers have been inserted into the references. --- Contributors All researchers contributed to the design of the study and interpretation of the data. AW and UI have completed courses on conducting qualitative research. AW conducted the interviews. AW and UI were responsible for the main data analysis. All researchers contributed to and approved the final manuscript. --- Competing interests None declared. Ethics approval This study has been approved by the Dutch Association for Medical Education Ethical Review Board , dossier number 408. Provenance and peer review Not commissioned; externally peer reviewed. Data sharing statement Anonymised data set available from the main author upon request.

Objective To explore high school students' motivation for applying to study medicine and the factors that influence this. To find explanations for under-representation of minority students in medical education, descriptions of motivation of students with different background characteristics were compared. Design Qualitative phenomenological study using semistructured one-on-one interviews. Setting One predominantly white and one mixed high school in a large multicultural city in the Netherlands. The study was conducted in March-December 2015. Participants Twenty-four high school students, purposively sampled for demographic characteristics. Methods The analysis consisted of the coding of data using a template based on the motivation types (autonomous and controlled motivation) described by self-determination theory and open coding for factors that influence motivation.The main reasons for pursuing a medical career pertained to autonomous motivation (interest in science and helping people), but controlled motivation (eg, parental pressure, prestige) was also mentioned. Experiences with healthcare and patients positively influenced students' autonomous motivation and served as a reality check for students' expectations. Having to go through a selection process was an important demotivating factor, but did not prevent most students from applying. Having medical professionals in their network also sparked students' interest, while facilitating easier access to healthcare experiences. Conclusions The findings showed a complex interplay between healthcare experiences, growing up in a medical family, selection processes and motivation. Healthcare experiences, often one of the selection criteria, help students to form autonomous motivation for studying medicine. However, such experiences as well as support in the selection process seem unequally accessible to students. As a result, under-represented students' motivation decreases. Medical schools should be aware of this and could create opportunities to acquire healthcare experiences. High schools could incorporate internships as part of their study counselling programmes and offer tailor-made guidance to each individual student.Concerns exist that the medical profession insufficiently reflects the society it serves. 1 Debates increasingly focus on the disadvantage of under-represented socio-demographic groups due to the use of selection processes to decide who gets admitted to medical programmes. Under-representation of ethnic minority students, 2 3 students without a medical family background, [4][5][6] students who are the first in their family to go to a university, 5 7 8 lower socioeconomic status (SES) students 3 5 9 and males 9 10 in the medical student population can, to some extent, be ascribed to biased selection processes. 11-24 However, under-representation is not always due to a bias in selection processes 4 6 14 16 21 25-27 and widening-access efforts targeting selection bias have been till date only moderately successful in resolving disparities. 15 21 22 24 28 This stresses the need for other measures. A self-selection effect, which refers to students deciding whether to apply based on the information they have, for example, about the selection process, 16 29 has been suggested

Introduction Social support is widely understood to have positive effects on health and wellbeing [1][2][3][4]. Although there is some debate surrounding how it can best be defined [1], social support comprises at least three types of support; tangible or instrumental support , informational support and emotional support . At its core, social support is based upon personal relationships in which people believe they are cared for and valued, and belong to a network of communication and mutual obligation [6]. Numerous studies have demonstrated associations between social connectedness and engagement and both physical and psychological wellbeing [7,8]. Social support is particularly recognised as an important resource for parents of young children. The transition to parenthood can be a challenging period for many parents, of enduring stress, financial adjustments, upheaval of life, new responsibility and sleep deprivation. In addition, loneliness or social isolation may occur [9]. In this context, the availability of social support may help to buffer against the effects of stress [8]. The support may be delivered in terms of informal child care or financial support by relatives , advice about parenting practice or through social ties and interpersonal relationships [10]. As in the broader social support literature, the benefits of social support for parents in particular are well recognised. For example, higher levels of social support have been linked with better health for women pre-and post-natally [11], lower rates of depression and stress [12,13], increased parent self-efficacy [13,14], and more secure mother-infant attachments [15]. Despite knowledge about the value of social support for new parents, and mothers in particular, Parry et al. [16] suggested that the transition into motherhood is more socially isolating in modern times than in the past, with factors such as increased workforce participation, increased geographic mobility and distance from family members, and higher levels of lone parenthood resulting in more sole parents parenting with less social support. These societal changes, in conjunction with the value of social support, mean that programs or services that promote the development of social networks for parents, particularly parents who are socially isolated, can therefore be a valuable tool for new parents. A range of opportunities to expand social networks may be available to parents of young children, including child care or early learning centres, preschools, community events, mothers groups and playgroups. Research from the United States has shown that enrolling children in child care centres may lead to the development of new friendships and extended networks, compared with mothers who do not use child care services, particularly if those centres promote socialisation among parents [17]. However, as child care participation is strongly linked with labour force status and affordability [18], not all parents can access formal child care. Australian estimates of formal child care participation suggest around 10% of 0-1 year olds and 40% of 2-3 year olds attended formal care of varying duration, and most were attending so that caregivers could meet their work or study commitments [19]. Though some Australian parents will likely form social support networks through child care participation, as found in the US, a large proportion do not have this opportunity. In Australia, an alternative, targeted opportunity for parents of young children to build their social networks is playgroup participation. Playgroups, a common way for new parents to engage with others, are regular gatherings of parents and children under the age of five. Though formats can vary, they are typically held once a week for around 2 hours in a variety of locations, including the homes of participants, in schools or community halls, or parks and playgrounds. Importantly, playgroups are distinct from child care or crèche arrangements, as caregivers stay for the duration of the playgroup and participate in activities with their children and socialise with other caregivers. Two broad playgroup models operate in Australia, community and supported playgroups. Community playgroups, which are parent-led and managed by participants, are the most common. Supported playgroups are developed and supported by State and Territory playgroup associations and other not-for-profit agencies, where a facilitator is employed to organise the activities undertaken at the playgroup. These playgroups are supported by funding from both federal and state government bodies, and are often offered to disadvantaged communities where the development and management of playgroups can be difficult. These playgroups are supported in recognition of the broad objectives that playgroups aim to achieve: to improve the wellbeing of parents and children, to improve parenting skills and family functioning, and to develop stronger communities. Evidence regarding the extent to which playgroups meet these objectives is relatively scant, despite the large proportion of Australian families that access them. Prior research using data from the Longitudinal Study of Australian Children showed over 60% of the study children born in 2004-2005 had accessed a playgroup at least once by the age of 4-5 years [20]. This study also showed that children from disadvantaged families performed significantly better on measures of learning competency and social and emotional wellbeing if they persistently attended playgroup across the ages of 0 to 3 years than disadvantaged children who never attended a playgroup. Yet, disadvantaged families were significantly less likely to participate in playgroups than their non-disadvantaged counterparts. In an Australian qualitative study, Strange et al. [21] found that mothers of young children in newer residential areas reported that through playgroup attendance they were able to form friendships, build a supportive network, and had an increased sense of community connectedness. Other studies have shown that parent group experiences are not always positive for those who participate. In a Canadian qualitative study, for example, Mulcahy et al. [22] found that participating in an informal mothers group enabled some mothers to 'get together, get by and get ahead' whilst others 'get left out, get judged and get gendered'. Some mothers reported excluding some participants because of differing attitudes towards parenting. Thus, where differences exist amongst members of the group, the experience may be less positive overall. This is in line with research showing that mothers tend to seek support from others of similar backgrounds [14]. While playgroups are more structured and organised than informal mothers groups, and can provide access to other services including conflict mediation that may help to ameliorate relationship issues, it is likely that the social benefits of playgroups are influenced by both individual factors and group dynamics, as is the case with other social networks [23]. The aim of this study was to examine the extent to which playgroup participation relates to social support outcomes for Australian mothers, and to support received from friends in particular. This study expands on the previous research of Hancock et al. [20], using data collected for the Longitudinal Study of Australian Children . We examined the likelihood of mothers having no support from friends when their child was aged 4-5 years according to the child's participation in playgroup at age 3-19 months and 2-3 years. We also assessed the same friendship support outcome for mothers when their child was 8-9 years to examine the extent to which the friendship supports endured longer-term. Given the previous research demonstrating poorer friendship outcomes for mothers whose children did not attend child care [17]-organisations that are arguably less accessible and less focussed on promoting social networks than playgroups-we hypothesised that mothers who do not participate in playgroups will have a greater risk of having no support from friends, relative to mothers who do participate, both in the short-term and long-term, after controlling for initial levels of support reported by mothers when their child was aged 3-19 months and other socio-demographic factors. --- Methods --- --- Ethics Statement The Longitudinal Study of Australian Children is conducted in a partnership between the Department of Social Services, the Australian Institute of Family Studies and the Australian Bureau of Statistics. The study has ethics approval from the Australian Institute of Family Studies Ethics Committee. The Ethics Committee is registered with the Australian Health Ethics Committee, a subcommittee of the National Health and Medical Research Council . As the study children were all minors at the time these data were collected, written informed consent was obtained from the caregiver on behalf of each of the study children. The signed consent forms are retained by the field agency. --- Data Collection Data were collected at each wave from multiple informants, using a variety of methods. The primary caregiver of the study child was the main provider of information, who in most cases was the biological mother of the study child . Parent 1 was asked to complete an in-home interview as well as a self-complete questionnaire at each wave. At Wave 2, the questionnaire was divided into two surveys, one to be completed during the home visit , and the other to be completed and returned at a later time . Parent 1 response rates on the self-complete questionnaires were 85% at Wave 1, 98% for the in-home survey and 76.8% for the leavebehind survey at Wave 2, and 87.4% at Wave 3. At Wave 5 the self-complete questionnaires were replaced with computer-assisted self-interviews that were completed during the home visit with a 98% response rate. Questionnaires were also completed by Parent 2, the study child themselves, parents living elsewhere, teachers and child care workers where appropriate. As this study was concerned with support outcomes for mothers, and because the vast majority of primary caregivers in the study were mothers of the study children, analyses were restricted to cases where social support information was provided by mothers. --- Measures Support from friends. The key outcome measure was support received from friends, which was a single-item measure collected at Wave 3 and Wave 5 . Mothers were asked in the self-complete surveys "how often do you receive support from friends in raising study child?" Responses included 'always', 'often', 'sometimes', 'rarely' or 'never', with an additional option for 'don't have friends'. To address the hypothesis that mothers who do not participate in playgroup are at greater risk of having no support from friends, we identified mothers as having no support if they had responded that they never received support from friends, or did not have friends. All other responses were coded as having at least some support. The selected measure addresses the frequency of support received from friends, however it does not capture the type of support , the strength of attachment, or whether mothers actually need any social support from friends. For example, one mother may frequently receive low-level support , where another may only need occasional, but more invested support, such as emergency child care. Other assessments used in the literature typically capture different types of support, for example the Medical Outcomes Study Social Support Survey [28], but they do not determine the source of social support. Irrespective of support needs, support type or quality of support provided, mothers who indicate that they do not have friends or who say they never receive support will not receive any type of support from friends. To the extent that regular playgroup participation would be associated with the availability of friendship supports-a necessary, though not sufficient requisite for getting support-the selected measure was considered appropriate for examining the relationship between playgroup participation and the availability of social support, or lack thereof, from friends. As the measure of support from friends consisted of only one item, the measure was compared to other broad indicators of friendship and support, including the frequency of contact with friends, and how often mothers felt they needed support but couldn't get it. When asked, "How often do you see, talk to or email your friends?" at both Wave 3 and 5 a substantially lower proportion of mothers with no support from friends said they were in contact with friends at least once a week compared with mothers with some support . Similarly, a higher proportion of mothers with no support said they were in contact with friends fewer than a few times year . Regarding perceived support needs, a higher proportion of mothers with no support from friends reported they often or very often needed support but couldn't get it from anyone compared with mothers with at least some support . Friendship networks. It is possible that mothers who did not have friendship networks around the time the study child was born may also have lacked the social confidence, awareness or motivation that might otherwise encourage them to participate in a playgroup. As such, any association between playgroup participation and social support outcomes in later years may simply reflect a situation where mothers do not participate in playgroups because they didn't have the initial support or confidence. Therefore we assessed indicators of friendship and support reported by mothers at Wave 1 . At Wave 1 the survey items relating to social support did not differentiate whether the support was provided by family or friends. However, three items assessing attachment to friends were included. While having friendship attachments does not mean that social support is necessarily provided by those friendships, a lack of friends or friendship attachments likely precludes receiving social support from friends. Therefore the attachment to friends measure was used as a proxy of social support from friends at Wave 1. The attachment to friends measure was included in the Wave 1 leave-behind questionnaire and consisted of 3 items: 'I feel closely attached to my friends', 'my friends take notice of my opinions', and 'sometimes I feel excluded among my friends'. Responses ranged from 1 = totally agree to 5 = totally disagree, with a further option for "no friends". Responses were summed, with the third item concerning exclusion being reverse coded. Mothers with a total score of 12 or above-meaning they disagreed with all three items, or if they indicated they did not have friends-were coded as having poor friendship attachment. All other mothers were coded as having at least some attachment to friends. As the measure of social support at Wave 1 relates more to the absence of social support than its presence, the attachment to friends measure was compared to other broader social support indicators to assess its validity as a proxy measure. Compared with mothers with adequate friendship attachments, at Wave 1 mothers with poor attachment were more likely to report that they did not receive enough support from family and friends at , and that they often or very often needed support but could not get it . Mothers with poor attachment to friends were also less likely than mothers with adequate attachment to nominate friends as a top 3 source of parenting information , practical help , emotional support and financial help . Playgroup participation. At Waves 1, 2 and 3 Parent 1 was asked "In the past 12 months, have you used any of the following services for the study child. . . Playgroups or parent-child groups?" and could respond either 'yes' or 'no'. The items were collected from the self-complete survey at Wave 1, the self-complete leave-behind survey at Wave 2, and the face-to-face interview at Wave 3. As there were no further questions on the type of playgroup attended, or the frequency of participation, we cannot distinguish between children who attended playgroup consistently throughout the previous 12 months and those who attended infrequently. We also could not ascertain who participated in playgroups with the study child, but assume that the majority of children attending playgroup did so with their mother. We also could not distinguish between the types of playgroups that families were attending. Previous research using the same data set showed that 40% of respondents indicated the study child had used a playgroup at age 0-1 year , 53% had participated at age 2-3 years and 25% at age 4-5 years [20]. For consistency with the previously published work our measure of playgroup participation excluded participation at Wave 3, as the study children were moving into more formal education settings by age 4. Thus, this study focused on the pattern of playgroup participation at ages 3-19 months and 2-3 years . Socio-demographic covariates. Data on a range of covariates were also collected via parent report, including: mother's age at birth of first child, mother's employment status, mother's highest level of education, family structure, language spoken in the home, household income and maternal mental health. At Wave 1, household income was collected by asking respondents to select from a category . These categories were then broadly grouped into quartiles. At subsequent waves more detailed information was collected allowing the calculation of equivalised household income, adjusting for the number of people in the household. Maternal mental health was assessed using the Kessler 6 scale . Scores of 13 and above indicate a probable serious mental illness [29]. As the LSAC is a non-clinical sample, very few mothers fell above this cut-point. In concert with other studies [30,31] we used a lower cut-point of 8+ to classify mothers as having elevated non-specific psychological distress. When considering the families that participated not only in all relevant interviews, but also completed and returned all of the self-complete questionnaires, and where respondents were mothers of the study child, the final sample size consisted of 2,617 mothers at Wave 3 and 2,576 at Wave 5 . To assess potential bias due to attrition and nonresponse, the Wave 1 characteristics of the families contributing to the Wave 3 and Wave 5 analyses were compared with the characteristics of the full Wave 1 sample . The analytic samples at Wave 3 and Wave 5 were slightly overrepresented in terms of labour force participation, being older first-time mothers, having higher household incomes, two-parent families, and speaking English in the home. Minor differences in distributions were observed for Wave 1 attachment to friends, needing support, mother's highest education mother's highest education, and mother's mental health. --- Statistical Analysis Firstly, the proportion of mothers with no support from friends at Waves 3 and 5 was cross-tabulated with the playgroup participation variable and other covariates to explore the distribution patterns and to determine which covariates would then be included in the multivariate analysis. Next, we used multivariate log-binomial regression to generate the adjusted relative risk of mothers having no support from friends according to playgroup attendance pattern. Separate models were used to estimate the adjusted relative risk of having no support at Wave 3 and Wave 5 . In addition to playgroup attendance, Wave 1 attachment to friends and the frequency of needing but not getting support were included as controls for initial measures of friendship networks and general support. All covariates that were significantly associated with Wave 3 or Wave 5 support outcomes in the exploratory analysis were included as covariates in the regression analyses. Results are reported as risk ratios and 95% confidence intervals. The adjusted probabilities of having no support from friends for each level of playgroup participation were also estimated. SAS 9.4 [32] was the statistical software package used for all analyses, and longitudinal weights were used to account for sample attrition bias across waves. The log-binomial regressions were estimated using the GENMOD procedure. Correlation among families living in the same postcode was accounted for using the 'repeated' statement. --- Results Table 2 displays the proportions of mothers reporting no support from friends at Wave 3 and Wave 5 according to the pattern of playgroup participation, as well as the demographic characteristics of the sample. In total, 12.7% of mothers reported having no support from friends at Wave 3, and 14.5% at Wave 5. For mothers who reported poor attachment to friends at Wave 1, these proportions were significantly higher , and likewise for mothers who had reported they often needed support but couldn't get it . The descriptive analyses showed that a significantly higher proportion of mothers had no support from friends if they never participated in playgroup compared with mothers who persistently attended when their child was aged 3-19 months and 2-3 years . For the socio-demographic variables, a higher proportion of mothers reported having no support where they had lower levels of education, if they were not employed part-time, if they had lower levels of income, were single parents , if they did not speak English in the home, or had likely elevated psychological distress . Two log-binomial regression models were then fitted to further examine the association between playgroup participation and the risk ratios of having no support from friends when their child was aged 4-5 years and 8-9 years . After adjusting for initial friendship networks, social support needs and socio-demographic characteristics, the relative risk of having no support from friends at Wave 3 was 1.7 times higher for mothers who never participated in playgroup, and 1.8 times higher for mothers who participated at either Wave 1 or Wave 2 , relative to mothers who persistently participated at both Wave 1 and Wave 2. The adjusted proportions of mothers having no support from friends at Wave 3 was 12.8% for those who never participated, 13.6% for those who participated at one wave only, and 7.7% for those who participated in playgroup at b Measures taken from the concurrent Wave . c Excluded for small cell size at Wave 5. doi:10.1371/journal.pone.0133007.t003 both waves. At Wave 5, approximately four years later, the relative risks of having no support from friends at Wave 5 were 2.0 times higher for both mothers who either never participated in playgroup and those who participated at only Wave 1 or 2. The adjusted proportions of mothers having no support from friends at Wave 5 was 14.7% for those who never participated, 14.4% for those who participated at one wave only, and 7.3% for those who participated in playgroup at both waves. Post-hoc analyses using the ESTIMATE statement indicated there was no difference in the relative risk of having no support from friends at either Wave 3 or Wave 5 when comparing mothers who never participated in playgroup with those who participated at one wave only . Having poor attachment to friends at Wave 1 , our proxy control for initial indicators of support from friends, was strongly associated with the risk of having no support from friends in later years, where these mothers were 2.1 times as likely as mothers with adequate friendship attachments to have no support from friends at both Wave 3 and Wave 5 . The frequency of needing support at Wave 1 but not getting it was not associated with social support from friends at either Wave 3 or Wave 5. Few of the socio-demographic characteristics considered were independently associated with the risk of not having support from friends in the multivariate models . Mother's employment status was associated with support outcomes when the study child was aged 4-5 years , but not at age 8-9 years . Mothers who worked full-time at Wave 3 were 1.5 times as likely to have no support from friends, and mothers employed on maternity leave 2.2 times as likely to have no support from friends compared with mothers who worked part-time. Mothers who did not speak English in the home were 1.8 times as likely as mothers who spoke English at home to have no support from friends at Wave 5 . In the multivariate models, frequency of needing support but not getting it, mother's highest level of education, equivalised household income, family structure and mother's mental health were not independently associated with support from friends at either Wave 3 or Wave 5. --- Discussion This study examined the associations between playgroup attendance and the social supportor lack thereof-that mothers received from friends in later years. We found that persistent playgroup participation was a protective factor against having no support from friends in helping to raise the study child, both when children were aged 4-5 years and four years later at 8-9 years . Compared to mothers whose child participated in playgroup at both Wave 1 and Wave 2 , mothers whose child did not participate in playgroup, or who participated for only one wave, were almost twice as likely to report not receiving support from friends when their child was 4-5 years old , and twice as likely when their child was aged 8-9 years . These findings were independent of confounding variables including the mother's initial attachment to friends and ability to obtain social support, education level, employment status, household income, family structure, language spoken at home and mental health status. As this was an observational study, the observed associations between playgroup participation and receiving support from friends could be explained in a number of ways. One is that persistent playgroup participation leads to the development of friendship networks from which mothers can draw social support in later years. Participating in playgroups over a long-term period might also allow the opportunity to develop or reinforce the social confidence to engage with parents in other settings, for example, with the parents of their child's school friends, which in turn may reinforce their ability to draw on social support from friends. Another possible explanation is that mothers with a degree of social confidence and who are good at maintaining their friendships and networks are simply more likely to participate-and continue to participate-in playgroups. They may have chosen to attend a playgroup that already included friends or family members. Conversely, individuals who find it difficult to establish and maintain relationships, be that due to geography, circumstance or personality, may find it difficult to participate in social settings such as playgroups and to maintain social support resources over time. These mothers would therefore have less support from friends in the long term. However, our results remained after controlling for initial friendship attachments, suggesting that the relationship between playgroup participation and later friendship supports is unlikely to solely reflect the propensity of mothers to have friends or social supports and to participate in playgroups. Even for mothers who are socially inclined to participate in playgroup, issues around equity of access may be a contributing factor. Access to playgroup is impacted by the availability of transport, appropriate venues, and most importantly, other families. In growth corridors or new residential areas, a lag in social infrastructure and family relocation can potentially result in the isolation of families with young children [33]. Disadvantaged families, previously shown to have lower rates of participation in playgroup [20], may also be excluded from certain playgroups where payment from participants is required to cover the cost of venue hire and other materials. Therefore some socially isolated mothers may have limited friendship support outcomes because playgroups are not accessible in some areas. Our findings showed that the risks of having no support from friends were higher both for mothers who never participated in playgroup and those who participated at only one wave. Differences in outcomes only emerged for those mothers who persistently attended both at Wave 1 and 2, and even for this group of mothers 8% still reported having no support from friends in later years. It may be that some mothers attended several playgroups over the period, and did not establish friendships. Friendships, and the support that may stem from such friendships, are often not immediately established. Rather, it takes time for mutual trust, respect and affinity that stem from friendships to be established. As with any social network, there are likely to be a multitude of individual and group factors which will influence the level and types of social support that will stem from playgroup attendance [23]. Future research could examine whether parenting style and other factors such as personality contribute to the observed relationship between playgroup participation and social support outcomes. In addition to the study child's playgroup participation, mothers' social support outcomes were also associated with a number of other socio-demographic characteristics in the unadjusted analyses. Mothers with elevated psychological distress were more likely to report not having support from friends compared with mothers without elevated levels of distress. This finding is consistent with previous literature demonstrating that poor social support is associated with an increased risk of maternal depression [34,35], and that mothers experiencing mental health problems are less likely to seek and/or recognise that they are in need of support [36,37]. Poorer social support outcomes were observed for mothers who worked full-time rather than part-time, and those with lower levels of education. Educational attainment, employment status and social support outcomes are typically associated in the wider literature [38]. The finding that single-parent families had poorer friendship support outcomes than two-parent families is also consistent with the wider literature. For example, Cairney et al. [39] found that, compared to coupled mothers, single mothers were more likely to report lower levels of perceived social support, social involvement and contact. The literature concerning sources of social support for mothers of young children has previously examined a number of ways that mothers might connect with others and develop social support networks, including through organisations such as child care centres or preschool centres [17]. However, many Australian families either do not need, or cannot access, child care services or may have children who are too young for preschool, and therefore miss out on such opportunities. Additionally, by their nature, child care centres and preschools provide care and education services for children so that caregivers can attend to other commitments, such as work or study [19]. Playgroups, in contrast, actively promote the development of social support for the families who participate, where caregivers are expected to be actively involved in each session, interacting with both their child and with other caregivers. The active participation of caregivers also means that playgroups can address the social support needs of caregivers much earlier than these other sources, particularly for new parents experiencing a significant life transition with the arrival of a child. In the context of previous research therefore, our findings are not unexpected. However, research on the associated benefits of playgroup participation in particular, has been limited. As such, this paper contributes to the limited playgroup literature by demonstrating that persistent participation in playgroup is positively associated with friendship support outcomes. Our findings also support the approach of playgroup organisations in promoting access to playgroups for disadvantaged and socially isolated mothers of young children. These results extend the findings of Hancock et al. [20] who found that children from disadvantaged families who attended playgroup had better social-emotional functioning and learning competence compared to children from disadvantaged families who did not attend playgroup. This study shows the value of playgroups for the development of parenting support networks, as well as the value of longitudinal data for examining such relationships. This study has several limitations. As noted, the LSAC is an observational study, and hence we are unable to conclude whether there is a causal association between playgroup participation and social support outcomes. This study only used a broad level indicator of playgroup participation, thus we lacked information on the amount of playgroup attendance. In addition, there are many different types of playgroup, comprising different people with different likes, dislikes, parenting styles and personalities. As such, playgroups are not a guaranteed source of social support, and it is possible that even mothers who persistently attend playgroup, be it the same or a variety of playgroups, will not establish friendships that extend beyond the playgroup setting. It is possible that, with more comprehensive data on frequency of attendance as well as type of playgroup attended, more informative results may have emerged. Our indicator of social support from friends may also be somewhat limited, being a singleitem measure. As such, some uncertainty may exist as to the extent to which the item captures social support from friends. However comparisons of how the item compared with related constructs, including friendship attachments, how often they saw their friends and how often they needed support but could not get it, suggested that the measure was a valid proxy. In our study, we make the assumption that the mother would usually be the caregiver accompanying the child to playgroup. There is no Australian data available to determine which caregiver participates in playgroups with their child, but they can include mothers, fathers, grandparents and paid carers such as nannies. However, given that over 95% of the primary carers of the children in this study were mothers, we therefore assume that in most cases the mother attended playgroup with the study child. Had we been able to restrict analyses to cases where mothers accompanied the child to playgroup, it is possible that different associations may have been observed. Despite these limitations, the findings provide evidence that playgroup participation is positively associated with friendship support outcomes for mothers. While social support can be provided from many sources, including partners, family, neighbours or work colleagues, mothers with extensive support networks may still find that playgroups help to widen and reinforce their support networks for times where other supports are unavailable. For mothers with limited social support options or those who are socially isolated, playgroups may be a helpful resource to establish friendships from which they can potentially draw social support. --- This study uses data from the Longitudinal Study of Australian Children. The survey data remain the intellectual property of the Australian Government Department of Social Services . The authors are licensed users of the data, and under the licensing agreement are not permitted to share or distribute the data. The DSS make data available, for a small fee, to researchers who apply for an individual or organisational license. Researchers wishing to apply to become a licensed user can contact the DSS by emailing [email protected]. Further information about accessing the data can be found at https:// www.dss.gov.au/our-responsibilities/familiesandchildren/programmes-services/access-to-dsslongitudinal-datasets. ---

This study aimed to examine friendship networks and social support outcomes for mothers according to patterns of playgroup participation.Data from the Longitudinal Study of Australian Children were used to examine the extent to which patterns of playgroup participation across the ages of 3-19 months (Wave 1) and 2-3 years (Wave 2) were associated with social support outcomes for mothers at Wave 3 (4-5 years) and four years later at Wave 5 (8-9 years). Analyses were adjusted for initial friendship attachments at Wave 1 and other socio-demographic characteristics.Log-binomial regression models estimating relative risks showed that mothers who never participated in a playgroup, or who participated at either Wave 1 or Wave 2 only, were 1.7 and 1.8 times as likely to report having no support from friends when the child was 4-5 years, and 2.0 times as likely to have no support at age 8-9 years, compared with mothers who persistently participated in playgroup at both Wave 1 and Wave 2.These results provide evidence that persistent playgroup participation may acts as a protective factor against poor social support outcomes. Socially isolated parents may find playgroups a useful resource to build their social support networks.

Introduction Population-wide mass media can use the combination of television, radio and print. This includes television and radio, episodes as well as inserts in key newspapers during each year of intervention, and the campaigns are repeated annually. Bertrand et al. [1] examined the effectiveness of 24 mass media interventions on changing HIV-related knowledge, attitudes-and behaviours in low-and middleincome countries. The most frequently reported outcomes were condom use and knowledge of modes of HIV transmission , followed by reduction in highrisk sexual behaviour , perceived risk of contracting HIV/AIDS , interpersonal communication about AIDS or condom use , self-efficacy to negotiate condom use , and abstaining from sexual relations . The studies yielded mixed results, and where statistically significant, the effect size was small to moderate . On two of the seven outcomes, at least half of the studies did show a positive impact of the mass media in the improvement of knowledge of HIV transmission and reduction of high-risk sexual behaviour [1]. In a more recent review, Noar et al. [2] evaluated 34 distinct HIV/AIDS campaign efforts conducted in 23 countries and found among the 30% of studies with stronger outcome evaluation designs, that the vast majority of the well-controlled studies demonstrated effects on knowledge , behaviour , or behavioural intentions . In addition, several studies [3][4][5] found that exposure to HIV-related communication on the media was significantly related to less stigmatizing attitude toward people living with HIV and AIDS . In South Africa social and behavioral communication interventions are a critical component of HIV/AIDS prevention [6]. Numerous communication campaigns have been implemented intensively across the country through 2 The Scientific World Journal government initiatives and nongovernmental organisations over the past decade [6][7][8]. The purpose is to help combat the spread of HIV and AIDS by improving knowledge about modes of transmission, risk perceptions, changing sexual behaviours, questioning potential harmful social norms, and promoting resources and services that support prevention [6,9]. Campaigns utilise radio, television, and other outlets to address prevention of HIV/AIDS by conveying messages through interpersonal, community, and national channels [7,8]. Four major national-level HIV and AIDS communication programmes using mass media and other communication channels include Khomani, meaning "caring together;" LoveLife, which focuses mainly on teenagers, Soul City and Soul Buddyz which address adults and children, respectively, and the 46664 campaign which aims to promote HIV prevention through a series of events and activities linked to the ethos and values of Nelson Mandela. This latter campaign is conducted nationally, but also extends globally [6,8]. National HIV/AIDS communication surveys demonstrated impacts in the improvement of knowledge and awareness and have illustrated outcomes related to HIV prevention [7][8][9]. An evaluation of Soul City in South Africa illustrates the mediation of community change as a product of communication programming [10]. In a national communication survey in 2009, Johnson et al. [6] found that HIV communication programmes as a whole brought about positive changes in condom use and HIV testing but did not impact on reduction of multiple sexual partners. In an evaluation of a youth programme in South Africa, Pettifor et al. [11] found that the participation in loveLife reduced chances of being infected with HIV, and youth were more likely to report condom use at last sex and used condoms more consistently. The aim of this paper is to assess the reach of these communication campaigns in conjunction with contributions to knowledge, attitudes, and HIV risk behaviours in the general population in South Africa. --- Methods --- Sample and Procedure. The survey targeted all persons over 2 years of age living in South Africa and residing in homes . A multistage cluster stratified sample stratified by province, settlement geography , and predominant population group in each area was used. A systematic sample of 15 households was drawn from each of 1 000 census enumeration areas . In each household, one person was randomly selected in each of the four mutually exclusive age groups . Sociodemographic and behavioural information was collected with questionnaires administered by trained field workers . The age group selected for analysis in this paper was 13234 people aged 15-55 years. This age range was chosen because of greater HIV risk. Ethical approval for conducting the study was obtained from the Human Sciences Research Council's Ethics Committee . Informed consent was obtained for agreeing to participate in the interview. Registered professional nurses were trained to conduct interviews. --- Measure. A structured questionnaire was used to determine the reach of national mass media communication campaigns that promote HIV/AIDS education in South Africa. Respondents were asked questions related to mass media reach as well as their knowledge of campaigns and perceptions of HIV/AIDS communication in general. The sections of the questionnaire analysed here included access to media channels and exposure to 18 different HIV mass communication programmes in the past 12 months in South Africa. Impact of the HIV mass communication programmes was measured in terms of HIV knowledge , sexual risk behaviour , and HIV stigma attitudes . --- Data Analysis To create a joint index of HIV mass communication exposure, each programme was coded 0/1 for recall and then summed to create scale from 0 to 18. By taking the median of 6 as a cut-off point, this scale was dichotomised as low or high exposure, and it was further grouped into three groups of 1 = 0-3 exposures, 4-8, and 9 or more exposures to mass communication programmes. Other dependent variables included HIV knowledge, which was dichotomised into 1 = 14-18 correct responses and 0 = 0-13 correct responses and HIV/AIDS stigma, which was dichotomised into 1 = 2-5 scores and 0 = 0-1 scores. Accounting for complex sampling design, a weighted analysis of the outcome of interest was carried out for the main reporting domains. Weighted data were analyzed using STATA 10.0 software. STATA suite of survey commands was used to obtain estimates with 95% confidence intervals that took into account the survey design. Adjusted odds ratios are reported to indicate the strength and direction of association. A P value less than 5% is used to indicate statistical significance. The Scientific World Journal Adjusted odds ratio: AOR. 1 Adjusted by media access . Significance tested using logistic regression at * * * P < 0.001; * * P < 0.01; * P < 0.05. --- Results --- Access to Media. Radio and television are the most popular mass media consumed by people in South Africa, with 81.7% of people watching TV and 83.6% listening to the radio. While 66.9% and 60.7% of people reported reading a newspaper and a magazine, respectively, only 22.4% and 12.9% reported reading a newspaper and a magazine, respectively, every day. The internet was the least accessed channel, with 17.7% of the population accessing it . --- Exposure to HIV Mass Communication Programmes. Exposure is defined as have heard or seen at least one of any HIV mass communication programme such as billboard, radio, TV, or other components in the past 12 months prior to the survey interview. Overall, a high exposure to 18 different HIV communication programmes across different age groups was found. Most programmes were more often seen or heard by youth , and this is in addition to programmes specifically designed for youth such as loveLife and Soul Buddyz . The median exposure to HIV mass communication among participants was 6 programmes, IQR 3-9. In the multivariate analysis, younger age, higher education, urban formal residence, and the African population group were associated with high HIV mass communication exposure . --- Association between HIV Mass Communication Programmes and HIV Risk Behaviour. In multivariate analysis, greater exposure to HIV mass communication programmes was associated with greater HIV knowledge, condom use at last sex, having been tested for HIV in the past 12 months, and less stigmatizing attitude toward PLWHA . --- Discussion Overall the study found that there was high exposure to the 18 different HIV communication programmes among the general population, especially among youth. Higher HIV mass communication exposure was in this study associated with improved HIV knowledge and lowered HIV/AIDS stigma, as found in previous studies [3][4][5][12][13][14][15]. Further, higher HIV mass communication exposure was in this study, in agreement with other studies, associated with the reduction of HIV risk behaviour [20][21][22][23]. The study did not find any association between higher HIV mass communication exposure and reduction of the number of sexual partners, as found in some other studies [18,24,25]. Trend data from populationbased surveys in South Africa seem to confirm an increase in condom use, in particular among young people, but not a reduction in the number of sexual partners. Overall, the current study adds to the growing literature suggesting that mass media campaigns can be effective in changing HIV risk behaviour and attitudes, at least over the short term [2,19]. This finding confirms the development of a shift in the purpose of campaigns, from simply aiming to raise awareness about HIV and AIDS to attempting to impact safer sexual behaviours [2]. --- Study Limitations The study design used was a weak outcome evaluation design; more rigorous quasi-experimental designs are needed [2] including multiple assessments [19]. Exposure to communication campaigns relied on selfreporting, and it was not possible to measure the intensity of exposure, which should be included in further studies. The communication campaigns were not the only potential exposure to HIV and AIDS information among respondents, and direct and exclusive contributions to change cannot therefore be claimed. Further, the study found a low reliability of the HIV/AIDS stigma index used; therefore, if the stigma scale had been used as a scale, there would have been some limitation in validity of results. Finally, there are several biomedical and structural issues which the mass media could well have been instrumental for; their utilization in sub-Saharan Africa seems to be limited to behavioural interventions [26]. --- Conclusions Exposure to communication campaigns in South Africa is associated with a range of outcomes relevant to addressing HIV and AIDS in relation to attitudinal aspects of the disease-particularly stigma reduction-and for HIV prevention. Communication campaigns thus provide a broad backdrop of support to HIV prevention activities and are likely to be strengthened via coherent and systematic prevention activities conducted at the community level. --- Submit your manuscripts at http://www.hindawi.com --- Stem Cells International Hindawi

In South Africa social and behavioural communication interventions are a critical component of HIV/AIDS prevention, and numerous communication campaigns have been implemented intensively across the country through government initiatives and nongovernmental organisations over the past decade. The aim of this paper is to assess the reach of HIV and AIDS communication campaigns in conjunction with contributions to knowledge, attitudes, and HIV risk behaviours in the general population in South Africa. The sample included in this nationally representative cross-sectional survey was 13234 people aged 15-55 years. Overall, the study found that there was high exposure to 18 different HIV communication programmes (median 6 programmes and 14 programmes more than 30%) across different age groups. Most programmes were more often seen or heard by young people aged between 15 and 24 years. In multivariate analysis, greater exposure to HIV mass communication programmes was associated with greater HIV knowledge, condom use at last sex, having tested for HIV in the past 12 months, and less stigmatizing attitude toward PLWHA.

Introduction The outbreak of the COVID-19 pandemic dramatically influenced the personal lives of many people worldwide and led them to face unprecedented challenges. This new acute respiratory syndrome created havoc for societies, economies, communities, and individuals [1]. National governments imposed urgent measures limiting individual freedom and required behaviors that went against shared social norms, such as maintaining physical distance from the members of in-groups, including family and friends [2]. The success of the fight against the virus greatly depended on citizens' adherence to such restrictive policies and their intentions to engage in protective behaviors, such as getting vaccinated. The pandemic gave rise to many questions about the virus, its origin, and its ways of transmission, some of which were left unanswered for several months. Moreover, it was accompanied by a flourishing of conspiracy theories offering explanations for the spreading of the pandemic and its management by governments [3,4]. When laypeople lack knowledge of hazards such as COVID-19, they need to rely on "trusted others" to assess the risk and to implement correct behaviors. Trust represents a "social glue" in societies and is a crucial element for social functioning, as it helps individuals act in a complex environment and facilitates the pursuit of collective goals [5,6]. Trust in those who have the authority and responsibility for making decisions, such as scientists and government agencies , as well as general trust-that is, the belief that "most people are trustworthy most of the time" [8]-can play a critical role in facing uncertain and complex situations such as the COVID-19 pandemic. Indeed, studies conducted during previous pandemics such as H1N1 or Ebola and during the COVID-19 pandemic highlighted that trust in governments and other institutions, citizens, and science can favor the acceptance of prescribed behaviors and engagement in prosocial behaviors [9][10][11]. The present study, which was conducted in Italy during the first months of the COVID-19 outbreak and the national lockdown, examines the association between social and general trust and individuals' adoption of coping strategies to face the coronavirus pandemic. Specifically, it focuses on trust in the different institutions and governments that were in charge of managing the crisis or that played an important role in supporting people as well as generalized trust in other Italians and humankind. --- Social and General Trust during the Outbreak of COVID-19 Trust is a critical element for social functioning at both micro and macro levels: it increases interpersonal cooperation and makes social and political institutions more effective and economic activities more efficient [12,13]. Trusting other people, groups, or institutions can also be perceived as a risk [14], but individuals accept this vulnerability based upon their positive expectations of the intentions or behavior of who they are trusting [15]. As mentioned, research on risk management suggested distinguishing between social and general trust [5]. Social trust can be defined as the willingness to rely on those who have the responsibility for taking actions related to technology, environment, medicine, public health, and safety [7]. This kind of trust is particularly relevant in situations where the individual has insufficient knowledge or an insufficient technical background to make decisions and take action: trusting scientists and government agencies represents a way to reduce the complexities and fears that people face. Research showed that social trust influences, for example, the perceived risks concerning pesticides, nuclear power, and artificial sweeteners [7] and the likelihood of accepting a referendum to site a hazardous waste disposal facility [16] or irradiated food [17]. In the health domain, studies on H1N1 influenza found that trust in the government, medical organizations, and the media plays a crucial role in promoting the acceptance of the recommended behavior to control the spread of the disease [11,18]. Moreover, social trust positively influences the intention to be vaccinated against this virus inflection [19]. The literature on social capital [20] suggested that general trust is related to "how much someone trusts people whom one meets for the first time" [8] . This belief in the benevolence of human nature is not limited to a particular group or organization, but it is a default expectation of others' trustworthiness [21]. General trust is essential for social cooperation and economic exchanges [12]. Moreover, it plays an important role in digitalized societies, where people can easily form new social relationships that are no longer bound by interpersonal social networks [22]. Research focusing on the COVID-19 pandemic examined trust from two perspectives. On one side, studies pointed out an increase in social trust as a consequence of the pandemic. Evidence from European countries demonstrated that the rise of the pandemic increased social trust, particularly trust in the government because of "the rally-around-theflag-effect" [23,24]. According to Schraff [25], anxiety related to and fear of the contagion and its consequences rallied people around political institutions to increase security in an unmoored situation. In addition, several studies showed that the outbreak of COVID-19 increased trust in science or scientists [26]. Little attention was, instead, devoted to the impact of COVID-19 on general trust. A study conducted in Norway found no overall increase in generalized trust in the early phase of the pandemic, which was, however, higher for individuals who had a direct experience with the disease compared to those who perceived themselves as at risk or were worried about the pandemic [27]. Research [28] found a generalized increase in interpersonal trust during the COVID-19 pandemic compared to the pre-pandemic level among the Italian population and increased levels of trust in strangers among people who caught COVID-19. Different results for social and political trust emerged in China, where greater exposure to COVID-19 risks significantly decreased general and political trust among the adult population [29]. On the other side, a consistent corpus of the literature is focused on the impact of trust on the measures to reduce infection. Researchers [8] found that individuals who trusted the government and the pharmaceutical industry regarding SARS-CoV-2 tended to perceive more risks and showed more acceptance of containment measures than participants with lower social trust. Moreover, individuals with higher levels of general trust perceived less health risk and were less inclined to accept the related measures, probably because they believed in the benevolence of others and, therefore, found it difficult to perceive others as risk factors of contagion. The positive influence of social trust found further support. A higher level of government trust and trust in its ability to control COVID-19 were significantly associated with greater compliance with protective behaviors such as frequent handwashing, avoidance of crowded spaces, and social isolation or quarantine [30,31]. Similarly, greater trust in science and scientists increased individuals' tendency to follow the COVID-19 prevention guidelines [32] and to accept the lockdown [33]. A recent study [10] found that trust in governments, science, and fellow citizens was a better predictor of compliance with prescribed behaviors than information related to the actual threat of the virus. In this case, differently from Siegrist and colleagues [8], general trust positively influenced the acceptance of the containment measures. However, studies [34,35] highlighted the apparently paradoxical effects of generalized trust, which was related to higher compliance with vaccination but lower support for nonpharmaceutical interventions such as wearing masks. This might be due to people's beliefs about the ability of others to respect social distancing or to the fact that those with a high level of general trust fail to acknowledge that all people can pose a health risk. Other studies [4] highlighted the role of trust in mediating the effects of identification with Italians and the beliefs of conspiracy theories on the wish for a strong leader to deal with the pandemic. Individuals' identification with the national group was related to a lower wish for a strong leader through the mediation of higher trust. In contrast, the belief in conspiracy theories was associated with decreased trust and, through it, with an increased need for a strong leader. However, in Moscatelli and colleagues' study [4], it was impossible to disentangle the effect of social and general trust since the authors employed a combined measure. Considering the role of trust in fostering well-being, Paolini and colleagues [36] reported that Italians' trust in social and political actors was positively related to well-being during the first phase of the COVID-19 pandemic, and this effect was mediated by the identification with Italians and humankind. Moreover, Roccato and colleagues [37] found that increased trust in political, super partes, and international institutions positively influenced well-being by reducing anxiety, collective angst, and anger via the mediation of participants' perceived control over their lives. Finally, Moscatelli and colleagues [9] reported that Italians' higher identification with their country and the EU was related to increased levels of trust in the Italian and European governments, which, in turn, accounted for greater expectations that the pandemic crisis would eventually have positive outcomes, by improving Italian politics, the EU, and humankind. However, as far as we know, no research has so far examined the impact of social and general trust on predicting adaptive coping strategies in response to pandemic challenges, adaptive coping strategies that are crucial to maintaining individuals' physical and psychological well-being during such a crisis. --- The Role of Coping Strategies Dealing with COVID-19 Coping strategies, that is, the set of cognitive, emotional, and behavioral strategies that individuals use to master, tolerate, reduce, or minimize especially stressful events [38], are crucial to buffer adversities due to the virus. Coping strategies can be divided into approach and avoidance coping strategies [39]. The approach strategies are represented by actions aimed at directly alleviating the problem . In contrast, avoidance strategies aim at distancing oneself from the problem . Traditionally, avoidance coping was considered an unsuccessful strategy, especially in the long term [39][40][41][42]. Research showed that different strategies fulfill different functions [38]: emotion-focused strategies serve to manage and reduce emotional distress, whereas problemfocused strategies aim to solve or remove the source of stress. Of interest here, Foà and colleagues [43], based on the work of Carver and colleagues [44], proposed and validated a measure of coping strategies based on five critical dimensions. Among them, three dimensions are related to active coping strategies: problem-solving orientation, that is, the tendency to search for information and plan solutions while suppressing other activities that can interfere with problem solving; positive attitude and reinterpretation, intended as the tendency to accept and reinterpret events in order to transform them in opportunities of growth; and avoidance, which represents the attempt to ignore and deny the stressful event. The other two strategies seem less focused on the stressor and are related to seeking the help of others and of higher entities, that is, seeking social support and transcendent orientation, which is the tendency to seek comfort in religion [43]. In general terms, research pointed out that the way individuals react to stressful events can have long-term psychological and health effects. For instance, using avoidance strategies is likely related to depression and anxiety [45][46][47]. Conversely, engaging in cognitive reappraisal and problem-solving strategies is conducive to better mental health outcomes and well-being [48][49][50]. Similarly, an approach coping strategy was found to be related to a higher psychological health and a better quality of life [51,52]. The different roles of approach and avoidance coping strategies emerged during the COVID-19 pandemic. Approach strategies such as positive reinterpretation and planning strategies were protective against anxiety and depressive symptoms, whereas avoidance coping strategies predicted higher levels of anxiety and depression [53]. Similarly, a study conducted in Austria during the 4 weeks of the COVID-19 lockdown [54] revealed that coping strategies predicted a set of mental health indicators. Specifically, individuals who engaged in positive thinking, active stress coping, and social support strategies reported a higher psychological life quality, higher levels of well-being, and lower levels of stress, depression, anxiety, and insomnia. In particular, positive thinking was the strongest predictor, followed by social support. This result aligns with previous evidence as well as with other studies conducted during the pandemic, which found that positive thinking was related to lower depressive symptoms and lower distress in Italian health care professionals [55] and in the Greek population [56]. However, a study conducted among Lebanon adults reported that both approach and avoidance coping strategies predicted higher general well-being during the COVID-19 pandemic [57]. The positive role played by avoidance strategies might be explained by the specific nature of the stressor. Indeed, it is possible that people benefitted from taking a break from the threat and the several issues related to the pandemic, taking psychological distance from them while regaining energy to more directly address them [57]. Interestingly, for our purpose, Fluharty and colleagues [58] explored predictors of coping strategies during the COVID-19 pandemic in UK adults. First, they found clear differences due to demographic and social background factors. Problem-focused and emotion-focused coping strategies were used more by women, older people, and more educated and higher income people, whereas they were less influenced by factors such as living conditions. Women, people with more education and a higher income, younger people, and people living with others were also more likely to search for social support, whereas women and people from more disadvantaged groups showed a tendency to engage in avoidance strategies. Since religion was not measured independently from emotion-focused coping, no specific conclusion can be drawn in this respect. This study allowed for the understanding of the variations in responses to the COVID-19 pandemic, while highlighting which groups could benefit from interventions to improve their coping abilities, at the same time the focus on demographic variables and living conditions left more psychosocial variables in the shadows. Thus, it is important to examine how trust in different groups and institutions, as well as generalized trust in others, relates to the reliance on different coping strategies during a long-term stressor such as the COVID-19 pandemic. --- Research Overview The present study investigated whether social and general trust were associated with different coping strategies to face the COVID-19 pandemic. It was carried out during the first months of the pandemic, when Italy was the worst-hit country after China and the first Western country to enter a national lockdown, and there were no proven treatments for COVID-19. In this situation of great uncertainty, people had to turn to political and scientific authorities for information and guidance to prevent the spread of the infection. Regarding social trust, we focused our attention on trust in several institutions that, at different levels, were involved in managing the pandemic and containment policies: the Italian government, the government of one's region, the Civil Protection, and the European Union. Furthermore, we considered trust in the scientific community and the Catholic Church. In fact, the scientific community was highly salient at the time, since scientists were moving at a record speed to find effective treatments and create new vaccines and were often required to comment upon the pandemic on mass and social media. For its part, the Catholic Church played a crucial role in supporting people facing emergencies with existential reasons and psychological resources [59]. The efficacy of the containment measures depended on people's adherence to such restrictions. General trust concerned the belief that Italian citizens and humankind would do their best to face the COVID-19 pandemic, for instance, by respecting the rules imposed by the government and scientific authorities. In general terms, we expected that social and general trust were related to the adoption of adaptive coping strategies when facing uncertainty during pandemic times. Specifically, we considered the following adapting strategies: seeking social support, which implies seeking information and support from significant others; positive reinterpretation and growth, which implies the acceptance, containment, and positive interpretation of the situation; problem-solving orientation, which implies focusing on the problem, suppressing competitive activities, planning strategies to overcome the problem, and acting; transcendence orientation, which implies turning to God and praying for help and support. Moreover, we expected that trust would be negatively associated with the recourse to the avoidance strategy, which implies negation and behavioral and mental detachment from the problem. --- Materials and Methods --- Method 2.1.1. Participants A total of 1146 Italian participants were recruited through social networking sites and snowball sampling and volunteered for the study. Inclusion criteria were the following: living in Italy, being 18 years or older, and filling out the informed consent form. While age was registered at the end of the questionnaire, participants were informed that the questionnaire was intended for adults only in the initial consent statement. In addition, participants were excluded if they did not complete the questionnaire. Four participants were excluded as they did not provide consent, while further 259 did not complete the questionnaire, leaving a final sample of 883 participants . A total of 229 participants were categorized as having had "personal experience with COVID-19", as they reported having contracted the coronavirus and/or indicated that a family member or a close friend had contracted it . --- Procedure The Bioethical Committee of the University of Bologna approved the project. The questionnaire was administered via Qualtrics between 15 and 30 April 2020 and included further measures that were published elsewhere [9,60] and are not described in this paper. Respondents were first presented with measures of trust in different institutions and people; then, they filled in the measure of coping. Afterwards, respondents were asked whether they, a member of their family, or a close friend had contracted the coronavirus and indicated their religion . The great majority of respondents reported to be Catholic; 202 respondents indicated that they did not believe in any religion; 13 respondents indicated Buddhism, 37 respondents choose "Other religion", and 96 indicated that they believed in a Higher Entity without identifying with any religion. Since we were interested in trust in the Catholic Church as predictor of coping, we categorized the respondents distinguishing between non-Catholic and Catholic . Afterward, respondents reported their political orientation on an 11-point left-right scale and provided demographic information, including their region of residence. Since, at the time of data collection, Italy's northern regions were more severely hit by the coronavirus in comparison with central and southern regions, we categorized respondents' places of residence in northern regions and central-southern regions. --- Measures Trust. Participants read "Thinking to the following groups and institutions, please indicate how much they are able to face the COVID-19 pandemic." The instructions were followed by a list of groups and institutions. For social trust, the institutions were "the government of my region"; "the Italian government"; "the Civil Protection"; "the European Union"; "the Catholic Church"; and "the scientific community". For general trust, "the Italian citizens" and "humankind" were considered. The answers were given on a 5-point scale ranging from 1 to 5 . Coping strategies. We adopted the 25-item version of the Coping Orientation to the Problems Experienced-New Italian Version , which is derived from Carver's and colleagues [44] Coping measure . The COPE-NIV measure refers to five macro strategies of coping: seeking social support , positive reinterpretation and growth , problem-solving orientation , avoidance , and transcendent orientation . The answers were given on a 7-point scale ranging from 1 to 7 . We conducted a Confirmatory Factor Analysis with Mplus 8.4 to determine whether the items fitted the structure of the measure, as reported by [43]. To examine the model fit, we used various indices [62]: the comparative fit index and the Tucker-Lewis index , which should exceed 0.90 to be considered acceptable, and the root-mean-square error of approximation and the standardized root-meansquare residual , which should be less than 0.08 [63]. The fit of the model was initially not completely adequate according to the described fit criteria: CFI = 0.850; TLI = 0.831; RMSEA = 0.076, 95% CI [0.072, 0.079]; SRMR = 0.079. Based on the inspection of the modification indexes, we deleted 6 items . After deleting these items, fit indices resulted as acceptable: CFI = 0.932; TLI = 0.919; RMSEA = 0.061, 95% CI [0.056, 0.066]; SRMR = 0.052. The reliability indexes were acceptable: seeking social support , positive reinterpretation and growth , problem-solving orientation , avoidance , and transcendent orientation . --- Data Analysis All the analyses were run with IBM SPSS Statistics 23, except for the CFA on the measure of coping strategies, which was conducted with Mplus 8.4. To examine the hypothesized associations between social and general trust and the coping strategies, we conducted a series of bootstrapped hierarchical regression analyses with 5000 resamples, separately considering each coping strategy as the outcome variable. In Model 1, we accounted for the variability due to socio-demographic variables, personal experience with COVID-19, and political orientation. Specifically, in Model 1, we included age, gender , place of residence , personal experience with COVID-19 , political orientation, and Catholic religion . To analyze the relationship between different forms of trust and coping strategies, in Model 2, we added trust measures. Bivariate Pearson correlations were also conducted to determine linear relationships among all continuous variables. To gain an overview of participants' levels of social and general trust in different institutions and groups, we ran pairwise t-tests comparing different trust measures. To adjust for multiple comparisons and decrease the likelihood of committing a Type I error , we applied the Bonferroni correction for multiple testing and adopted more stringent levels of significance. This quite conservative correction consists of dividing the nominal significance level of the α test by the number of tests [64]. For pairwise comparisons, the highest number of tests is 27 ; accordingly, the significance level was set at p = 0.002 . For independent-sample comparisons, since we compared the groups with respect to 13 variables, the significance level was set at p = 0.004 . --- Results --- Descriptive Statistics Table 1 reports the means and standard deviations of the trust and the coping strategies measures for the total sample and as a function of gender, place of residence personal experience with COVID-19 infection , and Catholic religious affiliation . --- Social and General Trust First, a series of pairwise t-tests was run to compare respondents' level of social and general trust in different institutions and groups. Respondents turned out to trust the scientific community more than all the other groups and institutions considered, ts > 11.85, ps < 0.001, ds > 0.41. Respondents trusted the Civil Protection more than all the other institutions and groups , ts > 7.40, ps < 0.001, ds > 0.25. Trust in the regional government was higher than trust in the Italian government, the EU, the Catholic Church, Italian citizens, and humankind, ts > 7.31, ps < 0.001, ds > 0.25. Respondents reported higher trust in the Italian government compared to the EU and humankind, ts > 5.63, ps < 0.001, ds > 0.19, while they reported trusting humankind more than Italian citizens, the Catholic Church, and the EU, ts > 4.71, ps < 0.001, ds > 0.16. Trust in Italian citizens was higher than trust in the EU or the Catholic Church, ts > 7.94, ps < 0.001, ds > 0.27, whereas trust in the EU and trust in the Catholic Church scored lower than trust in all the other groups and institutions and did not differ from each other, t = -0.64, p = 0.521. Second, we conducted a series of independent sample t-tests to examine whether trust in the considered institutions and groups differed as a function of respondents' gender, place of residence, personal experience with COVID-19, and religion . The findings revealed that female respondents showed higher trust than men did in the Civil Protection, t = 4.79, p < 0.001, d = 0.36, and the Catholic Church, t = 3.30, p < 0.001, d = 0.25. Since, as mentioned, a significance level of p < 0.004 was set to adjust for multiple comparisons, no other comparisons between men and women can be considered significant, ps > 0.008. Respondents living in the north of Italy reported higher trust in the regional government, t = 8.19, p < 0.001, compared to those who lived in the central-southern regions. No other comparisons between respondents living in Northern vs. Central-Southern Italy were significant, ps > 0.011. There were no significant differences due to personal experience with COVID-19, ps > 0.060. Finally, respondents who identified as Catholic reported higher trust in the Catholic Church, t = 10.49, p < 0.001, d = 0.72, compared to non-Catholic respondents. The analyses showed no other effects due to religion, ps > 0.013. --- Coping Strategies Pairwise t-tests highlighted some differences in the recourse to the various coping strategies. Positive reinterpretation and growth turned out to be the most relevant coping strategy, with scores being higher than those of all other strategies, ts > 31.44, ps < 0.001, ds > 1.08. Seeking social support scored higher than avoidance and transcendent orientation, ts > 18.01, ps < 0.001, ds > 0.62, while it scored lower than problem-solving orientation, t = -5.42, p < 0.001, d = 0.19. Finally, the scores of problem-solving orientation were higher than those of avoidance and transcendent orientation, ts > 21.50, ps < 0.001, ds > 0.74, while the scores of avoidance were lower than those of transcendent orientation, t = -5.97, p < 0.001, d = 0.21. Independent sample t-tests revealed that women scored lower than men on the avoidance strategy, t = -3.47, p < 0.001, d = -0.26, while they scored higher than men on all other strategies, ts > 3.47, ps < 0.001, ds > 0.26. Participants living in Central-Southern Italy reported higher levels of the transcendent orientation strategy compared to participants from Northern Italy, t = 7.21, p < 0.001, d = 0.50. No other comparison between the two areas of residence reached the statistical level of significance set for this analysis, ts < 2.45, ps > 0.014, ds < 0.17. Regarding experience with COVID-19, those who had no personal experience made a greater recourse to the avoidance strategy, t = 3.95, p < 0.001, d = 0.30, with a nearly significant effect on the transcendent orientation strategy, t = 2.87, p = 0.004, d = 0.22. There was no other significant effect for experience with COVID-19, ts < -1.97, ps > 0.049, ds < 0.15. Finally, participants who identified as Catholic scored higher than non-Catholic participants on the transcendent orientation strategy, t = 19.49, d = 1.34. No other comparison between Catholic and non-Catholic participants reached statistical significance, ts < 2.39, ps > 0.008, ds < 0.16. --- Hierarchical Regression Analyses Table 2 presents bivariate correlations among the measures of trust and coping strategies. Overall, the results confirmed the existence of significant correlations between coping strategies and the dimensions of social and general trust. However, it is noteworthy to point out a different pattern of correlation between the different forms of social trust and the coping strategies. Seeking social support was positively correlated with trust in the Italian government, in the Civil Protection, in the EU, in the Catholic Church, and in the scientific community. No significant correlation emerged between seeking social support strategy and trust in the regional government. Positive reinterpretation and growth was correlated with all the different social trust measures, while problem-solving orientation positively correlated only with trust in the Civil Protection, in the Catholic Church, and in the scientific community. Trust in the regional government, in the Italian government, in the Civil Protection, in the Catholic Church, and in the scientific community was negatively correlated with the avoidance strategy. The analysis did not reveal any significant correlation between avoidance and trust in the EU. Transcendent orientation was positively correlated with trust in the Catholic Church and negatively correlated with trust in the regional government and in the EU. Trust in Italian citizens and in humankind, as the dimension of general trust, was positively correlated with seeking social support, positive reinterpretation, problem-solving orientation, and transcendent orientation. Regarding the avoidance coping strategy, the analyses revealed only a negative correlation with trust in Italian citizens. As explained before, to test for the associations among different forms of trust and coping strategies, we conducted a series of hierarchical regression analyses, whereby age, gender, place of residence, personal experience with COVID-19, political orientation, and Catholic religion were entered in Model 1, and the trust measures were entered in Model 2. Coping strategies were entered as outcome variables. Table 3 shows the results of these models. The analyses revealed, for Model 1, the significant effects of age and gender on the seeking social support strategy. Being younger and being female were associated with a stronger reliance on this strategy. When the trust measures were entered in Model 2, stronger trust in the Catholic Church and in the scientific community were significantly associated with a greater tendency to seek support to face the pandemic. Second, in Model 1, the positive reinterpretation and growth strategy was positively associated with being female and living in Central-Southern Italy. In Model 2, reliance on this strategy was negatively related to being Catholic and trusting the EU. Conversely, trust in the Catholic Church, in the scientific community, and in humankind was positively associated with reliance on the positive reinterpretation and growth strategy, while trust in the EU was negatively associated with such a strategy. In Model 1, the analysis concerning the problem-solving orientation strategy revealed, again, positive associations with being female and living in Central-Southern Italy. Respondents who reported a personal experience with COVID-19 also scored higher on such a strategy. In Model 2, trust in the Catholic Church and trust in humankind were significantly related to problem-solving orientation. In Model 1, gender, age, experience with COVID-19, and political orientation had significant associations with the avoidance strategy. Specifically, being female, being older, and having had personal experience with COVID-19 were associated with a lower reliance on the avoidance strategy. Right-wing political orientation was, instead, positively associated with this strategy. In Model 2, trust in the EU was positively associated with the reliance on the avoidance strategy, which was, instead, negatively related to trust in the Catholic Church and trust in the scientific community. Note: Parameters are beta weights. Significant parameters are in bold. Gender was coded 0 = men, and 1 = women. Place of residence was coded 0 = northern regions, and 1 = central-southern regions. Experience with COVID-19 was coded . Catholic religion was coded . * p < 0.05, ** p < 0.01, *** p < 0.001. Finally, in Model 1, being older, being female, living in Central-Southern Italy, being right-wing-oriented, and being Catholic were associated with a greater recourse to the transcendent orientation strategy. In Model 2, trust in the regional government, in the Italian government, and in the EU was negatively related to this strategy, which was instead positively associated with trust in the Catholic Church and humankind. --- Discussion The present research aimed to verify the role of trust in helping individuals to face the strains and uncertainties due to COVID-19. In particular, we focused on social trust, that is, the confidence that authorities and institutional organizations would be able to manage the pandemic, and general trust, which represents the confidence that other human beings would do their best to limit the spread of disease. To reach this goal, we conducted a study during the first phase of the pandemic evolution in Italy, focusing on different aspects of social and general trust regarding the adoption of individual coping strategies against COVID-19. Specifically, we considered four adaptive strategies of coping, that is, seeking social support, positive reinterpretation and growth, problem-solving orientation, and transcendent orientation, and one maladaptive form of coping, that is, avoidance. Overall, the findings supported the general expectation that social trust and general trust were associated with individuals' ability to employ different coping strategies. First, the seeking social support coping strategy underlines the importance of significant others and one's own social network for information seeking and understanding and for emotional support [65]. This adaptive strategy implies the perception of being valued by others and of being part of a social network [66] and concerns humans' fundamental need for belongingness [67]. Our results confirmed the importance of seeking social support and showed that trust in the Catholic Church and in the scientific community was associated with a stronger tendency to seek information and emotional support from others. It is noteworthy that despite having different scopes and often being considered to be in conflict with each other [68], both institutions promoted an adaptive coping strategy that helps individuals focus on the problem's solution and the relief from emotional stress. It should also be noted that, in general terms, the tendency to seek social support seems to decrease with age. A possible explanation may rely on the peculiar situation due to the pandemic, since the lockdown challenged the way of maintaining social connections. Digital and online technologies became crucial to maintain interpersonal relationships, but, at the same time, they prevented older adults from staying connected with others since some older people were unable or reluctant to use technology, leaving them vulnerable to social isolation [69]. Positive reinterpretation and growth is a coping strategy that refers to the acceptance, containment, and positive reframing of stressful events [61]. It is considered an adaptive strategy that helps individuals face negative emotions by construing a stressful situation in positive terms and focusing on one's own emotional growth. This strategy was strongly employed by our participants. Concerning the effect that social trust exerted on this coping strategy, the results confirmed the positive association between trust in the Catholic Church and in the scientific community. The higher the trust in these different institutions, the more participants considered the pandemic as an opportunity for personal growth. Interestingly, trust in the European Union was negatively associated with this coping strategy. We suspect that, at the very beginning of the pandemic, Italians' trust in the EU might have been under strain due to the lack of a joint and coordinated reaction by European countries, as well as by the awareness that Italy was the only EU country to be severely hit by COVID-19 [70,71]. This result can be interpreted in terms of compensatory processes: the decrease in trust in Europe could have enhanced participants' need to restore control over uncertain and unpredictable events by positively reframing stressful events. Moreover, identifying oneself as Catholic had a negative relation with the positive reinterpretation of COVID-19 events. This result seems to contradict the role of trust in the Catholic Church. A possible explanation relies on the unique Italian situation, especially in comparison to other European nations. A vast majority of Italians continue to consider themselves Catholic, even if this affiliation is now expressed in a variety of forms, with different levels of intensity, and, in many cases, it is characterized by contradictions and feelings of ambivalence [72]. Moreover, compared to the past, regularly practicing Catholics are now a minority in Italy, even if they are far more numerous than in other European nations. In our opinion, the measure of trust in the Catholic Church captured participants' tendency to identify with the institution and consider it as an epistemic authority, while the self-definition of being Catholic is a broad category characterized by pluralistic points of view. Such a distinction can help understand the seemingly contradictory findings concerning trust in the Catholic Church and religiosity. Finally, our findings highlighted that general trust-specifically, trust in humankindwas associated with the positive reinterpretation coping strategy. Indeed, the confidence that other human beings would be able to properly act to limit the spread of the pandemic increased the positive interpretation of this dramatic event. Problem-solving orientation is a problem-focused coping strategy characterized by the employment of active and interpretative strategies to change or eliminate the underlying causes of stress via individual behavior [61]. It is considered an adaptive coping strategy that helps individuals focus on the root of the problem. Our results showed that this strategy was particularly employed by people living in the center-south of Italy and by those who had experience with COVID-19. Again, trust in the Catholic Church and trust in humankind were positively related with the employment of this coping strategy: the higher the trust in the Catholic Church institution and in other human beings was, the more our participants used planning and active strategies to face the pandemic's strains. Avoidance is a coping strategy aimed at escaping stressful situations, experiences, or difficult thoughts and feelings rather than dealing with them [61]. It is generally considered as a maladaptive form of coping because it does not address the sources of stress, and it tends to increase stress and anxiety when overused. The regression analysis revealed that age and having experience with COVID-19 were negatively associated with the use of the avoidance strategy: older participants and those who had personal experience with the disease tended to limit the employment of this strategy. Moreover, the analysis showed interesting results for social trust: considering the Catholic Church and the scientific community as a resource to fight the pandemic represented a protective factor that limited the tendency to escape from the problem. Analogous to the positive interpretation strategy, trust in the European Union did not seem to help individuals employ adaptive coping strategies: the higher the trust in the EU was, the more participants relied on avoidanceoriented coping strategies to face the pandemic's stressful events. Finally, transcendent orientation is an emotion-focused coping strategy that refers to turning to religion to find comfort through praying or meditating. The literature showed that in Italy there was an increase in religiousness during the COVID-19 crisis: people derived more comfort in religious activities during the hard times of the pandemic [73][74][75]. Our findings showed that place of residence, political orientation, and religious beliefs were positively related to the employment of this strategy. Participants who lived in the center-south, who had right-wing beliefs, and who defined themselves as Catholic tended to refer to religion to find relief from COVID-19's strains and uncertainties. Interestingly, the results concerning social trust highlighted a different pattern of results. Trust in civil authorities such as the regional government, the Italian government, and the European Union reduced the use of this coping strategy, while trust in the Catholic Church was positively associated with its employment. Concerning general trust, our findings showed that trust in humankind seems to lead to relying on religion and prayer to face a stressful situation. An important strength of the current study is that we considered different aspects of social trust. In particular, different authorities that played an important role in managing the pandemic were considered: several national and European political actors, the Catholic Church, and the scientific community. The results from the correlation analyses showed that the different dimensions of social trust correlated differently with the coping strategies. For example, the higher the trust in the Civil Protection, in the Catholic Church, and in the scientific community was, the higher the individuals' employment of problem-solving orientation coping strategies. The regional, Italian, and European authorities did not seem to play an important role in helping individuals to plan strategies to overcome the situation. This was probably due to the peculiar situation, since, in the first phase of the pandemic, local, national, and European authorities might have been under strain due to the lack of information and certainty about the infection, and citizens considered the Civil Protection, the Catholic Church, and the scientific community as more trustworthy epistemic authorities. At the same time, all the social trust dimensions, except for trust in the EU, negatively correlated with the avoidance coping strategy. The higher the trust in these authorities was, the lower the employment of this maladaptive coping strategy was. As previously pointed out, at that time, when Italy was the first European nation dealing with COVID-19, the lack of clear support could diminish trust in the European Union. These results and the evidence gathered from the regression analyses showed that social trust dimensions are not all alike in helping individuals to adopt adaptive coping strategies. Moreover, they highlight that social trust cannot be taken for granted, but the level of trust increases or decreases as a consequence of the authorities' ability to deal with dramatic events. Interestingly, our results showed significant gender differences in coping mechanisms, with women using more adaptive coping strategies than men. According to the literature, women tend to use more emotion-focused coping, whereas men use more problem-focused coping strategies, when dealing with stressful events . Our results showed that, irrespective of the focus of the coping strategies , our female participants were able to employ adaptive coping strategies more often than male participants. Furthermore, women used avoidant coping strategies less often than men. --- Limitations and Future Directions Some limitations of the present study should be acknowledged. Firstly, the participants' recruitment adopted the snowball sampling strategy, which is not based on a random selection of the sample. Therefore, the study's sample did not reflect the actual pattern of the general Italian population. In addition, the online administration of the questionnaire could also have undermined the sample representativeness. Secondly, the cross-sectional design limited the possibility of solidly establishing cause-and-effect relationships between the examined variables. Future studies could investigate in depth the cause-and-effect relationships between the different forms of trust and the employment of adaptive coping strategies. --- Conclusions Overall, these findings contribute to the literature on the antecedents of coping strategies as well as to the growing evidence concerning individuals' responses to an unprecedented global event such as the COVID-19 pandemic. In particular, they highlight how, in the first phase of the pandemic-when uncertainty about the causes and the better ways to deal with the situations was at its most-feelings of trust with different institutions as well as of general trust in others played a protective role and positively influenced individuals' coping. Given that adopting adaptive coping strategies-such as problem solving, positive reinterpretation, transcendent orientation, and seeking social support-was found to be conducive to a higher level of well-being and fewer depression and anxiety symptoms [48][49][50], interventions to improve resilience and mental health during stressful situations should consider the important role played by social, political, and religious institutions. In general terms, it is important that practitioners and other professionals are aware that social trust and general trust in others are critical factors in sustaining people during tough situations. Our results showed that individuals' feelings that they can rely on institutions as well as on the scientific community as resources for dealing with crises can be crucial in dramatic situations. Of course, political, social, religious, and scientific agencies should build the basis for social trust before an emergency event occurs, because unreliable authorities may not be trusted by citizens. Consequently, citizens could be distrustful of the adoption of authorities' recommended behaviors [8]. National and European political leaders as well as the scientific community should be aware of the importance of nurturing citizens' social trust by employing effective policy and clear communication strategies. --- Data Availability Statement: Data available on request from the authors. --- Acknowledgments: The authors thank all the participants who helped with the data collection. --- Author Contributions: Conceptualization, A.R.G., S.M. and L.B; methodology, A.R.G., S.M., L.B. and M.M.; data curation, S.M.; writing-original draft preparation, A.R.G.; writing-review and editing, all authors. All authors have read and agreed to the published version of the manuscript. ---

Research in the field shows the crucial role of trust in the functioning of many aspects of social life, especially when dealing with emergencies. We report the results of a study (N = 883) carried out in Italy during the first phase of the COVID-19 crisis to assess whether and how social trust (i.e., trust in those who have the authority and responsibility for making decisions, such as the Italian government, the regional government, the Civil Protection, the European Union, the Catholic Church, and the scientific community) and general trust (i.e., trust in the trustworthiness of other individuals, such as Italians and humankind) are associated with the employment of coping strategies in dealing with the challenges of the pandemic. Overall, the results highlight that trust in different authorities and institutions and trust in other human beings are positively associated with the adoption of adaptive coping strategies such as seeking social support, positive reinterpretation and growth, problem-solving orientation, and transcendent orientation. In contrast, they are negatively related to the adoption of maladaptive coping strategies such as avoidance. These findings point out the benefits of various types of trust in helping individuals deal with crises, such as the COVID-19 pandemic.

Introduction Unintended pregnancies are pregnancies that are unwanted which occurs with no desire to have a child, mistimed which occurs before the desired time or unplanned at a time of conception [1]. Worldwide, an estimated 33 million unintended pregnancies are a result of contraceptive failure or incorrect use [2]. In developing countries, the majority of unintended pregnancy occurs due to using traditional family planning methods or not using any type of modern contraceptives [3], and it's the main reason for induced abortion [4]. In Ethiopia, 24.7 and 42% of unintended pregnancy caused by contraceptive failure and not using contraceptive methods, respectively [5]. Globally, each day about 100 million sexual intercourse takes place, of these 1 million conceptions occur [6]. Among 208.2 million pregnancies of the world, 41% are unintended [7]. In addition, 213 million pregnancies occur each year, 89% of this occurs in developing countries, and 40% of these are unintended pregnancies [8]. Unintended pregnancy is the major sexual and reproductive health problems that impose to substantial health, economical and psychosocial costs to individual and society as well as significant emotional distress to women, families, and society [3,9,10]. Besides, contributing to late antenatal care visit, increase exposure to the substance, less care for their child, and experiencing physical and psychological violence [11]. The impact of unintended pregnancy is higher during the adolescent period that levy to dropping out of school, unstable and lack of proper management of family relationships [12]. In addition, children born to teenage mothers are much more likely to experience a range of negative outcomes in later life, such as developmental disabilities, behavioral issues and poor academic performance [13]. About half of unintended pregnancies in developing countries result in unsafe abortion which accounts for 13% of maternal deaths [14]. Poor educational status, lack of access to health services and health education, poor economic status, single in marital status, peer pressure, sexual violence, and family planning failure were exposed to unintended pregnancy [9,15,16]. Unintended pregnancy is one of the most evident for the violation of women's sexual and reproductive rights in developing countries. Despite the availability of highly effective methods of contraception, different studies in Ethiopia revealed that there is a high level of unintended pregnancy [17][18][19][20]. Therefore, this study aimed at assessing the magnitude and associated factors of unintended pregnancy among pregnant women attending antenatal care in Maichew town, northern Ethiopia. --- Main text --- Methods --- Study design and period An institution based cross-sectional study was conducted from April 5 to May 4, 2017. --- Study area The study was conducted in Maichew town, Tigray region, northern Ethiopia. Maichew town is found 665 km far from Addis Ababa, the capital of Ethiopia. According to the Central Statistical Agency of Ethiopia , the total population of 23,395 lives in Maichew town, of which 12,395 are females. The town has one general hospital, two health centers, and five private clinics. --- Study population All pregnant women who had ANC follow up during data collection were included in the study. However, pregnant women who were seriously ill during data collection were excluded. --- Sample size determination and sampling procedure The required sample size was determined using a single population proportion formula by considering the proportion of unintended pregnancy in Welkaite, 26.4% [17], a 95% confidence interval and 5% of margin of error. By adding 10% for the non-response rate, the final sample was 329 women. The data were collected from all governmental health institutions in Maichew town. Based on the previous 2 months of the Maichew Town Health Administrative report, 325, 189 and 125 pregnant women had ANC follow up in Lemlem Karl hospital, Maichew health center and Semere Melese health center, respectively. Then, 167 women from Lemlem Karl hospital, 68 women from Maichew health center, and 97 women from Semere Melese health center were selected using systematic sampling and proportional allocation technique. --- Data collection tools and techniques A structured interviewer-administered face to face interview was used to collect data. The questionnaire was consisted of sociodemographic and economic characteristics, sexual behaviors, and behavioral characteristics of participants which was initially prepared in English and translated into local languages and back-translated to the English language by three language experts to check the consistency. A total of three BSc. Midwifery data collectors involved in the study. Data collectors received 3 days of training prior to data collection. A pretest was done on five percent of the sample on the Garjale health center . Finally, the collected data were checked for its consistency and completeness in each day before compiling. --- Data processing and analysis The collected data were checked and entered into Epi-Data version 3.10 and exported into SPSS version 20 statistical software for analysis. Descriptive statistics were carried out and the result was presented in text and table. Bivariate logistic regression analysis was used to identify factors associated with an unintended pregnancy. Crude odds ratio and adjusted odds ratio with 95% confidence interval was calculated to identify factors associated with unintended pregnancy. Variables with a p-value less than 0.2 in the bivariable logistic regression analysis were retained into a multivariable logistic regression analysis to control the potential confounders. In the multivariable logistic regression analysis, variables with p-value < 0.05 at 95% confidence interval were considered as statistically significant. --- Result --- Socio-demographic and economic characteristics A total of 313 pregnant women participated in the study making the response rate 95.14%. The median age of the respondents was 26 years with an inter-quartile range of 8 years. The majority of women were in the age range of 18-34 years. Nearly two-thirds of women lived in an urban area. Moreover, 33.1, 11.8 and 8% of women had no formal education, lived alone and had five or more children, respectively . --- Sexual and behavioral characteristics About, 18.8% of women had multiple sexual partners. A substantial proportion of women discussed with their partner on pregnancy-related issues. The most and 53% of women heard about and reported that school was a major source of information on contraceptive methods, respectively. Nearly one-third of the participants had good knowledge of contraceptive methods . --- Prevalence of unintended pregnancy 93012 The overall prevalence of unintended pregnancy was 29.7% . --- Factors associated with unintended pregnancy Both bivariable and multivariable logistic regression analyses were done to see the effects of the selected characteristics on unintended pregnancy. As it is presented in Table 3, knowledge of family planning, economic status, participant's age, living arrangements, number of children, mother's occupation, husband's occupation, and marital status were factors associated with an unintended pregnancy in the bivariable analysis. Moreover, the result of multivariable logistic regression analysis revealed that marital status, number of children and living arrangement were significantly associated with unintended pregnancy. In this study, the high odds of unintended pregnancy were observed among women who lived alone and had three or four children . Moreover, the odds of unintended pregnancy were 38.6 times as high among women who were single compared to married . --- Discussion This study was conducted to assess the prevalence and associated factors of unintended pregnancy among women attending antenatal care in Maichew town, northern Ethiopia. The prevalence of unintended pregnancy among women attending ANC in Maichew town was found to be 29.7% . The finding is in line with other studies done in Hosanna [21], Walkait [22] and Mekelle town [23]. But compared to other studies, this finding is lower than those of studies done in Addis Ababa [3], Kenya [24], and Wolaita zone [25]. The possible reason for the difference might be due to the socio-demographic difference of participants, particularly the aforementioned studies were conducted in capital cities whereas the current study was conducted in district town having poor awareness of family planning, and sexual and reproductive health rights. Moreover, the prevalence of this study is higher than those of other findings in Arbaminch town [26], South Africa [27] and India [21]. This might be due to the difference in the educational status of the participants having more awareness of family planning methods. The difference in the availability and accessibility of family planning methods may contribute to the difference. Pregnant women who were single and living alone were independently associated with an unintended pregnancy. Single women were more likely to develop untended pregnancy than a married one. This study is supported by studies done in Gelemso [9], Wolaita zone [25], Mulago hospital [28], Ghana [29] and Kenya [24]. This might be due to women who are single or living alone are prone to unsafe sexual acts because of parents or families are important to monitor and support the behavior, and sexual and reproductive health of women. Moreover, it might be due to social influence or bad taboos that prevent women to use family planning methods. In this study, women who had three or more children were more likely to develop unintended pregnancy as compared to their counterpart. The same finding was observed in Felegehiwote, Bahir Dar [30], Gelemso [31], Hosanna town [21], rural Ghana [29] and Zambia [32] studies. This might be due to women having a large family size spend time to take care of their children which may result in missing appointments and even delay to seek maternal health services. --- Conclusion The prevalence of unintended pregnancy was higher in the study area which confirmed the major public health problem. Marital status, the number of children and living arrangement were significantly associated with an unintended pregnancy. Hence, creating awareness on sexual and reproductive health rights, increasing the accessibility of maternal health services for women are recommended to reduce unintended pregnancy. --- Limitation of the study The major limitation of this study was the use of small study participants. It's recommended to use a large sample size for future studies. --- --- Abbreviations ANC: antenatal care; AOR: adjusted odds ratio; CSA: Central Statistics Agency; CI: confidence interval; COR: crude odds ratio; EDHS: Ethiopian Demographic Authors' contributions EAK, BGG and HIO developed the concept of the research, acquired the research proposal, facilitated data collection and data entry. EAK and LBZ were prepared the first daft of this manuscript. EAK, BGG, HIO, LBZ, ADD, HAD, MWA and GGA participated in statistical analyses and critically reviewed the manuscript. All authors read and approved the final manuscript. --- --- --- Competing interests The authors declare that they have no competing interests. ---

Objective: Unintended pregnancy is one of the most public health issues in the world, and it is the major sexual and reproductive health problem which carries a higher risk of morbidity and mortality for women, often due to unsafe abortion. Even though family planning services are effective and available than ever before, unintended pregnancy and unsafe abortion are the major public health problems in the study area. Therefore, this study aimed at assessing the magnitude and associated factors of unintended pregnancy among pregnant women attending antenatal care follow up in Maichew town, northern Ethiopia. An institution based cross-sectional study was conducted on 329 pregnant women selected with a systematic sampling technique from April 5 to May 4, 2017.The magnitude of unintended pregnancy among pregnant women attending antenatal care in Maichew was found to be 29.7% (95% CI 24.30, 35.50). On the other hand, single in marital status (AOR = 38.6, 95% CI 10.07, 148.01), living alone (AOR = 9.9, 95% CI 1.80, 53.40) and having three or four children (AOR = 3.5, 95% CI 1.10, 11.04) were factors associated with an unintended pregnancy. Creating awareness about unintended pregnancy associated factors and implication of unintended pregnancy is highly recommended.

Introduction Mental health problems affect at least one in four people worldwide at some time in their lives [1]. Neuropsychiatric disorders are the second greatest cause of the burden of disease in Europe and are the greatest cause of years lived with disability [1]. Since the beginning of the economic recession in Europe, various studies have been pointing out the relationship between crisis and poor mental health [2], as well as the various determinants that may be affecting it [3][4][5][6][7]. The majority of studies on the subject have focused on the analysis of psychological and behavioural morbidity, with predominance of countercyclical studies on depression, anxiety, or violent behaviour and their relation with job loss as a risk factor. Additionally, other studies have analysed changes in some behavioural risk factors, suggesting that in a situation of prolonged economic depression, it is likely for higher rates of alcoholism and smoking or substance abuse to be observed [7][8][9]. On the other hand, some authors have also described the impact of the economic crisis on general mortality, as well as on some specific causes, such as suicides or external causes, suggesting countercyclical as well as procyclical effects [7,[10][11][12][13]. In Spain, following this trend, a number of studies have been published in recent years describing the association and impact of the economic crisis and poor mental health [14][15][16][17][18]. In general, it has been identified that economic changes have intensified the social exclusion and marginalisation of people with mental health problems, especially in men and, to a greater extent, in those with low levels of education [11,14,19], in people affected by mortgage-related financial difficulties or evictions [9,17], as well as among immigrants [20][21][22]. Unemployment has also been shown to have a significant negative impact on overall health and mental health and has been described as the main risk factor for mental disorders [16,17]. Despite the recent literature on this subject, several limitations have been described in the studies, as well as the need to generate new knowledge from the basic sciences, and the epidemiological method to establish the association between economic decline and the effect on mental, behavioural, or somatic health has also been highlighted [7,18,23]. Finally, the impact of the financial crisis on Spain, unlike other countries was delayed until the end of 2011. This was due partly to the cushioning of the highly developed popular social support network and partly to the governmental contentious strategy that, despite the increase of the debt, allowed to maintain the benefits of the welfare state and national social protection [11]. Nevertheless, Spain has also been described as one of the countries of the European Union enduring the worst consequences due to the weakening of its economic activity and the deterioration of its public finances [9]. The Valencian Community is one of the 17 regions in which Spain is structured, located in the Mediterranean coast. It has been one of the regions most affected by the crisis, along with Andalusia, Castilla-La Mancha, and Murcia, unlike other, less affected territories such as Navarre, La Rioja, and the Basque Country. While poverty has grown in Spain to 8% between 2008 and 2011, in the VC it has increased up to 18%, ranking second, only behind the Canary Islands [24]. On the other hand, although the prevalence of mental disorders in Mediterranean countries has been described as inferior to the countries of the north of Europe [25], it is necessary to find out the impact on the health due to the much more pronounced economic changes in these countries. Given the context of changes in Spain in recent years-related to the economic crisis-and the lack of evidence from studies that have suggested a specific methodology to evaluate the contribution of socioeconomic changes introduced since the beginning of the economic crisis, the objectives of this study are to analyse the association of the risk of poor mental health with various demographic, socioeconomic, health status, quality of life, and social support variables; and, specifically, to evaluate the contribution of socioeconomic variables most affected by the economic situation on changes in the prevalence of poor mental health risk between 2005 and 2010 in the general adult population of the VC, a Mediterranean region of the Spain. --- Materials and Methods --- Design, Population, and Sample This is a study of prevalence evolution of poor mental health risk and associated factors in adult population, over 15 years of age, between 2005 and 2010. The total sample sizes were 5781 subjects in 2005 and 3479 in 2010, living in the VC, an autonomous community with just over five million inhabitants in 2008. The samples corresponded to the adults of the Health Surveys of the Valencian Community carried out every year under study, being representative of the non-institutionalised general adult population of the VC. The subjects of the samples were selected using a complex sampling design that assigned each subject a weighting according to their representativeness. The weights were included in the ESCV databases provided by the Health Plan Service of the Conselleria de Sanitat of the Generalitat Valenciana . Details about the survey methodology have been published elsewhere [26,27]. This research uses two transversal observational studies based on administrative data. An informed consent was required from every participant in the Health Surveys of the Valencian Community by the Valencian Health Authorities. According to national regulations, data from National or Regional Health Surveys [28] are public in Spain and the Valencian Health Authorities are responsible for and guarantee confidentiality and anonymity, making the approval of an ethics committee unnecessary. The researchers only had access to public data that had been rendered anonymous, and so this research poses no ethical issues. --- Variables The variable answer was 'Case at risk of poor mental health' with yes or no as possible results. In order to construct this variable, the questions corresponding to the 12-item General Health Questionnaire were used, assigning to each of the items that compose the score 0 if the answer was 0 or 1, and 1 if the answer was 2 or 3. The subject at risk of poor health was classified if the sum of the scores of the 12 items was equal to or greater than 3. In addition, demographic explanatory variables were included for both years: Sex , age ; variables of socioeconomic level: Country of birth , level of education , employment status , personal self-perceived income level , occupational class ; Health status variables: Presence of a nonmental chronic disease , number of nonmental chronic diseases, presence of a disability , score of self-perceived quality of life questionnaire or EuroQoL-5D [29]; and variables related to social support: Marital status , cohabitation with a partner . Both the response variable and all the explanatory variables were measured equally in both the 2005 and 2010 surveys. --- Methods of Analysis Number, percentage, and 95% confidence interval of subjects in each category of the explanatory variables were calculated for the qualitative variables and for each year of the survey. Prevalence and 95%CI of the GHQ+ result altogether and in each category of the explanatory variables were calculated by analysing with the Chi-Square test the significance of the association between GHQ+ and each variable, for each sex, separately. For the quantitative variables and for each year, the number, average value, and 95%CI in each category of the response were calculated, analysing with the t-test the significance of the differences of averages, separating by sex. For the analysis of association of GHQ+ with the explanatory variables, logistic regression models for both sexes were adjusted together. As a measure of association, the Odds Ratios were calculated between the GHQ score and the explanatory variables, as well as their corresponding 95%CI, first in simple analysis and then adjusting for all variables. The statistical significance of the interaction of all variables with sex was checked to verify the homogeneity of the models in both sexes. All the analyses have taken into account the complex sampling design, using the weighting of the subjects of each sample. Statistical analysis was performed using the SPSS v.18 program. The level of significance was 0.05 in all analyses. In order to study the extent to which changes in the employment situation and income could have affected the observed changes in the prevalence of risk of poor mental health from 2005 to 2010, the percentage distribution of the sample by these variables in 2005 was projected on each one of the levels and sublevels of the categories of variables in 2010, adjusting for age, sex, and country of birth. The variables included in this analysis were sex , age , country of birth , income and employment status , giving rise to a total of 72 strata or different levels . The projection was made as follows: -Calculation of the frequencies that would have been observed in the 2010 sample in every stratum, in case the percentage distribution observed in 2005 would have not changed: P 2010 e i = P 2005 i ∑ P 2005 i ∑ P 2010 i ; i = 1, 2, 3 . . . , 72, where: -Finally, the overall expected prevalence in 2010 was calculated if the percentage distribution with respect to 2005, disaggregated by sex, had not been changed, adding all previous cells, as follows: P 2005 i = prev 2010 e,total = ∑ n e_GHQ+ 2010 i ∑ P 2010 e i . When comparing observed versus expected prevalence, in 2010, this standardisation procedure would make it possible to distinguish between: The change in overall prevalence due to variations in population structure from 2005 to 2010 in terms of variables considered, and the change due to the different period considered that would not be justified by the variations in the population structure. --- Results --- Risk of Poor Mental Health and Associated Variables For 2005, 5781 subjects were analysed, 2855 of which were men and 2926 were women. The prevalence of risk of poor mental health was 20.0%, being 16.2% in men and 24.0% in women. In 2010, we analysed 3479 subjects, 1702 of which were men and 1777 women. The prevalence of risk of poor mental health increased to 27.8%, being 25.4% in men and 30.6% in women. In 2005, the highest prevalence of GHQ+ was found among older men, without studies, no employment status, low income level, presence of some chronic nonmental illness, presence of some disabilities, marital status separated, divorced, or widowed, and those with poor social support. In 2010, the most frequent profile was those born outside Spain, unemployed labour status, low income level, manual workers, presence of some chronic nonmental illness, presence of a disability, separated civil status, divorced or widowed, single, and those with poor social support . Regarding women , the profile of those with the highest prevalence of poor mental health risk in 2005 was of an older person, without studies, unemployed, low income, presence of chronic nonmental illness, presence of a disability, widowed, single, and those with poor social support. In 2010, they were the most elderly, without studies, unemployed, low income level, presence of some chronic mental illness, presence of chronic nonmental illness, presence of a disability, widowed, and those with poor social support. For both years, in both men and women, quantitative explanatory variables present significantly worse average values in subjects at risk of poor mental health . Overall, for both men and women, prevalence of poor mental health risk increased from 2005 to 2010 in most of the categories of variables studied, in line with the overall increase in prevalence. Table 4 shows the ORs and 95%CI risk association of the poor mental health with the other variables , for both sexes together. In the two years, a similar logistic multivariate model is reached. It can be observed that the association of the labour situation and the level of income remain significant after adjusting for the remaining variables in both years, with high ORs for the categories of 'unemployed' and 'low income'. The country of birth was not significant in 2005 after adjusting for the rest, but was so in 2010. --- Impact of the Crisis and Changes in Prevalence Table 5 shows the results corresponding to the change in the prevalence of poor mental health between the years 2005 and 2010. It is observed that if the population structure in terms of age, country of birth, income level, and employment status had not changed, the total prevalence of poor mental health expected in 2010 would be 21.4% in men and 28.5% in women, compared to the 25.4% in men and 30.6% in women actually observed. Consequently, the difference between these expected prevalence values and those actually observed in 2010 would be related to changes in population structure. Given that the prevalence of the risk of poor mental health observed in 2010 was 25.4% in men and 30.5% in women, the difference was higher than expected and, therefore, the observed prevalence excess attributable to changes in population structure would be around 4.0% in men and 2.1% in women in absolute terms. These values would translate, in relative terms, into a contribution of 43.5% and 31.8% of the increase in prevalence as attributable to changes in population structure, while the rest of the increases would be attributable to changes in other explanatory factors of the risk of poor mental health. In order to explain the changes in population structure, Table 5 shows its distribution in each of the categories of variables considered. It can be seen that the distribution in 2005 and 2010 is similar by age group, in both men and women. In the case of the country of birth, there is an increase in absolute terms of 4.2% of the foreign population in men and around 2.0% in the case of women . However, the greatest differences are detected in variables such as employment status and income. In the case of income, there is an increase in the population with less income of 11.1% in men and 9.3% in women in absolute terms, whereas for the labour situation, there is an increase in the unemployed population of 11.9% in men and 6.3% in women. It would be worth noting that the prevalence of poor mental health associated with these levels is the highest for the corresponding variable. In men, it can be observed that in the low-income category, the observed prevalence is 23.5% in 2005 and 33.5% in 2010 , while regarding those in unemployment it is 34.4% in 2005 and 37.3% in 2010 . Something similar occurs in the case of women, where the prevalence of poor mental health in the low-income categories is 33.6% in 2005 and 38.5% in 2010 and in unemployed women of 27.5% and 44.1%, respectively. --- Discussion This study has shown, firstly, that various variables regarding the demographic and socioeconomic context, as well as the areas of social support and health and quality of life were closely associated with the risk of poor mental health. This is particularly so regarding: Sex , age , country of birth , number of chronic nonmental illnesses , disability , quality of life , social support , cohabitation in couple , employment situation , and income . These results coincide with some studies in Spain [5,6,30] that found the strongest predictors associated with poor mental health to be: Being a woman, having chronic illness, and having poor perception of health and quality of life and limited activity. However, the age in some studies had an inverse sense [5,30]. Regarding changes in the prevalence in the two periods , this situation of increased risk has already been described in other studies that evaluate the impact on mental health in periods of economic recession in various countries of the world [18] as well as in Europe [31,32] or in Spain [9,10,15]. Our data reflect much higher prevalence than that found in England [32], and the latest epidemiological studies in Spain hardly show any changes between periods [14,16]. They display only a small increase in the prevalence of ill-health in men, passing from 14.7% in 2006 to 16.9% in 2011-2012 and, in the opposite direction, a reduction in women from 24.6% in 2006 to 22.7% in 2011-2012 is also described [14]. These differences could be explained in part because of the sensitivity of the instrument to collect the disorders, age periods, and intervals analysed and variation of diagnostic categories, or also due to selection or recall bias. They could also be explained by the impact of the crisis on pre-existing regional inequalities, the dampening of the social apparatus, the differences in unemployment rates since the first years of the crisis, the public indebtedness and the type of economy in the region, etc. It should be noted that the VC was already among the regions with the highest prevalence of risk of poor mental health in Spain. This can be partly explained by characteristics such as socioeconomic level, level of education, unemployment and immigration rates, and development of mental health care resources [33]. Regarding sex, women generally had worse results; however, when taking into account the changes in the period, there is a greater increase in the prevalence in men. In the literature, it has been described that in periods of crisis or recession, men show a greater increase in risk compared to women [7,32,34] and partly explained due to the impact of unemployment as a risk factor of mental health worsening [14]. These differences may be attributed to the relation between work and social role of the man as main supporter with high family burdens [35]. Despite this, it is necessary to highlight the starting high prevalence in women, possibly explained by factors such as the status of women in society, their workload, lower economic resources, lack of autonomy, lack of social support, and in some cases, the violence, overburden, and stress that they experience and that contribute to their poor health [36]. Studies in Spain have shown worse mental health in older women, immigrants from a low-income country [33], from rural areas [37], with increase in family burden [38], with obesity [39], and who have experienced different types of intimate partner violence [40]. In our study, we have found among unemployed women the highest prevalence of poor mental health and the greatest increases over the expected prevalence. It is necessary to take into account this aspect, since in Spain a greater impact of the crisis has been described in women with higher rates of unemployment, part-time employment, and precarious and lower incomes, in comparison with men's similar working hours [41]. When analyzing the country of birth, our results coincide with what is described in other studies. These describe an increased risk of poor mental health among immigrants in the wake of the crisis [20][21][22]. This is partly due to the fact that they are a group with high social and economic vulnerability, as evidenced by the very high rates of unemployment among the foreign population in Spain and in the VC. Unemployment reach quotas of up to 41% among immigrants from non-EU countries [42]. To this, the persistence of occupational segregation should be added, as well as economic and labour instability, family burdens, discrimination, lack of social support, and precarious income, among others. Second, this study has demonstrated the importance of the contribution of the socioeconomic variables most affected by the onset of the economic crisis on changes in the prevalence of the risk of poor mental health between 2005 and 2010 in the population of the VC. Thus, the employment situation and the rent presented substantial changes from 2005 to 2010, increasing the number of unemployed and low income. This study presents unpublished results rarely evaluated in the current studies on this subject. Unlike other studies that have used "the pooled data" of the two health surveys analysed in their models to evaluate changes in mental health prevalence in two periods in Spain [14,16], we wanted to analyse the contribution of the effect of changes in the population structure on the excess prevalence of GHQ+. The analysis has taken into account the variables that have modified their distribution among the population in the second survey and coinciding with the changes in the economic and social situation of the VC . In this regard, this analysis has made it possible to quantify the contribution, adjusted by age and sex, of these variables to the increases in the prevalence of risk of poor mental health from 2005 to 2010 in men and women in 43.5% and 31.8%, respectively, of the total increase that occurred in the prevalence. However, the rest should be explained by other factors. This could confirm the sensitivity of mental health to the economic deterioration caused by the crisis. According to a WHO report, the crisis can deteriorate health through reductions in household financial security, especially as a result of job losses [8]. It has been described that unemployment is the determinant with the most stressing effect in life [43], which carries multiple health risks [7,12,19,34], and as the main factor in the appearance of problems such as anxiety, insomnia, depression, and dissociative and self-injurious behaviours that can cause the first mental health problems in a healthy person [34]. In addition, several studies have shown an increase of a 2-7-fold risk of suffering problems or symptoms associated with depression and anxiety about loss of work [7,19], as well as its detrimental effect over time [44]. One of the main consequences of the economic crisis in Spain has been the increase in unemployment. To date, Spain occupies the first position in relation to the unemployment rate vis-à-vis all EU countries since 2008 [45]. In the VC, the evolution of unemployment has shown a trend similar to the general Spanish one, although with rates higher than the national average [42,46], which shows a continued and chronic trend of unemployment both in Spain and in the VC. Although our data are in line with the results of other studies in Spain, which have confirmed that unemployed people have higher levels of depression than the employed [47] and that the increase in unemployment is an important risk factor that could be related to the increase in demand for primary care [9,15], our findings do not coincide with those found in England, which found that changes in the mental health of the population do not seem to be entirely mediated by changes in the unemployment rate or household income [32]. One possible explanation could be that in the VC, there were much more drastic changes in unemployment rates , a reduction in social and health services as part of the austerity policy to reduce debt, and a delay in government intervention strategies that made the impact more pronounced than in England. --- Strengths and Limitations One of the strengths of this study has been its sample size and the representativeness of the sample with respect to the general population. Furthermore, the study refers to the general population of over 15 years, covering a wide range of age. Another strength of the study is the use of data from the 2005 and 2010 ESCV survey, designed and validated to obtain population information on the variables studied and with little lack of response. This study does not include an institutionalised population, so there may be an underestimation of mental disorders, since a high prevalence of mental health problems in nursing homes and residences has been described [30]. On the other hand, cross-sectional studies cannot identify the direction of associations, for example, between mental health and work status or other variables. Regarding the instrument, it should be emphasised that the General Health Questionnaire is not suitable for assessing chronic disorders, but it does allow certain "mental health problems" [48] to be identified. Other studies also point out that since GHQ is a screening instrument and not a diagnostic tool, and more sensitive than specific, it may overestimate the existence of mental health problems [33]. Other variables described in the literature and related to poor mental health have not been taken into account in the analyses, which may also be explaining a greater risk due in part to the fact that the intention of the choice of variables was directed to those related with the socioeconomic scope of the current crisis. --- Conclusions The prevalence of poor mental health risk increased substantially in the VC from 2005 to 2010. Several variables were closely associated with this: Sex, age, country of birth, chronic diseases, disability, quality of life and social support, and employment status and income. Nevertheless, there was no interaction of any variable with sex. On the other hand, employment situation and rent presented substantial changes from 2005 to 2010, increasing the number of unemployed and the low income. Thus, the contribution, adjusted for age and sex, of these variables to increases in the prevalence of risk of poor mental health, from 2005 to 2010 can be quantified in men and women by 43.5% and 31.8%, respectively, of the total increase that occurred in the prevalence. The rest should be explained by other factors. As a consequence, this could confirm the sensitivity of mental health to the economic deteriorations caused by crises. --- Author Contributions: A.N. and N.T.-F. designed the protocol and led the project; C.B., N.T.-F., R.M., and M.Á.I. collected the necessary data; P.P.-Z. and J.M. performed the statistical analysis. All authors contributed to the interpretation of results; N.T.-F. and A.N. wrote the first draft, with all authors providing critical comments. All authors read and approved the final manuscript. --- Abbreviations --- VC Valencian Community ESCV Health Survey of Valencian Community

Previous research suggests that the economic crisis can affect mental health. The purpose of this study was to analyse the association of risk of poor mental health with various socioeconomic, demographic, health, quality of life, and social support variables; and to evaluate the contribution of socioeconomic variables most affected by the beginning of the economic crisis (employment situation and income) on the changes in the prevalence of the risk of poor mental health between 2005 and 2010. A study of prevalence evolution in adult population residents of the Valencian Community in the Spanish Mediterranean was conducted. We studied 5781 subjects in 2005 and 3479 in 2010. Logistic regression models have been adjusted to analyse the association between variables. A standardisation procedure was carried out to evaluate which part of the changes in overall prevalence could be attributed to variations in the population structure by age, sex, employment status, and income between the years under study. The prevalence of GHQ+ increased from 2005 to 2010, in both men and women. Several variables were closely associated with the risk of poor mental health (sex, age, country of birth, number of nonmental chronic diseases, social support, disability, cohabitation in couple, employment status, and income). The changes produced as a result of the onset of the economic crisis in income and unemployment (increase in low income and in unemployment rates) contributed to the increase of poor mental health risk. This could confirm the sensitivity of mental health to the economic deterioration caused by the crisis.

Introduction The COVID-19 pandemic has drastically affected individuals, families and communities worldwide [1,2]. Within Australia, community service organisations, alongside government initiatives, have played a notable role in responding to the social impacts of the pandemic such as increased unemployment, mental health issues and social isolation [3,4]. While working to meet increasing community needs, organisations were also required to rapidly restructure service arrangements in accordance with social distancing guidelines [3,5]. This led many to implement telehealth as a means of ongoing service delivery [3]. Telehealth is defined as the "delivery of health care services, where patients and providers are separated by distance" [6] . Wade [7] distinguishes between two forms of telehealth: asynchronous and synchronous; the former involves no 'real-time' interaction between provider and client , while the latter is characterised by simultaneous interactions . Although telehealth is commonly conceptualised in a medical context, its meaning has extended over time to include broader understandings of health and well-being [5]. Alternative phrases have emerged with this expansion; telepractice is regarded as more general term that encompasses the use of information and communications technology [ICT] to deliver services across a range of fields [8]. In this study, the terms telehealth and telepractice are used interchangeably to mean the use of synchronous communications, particularly teleconferencing and telephone, to deliver community services. Community services can be broadly defined as agencies that assist people who may be experiencing vulnerability by supporting their well-being and relationships, access to resources, and social participation [9]. Telepractice has been widely studied across multiple disciplines; over the decades, substantial literature has emerged commenting on its acceptability and efficacy [8,10]. A recent review of telehealth literature from the last decade found that although some studies reported telepractice outcomes to be equally, if not more, effective, the "strength of this evidence is unclear" [8] . Even so, it is widely agreed that telehealth is a promising mode of service delivery with distinct advantages [7,8,[11][12][13]. Commonly reported benefits for service users include improved access to services; enhanced quality, frequency or timeliness of services; reduced costs; and increased convenience [8, [11][12][13]. Telepractice has also been found to benefit practitioners, allowing greater "access to continuing education and professional development activities, [and] the ability to provide an enhanced local service" [12] . Along with advantages, the barriers and challenges associated with telehealth have been described within the literature. These have notably included clinician and provider preferences for in-person services, limited resources and funding, insufficient infrastructure, lack of reliable connection or broadband, and ethical and legal concerns [8,11]. Within the Australian context, research into telehealth holds particular importance for those residing in rural areas [11]. Existing studies have established that rural, regional and remote populations can benefit from using telepractice to access a range of medical and allied health services [12]. However, there are also distinct challenges for those living rurally, with Thomas et al. [14] reporting a considerable "Capital-Country Gap" in digital inclusion: "whether a person can Access, Afford and have the Digital Ability to connect and use online technologies effectively". Considering this subset of the population stands the most to gain from telepractice [15], ongoing research into the perspectives of stakeholders in regional areas is crucial. Despite the extensive literature published about telehealth, its actual implementation was still in its "infancy stage" prior to COVID-19 [16] . The pandemic and subsequent restrictions catalysed digitalisation and adaptation across numerous sectors, with telehealth operating at a "much larger scale than before" [17] . This markedly impacted the Australian community services landscape, with 96% of services shifting some aspect of their service delivery from face-to-face practice to other modes such as telepractice [3]. According to the recent Australian Community Sector Survey, almost a quarter of services "reported that their entire service shifted from face-to-face" [3] . This rapid transition was commented on by Sanders [5], who reported on the experiences of human service workers in regional Victoria; practitioners expressed both positive and negative aspects of shifting to telepractice during the pandemic. The workers commented on the benefits of increased accessibility, improved communication between agencies and professionals, and a greater work/life balance when using flexible modes of service delivery [5]. However, practitioners also had concerns about ICT issues, communication with service users, and the overall quality of services [5]. While Sanders [5] offers insight into the experiences of practitioners, the paper does not provide information about the service areas under comment. As the human service sector is diverse in terms of practice and client needs, research which explores the contextualised experiences of service workers could contribute to the ongoing discussion of telepractice in this field. In addition, there is a need for further qualitative research into clients' experiences of accessing community services through telepractice. Callis et al.'s. [1] recent 100 Families WA COVID-19 report found "there was substantial variation in people's experiences of services during COVID-19". Respondents' perceptions of changes to community services were divided relatively evenly between positive and negative; 46% of families found the shifts either more or wholly positive, while 54% indicated that they were more or wholly negative [1]. This is noteworthy as other fields and disciplines, such as medicine, have noted that service users, alongside providers, have widely embraced telehealth [18,19]. An in-depth exploration of individuals' experiences of using telepractice to access community services could provide insight into this disparity and the unique nature of telehealth in this context. To date, most of the research into telehealth has focused on medical, psychiatric, nursing, and allied health applications. Few studies have examined the use of telehealth by community service organisations, especially within the regional Australian context. Centacare Family Services-Geraldton [CFS], the setting for the current study, was uniquely positioned as a regional organisation which offers multiple community services including counselling, family and parenting support, family separation services, and emergency relief [20]. CFS is a Western Australian [WA] not-for-profit agency that services a large catchment area across the Midwest and Gascoyne regions. On the 16 March 2020, it became one of the first organisations in Geraldton to transition almost entirely to telepractice. While CFS, like many regional agencies, had some elements of telehealth established prior to COVID-19, this was limited in scope. Telephone was mainly used to work with clients residing outside of Greater Geraldton or the Midwest region; apart from this, policies, practices, and infrastructure related to telehealth were minimal. As such, many staff and clients had little prior experience using telepractice to deliver and access community services. Although the direct impacts of COVID-19 and the subsequent restrictions were relatively short-lived in WA, the service landscape has been markedly affected. Thus, an examination of the use and responses to telehealth within this context can add to the growing evidence base surrounding the effectiveness, complexities, and implications of telepractice. Considering this, the aim of this study was to investigate the service response to COVID-19, and how it impacted staff and clients' experiences of providing and accessing regional community services. As telepractice emerged as a key means of ongoing service delivery, this paper captures the response to two core objectives: to explore clients' experiences of using telehealth to access community services; and to understand staff's experiences of delivering community services via telehealth. The findings indicate the need for more tailored service delivery, specified training for staff, and organisational mechanisms for client collaboration and choice. --- Methods --- Research Design and Methodology This study used a qualitative approach with insights from an interpretivist paradigm, which recognises the "importance of people's experiences, meanings, understandings and interpretation of events" [21] . This approach suited the research which aimed to examine participants' experiences and perceptions in relation to the uptake and use of telehealth for community service provision. In-depth interviewing-a method which aligns closely with interpretivism [22]-was conducted to elicit thick descriptions and allow participants to articulate their experiences in their own words. Data were then inductively coded and thematically analysed, drawing on Brooks and King's [23] template style of analysis. Basic statistical analysis was also used to capture participants' demographic details. --- Sampling and Participants Participants were sourced using purposive maximum variation sampling, which sought to include service providers and clients from different service areas: counselling, family support, family separation, and emergency relief. This sampling framework, a "more flexible variant of stratified purposive sampling" [24] , was suited to the research as it gathered experiences from multiple subgroups and facilitated comparative analysis. Recruitment of service providers involved sending an email invitation which included a plain language description of this study and copies of the participant information and consent forms. Clients who met the inclusion criterion were informed of the research by the organisation's clinical and administrative staff. The inclusion criterion was any client aged ≥18 years who engaged with the organisation between March and May 2020; this was deliberately broad to elicit a wide range of client voices. With their consent, staff provided interested clients' contact details to L.A. We then called or emailed these clients to discuss this study further and provide copies of the participant information and consent forms. --- Data Collection Separate interview guides were constructed for the client and service provider subgroups, ensuring questions were relevant and appropriate for each. Interviews with service providers focused on their perceptions of how clinical practice and service delivery changed, any related benefits or challenges, client reactions, and their experience of the organisation's response. Interviews with clients centred around their experiences of accessing telehealth services, anything helpful or challenging about the changes to service delivery, and what they would like to see in relation to future service provision. Overall, 26 semi-structured in-depth interviews were conducted by L.A. over a fivemonth period to elicit detailed accounts of each participant's experience. Interviews were scheduled at the convenience of participants and were conducted by telephone, videoconferencing, or in-person, generally lasting approximately 30 min. All interviews were audio recorded for accuracy and transcribed verbatim; verbal consent was given prior to data collection and captured on tape. Participants were offered the opportunity to review their transcripts; two participants did so, with one providing comments and revisions. Demographic details were sought from participants. While the original intent had been to include this information, it was felt that certain details such as cultural identity would make participants and especially participating service providers readily identifiable within the organisation. Hence, to protect the privacy of participants, identifying information was omitted or changed in reporting and pseudonyms used to ensure client and service provider anonymity. The duration of data collection was guided by inductive thematic saturation-a form of saturation that requires ongoing critical reflection on the generation of new codes and themes from the data [25]. When used in this way, saturation is conceptualised as an incremental process rather than a fixed point [25]. We considered data collection sufficient when a rich network of themes was generated which identified commonalities in meanings and experiences, as well as complexities and differences depending on participants' service area. This use of saturation was appropriate to the study design as it aligns with its inductive methods [25]. --- Data Analysis Inductive coding and thematic analysis were used to manually organise, interpret, and generate themes from the data. This method was selected for its flexibility and alignment with this study's aims; Nowell et al. [26] note TA is especially useful for examining participants' perceptions and exploring the commonalities and differences across data. This fitted well with our research objectives and focus on multiple subgroups' experiences. We were guided by Template Analysis-a 'codebook' style of TA which offers a systematic yet flexible approach within a qualitative paradigm [27,28]. As per Brooks and King's [23] procedure for Template Analysis, our process began with data familiarisation. This involved reading and re-reading transcripts as well as re-listening to segments of the audio-recordings [29]. An inductive method was then used to code the data section by section, with initial codes ranging from a few words to several sentences in length. Coding techniques included manually underlining segments or ideas that were repeated or emphasised by the participant and noting codes in the margins. After preliminarily coding all transcripts, codes were compared and further developed to produce an initial template; this included broad overarching themes and successive themes, which were narrower and more focused. The refinement of the template occurred iteratively as it was applied to the data and modified accordingly. This involved reorganising existing themes, generating new themes, and removing those that lacked relevance. The template was reviewed and discussed by L.A., L.P., and S.T. at each stage of refinement to facilitate critical exploration of the themes generated and discussion of alternate interpretations. A final template focusing on participants' experiences of telehealth service delivery was agreed upon and applied to the data set. This served as the basis for analysis and was supported by excepts from the data. The quotations were identified and selected based on being richly detailed or elucidating the meaning or context of the themes. Basic descriptive statistical analysis-mean and standard deviation-were used to provide an overview of participant characteristics. --- Methodological Rigour Several strategies were employed to promote rigour; ongoing peer debriefings were utilised at all stages of the research to facilitate reflection on researcher subjectivity and bias. Thick and contextualised descriptions of participant's perceptions and experiences were also sought, ensuring sufficient depth was reached for analysis, promoting the credibility of the research [30]. Validity was also promoted through data triangulation [31]; multiple perspectives were gained from service providers and clients from different service areas which allowed for a more comprehensive understanding of experiences. Analysis was refined collaboratively through team meetings to further increase the soundness of the interpretive approach. --- Ethical Considerations The research project was approved by the University of Western Australian Human Research Ethics Committee, project number RA/4/20/6186. The voluntary nature of this study was made explicit to all participants. Both service providers and clients were assured that their decision to participate or decline participation would not result in any negative consequences such as decrease in support or disadvantage in employment. While the director of the organisation was an investigator for this study, she was appropriately distanced from data collection and did not have access to staff or client interviews or transcripts. --- Results --- Participant Characterisics Of the 29 staff at CFS, 15 expressed interest in this study, and all participated. The majority were female, and providers had an average age of 48 years = 15.1). Staff were from multiple service areas: counselling , family and parenting support , family separation services , and emergency relief [ER] . Two worked across more than one service area. At the time of interview, nine staff were working as practitioners delivering services directly to clients, three were in support roles that involved client contact, and three were not in client-facing roles. The length staff had been employed at CFS ranged from less than a year to over 25 years. Fifteen clients expressed interest in this study, with 11 consenting to be interviewed. Again, most were female , with an average age of 41 years . Two clients lived outside of the Midwest region; both were accessing family separation services and had ex-partners living in Greater Geraldton. One client did not access services directly but reported on the experiences of her children. Clients utilised a range of services: counselling , family and parenting support , and family separation services . The participant characteristics for both client and staff samples are summarised in Table 1. 20) 1 The ages of two client participants have been excluded due to imprecision . --- Staff Experiences of Using Telepractice to Deliver Community Services Four main themes were generated from staff experiences of delivering community services by telehealth: benefits and opportunities arising from telepractice; challenges and barriers encountered by providers; staff attitudes and the perceived value of faceto-face practice; and complexities and tensions within the community services context. --- Benefits and Opportunities Arising from Telepractice --- • Increased Accessibility for Clients Staff of most service areas discussed the value of telepractice in increasing the accessibility of community services for clients, especially those living outside town. This finding is consistent with previous research which outlines the benefits of telehealth for rural and remote Australians [12]. Several staff spoke highly of telehealth's potential to facilitate greater support for individuals and families residing on farms or in remote communities, and its usefulness even beyond the context of COVID-19. There are people out in Meekatharra and on stations that are a thousand kilometres out who can't possibly get in for counselling. [...] So now that we've got this system working so well, I can see us offering this to those external communities . This issue of accessibility cannot be understated within the Australian context, with WA having one of the lowest population densities globally [32]. Locality can profoundly impact an individual's ability to engage with services, and staff identified long travel distances and insufficient transport as barriers for rural clients. As one provider succinctly put it: "it might be telephone or nothing" . Staff believed other client groups, such as those with chronic conditions, mobility issues, or mental health conditions also benefitted from increased accessibility of services. . . . for a client who might not have enough energy to get up, get dressed and go to their appointment-now they can do it from their lounge, it's easier for them . Telehealth was also useful for families facing structural barriers such as lack of private transport "there are clients that don't have their own transport, who do rely on public transport [...] so, you know, getting to an appointment is harder for some people" . Those clients accessing emergency food relief and financial support services were identified as particularly benefiting from this mode of practice. Others benefitting were clients who otherwise struggled to access services due to their schedules, including shift workers and parents, with some requesting to continue telehealth beyond the mandatory lockdown period. --- • Increased Options for Practice Providers identified ways to extend and enhance service delivery through telehealth which was seen as an additional 'tool' which could be integrated into practice, providing both clients and practitioners with more options. One staff member involved in a parenting information and support group explained how telehealth could be used to connect with a greater range of speakers and experts "[...] we realised how easy Zoom was and that we can utilise guest speakers from interstate or in the city" . Given the limited number of services and professionals within regional and rural contexts, this was considered invaluable. The ability to connect with professionals, who would otherwise be inaccessible, was viewed as an opportunity to provide clients with both a broader range of information and specialised expertise. Staff involved in family dispute resolution [FDR]-a mediation service for separating families-found telepractice especially advantageous. One topic repeatedly raised was the way telehealth could be implemented to increase safety for participants and negotiate power disparities between those mediating. . . . running mediations by Zoom can be quite good when there is a risk of the parties becoming heightened or you know, one party's perhaps going to dominate or talk over the other because you can mute them immediately [...] you can turn off their videos if they start to become threatening or they look like they're intimidating the other party [...]. You can do that with a click of the button so-as a facilitator-that's a handy tool to have . Telehealth also offered FDR practitioners an alternative to traditional 'shuttle' mediation, a form of FDR where clients deemed unsuitable for face-to-face services are in separate rooms with the facilitator 'shuttling' between them. Telepractice was seen as an opportunity to increase the safety and effectiveness of mediation in cases of high conflict or risk. I found that [telepractice] is quite a good platform for some mediation, where it's not quite suitable for face-to-face [...] but better than doing shuttle. So I think it opened a lot of new opportunities to think that some mediations could be done by Zoom as a safety factor or a different opportunity rather than just face-to-face or shuttle . Providers delivering FDR identified distinct benefits of videoconferencing over phone calls reporting it allowed them to gain a better sense of clients' comfort levels and work more collaboratively using online functions such as screen sharing or writing on a digital whiteboard. You can pick up on a lot of that body language [...] you can sort of see if someone's getting agitated, if someone's not coping very well, you know. There's a lot of signals there. [...] we can share the whiteboard or the screen we're working on. And that makes a huge difference because the people can actually see what we're writing, they get to read it. Whereas on the phone, it's just verbal, and I think that makes it really quite more difficult . These features were used to enhance the service for clients who lived rurally or outside the Midwest. Where staff had only used phone in the past, use of videoconferencing during COVID-19 highlighted the potential for improving and extending service delivery by embracing new platforms. --- Challenges and Barriers Encountered by Providers --- • Technical barriers and disparities ICT issues were a significant barrier for some providers, impacting on the quality of their service and, in some cases, inhibiting them from working with clients altogether. [...] the biggest issue was poor internet. So that would sabotage the session [...] the picture would freeze, the picture would fade, they'd drop out, I'd miss some of the audio [...] I think 50% of my cases that I did visual audio had some sort of interference . This echoes findings from existing research, with technological issues as a barrier to the implementation and ongoing use of telepractice [33,34]. However, this was not the case for all, with some reporting little to no technical issues "ninety-nine times out of one hundred it worked perfectly well. Just straight on to the client, the client can see you easily . . . we're both side-by-side on the screen" . This disparity may reflect clients' broadband access, with providers explaining there was little they could do when clients had poor connection or reception "We could provide the service but if they [clients] had internet issues, if they had phone line issues, that was really out of our control" . Reliable internet is a known key determinant for access to telehealth services [16]. Thus, while telepractice can increase accessibility for some, others may be excluded from this mode of service delivery. --- • Clients' Digital Abilities and Hesitancy towards Telepractice Another challenge was clients' competence and confidence with using technology. Practitioners explained that digital ability, the skills required to effectively navigate and use technology, varied considerably among clients "[...] some people [clients] indicated quite clearly some distress about being competent to use the technology. Then you had others where it wasn't an issue, it was seamless" . Providers felt the transition to telehealth was most challenging for clients who had high levels of vulnerability or difficulties with ICT. "I know for some clients though who were really vulnerable it was really challenging, particularly if they had issues with technology or things, even just setting up the phone sessions could be a challenge [...]" . This aligns with existing literature which asserts basic ICT skills and literacy are requisites for people to make use of digital technologies of any kind, including telehealth [14]. To navigate these challenges, some providers offered practice sessions with the aim of increasing clients' confidence and skills with relevant technologies. [...] we talked them through how easy it was, you download the app, it's free, we can do dummy runs with you beforehand so you're comfortable with it when you get to the appointment stage . Some staff identified clients' attitudes towards telepractice as another challenge. Providers reported that some clients expressed hesitation or resistance to engaging in telehealth. This occurred predominately for those involved in counselling services; telepractice was considered a major obstacle and distressing for some clients. As a result, some counsellors noted decreases in their caseloads. I would say, out of all the clients I had, probably less than half chose to continue with online or telephone. Most people just chose not to engage during that time . Counselling staff also reported higher rates of cancellation during this period. As one provider commented: . . . there was a definite drop in clients [...] we found that a lot of people would actually rather cancel their appointments then do it by phone. Actually, I can remember a comment made by one gentleman who said, 'it is hard enough opening up to the counsellor sitting across from them, I can't imagine how I would open up on the phone, not seeing their face and everything' [...] I think that really stuck with me because I think that's how a lot of people felt . While such experiences were common for counselling staff, other service areas found clients relatively accepting of telehealth. In the area of emergency relief and food assistance, replacing in-person appointments with telepractice was considered a relief for some clients. "They [emergency relief clients] are just more comfortable, they're able to tell you maybe a little bit more than what they might tell you if they're in the office" . Similarly, clients engaging in a support group for new parents were described as enthusiastic for the online service. "I contacted them and asked them their thoughts on 'how do you think the group would be if we did it via Zoom?' They all responded 'yep, do it, do it, do it'. [...] So, I put it out there [...] and we were inundated" . This highlights that clients held widely varying attitudes towards telehealth which impacted their engagement with services. Client apprehension was mainly identified by those within mental health services and was a significant challenge for staff. --- Staff Attitudes and the Perceived Value of Face-to-Face Services While most staff recognised both the challenges and benefits of telehealth, attitudes and perceptions varied. Those who were less accepting of the changes focused more on limitations and viewed telehealth as an interim substitute or a substandard mode of practice. I believe that there is always going to be a place for face-to-face-that it will always take precedence [...]. When there's a choice, face-to-face is going to be the number one way to connect with clients. . Conversely, those more accepting of telehealth spent time discussing the benefits and possibilities for practice. Many factors shaped providers' attitudes: familiarity with telepractice and technology, perceptions of efficacy, and personal characteristics. For some, prior telehealth experience appeared to influence levels of comfort and acceptance. Those familiar with telepractice described 'slipping' back into it with less difficulty. Some suggested that working regionally and servicing a large catchment area supported their uptake of telehealth "It was an easy transition because, being a regional service, we have a lot of phone clients anyway [...] so it was a skill I already had" . However, this comfort level did not always translate to other modes of telehealth; some practitioners, despite extensive phone experience expressed apprehension towards videoconferencing, a less familiar platform. Formal training was identified as important to improving providers' comfort with telehealth. One staff member explained the limitations arising from the sudden transition. "I received absolutely no training in it [...] it's self-learning what we did, and yeah, I think that [training] could be a help" . Providers' uptake of telehealth was also impacted by their perceptions of its efficacy and usefulness, with those who saw benefits for clients discussing its continuation beyond COVID-19. They suggested ways telehealth could be used to improve processes and identified changes they would like to see implemented on a more permanent basis. [...] we were able to do phone registrations at that time and I really think it was a bit better for some of our clients [...] so I think the phone regos are definitely something I'd like to see as an option for people . Equally, where telehealth was considered less effective for clients, it influenced the provider's attitude, leading them to express greater ambivalence and reluctance. I can't really think of anything beneficial, you know, to come out of it [...] I think we need the body language to see what kind of situation and state and . . . yeah. It was just not really working . Moreover, personal characteristics such as openness to change appeared to inform some providers' attitudes. Some staff indicated being more comfortable with established modes of practice. "I was so excited to just go back to the way things were, but that's me. That's my personality-I'm a creature of habit, not being as open to adjustment" . Others seemed excited by the prospect of change "we need to make certain that we're maintaining technology [...] and we're willing to be experimental" . Although staff attitudes varied, a common thread was the perceived value of face-toface practice. Even those who embraced telehealth discussed the importance of working in person with clients, citing it as an intrinsic aspect of community service provision "there's nothing like a face-to-face interaction [...] you know just tracking all those thoughts and emotions and body language that you get from a client, it's just invaluable" . The importance of non-verbal cues was recurrently mentioned by providers who noted that face-to-face contact allowed them to better connect with clients and 'read' their emotions. Additionally, the importance of the practitioner-client relationship was reiterated. My main focus with clients is always the relationship and making sure the client feels comfortable and is in a position to trust what's going on-that's most important. That could work better face-to-face than remote. I think it'll take a lot more before the client can really allow themselves to become more vulnerable . The relationship between providers and clients holds particular importance in the community services sector, often forming the basis for service provision. As such, establishing trust and rapport with clients is vital. Staff's comments around this aligned with recent dialogue in the field which has considered the "intangible benefits of in-person, human contact" and the importance of maintaining quality relationships [35] . --- The Complexities and Tensions within the Community Services Context In line with existing literature, staff discussed additional complexities and tensions that arose when delivering services remotely through telehealth. A recurring concern was the privacy of clients and their ability to maintain confidentiality when discussing sensitive matters. This was emphasised by counselling staff. The biggest issue with people doing counselling at home is I can't see who is in the room [...] I rely on my client to tell me they're in a safe, private environment, but I can't know for sure. That's probably the worst-my biggest concern . Providers also discussed the added layer of complexity when working with clients who were particularly vulnerable, such as children. Practitioners described needing to have a heightened awareness of who else may be present during appointments. In terms of confidentiality, the other big issue was around kids. So, a lot of practitioners, I guess, just made sure they were aware that kids might have parents or carers listening in and really changed the way that they did their sessions . In addition to privacy, staff mentioned navigating issues surrounding client safety. As experiences of family and domestic violence [FDV] are not uncommon for clients accessing community services, providers needed to ensure telehealth would be safe for victim-survivors who were residing with perpetrators. "I was dealing with one-more than one client that was experiencing high family violence and I did feel it was especially pertinent to be more mindful of the risk people experiencing family violence would be at .. --- ." . To manage this risk, some necessary precautions were outlined: not messaging, calling or emailing the client without explicit permission; and only contacting the client when they were already in a position of safety . However, in some cases, telehealth was not suitable for people experiencing FDV or abuse. When discerning the appropriateness of telehealth, providers prioritised clients' safety and well-being, taking into account the person's circumstances. Obviously one of the big vulnerabilities and issues we were really concerned about was clients in DV situations, clients who, you know, we might have known had a partner in the home who was abusive or things like that, and not being able to have those phone sessions or things. I guess one way we sort of navigated that was by having the option to come in still [...] each practitioner just did the best that they could and offered what they could . --- Client Experiences of Using Telepractice to Access Community Services Four central themes were generated from client interviews to capture the experiences of individuals from different service areas: accessibility of telepractice; suitability of telepractice; communication and connection; and client perceptions and attitudes towards telepractice. --- Accessibility of Telepractice Notions of accessibility were recurrently discussed by clients who overall found telehealth a relatively accessible mode of practice. While staff identified technological challenges such as unstable internet as an impediment, all except one client described the technical aspects of accessing telehealth to be acceptable; in that case, the client described "a few hiccups with the initial phone calls" . Overall, service users suggested that telepractice was easy to use "It was simple. I just called prior to my appointment, made my payment. They would check me in, I think, and then [the provider] would give me a call in a few minutes" . Despite this, some reflected on the barriers that exist for others in their community. "[...] the technology's not going to work for lower socio-economic groups either. Internet connection and, I mean, I . . . a lot of people still don't have a webcam you know" . That said, some clients argued telehealth can increase accessibility for vulnerable individuals and families. The clients of a parenting support group in particular considered telepractice to be beneficial for geographically isolated women and suggested that an online platform may be more inclusive of those living rurally "especially for the poor ladies that are isolated and out of town. [...] All the ones stuck on farms and like, north or Northampton that might not have access to stuff like this" . These clients also discussed use of telepractice beyond COVID-19 to increase the accessibility of information for new parents. Concrete examples were given, such as recording guest speakers during sessions and uploading content to social media for a greater audience to access. --- Suitability of Telepractice While most clients reported being able to access telehealth, there was a range of responses regarding its suitability. Some felt telepractice was an appropriate fit, while others found that it was incongruous with their needs or personal circumstances. For those engaged with the parenting support group, the flexibility and informality of telehealth were positives. Clients of this service, who were all mothers with young children, articulated this. [...] it was quite nice, you know, like not having to dress up to go out of the house. You could go off and get yourself a cup of tea in your own home . . . so it was a lot easier to manage . Despite this benefit, some clients made the distinction between the online and inperson support groups "I think the real-life mother's group is for mum and baby, whereas the virtual one is very much for mum" . Speaking to this difference, one client explained that while online groups offer women a platform for connecting and seeking information, they are incapable of providing the same level of socialisation for infants. [...] my daughter, she's a very sort of . . . high needs baby and likes a lot of stimulation [...] So I'd preferentially sort of go for a walk or something with her. So even though I wanted to go, you know, to hear the speaker or to see the person, it didn't really work for me . Similarly, issues were identified regarding the suitability of online counselling for children. In one case, despite having the ability to engage in telepractice, the family decided to suspend the children's sessions "We ended up stopping the counselling. [...] Mainly because I knew the girls wouldn't sit still long enough to look at the screen or talk on the phone" . Other issues raised by counselling clients were having a suitable environment for sessions and privacy. Like staff, some clients felt they did not have an appropriate space to discuss sensitive matters that would be free of disruptions. Even though I . . . I sort of said to my family, 'Hey I'm, you know, talking to my counsellor at this time,' there were interruptions [...] So yeah, that was a downside of it as well: unforeseen interruptions" . Some clients also felt telehealth was incompatible with their individual needs. It was suggested that the personal circumstances that brought some clients to the service in turn shaped their experiences of telepractice. One client emphasised this when discussing their preference for face-to-face counselling. I felt like [phone counselling] still helped me, but I think the face-to-face helped me more. Yeah, purely because I don't see a lot of people, yeah, so maybe if I had more family and more support around me, it might not have been that bad to have a phone call, but because that's life for me [...] I guess I kind of look forward to seeing the counsellor to get things off my chest in person . Another client felt telehealth was unsuitable due to their experience of anxiety and trauma, which meant they felt distressed by phone calls "I hate phones . . . absolutely hate them [...] I start getting very panicky. I suffer from panic attacks and anxiety and with the phone I get very very very stressed out" . --- Communication and Connection Most clients discussed the impact telehealth had on their ability to connect with practitioners or, in the case of group programs, other service users. Despite recognising the benefits, many clients felt more comfortable with face-to-face interactions. There was a perception among clients that telepractice was less personal and some found it difficult to communicate over phone or videoconference. [...] not having the face-to-face really was the most difficult part [...] you lose that sort of emotion side of it and the genuine conversation with somebody and the focus. That's the sort of thing that means it's sort of harder for me to open up . This concept of 'opening up' was repeatedly mentioned, with clients suggesting that they were more hesitant to discuss their personal matters when accessing services via telehealth. "[...] it's a bit harder to open up [...] maybe you might not let everything out that's bothering you as much" . These experiences were expressed most by those accessing individual counselling or other forms of emotional support. For these clients, non-verbal cues were considered essential for effective communication. Many found conversing over the phone to be challenging. I felt it was more difficult to convey exactly how I was feeling about things, whereas when you're face-to-face with someone you can read their emotions and yeah, it felt like I wasn't totally getting across what I was trying to say . Clients also reported that the lack of body language and facial expressions resulted in feelings of isolation and disconnect "You can't-I guess you can't sort of see the expression on their face . . . Yeah. It's just . . . it just seems isolated" . Yet despite the value placed on visual cues, many clients chose phone instead of videoconference-some for practical reasons but others perceived phone conversations as more casual and informal, "the phone felt more like . . . more of a normal conversation . . . whereas video felt like you were, I don't know, like it was an interview" . Having an established bond facilitated effective communication even in the absence of non-verbal cues: I guess because I had been seeing my counsellor for so long, they know me quite well and what my voice sounds like and the way my head works [laughing] basically . . . so I still felt like, because we had such a long relationship prior to that, going on the phone was fine because I was sure they'd be able to read my voice and what I was saying anyway" . However, others found communicating by phone challenging, even when they had an existing rapport with providers "[...] to me it's no reflection on my counsellor because I reckon I've got the best counsellor in-if not the world, it'd be Australia [...] It just keeps going back to I don't like phones" . The lack of body language and physical presence resulted in feelings of anxiety with clients describing feeling more relaxed and at ease when speaking to providers in person. Conversely, phone interactions seemed to carry an innate pressure to continue the conversation. When something was particularly emotional [...] it's easier to take a break when you're talking to the person than when you're on the phone. [...] You feel like you gotta-you can't have silences over the phone [...] like I felt like was under pressure to keep talking and not have that break . Previous literature has found that clients who perceive telehealth as less effective than face-to-face practice often cite communication as their primary reason [36]. While all client participants, except for one, had engaged with telehealth, the preference for face-to-face practice was clear. Clients from the parenting group program maintained a similar view. Despite recognising the benefits and suitability of telehealth, they discussed the value of connecting face to face. I don't know if [the group] could be a webcam thing [...] I think fundamentally the mothers' groups are about connecting with people and yeah . . . the online meetings were basically trying to make the best out of an odd situation . Limitations of virtual focus groups were noted, with clients wanting to have more personal interactions with peers. One of the downsides is that it's less personal [...] so you couldn't-it's not like in a group, where you've got like seventeen faces there, you can have a personal conversation [...] you couldn't sort of just have your one-on-one time which you might have done in a face-to-face setting . These clients also suggested that telehealth limited their ability to dialogue fully with the facilitator and guest speakers. For some, the chat function of the videoconferencing platform made communication more difficult. [...] you probably couldn't, like you couldn't ask the questions as well as what you wanted to because normally you can just interrupt and ask the questions, but it was a bit-a lot harder when you had to stop and type . --- Client Perceptions and Attitudes towards Telehealth Clients commonly viewed telehealth as being a temporary substitute for face-to-face services, rather than a distinct mode of practice. While a few clients discussed the benefits and uses of telepractice beyond COVID-19, many regarded it as secondary to traditional in-person services. A recurrent thread across clients' accounts was that telehealth was a way to 'make the best out of a bad situation'. When asked about whether they would like to see aspects of telehealth integrated into service delivery permanently, many reported they would only use it if they were unable to attend face to face. I guess . . . well, no. Not really, because, you know, the situation we were in was-they were making the best out of a situation [...] I guess [it could be used for] all that sort of stuff where obviously people can't make it or say their kids are sick or they're sick or something like that . In line with some staff attitudes, when given the choice, clients generally felt face-toface practice would be preferrable. This attitude was most strongly held by those accessing counselling services; in other service areas, views were more varied. Within the parenting support group, there were both those who believed in the primacy of face-to-face practice and those who saw the potential for telehealth to enhance services. Some suggested using telepractice to record information sessions to enable access for women living rurally. Clients who were engaged with FDR services were, for the most part, neutral towards telehealth. These individuals reported having more experience with telepractice prior to COVID-19 as either they or the person they were mediating with were located outside Geraldton. For these clients, the biggest shift was the increased use of videoconferencing, with previous telehealth being limited to phone. This change was viewed positively, some suggesting that it should be adopted into regular service delivery. They should keep Zoom. It works and it's good and it's fairly broadly used these days [...] the phone works alright. It's a little bit . . . Zoom tends to be better in the sense it makes it a little more . . . a little bit more welcoming, I suppose, or a little bit less clinical . --- Discussion This in-depth exploration in a regional setting was undertaken within a single community service agency. It considers the effects of adaptations implemented to enable the continuation of multiple programs during the rapid imposition of changes associated with COVID-19. Qualitative examination of organisational and client experiences of rapidly adapting to telehealth delivery shows that responses to, and perceptions of, telehealth varied substantially amongst both service providers and users. Many themes and subthemes identified are broadly consistent with existing literature. The benefits of telepractice, especially for regional and rural residents [8,12], the challenges associated with telehealth, such as technological barriers [5,8], and concerns around confidentiality, privacy, and safety [37] have been discussed in past studies. The current findings build on previous reports and point to the need for greater consideration of ways in which telepractice can be tailored to the preferences and requirements of different client groups. A discussion of the principal findings, implications for practice and research, and limitations of this study follows. --- . Variation in Experiences Experiences and perceptions of telehealth varied considerably, with some participants responding positively to the introduction of telepractice and others expressing concerns regarding its effectiveness within the community service space. This was common across staff and client subgroups with an array of attitudes identified, including enthusiasm for telepractice, neutrality, and a strong preference for traditional in-person services. Participants' experiences seemed, in part, to be shaped by previous experiences with telehealth and levels of comfort and ability using technology. This is congruent with recent research, such as Békés' and Aafjes-van Doorn's [38] study of psychotherapists' engagement with telehealth which found past experiences with online modalities were associated with positive attitudes towards telepractice. In the current study, staff members' attitudes also appeared to be shaped by their perception of whether telehealth was beneficial to clients, whether they encountered technological challenges, and personal characteristics such as openness to change. Experiences of telehealth were differentiated between service areas. Those involved in the parenting support group acknowledged both the benefits and challenges of online delivery. Face-to-face services provided clients with a stronger sense of connection, a greater ability to interact with one another and the facilitator, and more value for their child. However, the online platform extended the reach of the program which could enable geographically isolated women to access the service. The current study yielded some suggestions from clients about the structure of online support groups, such as decreasing the duration of sessions to account for the increased tiredness arising from videoconferencing [39] and decreasing the number of participants to allow for greater interaction. The dearth of studies about the use of telehealth to facilitate support groups for new and expecting parents indicates that further research is needed in this area. Regarding counselling, staff and clients for the most part indicated a preference for face-to-face service delivery, describing it as more effective and valuable for the therapeutic process. Clients from this area emphasised the importance of non-verbal communication, physical presence, and the personal nature of face-to-face services. Clients also suggested that they were more conscious of what they discussed with their counsellor when using telehealth and felt restricted in how open they could be. These results reflect those of Callis et al. [1], who found that over 60% of WA families who accessed mental health services during COVID-19 reported that services met their needs 'less' or 'much less' than before the pandemic. Callis et al. [1] compared this to other service types, such as food and financial services; 56.7% of WA families accessing food support reported that the services "met their needs just as well or more" than prior to COVID-19. The present study supports this finding; while clients from emergency relief [ER] were unable to be recruited as participants, staff involved in providing food assistance described clients' responses to telehealth as broadly positive and that clients were more comfortable and open when accessing ER through telehealth. There are several potential explanations for the notable variation in clients' experiences between service areas. In the case of ER, previous studies have found that individuals who access these services often report feeling discomfort and embarrassment when seeking support [40]. It is possible that the higher degree of anonymity afforded by telephone appointments offered clients a greater sense of comfort and encouraged self-disclosure. Literature also notes that some groups are more likely to access ER, including those from low-income households, people with disabilities, sole-parent families, and people receiving pension [41]. Individuals from these groups may experience difficulties accessing traditional services due the cost of fuel or transport, mobility issues, or caring responsibilities. Thus, it is possible that the transition to phone appointments was received positively by ER clients as it increased the convenience of accessing the service. Further research exploring the perspectives of ER clients should be undertaken to elicit a clearer understanding of the perceived effectiveness of telehealth in this area. Previous research has demonstrated an association between "a history of both depression and anxiety" and poorer telehealth experiences [36] . This correlation may in part explain the current findings which indicate that clients who accessed mental health support perceived telehealth as less effective than in-person visits. The reasons clients were accessing counselling appeared to influence their experiences and perceptions of telehealth. This points to the need for discussions about how providers and organisations can tailor service delivery to the needs of individuals. This echoes Bryant et al.'s [42] assertion that dialogue regarding telehealth must transcend the practicability of technology and consider how ICT can be utilised in a way that is "practice-based" and focused on the end service user. It is clear from the findings that there is no 'one size fits all approach' for using telepractice to deliver community services. Individuals' experiences varied widely, both within and across service areas. To cater for the multiplicity of needs and circumstances, providers must consider several key factors on a case-by-case basis. Existing literature [43] describes the importance of the following factors: clinical appropriateness, quality , safety , and practicalities . McClean and Fowler [44] add to this the significance of the service relationship, staff characteristics , and the characteristics of clients and their circumstances. The current study supports the pertinence of these elements, while also suggesting the importance of an additional factor: clients' self-assessment of service modalities. Clients articulated which modes of practice were appropriate for their needs, preferences, and specific context. This demonstrates that service users are an important source of knowledge for determining the suitability of telehealth for them and speaks to the need for collaborative processes which view clients as partners in service delivery rather than passive recipients. This study accords with others [45] in asserting that clients' views and preferences cannot be assumed, and that ongoing review of telehealth is needed to accommodate for service-users' changing situations and needs. --- Centrality of Communication and Relationship The importance of effective communication and relationships was raised across service areas by both staff and client subgroups. For some, the lack of certain cues, such as body language or facial expressions, greatly impacted their experiences. This was especially true for clients engaged in counselling or other emotional support services. For these respondents, there was a sense that telehealth services were less personal than face-to-face interactions. These results are consistent with Callis et al. [1] , who found WA families accessing community services during COVID-19 cited "the loss of face-to-face contact" as significantly negative. Moreover, the current study found that perceptions of communication and experiences of telehealth appeared to be related; if participants found communication to be effective, they viewed telehealth more positively whereas participants who described communication as more difficult or less personal spoke about telehealth more negatively. This finding is in accord with that of Isautier et al. [36] , who found that clients who reported telehealth was worse than in-person services recurrently cited "communication is not as effective as face-to-face visits" as the main reason. This is also consistent with McClean and Fowler's [44] rapid review of telepractice in human services which found telehealth "can affect the development and maintenance of personal trust and the formation of a therapeutic alliance [...] because technology generally limits non-verbal communication". Therefore, ways of communicating and engaging through telehealth to ensure effective working relationships must be a central consideration. The current study suggests the importance of relational elements when providing community services through telehealth. It is possible that further training focused specifically on remote engagement and communication could assist practitioners to connect with clients. This supports van Galen et al.'s [46] claim that "telehealth requires expansion of physicians' communication competencies training". While van Galen et al.'s [46] work is situated within a medical context, their assertion that telehealth requires different communication competencies and thus additional training appears to resonate with community services context. In particular, the current study points to the need for supporting clients to speak openly, ask questions, and feel comfortable during moments of silence. Building on recommendations made by Sansom-Daly and Bradford [45] and Gordon et al. [47], it appears important to address expectations of telehealth with clients prior to the first session, explicitly naming and exploring some of the above experiences . While some providers described using test sessions or 'dummy runs' to familiarise clients with the technology, practitioners should also utilise this opportunity to discuss any concerns the client may have about telehealth, explore their individual needs and circumstances, and create a plan for if they encounter technical difficulties during the session. --- Issues of Accessibility and Equity Both clients and staff noted the potential of telepractice to expand the reach of services as well as deliver support in a more flexible and accessible way. The benefits of this for vulnerable groups, such as those living rurally, are well documented . However, the rapid digitalisation catalysed by COVID-19 has also "highlighted the broad implications of the digital divide" [48] . Disparities in digital literacy, access to reliable internet, and ownership of devices or hardware exclude some people from accessing telehealth services [48]. This is of particular concern for regional and rural residents, who tend to fall into lower-income brackets [17] and thus may experience difficulties affording the technology or data needed to support online telehealth sessions [16,49]. While no client participants discussed these issues, such disparities appeared to have impacted on service delivery for providers. Staff described encountering numerous challenges due to clients' limited internet broadband or poor reception. The implications of this cannot be understated within the context of community service provision, where clients may be experiencing multiple vulnerabilities preventing them from accessing high-speed internet or the required equipment and facilities to enable their engagement in telehealth and without disruption. The Australian Council of Social Service articulated the significance of this issue in a recent report, suggesting that "digital access and the digital divide have been at the forefront of concerns for the community sector" [49] . With anticipations that telehealth will be delivered at higher rates in the foreseeable future [14], it will be critical for community service organisations and providers to carefully consider the digital divide and issues of equity moving forward. This is also an important issue for future research; additional studies exploring clients' experiences are needed to develop of full picture of how technological equity can be supported within the community service space. In addition, partnering with Consumer Advisory Committees may offer valuable insights about specific community and place-based issues, allowing organisations to examine telehealth through an equity lens and facilitate greater accessibility. --- Telepractice beyond COVID-19 Two divergent discourses emerged regarding the future role of telehealth within the community service sector. Among participants, there were some that viewed telepractice as an opportunity to enhance services and increase options for clients and practitioners. Others, however, expressed relief that service delivery could revert to traditional face-toface modalities, indicating that they would sparsely utilise telehealth in the future. These opposing perspectives are evident in recent literature, with Sanders [5] reporting that regional human service workers were divided when discussing the utility of telehealth. The current study reflects this, with some providers confident that an increased use of telehealth would benefit clients. Staff and clients involved in FDR discussed the value of integrating videoconferencing into service provision beyond COVID-19, conceptualising it as an alternative to phone mediations for those living rurally to offer the added benefit of visuals. It was also seen as a tool for increasing client safety when mediating parties have high levels of conflict. While some participants were open to integrating telepractice into routine service delivery, others expressed ambivalence, describing it as a modality which is 'better than nothing'; that is, that telehealth would be a suitable option only when face-to-face practice is unavailable or if extenuating circumstances necessitated it. This view of telehealth has been observed across numerous professions, with Slavova-Azmanova et al. [50] suggesting that there is a "dichotomy of acceptance for telehealth but preference for face-to-face", which is particularly evident in rural settings. This is shown in the present findings, with some clients and staff expressing a strong preference for in-person services despite their recognition of telehealth's benefits. Overall, the current study supports the assertion made by Sansom-Daly and Bradford [45] that "each modality [telehealth and face-to-face] shines and falters in unique ways", and that a considered integration of both may be beneficial to cater for the diversity of clients' needs. To accommodate for the range of experiences and perceptions of telehealth, consumer choice should be prioritised. The current study was undertaken in the context of telehealth being rapidly and 'forcibly' implemented; as transition to a state of 'COVID-normal' occurs, it will be essential for organisations to re-examine telehealth systems using a client-centred lens. Additionally, to the best of the authors' knowledge, there are yet to be guidelines developed regarding telepractice in the context of community services. The creation of such a resource may be beneficial for providers in delivering telepractice beyond COVID-19. --- Limitations of This Study The findings of this study should be considered in the context of several limitations. Firstly, the research focused on the experiences of staff and clients from a single community service organisation within regional Australia. While this enabled an in-depth exploration and comparison of multiple service areas, the results may not be transferrable to other populations. In addition, the forms of telehealth under inquiry were primarily videoconferencing and telephone; thus the findings may not be transferrable to other systems of telehealth or uses of ICT. It should also be noted that most interviews were conducted by telephone due to social distancing measures being in place at the onset of this study. This may have impacted the responses of participants and unintentionally excluded individuals who lacked adequate reception or preferred face-to-face interactions. Additionally, the current study was not designed to examine how demographic characteristics, such as age, gender, and cultural background, impacted staff and clients' experiences. While the findings suggest that familiarity with telehealth and some aspects of personality, such as openness to change, may affect perceptions and experiences, further research would be needed to examine demographic factors. Finally, like most qualitative studies, the numbers interviewed were relatively small and caution should be used when considering the transferability of the results. That said, conceptual depth was achieved, with thick description elicited from individual interviews and the triangulation of client and provider experiences contributing in-depth knowledge to the growing evidence base surrounding telehealth. --- Conclusions This study explored staff and clients' experiences of using telehealth in a regional setting to provide and access community services during the COVID-19 pandemic. There was substantial variation in participants' experiences and perceptions of telehealth. Staff found the rapid digitalisation to be beneficial in some respects, such as increasing the accessibility of services for some clients and providing more options for practice. However, this was accompanied by an acknowledgement of challenges and complexities, such as safety concerns related to family violence and confidentiality, issues of equity for clients who lack ICT or reliable broadband and navigating client hesitance. For both staff and clients, emphasis was placed on the importance of relationships and communication. Clients, for the most part, found telehealth to be accessible; however, they raised issues regarding the suitability of telepractice for their needs and circumstances. Based on these results, several suggestions were made for providers, organisations, and researchers to move towards a greater individualisation of telepractice which caters for clients' diverse contexts and multiplicity of needs. --- Data Availability Statement: Data sharing not applicable. --- Acknowledgments: The authors wish to acknowledge the collaboration of Centacare Family Services and the support of staff. Furthermore, we extend our thanks to all participants for sharing their knowledge and experiences with the research team. --- Author Contributions: Conceptualisation, L.A., L.P. and P.S.; methodology, L.A., L.P., P.S. and S.C.T.; data collection and formal analysis, L.A.; data curation, L.A.; writing-original draft preparation, L.A.; writing-review and editing, L.A., L.P., P.S. and S.C.T.; supervision, L.P., P.S. and S.C.T.; project administration, L.A., L.P. and P.S. All authors have read and agreed to the published version of the manuscript. Conflicts of Interest: P.S. is the Director of the service agency and L.A. has undertaken some parttime work for the agency remotely. L.P. and S.C.T. declare no conflict of interest.

Community services have played a significant role in supporting the psychosocial health and well-being of vulnerable populations during the SARS-CoV-2 (COVID-19) pandemic. To meet increased community needs, organisations were required to rapidly modify service provision, often using remote delivery systems. This in-depth study, undertaken early in the pandemic, explored staff and clients' experiences of adapting to using telehealth to provide and access services in one regional social services agency. Semi-structured interviews from 15 staff and 11 clients from a regional not-for-profit agency in Western Australia were recorded and transcribed. Inductive coding, and thematic analysis identified eight subthemes, with experiences and perceptions of telehealth varying substantially among staff and client groups. Distinct benefits and challenges were associated with telehealth. Participants highlighted tensions and complexities and commented on the place of telehealth in the community service sector. Clients expressed the importance of relationships and communication. This study provides in-depth insights into the contextualised experiences of staff and clients during a time at which change was both enabled and necessary. The findings highlight the need for tailored service delivery; choice; client collaboration; ongoing staff training relating to telehealth; and guidelines specific to telehealth in the community service sector.

Background Conflicts and crises increasingly affect middle-income countries that have already experienced an epidemiological shift from infectious disease to Non-Communicable Diseases [1]. As physical injuries, forced displacement, degradation of living conditions and interruption of care caused by humanitarian crisis situations can lead to the severe deterioration of chronic disease, NCD care should be included in both emergency and protracted crisis contexts [2][3][4]. Health actors are increasingly aware that health care for refugees needs to focus on integrating care and health support beyond acute emergencies, as well as on primary care [5]. Integration of care is particularly challenging when refugees live in non-camp settings and access care through multiple providers, as is the case in Lebanon. In 2018, Lebanon hosted approximately 1.5 million Syrian refugees; one in four of its residents is a refugee [6] and most live in non-camp settings. Approximately 53 % of Syrian refugee households in Lebanon report having at least one member with a chronic illness [7]. NCDs, including diabetes, are highly prevalent among the Syrian refugee population [8,9]. Diabetes is a chronic disease that needs continuous, lifelong treatment, similar to other NCDs such as hypertension, or communicable diseases like HIV. If not followed up properly, kidney, eye, heart and nerve complications can lead to premature morbidity [10]. In humanitarian settings, additional morbidity and disability of diabetes could be avoided through adequate care provision [11]. The Lebanese health system is highly fragmented, with multiple providers including the Ministry of Public Health, Ministry of Social Affairs, political parties, nongovernmental organizations and faith-based organizations. It remains largely dominated by the private sector, which offers 85 % of hospital beds in the country, and most specialized care [12]. Health access for Syrian refugees in Lebanon went through several organizational and administrative changes since the onset of the Syrian Civil War in 2011. Before 2015, the UNHCR and NGOs acted separately from the national health system. Thereafter, with funding support of UNHCR, health care for Syrian refugees was integrated into the Lebanese national assistance system, engaging the Ministry of Public Health and providing Primary Health Care , including NCD screening and care, through an existing network of Primary Health Care Centers , most of them run by national NGOs [13,14]. The PHCCs charge a fee of approximately 10 USD per visit and distribute some essential drugs and vaccines free of charge for the patient. Not all types of diabetes medication are available [13]. Regarding secondary and tertiary care, UNHCR partially covers the cost of life-saving treatments in hospitals, but the remainder needs to be paid directly by the patient [12]. Outside the official health system, Syrian doctors and health care workers often provide less expensive services [15] and some pharmacies sell medicine for diabetes or antibiotics without prescription [16]. However, accessing diabetes care for most Syrian refugees requires out of pocket payments, particularly for hospitalization, medication and diagnostics [7,12,13]. Faith-based organizations and national and international NGOs including Médecins Sans Frontières strive to fill this gap and operate or support centers offering PHC free of charge or at low cost [17][18][19]. In this complex landscape of health care providers, Syrian refugees face challenges in accessing PHC, including care for NCDs [20][21][22][23]. While there are existing studies on patterns of health seeking behavior amongst Syrian refugees with chronic diseases in Lebanon, much of this research is quantitative or focused on health expenditure [7,20,[24][25][26][27]. Less is known about the personal experiences of refugees in accessing NCD care and their perspectives on choosing providers. MSF operates clinics in different regions of Lebanon, where it provides NCD care for vulnerable residents [17,28]. To inform improvements to the implementation of the NCD program, we conducted a qualitative study among MSF patients. We explored how Syrian refugees seek health care for diabetes in the multi-provider setting of Lebanon by examining determinants and pathways of health seeking behavior. --- Methods --- Study design and sampling This was a qualitative study, which was conducted in two sites in Lebanon in February and March 2019. Study participants included male and female Syrian refugees living with Diabetes Mellitus Type 1 or Diabetes Mellitus Type 2 who were receiving care from MSF clinics in Wadi Khaled or Shatila. We included adult refugees aged 18 years or above who resided in Wadi Khaled or the greater Beirut area and consented to be audiorecorded during the interview. Convenience sampling was used to recruit participants from the two clinics. We aimed at including a balance of male/female patients and those with different types of diabetes from both sites, as well as a varied age range. Patients present at the clinics during the study period were approached by a translator in the waiting room, informed about the study and invited to participate in an interview. --- Study setting and sites MSF offers basic NCD care free of charge at its clinics in rural Wadi Khaled, Akkar and in urban Shatila, Beirut. These services include diagnostic tests, clinical care, provision of medication, referral costs for emergency care, patient support, education and counseling services and social services [17]. In Wadi Khaled, there are nine PHCCs, four of which offer NCD services, as well as several private doctors and pharmacies. In the more urban area of Greater Beirut, a range of health-care options exist. Wadi Khaled is located in the north-east of Lebanon, on the border with Syria. Wadi Khaled has a registered Lebanese population of around 40,000 inhabitants, and approximately 25,000 registered refugees [29]. Wadi Khaled is among the poorest and most under-served regions of Lebanon. The Lebanese Armed Forces installed checkpoints upon entry to and exit from Wadi Khaled and refugees are often unable to cross. In June 2018, the active NCD cohort at the MSF clinic in Wadi Khaled included 31 patients with T1DM and 222 patients with T2DM. 3Shatila is originally a Palestinian refugee camp located in the capital city, Beirut. According to UNHCR estimates, Beirut and the surrounding Mount Lebanon area now host around 240,000 registered and 90,000 unregistered Syrian refugees, with the highest clusters in South Beirut [30]. Most Syrian NCD patients visiting the MSF clinic come from outside Shatila, largely from the Greater Beirut area. In December 2017, the active NCD cohort of Syrian patients at the MSF clinic in Shatila included 204 patients with T1DM and 1650 patients with T2DM. --- Data collection Interviews were conducted at the MSF clinic in each site or at the home of the interviewee, depending on the preference of the individual. The mean duration of interviews was 50 min. In-depth interviews were conducted by the female principal investigator using a piloted semi-structured interview guide with open-ended questions that explored the healthcare seeking experience of participants since their arrival in Lebanon, including information sources for health and health care providers, health care seeking pathways, and factors influencing choice of pathway . Interviews were stopped when saturation was reached. The PI was supported by trained translators who were fluent in Arabic and English. Neither the translator nor the PI were involved in direct patient care. Patients were reassured that their participation was voluntary, and that it was not linked to the care they were receiving from MSF. Several patients who were approached in the waiting room refused participation. If patients agreed, they were invited to the interview room where potential harms, voluntariness and confidentiality were discussed during the informed consent process. Two potential participants refused participation after the informed consent process. --- Data analysis All audio-recorded IDIs were transcribed and translated from Arabic to English by an external transcriber, reviewed by the PI then discussed with the translators and transcriber for verification. A coding framework was developed in NVivo , based upon content analysis of the data. Each code referred to an underlying theme, and themes were grouped into larger concepts that emerged from the data analysis. The PI coded the transcripts which were then reviewed by another co-investigator and discussed with the rest of the study team. In order to answer questions of factors influencing the choice of provider and health care seeking pathways among Syrian refugees in Lebanon with diabetes, we chose content analysis to allow themes to emerge from the transcribed data. Following the inductive coding of answers and the identification of themes, the analysis was then informed by theoretical approaches to factors influencing health seeking behavior to further interpret and categorize the answers. --- Results In total, 29 patients participated in the study and of those, 13 were female and 16 were male . Participants ranged in age from 19 to 80 years, with a median age of 50 years. The median age of the 11 participants with T1DM was 29 years, and for the 18 participants with T2DM, 57 years. The main themes emerging from the data were: 1. Awareness and understanding of diabetes treatment and services; 2. Factors influencing choice of provider; and 3. Different health care seeking pathways. --- Awareness and understanding of diabetes treatment and services Subjective knowledge of diabetes disease and care Participants described how their own and their family's knowledge about diabetes influenced which type of care provider they chose, and how they follow up and manage their disease. In general, participants had a good understanding of the bio-medical concepts of diabetes and the lifestyle factors influencing its progression, though some described a lack of knowledge, and how this then influenced the way they sought follow-up and care. Those unaware of the importance of regular monitoring reported delaying seeking care until experiencing acute symptoms or a worsening of their condition. "[My diabetes] wasn't stable when I was young. I told you, I didn't know about eating rice and white bread. All of this increased the diabetes levels. I didn't measure my glucose levels all the time. In Syria, I used to go to the PHC once every two to three years. Now, here with MSF, I am following up more. [In Syria], maybe it was a little bit of neglect from my parents and I was at school, I don't know. I did not know a lot about the disease. I did not know what it was, what it meant, and its complications. I used to know that I had to take the medication, eat normally and cut out sweets only. I did not know that carbohydrates increased the glucose level. I did not know any of this information. We did not follow up on the disease and we did not know that it had these complications. When we began to go and follow up with the doctors [at MSF], I had a better knowledge about diabetes." . --- Awareness of available services for diabetes care Participants in both Wadi Khaled and Shatila had differing degrees of knowledge of care providers. They described how their choice of care-provider was influenced by knowledge gained through their family and peer networks. Most interviewees learned about the existence of different health care providers from family members, Lebanese and Syrian neighbors and friends. They described how people who were already seeking care from a particular provider shared this information and made recommendations to them. "I was at my friend's house and there was a guy who came from Syria who has diabetes. So, my friend said that he had diabetes and it is improving; he also told us that he is going to MSF and they are giving him medication and he got better." . "When someone is sick, they would always ask people around them." . --- Awareness of available services for diabetes-related care Interviewees used the same channels of information that they used when seeking general diabetes care to find a specialist such as an ophthalmologist for diabetes-related complications or a gynecologist specialized in the pregnancy of women with diabetes. Three participants mentioned learning about providers through adverts on social media or posters, and home visits by MSF community outreach staff, while several found out about a new provider "by chance." "Concerning diabetes, I'm here [at the MSF clinic] and I would never change, but concerning other things, such as eye tests, we usually go to [a PHCC] on the old airport road, it costs very little. [I chose it] because I don't know anywhere else. I also know a girl who works there, she is Syrian and she told me it is very good." --- Factors influencing choice of provider Several factors influenced the patient's choice of health care provider for diabetes: quality, distance, economic factors and their legal status. --- Perceived quality Perceptions of the quality of care were important considerations when choosing a health-care provider. Perceptions relating to quality were shaped through people's own experiences as well as recommendations of specific doctors, tests, facilities and medication made by members of their social networks. "I care about the doctor being good medically, as well as people talking about his achievements… I care about the doctor's experience and his good reputation among people." . "There was a woman here that had diabetes, and she told us that [the doctor's] service is really good and that he takes good care of the patients… and he does a full checkup. To choose a doctor, I ask around. They might tell you that X and Y went and it was good. If it was bad, I don't try it. I try the one that they say is good. I did not really like [this PHCC]. Their treatment is not humanitarian. They don't treat you nicely. You feel that they barely give you the service, from the way of talking and treatment. I am talking about the employees, not the doctors. When [another PHCC] opened, I stopped going there." . Relevant considerations for assessing quality were the improvement of health status , communication and attitude of health care staff and the uninterrupted availability of medications. "It is not about choosing an option. I go try the doctor, if I see that I am getting results, I would carry on going to him." . "The only doctor that told me that diabetes doesn't go away is the one from [a city] in Syria. His treatment was really good, but the only negative thing was the insulin in Syria. The electricity was really weak and the insulin was not in the fridge. That's the only reason I came here [to the MSF clinic], it's better." "We used to go to [a PHCC], and every once in a while they gave us the medication. They stopped giving us the medication, so I stopped going. I want to be honest. I went the first time and they gave me the medication, the next time they gave me the medication was four months later." . Providers from the private sector were perceived to have the highest quality of care, especially those that were expensive, as well as MSF clinics. "He told me that [MSF] provided really good care and an amazing service. They treat you for whatever you have and they treat you like a private doctor." . "If you had a lot of money, you would have the best treatment. Everyone wishes that." . --- Economic factors Both direct health costs and indirect health costs were considered essential when deciding where to seek services for diabetes. In Wadi Khaled, several participants described having difficulties paying for transportation to the health facility, or finding affordable transportation. If services were "free of charge", this was an important determinant for choice, especially if people had co-morbidities. Several patients valued regular follow-ups with clinicians and monitoring tests, but when out of pocket payments were required, they would buy medication directly from pharmacies, change provider or miss routine visits instead. In some cases, economic challenges and lack of knowledge of the availability of free care services led patients to have periods without medication. Some interviewees mentioned financial factors as a reason for deferring care for diabetes-related complications. "I stopped taking medication for two years here in Lebanon, because it wasn't available for free. I couldn't afford it. When we first came here, we used to buy it. We had money. Later, I went to [a hospital] and got pills for two months, and when [MSF] opened here, I came here. I am taking my medication from here, monthly." . "Sometimes I couldn't buy it. I stopped taking it until I could buy it again, we have other priorities at home which are more important. So sometimes I would stop taking the medication for ten days until I could afford it again. My husband is a worker and sometimes he did not have the money for the medication, and we had children, so we had daily needs for the house which were more important." . "Back then, I was working, I used to make money and buy the medication. My children used to help me and there were some people that helped me by providing me with the medication… I told the doctor that I need the cheapest medication. The medication was for 5,000 Lebanese Lira . My wife's brother buys the diabetes medication that costs 80,000 Lebanese Lira . I prefer buying the cheapest one." . --- Distance and transportation Interviewees from urban Shatila did not mention distance or transportation as a barrier to accessing care, but several participants in rural Wadi Khaled described difficulties finding transportation, including in emergency situations or times of acute illness. If someone wasseverely ill and was unable to travel, they would consider changing provider and going to their nearest PHCC or buying medication directly from a pharmacy. In Wadi Khaled, several female interviewees highlighted the importance of proximity when choosing a PHCC, with many of them dependent on a male family member to accompany them to clinic visits. If the distance to the facility was large and took more time, male family members were less available to accompany their relatives as they could not afford to take time off work. "I only went once; I did not feel comfortable… The distance is far. And the transportation is hard. Whenever I think about it, I just stay at home and I do not go." . "I don't go if [my brother and father] don't take me, sometimes it is hard. If my brothers are at work, they need to take the day off to take me. Sometimes they cannot, because they need to work to make a living. We only go where it is near, because diabetes needs follow up so I always look for something close." . --- Legal factors The legal status of the Syrian interviewees also had an influence on where they chose to access health-care services. Some participants considered seeking care in Syria for diabetes-related complications but their legal status within Lebanon and official and unofficial border fees made this challenging. "I had a sponsor, so I used to go [to Syria] in an official way, but now I haven't had a sponsor for the last three months, so my papers are not official… I couldn't go [to Syria] for checkups every two months, it is cheaper, but I do not live there and I could not leave [Lebanon]." . "No, I never went back to visit my doctor in Syria. I came here [to Lebanon] legally but then I broke it. If I went to Syria and wanted to come back to Lebanon I had to pay 900 USD. I should pay because I am here for a visit and I stayed more. If I wanted to go back to the doctor, I would face obstacles." . In Wadi Khaled, legal status was also relevant, as some male participants would not cross a military checkpoint due to security concerns, which limited their access to care. "The issue is if you are not legal. I am not legal, but it doesn't stop me. To some people, it is a problem if you are illegal. It's rare that they [the military] take [detain] someone. If you are heading to the hospital, they won't stop you." . --- Health care seeking pathways Interviewees described seeking care from different 'combinations' of health-care providers, and rarely received all elements of required care from one provider. --- Seeking care on arrival Upon arrival in Lebanon, many participants described a period in which they "didn't know anything" about the health system and humanitarian assistance support available to them. This led to them buying their medication at a pharmacy in Lebanon or arranging its purchase in Syria and transportation to Lebanon. Patients would then ask their social networks of family members and peers for advice if they were no longer able to access or afford medication through this initial route. In other cases, interviewees learned about the availability of other providers through their social networks "by chance" and subsequently changed to a different facility: "I had the card that I used in Syria that had the insulin type and dosage on it. When I came [to Lebanon], I went to a pharmacy, I did not go to a diabetes doctor. I showed the pharmacist the card. I stayed like that for a while. We have been in Lebanon for 8 years, I stayed like that for almost 3-4 years. Then there were people that came to our building, and they told my daughter-in-law about MSF." . --- Seeking care from multiple providers Interviewees often sought care for diabetes from more than one provider at the same time, or changed providers based on challenges accessing their first choice of provider, unstable provision of medication or inability to pay. Consultations and prescriptions were often given at PHCC level, and medication then collected from a different pharmacy. In some cases, interviewees accessed a pharmacy as their principal provider and only went to a clinician in the case of acute symptoms. "I haven't visited [the doctor] in a long time. I stopped when my results became stable." . "I didn't see a doctor, I used to take my medication, and sometimes I used to go to the pharmacy to test my glucose. I only did the test for Hb1AC twice at the private doctor and that was two years ago… But in the last period, I felt tired and I thought that I had to do tests and see a doctor." . Sometimes, interviewees received a basic package of diabetes care free of charge or at a subsidized price at a PHCC, then visited a private specialist if they were pregnant or saw a change in their condition. Some described how they would attempt to obtain medications from other providers if drugs were not available at their usual PHCC. They also reported accessing services outside of their regular provider in the case of illness, when they would go to the closest facility. Seeing an improvement in their health was also reported as a factor for staying with a specific care provider for regular diabetes care, medication or care for diabetes complications. Three case stories of participants illustrate examples of health care seeking pathways. --- Case story of a 60 year old man with T2DM, living in Wadi Khaled AB is 60 years old. He came to Wadi Khaled from Syria in 2012. He has hypertension and T2DM. "When I came here, I had around three boxes of medication with me from Syria . They lasted me for five to six months. After that, the Lebanese people told us about the PHCC in the area to treat diabetes." He first went to a PHCC in F., quite far from his house, that provided care and medication free of charge. But after some years this facility closed. He changed to PHCC N, where consultation cost was low, but AB struggled to pay for tests and medication. When he learned that the MSF clinic provides comprehensive care free of charge, he accessed MSF's services instead. "I don't have money, if I go to a doctor or to the PHCC N, they would give me a prescription to go buy my own medication from the pharmacy. They don't give you medication for free. Since we came here, our financial situation is not really good. For the lab test, they wanted 40 USD; where do I get that from? If you don't have the money, they would not look at you." AB experienced an episode of acute illness that hindered his access to care at the MSF clinic: "I kept on going to MSF to get treated until one day I got sick. I couldn't go to MSF anymore. I gave my brother the card, MSF did not give him [the medication for me]. They told them that they wouldn't give it to him unless I was there. That is their right. I stopped going for four to five months and [my siblings] got me the medication from here and there. And people helped us. My siblings got me a box or two from a pharmacist. Sometimes [the pharmacist] did not take money, sometimes he did. I stayed like that for four months, and my diabetes was getting worse. I couldn't take it anymore; I came back to MSF." Now, AB combines the services of two providers. Once per month, he attends his regular appointment at the MSF clinic for consultations with the doctor, tests and the collection of his medication. If he feels acutely unwell and wants to check his blood pressure and glucose level, he goes to yet another PHCC near his house, where those tests are free of charge. His doctor at MSF told him about this option. If the test results are abnormal, he seeks immediate treatment at MSF. AB chose this combination because he struggles to pay for transportation to reach the MSF clinic. Case story of a 22-year-old woman with T1DM, living in Wadi Khaled Some years ago, PS felt unwell with diabetes symptoms. She visited three health care providers over the course of several months, but they all failed to diagnose her correctly. One provider asked her to pay for diagnostic tests, but her family could not afford them, so she did not return for follow up. Several months later she collapsed and was brought to the hospital outside Wadi Khaled for emergency care. T1DM was diagnosed and slow and rapid insulin was prescribed. PS then bought slow and rapid insulin at a pharmacy and sometimes went to PHCC X in Wadi Khaled where she received mixed insulin free of charge. The PHCC only provided mixed insulin but PS needed slow and rapid insulin, and the provision was not stable. At this PHCC, PS met Dr. Z who recommended that she go to the nearest MSF clinic because MSF would provide her with best available care at no charge. PS trusted Dr. Z and consulted her about any decisions that might impact the disease. After the long delay in diagnosis, the most important thing for PS at this point was to see a good doctor. She was willing to travel a long distance and pay for good care. She went to the MSF clinic near Tripoli twice, but stopped because she felt the distance was too far. She continued obtaining insulin from pharmacies and from the PHCC X, and followed up with Dr. Z over the phone for the period of one year. As soon as MSF opened an NCD clinic in Wadi Khaled, PS registered. For some time, MSF was her only diabetes care provider. Later, Dr. Z opened a private practice in Wadi Khaled near the patient's house. Now, PS feels comfortable with the care at MSF and she sees Dr. Z in case of acute illness or emergencies. Dr. Z offers her advice free of charge to PS. Case story of a 47-year-old woman with T2DM, living in Greater Beirut FM already lived in Lebanon when she first experienced symptoms. She went to a laboratory that is well known amongst her peers in the area she lived in, to get tested. After the positive test, she went to a PHCC and asked for referral to a specialist "doctor for diabetes". She chose that PHCC because it was near to her house, and cheaper than other places, "they treat people in a good way", and it has "all the doctors we need". FM says: "They have different doctors, and if the doctor that does not come to the center, we can go visit him/her in his/ her clinic after paying at the center." FM visited the diabetes specialist at her private clinic. She had a positive experience: the doctor was "really great, she treats people well, she takes care of the patient." Once FM's blood glucose levels were stable with medication, she stopped seeing the doctor. In the next 1.5 years, FM followed the doctor's prescription and bought medication from the pharmacy. Then she heard from peers that "at MSF there are doctors and they provide medication and they treat people really well". As FM suffers "from a chronic disease and always need[s] medication", she "can't always afford it". FM says: "When I first had diabetes only, it was okay. When I was diagnosed with hypertension, I had six different types of medication to take. So it was hard. And I have to constantly run tests." Now her only care provider for diabetes and hypertension is MSF, but recently she got diagnosed with anemia. This anemia medication is not available at MSF, so she buys it at a pharmacy. --- Discussion Our study adds a new perspective to the existing knowledge on health care access and the utilization of services within refugee populations, namely how people in a multi-provider health system choose to seek care and services for a chronic disease such as diabetes. The findings show that health seeking behaviour is influenced by an individual understanding of diabetes and its management, perceptions of quality of care and contextual constraints such as transport and legal barriers. Acute episodes and co-morbidity contribute to the variability in health seeking pathways. The subjective level of knowledge about diabetes and its management varied amongst the Syrian refugees in our study, which partially confirms other studies showing that diabetes patients from the Middle Eastern region have generally limited knowledge about disease management [31,32]. Other studies showed that Syrians who were diagnosed after the onset of the crisis have less knowledge than those diagnosed prior to displacement [33]. People with lower levels of knowledge about diabetes are known to be less adherent to treatment [34,35] and adequate self-management [36]. Irregularity of follow up visits also pose an increased risk for diabetes complications [37]. This study shows that Syrian refugees lack orientation in health-care services in Lebanon when they arrive, and may still rely on their social networks to obtain their medication from Syria before accessing the Lebanese health system. We found that Syrian refugees in Lebanon gain information about health services for diabetes largely from family and peers rather than through formal means, such as health-care providers or the UNHCR. Similar behaviors have been documented in refugee populations in other settings [38]. In the case of Syrian refugees in Lebanon, reasons for this might include inadequate information channels, experiences of unclear information, mistrust in public institutions and perceived discrimination in the health system [21]. Whilst the urban camp of Shatila and the rural setting of Wadi Khaled have different provider landscapes, we did not find pronounced differences in approaches to seeking care based to location. Similarly, we did not find pronounced differences between the health-seeking behavior of individuals with T1DM and T2DM, even though the disease types differ in pathology, progression and treatment. It seems that the other factors presented above are more prominent in determining care pathways, but more research is needed to better understand this issue. Among participants in this study, the parallel use of different combinations of health care providers was common as a result of limited knowledge, access barriers, and the structure of the health-care system. McNatt et al. found similar evidence in Jordan, where Syrian refugees usually visited more than three facilities for NCD care because of the narrow selection of services available at each institution [39]. In other settings, chronic disease and unsolved health problems [40] have been seen as positively associated with visiting multiple doctors for the same health problem, while trust in the current provider [41] and satisfaction with treatment contributes to not seeking care with other providers [42]. Receiving care from more than one provider in parallel can cause detrimental health outcomes, and attending more than one clinic for follow up has been found to be a risk factor for diabetes complications among Egyptian patients [37] and in other populations [43]. Socio-economic status and legal status posed structural barriers to accessing health care for many of our participants. Amongst the identified factors influencing choice of provider, we found that financial considerations were a main factor for those with co-morbidities and chronic conditions, who favoured providers that offered medication and treatment free of charge. Coping strategies for financial barriers, such as delaying doctor's visits and buying medication directly from pharmacies have also been documented among Syrian refugees in Jordan [39] and cost has been previously identified as the main barrier for accessing health care for chronic conditions for Syrian refugees in Lebanon [18,20,26]. Despite the efforts of many actors to enhance access to care, Syrian refugee's find it increasingly difficult to afford care [26], possibly because share of out of pocket payments is increasing, and economic conditions are further deteriorating. In 2019, about half of Syrian refugee households were living in extreme poverty [23]. In the light of high financial vulnerability and risk of catastrophic health expenditure, access to comprehensive care should be ensured with limited co-payments, especially for those with multimorbidities and chronic conditions that require long-term treatment. We found that Syrian refugees face difficulties navigating the Lebanese health system and accessing adequate care. On one hand, this is due to system level factors: the health system is fragmented and complex, and comprehensive information about the availability of and access to health services is difficult to provide [21]. Syrians refugees in Lebanon face an additional challenge as result of transitioning from a one-provider national health system in prewar Syria [44] to a fragmented multi-provider system. To facilitate the orientation of refugees in a new health system, information about health services needs to be easily accessible and understandable. The concept of "health literate health care organizations" can serve as suitable approach for health care organizations to address this issue on a system level [45]. On the other hand, the difficulties in navigating the health system are caused by individual factors, such as insufficient level of patient awareness and understanding of diabetes treatment and services in this specific context. A lack of health literacy amongst patients has been previously shown to contribute to the challenges of navigating the health system of a host country [38,46]. As suggested by others, [12,18] we find that Syrian refugees need better support to identify and access adequate NCD care. Health care providers should engage in raising awareness about existing services and directly assist Syrian refugees to identify and access care, as well as in improving the diabetes literacy of patients, for example through providing diabetes self-management, education and support programmes [33,47,48]. It is important for health-care providers to recognize that seeking treatment from multiple providers is common among this population. Including discussions about care received from other providers should be part of routine patient consultations, to reduce the high risk of drug interactions, to limit potentially harmful self-management strategies like adjustment of dosage, and to foster continuity of care among a patient group with low diabetes literacy. Further research is needed to fully understand the reasons for and implications of combining providers for diabetes care amongst Syrian refugees in Lebanon, as well as for refugee populations in other settings. In situations such as that of Syrian refugees in Lebanon, where a vulnerable population struggles to navigate a complex multi-provider system, health care providers need to further coordinate the services offered and organize care in a way that is easy for the patient to understand and access, whilst ensuring continuity of care. The following study limitations need to be taken into consideration. Firstly, we only recruited patients from MSF facilities, thus their views may not represent those of community members who do not access MSF's services. While results of this descriptive study are specific to the population and settings it was conducted in, findings may also shed light on related practices and experiences, for example for Syrian refugees accessing care for other chronic conditions in Lebanon, or vulnerable refugees accessing diabetes care in similarly fragmented health systems. Secondly, the PI and the translators who conducted interviews were affiliated with MSF, even though they were not directly providing care to patients. This means participants may not have felt comfortable to criticize MSF, thus bringing in a degree of social desirability bias. We strove to mitigate this risk by emphasizing the confidentiality of anything said within the interview during the informed consent process. Finally, the PI does not speak Arabic and was reliant upon translators. This limitation was mitigated through thoroughly reviewing the interviews and the transcripts to ensure that the loss of nuances during translation was minimal. --- Conclusions Syrian refugees with diabetes in Lebanon face considerable challenges in navigating the health care system which leads to reduced access and interrupted disease management. Continuity of care for this vulnerable population, and for people with other chronic diseases, needs special attention. Initial steps could be to support patients in Lebanon in their orientation within a multiprovider health system and encourage providers to ask patients about their pathways to care in order to help them access treatment and follow-up visits. --- --- Abbreviations DM: Diabetes mellitus; IDI: In-depth interview; MIC: Middle-income countries; MSF: Médecins Sans Frontières; NCD: Non-communicable disease; NGO: Nongovernmental organization; PHC: Primary health care; PHCC: Primary health care center; PI: Principal investigator; T1DM: Type 1 diabetes mellitus; T2DM: Type 2 diabetes mellitus; UNHCR: United Nations High Commissioner for Refugees --- Appendix 1 In-depth interview guide --- Introduction Introduce yourself and the translator. Thank you for agreeing to talk with me today. I would like to ask you some questions about your experiences regarding health care for diabetes. The interview will take approximately 30 min to 1 hour. There are no right answers or wrong answers to the questions I will ask you. If you don't feel comfortable answering some questions you have the right not to answer. We can also stop this interview at any time. --- --- Competing interests On behalf of all the authors, the corresponding author states that there is no competing interest. ---

Background: Navigating health systems in host countries can be a challenge for refugees, particularly in a multiprovider system such as Lebanon. Syrian refugees in Lebanon face a high burden of Non-Communicable Diseases (NCDs) including diabetes mellitus. Evidence on how refugees navigate the health system is essential to improve provision of NCD services. We conducted a qualitative study amongst Syrian diabetes patients visiting Médecins Sans Frontières (MSF) clinics in one urban and one rural setting in Lebanon to explore factors influencing choice of and pathways to diabetes care. Methods: In-depth interviews were conducted with male and female adult participants with DM type 1 or type 2 who were receiving treatment at MSF clinics. Participants were recruited using convenience sampling. Interviews were conducted in Arabic and directly transcribed and translated into English. Data were coded in NVivo and analyzed using an inductive thematic approach. Results: A total of 29 in-depth interviews were conducted with 13 men and 16 women. Knowledge and understanding of diabetes management differed among participants. Syrian refugees in Lebanon gathered information about health services for diabetes largely from social networks of family and peers rather than through formal means. Pathways to care included different combinations of providers such as clinics, pharmacists and informal providers. Conclusions: Syrian refugees with diabetes in Lebanon face considerable challenges in navigating the health care system due to their vulnerable status and limited knowledge of the host country systems. To ensure access to care for diabetes, efforts need to be made to support patients' orientation in the Lebanese health system.

It was extremely desirable, for the sake of those who may wish to study the evidence for Dr Tylor's conclusions, that the full information should be given as to the degree in which the customs of the tribes and races which are compared together are independent. It might be that some tribes derived them [i.e., the customs] from a common source so that they were duplicates of the same original. Galton's remark reproduced in Tylor Sir Francis Galton is known for his work in anthropology, for developing the correlation measure, and for devising a classification method for fingertips. He is also known for his celebrated comment on Dr. Tylor's research at an academic conference. As described in the introductory quote, Sir Galton challenged Tylor's theory about institutions across societies on the grounds that a high degree of social linkages or dependence between tribes undercuts theory validity. Sir Galton did not object to dependence between cases. Rather, he cautioned against researchers evoking cross-case evidence to support the veracity and generalizability of their claims, when knowledge about the degree of case dependence is limited. In multiple-case research, social dependence broadly refers to meaningful linkages between cases, such that conditions set for one case influence the other cases. One example is when managers who work in a case organization share social ties with members in another case organization. When dependence remains undisclosed or no attempt is made to unveil its role, researchers may treat two cases as highly different when these cases are in fact two illustrations of the same phenomenon. While case dependence might manifest in several ways, social linkages between members across cases are a pervasive and probably the most determinant form of case dependence . The researcher's knowledge about social linkages between cases is fundamental in establishing whether cases represent duplicates of the same original or different originals. The comparison and contrast of cases is a central feature in theory building using a comparative case design . The discussion about the researchers' assumptions about case dependence is relevant because case research designs are increasingly used to build theory . Comparative case research is suitable to examine emerging phenomena or aspects that are difficult to study . Broadly, the conjunction between the centrality of case comparison in inductive reasoning and the growth of comparative case research is such that the researchers' knowledge about whether the cases are linked in ways that are meaningful to the research question merits further scrutiny. To put it more directly, this article asks whether and under which conditions Galton's challenge is relevant for today's comparative case researchers. In this article, we build on research across the social sciences to clarifyand to some extent demystify-why and how issues of case dependence in theory building are distinct from those in statistical analyses. Following an inductive approach ourselves, we draw on research using comparative case design for the study of innovation diffusion to develop a systematic analysis of the role of case dependence. Our findings show two generalized research designs that, based on the analysis of prior research, relate to confusion-or incomplete reporting-rather than to variants of the comparative case design. One research design displays high dependence between cases , while the other design displays low case dependence . We provide illustrations of confusion in past research about the rationale for the use of comparative case research and evidence of case dependence through social linkages when building theory on innovation diffusion . To examine the degree of case dependence, we draw on past research on innovation diffusion using a comparative case design to develop an agentbased simulation that provides us with a "virtual lab" . The purpose of the simulation is to help us exploring how a researcher's assumptions affect what qualifies as cross-case evidence about innovation diffusion, according to varied levels of case dependence. Indeed, as we will elaborate in this article, we provide a novel way to use agent-based simulations as metatheoretical tools in sociological methods and research. Acknowledging the variety of approaches to theory building from multiple cases, we outline the general conditions under which case dependence undercuts the insights derived from cross-case comparisons. Specifically, we discuss implications about the veracity of the study's presumed causal or logical relationships and the extent to which the conclusions are generalizable or transferrable beyond the cases being examined . The experimental findings further aid us to devise actionable suggestions for studying and communicating case dependence in future research. Innovation diffusion provides an instructive context, but our insights enhance reasoning in the social sciences in general. This study is probably one of the first studies to take advantage of agent-based models to address central questions in research methods about qualitative research designs. 1 Our study also aims to facilitate transparency and promote the diversity of approaches to theory building in social sciences. This article proceeds as follows. It first presents the conceptual framework for the study and then shows the development of an ABM simulation grounded in prior research on innovation diffusion. Next, it reports the results and its consequences for research. Finally, it discusses the implications for researchers who are interested in multiple-case research. --- Conceptual Framework We synthesize the literature on the role of multiple cases in theory building followed by the use of the literature on innovation diffusion to illustrate the sources of social linkages between members of a social system. 2 Next, we propose that case dependence in building theory from multiple cases is best conceptualized as an issue about the researchers' bounded rationality. --- The Role of Case Comparisons in Theory Building Cases are central for theory building in comparative case research. However, the definition of a "case" is occasionally surrounded by confusion ). In an illuminating discussion, Ragin argues that what a case is might often be unclear to experienced researchers and novices alike. Ragin compares one study that conducts interviews about an organization's informal structure with another study that uses interviews about employees' job satisfaction. These two hypothetical studies are similar, but the former focuses on the organization while the latter concerns the employees . Several definitions of a case are available , but these definitions converge toward a class of, for example, events, organizations, and places that are of interest to a researcher. A case might be an organization , an episode , or a tribe, like the example from the introductory quote. 3 Why do case comparisons matter in theory building? Cross-case comparisons guide the researcher in identifying data patterns that support emerging propositions. Early methodological contributions stress the importance of case comparisons in theory building. This idea is present in Mill's method of agreement and method of difference . The comparative case study method entails a systematic analysis of a small number of cases, a number deemed too small for an analysis using statistical techniques . Case comparisons are useful to uncover explanatory linkages between concepts, to illustrate theory, and to compare and contrast different contexts . There is a wealth of approaches to comparative case research , but they unite around the importance of case comparisons as a prime source of theory building from cases . In a study of eight ailing organizations, Harris and Sutton address the aspects of case dependence by declaring the selection of organizations according to four categories: private, dependent; private, independent; public, dependent; and public, independent. As Harris and Sutton state, "[t]he varied sample served to strengthen theory-building about constant elements [ . . . ] and the similarities observed across a diverse sample offer firmer grounding for such propositions." The multiple-case study approach provides an instructive basis for discussion as probably one of the most frequently used case designs to build theory among case researchers . The central idea of multiple-case approach hinges on low degree of case dependence. Eisenhardt notes that "the juxtaposition of seemingly similar cases by a researcher looking for differences can break simplistic frames. In the same way, the search for similarity in a seemingly different pair can also lead to more sophisticated understanding." Furthermore, Eisenhardt and Graebner add that "each case serves as a distinct experiment that stands on its own as an analytic unit. Like a series of related laboratory experiments, multiple cases are discrete experiments that serve as replications, contrasts, and extensions to the emerging theory." The corollary is that case comparisons are instrumental for theory building by revealing similarities and differences between cases. However, the multiple-case study approach has also been criticized for attempting to mimic deductive research . At the core, still, case comparisons play a central role for theory building from cases, regardless of the exact approach followed by the researcher. Demystifying case dependence in multiple-case research. At first, case dependence would appear to be a matter for deductive research only. After all, inductive and deductive approaches are distinct and used for different purposes in social research. In deductive research , linkages between observations undermine statistical inference by violating the assumption of independence . In inductive research, cases are not used to test hypotheses in the statistical sense. Although uncontroversial, this observation should not result in the discharge of the role of case dependence in theory building. In referring to this point, George and Bennett conclude that "[t]he statistical version of the problem does not apply to case studies, but a more fundamental concern does." In particular, social linkages between case members are a typical and dominant manifestation of case dependence in comparative research . A low degree of case dependence ensures that theoretical propositions are substantiated through evidence from across cases . In contrast, cases with a high degree of dependence might actually be desirable to address specific research questions . For example, consider a study of the market-entry strategies that were adopted by several U.S.-based startup organizations in the software industry and how these strategies influence their financial performance. The financial performance of these startups might be influenced not only by different organizational strategies but also by the mobility of "star" engineers between start-ups and across neighboring states. Case dependence bears implications for internal validity and external validity of the answers being provided to research questions. As for internal validity, neglecting dependence among cases undermines the chain of evidence in support of the newly reported relationships between constructs. First, the presence of duplicates undercuts a tenet in comparative case research: Comparisons across cases strengthen the evidence of emerging relationships between concepts . New observations based on interconnected cases do not add as much new information as those from unrelated cases even if these observations are still be relevant . Second, a best practice in comparative case research is to rule out alternative explanations of the reported relationship between constructs . In qualitative research, more broadly, "the persuasiveness of the arguments is greatly strengthened if serious attention is given to alternative explanations" . The lack of awareness of dependence among cases restricts the range of distinct cross-case evidence available-where the choice for a multiple-case study approach is often justified by the interest in gathering extensive evidence from multiple cases-to rule out alternative explanations and bolster the evidence reported to support emerging insights and often formalized as propositions. As for external validity, information about case characteristics aids better specifying the transferability of the findings . A clarification of the degree to which the cases are dependent aids researchers to specify a nontrivial condition for whether the newly developed theory "accounts for phenomena not only in the setting in which they are studied, but also in other settings" ). By overlooking case dependence, researchers risk to omit a core case characteristic that defines the newly developed theory and its application to other similar contexts. As we show below, we can further address the concern above, asking how researchers might use case dependence to advance theory. --- Case Dependence: Studies on Innovation Diffusion as an Illustration Diffusion concerns how, why, and at what rate new ideas and technologies spread . Diffusion occurs through social influence among the members of a system . Rogers posits that "we must understand the nature of networks if we are to comprehend the diffusion of innovations fully." With such a focus, we identify three sources of social linkages between cases: the industry network, the diffusion process, and the organizational characteristics. Industry networks. Members of a system are embedded in industry networks. Industry and national contexts support ties among organizations and individuals . More specifically, countries and industries entail social institutions and values that are translated into regulations and industry standards that contribute to linkages among actors . Industry standards not only support information diffusion but also help mediate innovation activities between industry actors . Furthermore, the members of a social system maintain ties that support relational and economic exchanges. According to Alter and Hage , networks "constitute a basic social form that permits interorganizational interactions of exchange, concerted action and joint production." Industry networks underpin innovation diffusion . Examples of industry actors that support industry networks include supplier associations , industry regulators , certification agencies , and trade advisory boards . These actors also provide "networking" opportunities for managers to exchange knowledge and enact new practices . Prior literature further notes the role of personal networks in an industry, for example, based on graduate programs and communities of users . The high embeddedness of actors in industry networks promotes common values, practices, and standards that sustain linkages between members of a social system. Table 1 presents an overview of the sources of social linkages among cases. Diffusion process. The second source of social linkages among cases is the uncertainty of the diffusion process. Such uncertainty accrues from three factors: the technology , the actors , and the process . The term technology is given a broad meaning here; it refers to products, practices, ideas, and services. Technology's features add uncertainty to innovation diffusion. For instance, technologies that display a low degree of cumulativeness , maturity in the market , and applicability of routines add uncertainty about the extent to which system members adopt innovation. Dyer and Nobeoka , Hargadon and Fanelli , Keidel , and Yoshikawa, Tsui-Auch, and McGuire Structural and strategic characteristics of organizations create dependences between members The actors , such as organizations, and individuals, further influence the degree of uncertainty. They endorse values and norms that might change the process of diffusion. For instance, Toyota established a supplier association that was specifically designed to promote the exchange of technical information between Toyota and its suppliers . The adoption by specific actors , the extent to which they craft solutions that shape ideas about the technology , their preferences , and their reactions to conflicting institutional pressures introduce uncertainty linkages between members . The factors concerning uncertainty in the diffusion of innovations make it difficult for researchers to anticipate which system members share social linkages with . Organizational characteristics. The last source of social linkages between members of a system concerns organizations' structural and operational characteristics. Structural characteristics include organizational design features. For example, the degree of control-related work systems in the source company influences the diffusion and adoption of the recipient company . Other structural characteristics include work systems between parent companies and headquarters , resources and incentives shared between actors , and formalized interactions between actors . Operational characteristics support linkages between actors. A canonical example is an organization's set of buyer-supplier relationships. Other operational characteristics include the mobility of skilled engineers within and across organizations , joint teams between a focal organization and its suppliers , employees' career trajectories , talent recruitment strategies , and co-ownership agreements . Overall, the structural and operational characteristics of organizations define the extent to which members of a system are interconnected. As George and Bennett first conjectured, innovation diffusion is elucidative of the challenges that researchers face when accessing social linkages among case members. One may assume that information flows from the cases to the researchers without barriers. In this ideal position, it is plausible to form a realistic assessment of the degree of case dependence. One may also assume that researchers have the issue of the degree of case dependence in mind while designing and conducting their studies. The researcher could select, for example, theoretically distinct cases to serve as independent experiments. Thus, the researcher must explicitly consider dependence to be information worth gathering and meaningful for one's research. --- Case Dependence as a Researcher's Bounded Rationality Problem Innovation diffusion is a case in point to the challenges that researchers face when building theory from multiple cases and, more broadly, comparative case research. We submit that the issues around case dependence is best conceptualized as a bounded rationality problem. Drawing on behavioral science ), researchers' bounded rationality refers to the limits of their ability to identify optimally, sometimes even satisfactorily, the degree of case dependence in complex empirical settings. Researchers might intend to set up a comparative case design where cases share no meaningful linkages, but those linkages are often unknown to the researcher. We are concerned with the extent to which researchers' assumptions about case dependence fail to fit the empirical setting; we are not concerned with researchers' preference for any particular approach to comparative case design. We focus on researchers' challenge of identifying the degree of case dependence against their own aspirations and the misfit between researchers' aspirations and the actual degree of case dependence. In the widespread multiple-case research that foregrounds case comparison, this misfit presents threats to internal validity and external validity. 4 A conceptualization of the degree of case dependence as a bounded rationality problem contributes to a better understanding of social research methods. First, the notion of bounded rationality shifts the attention from an abstract debate about quantitative versus qualitative research or the preference for a specific version of comparative case to the challenges that researcher face when building theory. Second, by focusing on bounded rationality, the discussion moves forward by devising heuristics which future researchers can resort to align their research design with their declared rationale to use comparative case research and their research goals. Our call is not to standardize the research process. Rather, we aim to keep creativity as an essential element of discovery in social research. The central issue is the following: If we were to formulate a comparative case design in which high versus low degrees of case dependence were acknowledged, how would we revise the research process and evaluate newly developed theory against commonly held standards in qualitative research? --- Model Development Past literature aided us in identifying simulation parameters and value ranges for these parameters, thus lending realism to our model . The simulation model captured key features of innovation diffusion on the context of number of cases . This purpose fits broadly to what Edmonds et al. define as an illustration purpose. More specifically, we use the information gathered from the simulation model to learn about multiple-case research methods when it comes to the assumptions on dependence of cases. As we elaborate below, we use the model to examine whether, and when, case dependence could influence theoretical insights based on multiple-case research. Hence, the simulation model has a second-order , or meta-theoretical, purpose in that it results are informative to choices that might impact theory development efforts. 5 --- Comparative Case Research About Innovation Diffusion We first conducted a systematic review of the literature on innovation diffusion that uses comparative case research. This review aimed to identify current practices related to the degree of case dependence and to guide the design of a computational simulation. Review scope. We used a set of criteria to define the review scope. First, we selected the top-tier journals in general management and leading journals on innovation. 6 Second, we defined a set of search words used in the EBSCO Business Elite data set. We used "diffusion" OR "adoption" and "case" . Various labels have been used to refer to comparative case design . We searched for "case" in the full text. We manually removed single case studies. Finally, we placed no restriction on the start date; the first result in our search was Eilon and Elmaleh . The end date was December 2019. In total, we found 393 articles. Coding procedures. We assessed the relevance of articles against two criteria: They used any form of comparative case research, and they studied diffusion or adoption. Two independent coders assessed the relevance of the articles. Each coder received a coding booklet. This booklet included instructions for the coders, supported by examples. The coders compared their coding and settled any disagreements. We found very strong intercoder reliability . In total, we identified 40 relevant articles. Case dependence in past research. A few insights are worth noting with regard to the study of innovation diffusion using comparative case research. Table 2 presents an overview. First, we found sustained growth of the use of comparative case research to study innovation diffusion. Of 40 articles, 18 were published between 2009 and 2019. Table 2 shows that comparative case research has been used to study innovation diffusion across a wide range of national, industry, and organizational contexts. Second, researchers have studied various types of cases, such as firms , products , and teams . Finally, researchers have used six cases on average . The cases are the frequent number ; half of the studies use up to three cases . The use of comparative case research is overwhelmingly justified by the advantage of within-and cross-case comparisons to develop theoretical propositions . This suggests that prior research has largely followed Eisenhardt's contribution on the role of cross-case comparisons of distinct cases to build robust and generalizable theory. We found that studies provided limited information about the degree of case dependence. As Table 2 shows, 10 studies provided no information. Our review shows that 17 studies report information suggesting independence of the cases , while 13 studies reported information about case dependence. The dominant manifestation of case dependence was social linkages between case members . However, the role of case dependence was largely left open to interpretation. For example, it was unclear whether linkages across cases were of any relevance to the newly developed theory. It was also unclear whether the reported dependence played a part in the boundary conditions of the newly developed theory. Perhaps the dearth of information is a byproduct of articles citing methods from different intellectual traditions with little integration of ideas . 7 The insights from our literature review provided us with further assurance of the pertinence of examining the role of case dependence in theory building. In our review, we found two archetypes of comparative case research: high versus low dependence between cases. Dependence is primarily understood in terms of social linkages because it refers to the source of case dependence alluded in prior methods discussions and it dominates in prior comparative case research, as shown in our review. The complete research design involves a high degree of case dependence. Figure 1 illustrates the complete research design. By complete, we mean a high degree of social relationships across cases. The notion of a complete research design is consistent with a "collective case study" a The information reported in the articles was insufficient for an assessment of the degree of case dependence. We coded the articles using following categories: independent , dependent , and unclear . and an "embedded case" . Our notion of a "complete research design" specifically emphasizes the extent to which cases are connected through social linkages. In a study of architecture projects conducted by Gehry Partners, social relationships among individuals across projects influence practices of innovation diffusion . Other examples from our literature review include firms in Italy's aerospace industry , radical innovation in the aerospace industry , and computer games owned by the same firm . Ego research design features a low degree of dependence between cases. When cases under study do not share social linkages, we refer to them as using an ego research design. We borrow our terminology from sociological studies of networks. Similar to an ego network based on all ties by an "ego" , our notion of an ego research design describes a set of cases that are not connected through social ties. Figure 1 illustrates the ego research design; it entails a low degree of case dependence. The lines represent social ties and the node's entities within the case . Low case dependence occurs when cases are from different industries or distinct divisions in the organization . In general terms, "ego" research designs resemble conventional multiple-case research . In an "ego" network, the nodes other than the ego are called alters, and they usually have a direct connection with the ego. All other connections of the network-that is, outer connections of other alters with the ego's alters-are excluded. --- ABM Having identified two archetypes of comparative case designs, we faced the challenge of examining why and under which conditions each research design could influence understanding and analysis. As previously noted, we use the agent-based simulation in an indirect way. The purpose of our ABM is threefold: to explore the interaction between case dependence and emerging theoretical insights, to manipulate experimentally the degree of case dependence, and to explore why and under which conditions each research design could influence emerging theoretical insights. From many parts, computer simulation use has been advocated as an aid to theory building , when -researchers aim to analyze boundary conditions, -the landscape of potential alternatives is partially unknown, and -the impact of the construct under analysis is difficult to assess solely through the literature. The study of the degree of case dependence meets all three of these conditions. For this reason, we used an agent-based computational simulation to study experimentally the degree of case dependence. 8 Model specifications. The simulation model entails a nexus of agents that, in this case, represent organizations. We used our review of past research to develop realistic parameters in our simulation and align it to the standard mechanisms as usually coded in threshold models of diffusion . Below, each model parameter is described . In our computational model, 9 some agents are selected as case agents. Other agents are connected via links to each other and to the case agents. The other agents are part of the industry. The number of cases ranges between 2 and 50, while the other agents are kept constant at 300 . This figure of 50 cases is much higher than the average number of cases found in our literature review . We purposefully added a maximum of 50 cases to test the conditions of our model when pushed beyond the limits found in the literature . We also opted for a relatively large population of potential other organizations that might be suitable to develop theory on a specific issue. 10 Ultimately, the relevance and number of cases depend on the research question . Figure 2 depicts the simulation environment, developed via NetLogo Version 5.2 , an agent-based simulation software. Each agent is randomly located in a 3D space where the links among agents become visible. These links are established based on the parameter range that defines the Euclidean distance of r pixels around each agent, hence it specifies interaction based on proximity of other agents falling within the specified range. This is a typical feature of ABMs; it represents the idea that each agent has a limited reach and can only connect to those other agents that fall within its range of possible interactions . To some extent, the parameter range is a function that attributes uniqueness to each agent by helping define their interactions. In practice, --- Note: We aimed to develop a parsimonious model of innovation diffusion. Our approach does not preclude future researchers from adding parameters that suit their own interests in innovation diffusion. Our typology of sources of cross-case ties provides a starting point to tailor this model to specific research questions. even if the range is the same for all agents in the environment, the actual agent-based network takes a form that depends on the actual number and characteristics of other agents at a distance that is defined by this parameter. In other words, range represents the diverse extent to which organizations interact with other organizations in the environment. Therefore, it represents industries where companies interact more or less frequently-for example, a relatively low-interaction industry is that of dairy products and a relatively high-interaction industry is that of smart phones. When this parameter is set to 1, each agent screens the eight possible positions in the grid immediately around itself; when set to 2, there are 24 positions; when set to 3, there are 46 positions, and so on. The higher the range of interaction, the greater the opportunity for agents to establish links with other agents because the number of positions in the surroundings grows exponentially with every integer increase. The range values are set to 3, 6, 9, and 12 . Links between agents are established on the basis of a random number that has the range as its upper limit. For example, if range ¼ 6, then each agent connects to other agents that are approximately located at a distance given by a randomly selected number between 0 and 6. This parameter provides each network of organizations with different configurations. It further allows for changes in innovation diffusion on every run of the simulation, hence capturing the basic uncertainty of the diffusion process . Each agent is allocated a value of innovation; this value represents the extent to which the agent is inclined toward an innovation. This represents a general attitude toward innovation behaviors that an entrepreneur or a combination of decision makers in every firm has. Since our model represents an abstract diffusion of innovation without adhering to a particular function derived from existing data, values are initially assigned to agents using a random normal distribution % N . Our simulation features an overall precondition and two rules in which agents can adopt an innovation. In order to qualify as potential adopters, agents should be positively disposed toward it. This is given by the general innovation attitude that exceeds the 84th percentile of the distribution-this is the value 1, that is, at setup, equal to mean þ standard deviation . At the beginning of the simulation, these values represent Rogers's early adopters in the way specified below by the affiliation rule. When the precondition above is met, then each agent considers the network of close peers around it and compares their attitudes toward innovation to its own. If innovation exceeds the 84th percentile of innovation in the local network, then the agent adopts the innovation. This affiliation rule is designed to make adoption conditional on the dynamics relative to the local network, so that it represents the likelihood that each firm-agent is exposed to a proportion of peers as opposed to the entirety of agent-firms in the environment. Given the affiliation rule is particularly strict and it serves to define early adopters, the simulation includes a second rule based on agents' threshold levels , distributed according to a random uniform distribution with [0, 1). The threshold is a value such that when the number of locally available adopters is higher than the threshold value of the given agent, an innovation is adopted. A threshold is each agent's level of sensitivity to innovation. We conceived the level of sensitivity to innovation to be socially constructed through social linkages . The number of adopters considered by each agent are those around the range, as it is specified below. The affiliation rule works together with the threshold rule, in the sense that adoption may occur when the latter is satisfied, when the former is, or when both are. Screenshot of the diffusion model on NetLogo 3D Version 5.2 . Note: "Case" organizations are white colored and are larger while "other" organizations are yellow; when the organization adopts an innovation, the node shows red. Our simulation allows agents to be more likely to align their innovation attitudes to those of the case members. This is set to represent a dynamic adjustment of decision makers' attitudes that gets progressively close to those of the firm of reference. A case member is considered the equivalent of the focal firm in a business ecosystem or a supply chain . In other words, we assume that the multiple case researcher selects firms to be included in their study, partly because of their relatively influential role in the network. At every step of the simulation, noncase firms increase or decrease their innovation attitudes on the basis of: ASK noncase firm; D inn ¼ Inn C À Inn O ; SET Inn O ¼ Inn O þ D inn x D; where Inn C is the innovation of the case to which the noncase firm is connected, Inn O is the innovation attitude of the noncase firm , and a discount factor D is distributed uniformly at random between [0, 1). This factor D is necessary to allow for some noise in the adjustment, so that it is not excessively automatic. The pseudocode above operates to converge the inner level of the network around the case toward the attitudes of the case. The parameter mode refers to the two types of comparative case research designs: the ego mode and the com or complete mode. Under the com mode, the number of cases is randomly allocated in a 3D space and connects to other members who are close depending on range settings. Members are free to connect solely depending on their location. They may end up being connected to more than one case, hence serving as connectors or bridges between different cases. Under the ego mode, the noncase organizations tend to converge over the cases that are the closest. Thus, dependency among cases is less likely to occur because the network is more concentrated around cases. Finally, our simulation represents a multiple-case study designed by a researcher to study the diffusion and adoption of innovation as the main outcome . We thus study the extent to which the degree of case dependence has a bearing on the outcome of the researcher's interest. Due to bounded rationality, the researcher makes assumptions and selects cases on a possible ego or com mode but does not have complete knowledge of the world; hence, there is uncertainty about the cases that are actually ego or com dependent. To do so, our model runs with different configurations of parameters. We also use two measures of adoption. The first measure takes the case as a central point in the analysis; hence, we exclude the influences from members who simultaneously integrate two networks . The second measure takes into account all members contributing to the diffusion of innovation, regardless of whether they are part of more than one network . --- Simulation Procedures The configuration of parameters produced a factorial design of 4 Â 5 Â 2 for a total of 40, where range assumed four values, the number of case agents varied through five values, and there were two modes. We computed innovation for the ego mode and the com mode. This approach was desirable to capture the researcher's assumption of the degree of case dependence versus the actual degree of case dependence. We investigated whether the ego mode assumption made by a researcher was more suitable when studying networks that were structured in a way that seemed to be more in line with the ego mode and, vice versa, whether the com mode provided more robust results when the com mode was assumed by the researcher. Time was modeled explicitly by the so-called tick or steps . A good approximation in the case under analysis was that of days; thus, every tick represented a day. After a sensitivity analysis, ticks were set at 300 to cover approximately one year. In comparative case research, a full year is also suitable-it allows enough time for the diffusion to occur-and feasible-it allows the tractability of the data analysis. Multiple-case research often deals with multiyear time windows. Before launching the main experimental design, we engaged in a series of manipulation checks 11 to ensure that the assumptions were coded properly in our ABM simulation. They included an analysis of the characteristics of the network under "ego" and "com" modes, performed with measures of social network analysis . Namely, we used the mean path length between cases and the clustering coefficient . The "ego" and "com" modes differed significantly. We also did a preliminary study of graphical and analytical effects of parameters on the outcome variable. An interesting outcome of the analysis was that, at the upper bound of range , it is very difficult to establish whether a network is formed out of an "ego" or a "com" mode. However, this also makes that condition particularly interesting to observe, hence we decided not to exclude this element. Once settled on a number of conditions for the simulation and in following Snijders and Steglich's encouragement to combine inferential statistics and ABMs, we used statistical power analysis to estimate the number of runs for each configuration of parameters . Assuming that the effect size was small and setting the standard for the two statistical error types a and b at .01 and .05, respectively , we used the equation of Secchi and Seri and found that 130 runs reduced the potential incidence of Type II errors. In total, we had 5,200 runs. --- Results and Interpretation Results In the simulation, we explored the degree of case dependence by examining the effect of the model parameters on the diffusion of innovation, measured by the number of case organizations that adopted the innovation. Cases in the ego mode are largely independent of each other , while cases in the com mode feature high case dependence. Figures 3 and4 show an overview of the simulation findings. The y-axis of each plot is calculated on the difference between the proportions of case organizations that adopt the innovation subtracted from the proportion of other organizations in the system that also adopted. When the proportion of innovative case organizations is the same as that of noncase organizations, then there should be 0 ; when case organizations innovate at a rate higher than the other organizations, then the value should be above 0, and below 0 in the opposite situation. This measure repeats at every step of the simulation. The graph's lines show mean values over the 130 runs per each configuration of parameters . Figure 3 presents data on ego and com modes when there is low industry interconnectedness , while Figure 4 shows data when industry interconnectedness is high . The difference between the ego and com modes for low industry interconnectedness is remarkable. While there is a wide difference in adoption rates in the ego mode , there is relative convergence over À0.010 and 0.005 for organizations in the com mode. Configurations where there is a wide difference are those with a limited number of cases . In the com mode, if one excludes early oscillations for a number of cases of 2 and 5, the data pattern converges. Figure 4 shows an important insight. Under high interconnected industries , case and noncase organizations display similar patterns of technology adoption. However, there is a meaningful difference when the researchers draw on two cases, with an approximately 0.030 difference in the ego mode and approximately 0.020 in the com mode. The dependence between case organizations is a function of interconnectedness in the industry in which organizations operate. Although informative, Figures 3 and4 only provide a general overview of findings. To delve deeper into the results, we perform random coefficient panel regressions. This analysis aids us in comparing com and ego case organizations with their immediate and more distant network structures. Tables 4 and5 show the results of four models. In these models, the range is 3, 6, 9, and 12 per each mode when the case organizations are 2, 10, and 20, respectively . The outcome under analysis is the number of case organizations that adopt an innovation. The independent variables are the mean and standard deviation of innovation levels for case organizations and the number of other organizations that adopted in the inner network , in the second indirect order , and in the third . By including inner and outer circles, these regressions show the extent to which case adoption is a function of the actual network configuration and not just the other case organizations the researcher includes in the study. In Table 4 and Table 5 , a low number of cases always relates to an increase in the difficulty of explaining changes in the outcome variable. In fact, in this condition the R 2 is always the lowest for all regression models. This result is understandable because, with the number of the outer organizations kept constant, as the number of cases increases, the opportunity for contact and interaction also grows. While Table 4 shows that the estimated b coefficients show a satisfactorily low p value , on average, what happens in the inner network almost always has a larger effect than the outer circle on the outcome variable. There are exceptions in Model 1 and in Model 2, where the second-order network has a slightly more relevant impact . These regression models, when considered together, present results that may seem puzzling, at least at first. In fact, in both modes, adoption is a function of adoption in the system independent of whether these are considered ego or com. However, when a small number of case organizations are selected , the explanatory power of the regression models usually declines significantly. In the com mode, there is a gradual leap from models to models . In the ego mode, this leap is consistent as one moves from models to . In spite of low dependence, when cases increase five times, systemic interactions between the case organizations and the other organizations in the ecosystem drive most of case adoptions. When organizations are more closely connected , we find that the difference between assumptions of dependence or independence of case organizations is marginal. 1: Range ¼ 3 Model 2: Range ¼ 6 Model 3: Range ¼ 9 Model 4: Range ¼ 12 nc ¼ 2 nc ¼ 10 nc ¼ 20 nc ¼ 2 nc ¼ 10 nc ¼ 20 nc ¼ 2 nc ¼ 10 nc ¼ 20 nc ¼ 2 nc ¼ 10 nc ¼ 20 Range ¼ 3 Model 2: Range ¼ 6 Model 3: Range ¼ 9 Model 4: Range ¼ 12 nc ¼ 2 nc ¼ 10 nc ¼ 20 nc ¼ 2 nc ¼ 10 nc ¼ 20 nc ¼ 2 nc ¼ 10 nc ¼ 20 nc ¼ 2 nc ¼ 10 nc ¼20 --- Interpretation We use the simulation results to reflect on the degree of case dependence for theory building using comparative case research . Our findings suggest that dependence is not secondary when looking at the diffusion of innovation. Arguably, prior research details additional mechanisms that may affect the way in which an innovation propagates in a network but, if a network between firms is in place, dependence seems to be an essentially useful research frame. To help interpreting the results in a clear way and to connect them more directly to comparative case research, we use a fictitious situation in which a qualitative researcher approaches a study of 2-50 cases. As a fictional expedient to help our discussion, let us assume a researcher is conducting a study of the diffusion of innovation management techniques in startup organizations . In one scenario, the researcher has no interest in the degree of case dependence from a methodological or a conceptual viewpoint; thus, knowledge about linkages between case agents is irrelevant. Alternatively, the researcher might be interested in case dependence as a central aspect of building general theory. The researcher may have made assumptions about the degree of case dependence beforehand and selected the cases accordingly. The researcher accesses the range and innovation diffusion rates of the case members , together with the number of cases in the study. However, because of her or his bounded rationality, the researcher is not fully aware of other agents in the local network that may affect the case agents, especially in the com mode. The second scenario represents a typical instance where the researcher has limited ability to identify optimally, or sometimes even satisfactorily, the degree of case dependence in a complex empirical setting. There are four alternatives: 1. The researcher assumes high case dependence, and the cases relate to each other . 2. The researcher assumes low case dependence, and the cases do not relate to each other . 3. The researcher assumes high case dependence, and the cases do not relate to each other . 4. The researcher assumes low case dependence, and the cases relate to each other . We first discuss alternatives 1 and 2. These alternatives refer to instances in which the researcher's assumption matches the social linkages between cases. Hence, the main issue concerns the interpretation of findings. The researcher might identify propositions that take account of the high degree of case dependence. In a study of two start-up organizations with moderate network extension in the com mode, the researcher might observe the following: Proposition 1a: Higher levels of innovation result from interconnectedness between start-up organizations, which explains innovation in these start-ups. This finding is consistent with the sociological explanation of innovation diffusion as a process based on ties among members . From our results above, we know that interconnectedness is all the more relevant as the entire industry leans on these connections . Taking advantage of in-depth analyses, the researcher would then further examine how innovation actually diffused in these two start-ups. In contrast, in the ego mode , the researcher may interpret the following: Proposition 1b: A higher degree of innovation results from environmental pressures to which organizations react, and this reaction explains innovation in these start-ups. Unlike Proposition 1a, the claim in Proposition 1b stresses a contingency perspective, specifically how institutional pressure influences innovation diffusion . An in-depth comparative case study has the advantage of specifying how environmental pressures operate. In the com mode, the variation in the innovation average depends on the two connected case start-ups. Using the regression findings above as a basis, Figure 5B focuses on exactly this point by specifying that the innovation levels of the two case start-ups increase as the diffusion of innovation of the other organizations in their networks also increases. The regression lines are similar to those in the ego mode. Hence, in this hypothetical theorizing, either proposition displays low accuracy since innovation derives from the synergic effects of the inner and outer circles that jointly affect the innovation levels of the case start-ups. However, while in the com mode, these networks connect the two case start-ups, in the ego mode, the two case start-ups are unrelated because they share no social linkages. Assuming low dependence between case start-ups, the increasing levels of innovation accrue from pressure that other start-ups in the network exert on the case start-ups. As shown in Figure 5A, the adoption of innovation by case start-ups is likely to be determined by the mean innovation levels of the case studies. Moreover, the inner and outer network circles seem to relate to the way the case start-ups adopt the innovation. We then expand the example above to comparative case research with a low number of cases. The first general implication of our simulation for comparative case research is as follows: Implication 1: Researchers should always take into account the linkages between cases when building theory, specifically when drawing on a small set of cases . We now turn to alternatives 3 and 4 from the fictitious example above. These alternatives represent instances of a mismatch between the initial assumption made by the researcher and the social linkages among start-ups. Under alternative 3, a researcher would derive a conclusion similar to Proposition 1a, while under alternative 4, the researcher might propose a conclusion that is similar to Proposition 1b. However, either proposition would be based on an erroneous assumption about case dependence. For the sake of illustration, we take a potential comparative case study, this time with range ¼ 12 and case ¼ 20. Still building on the simulation results, Figure 6B shows regression lines for innovation levels and their impact on diffusion in the com mode, while Figure 6A provides the same information for the ego mode. In the latter, the slope of the lines is mild, where the innovation level of the case start-ups is unrelated to the outer and inner network circles. The case startups are actually disconnected from the network, as the combined innovation levels do not influence innovation significantly. In the com mode, the slope is positive and steeper than in the ego mode. Moreover, the direction of the outer and inner network circles' innovation is consistent with that of the case start-ups. However, the slope is particularly small, suggesting a minor impact. Based on our experimental findings, we argue that the social linkages between cases are essential for researchers to understand and analyze them as part of the process of building theory from multiple cases. The way in which an innovation is adopted is a function of the inner and outer networks that may link cases together. Hence, whenever theoretical proposition is developed from cross-case comparisons, the question of accesses to information about linkages between cases emerges. Still, there are instances where limited knowledge about case dependence is less problematic. The second implication of our study is as follows: Implication 2: When the number of cases increases , the assumption of a low degree of dependence is more accurate than assuming high degree of dependence between case firms. The experimental results show that when the researcher addresses a limited number of cases, the com mode should be the preferred choice regardless of how low or high interactions among cases are. If adoption depends on the average level of innovation in other cases, then high case dependence is more appropriate when either the com mode or the ego mode is assumed by the researcher . Instead, the ego mode assumption should be the preferred choice when the range of interactions is high and research is conducted with a relatively high number of cases. The relatively flat lines depicted in Figures 5 and6 suggest that the innovation levels of other cases do not influence the percentage of adopters. Another implication-a corollary of Implication 2 above, perhaps obvious to well-versed case study researchers, is that there is enough information on the dynamics in the industry already with 20 cases . In practice, researcher often have access to a few cases, but it is precisely when researchers draw on a few cases that information about case dependence is most valuable. Incomplete information about the degree of case dependence might undercut the specification and reporting of newly developed theory. Given the above evidence, pertinent questions arise: What are the implications of these results for theory building from multiple cases? How can researchers address the degree of dependence in future studies? --- Discussion The comparative case design is instrumental to aid researchers with building theory, but the role of case dependence in theory building appears to be surrounded in confusion in social science methods. To address this shortcoming, we study whether social linkages between cases when building novel insights through comparison and contrast of cases might influence the veracity and transferability of findings. We develop an agent-based simulation based on the survey of comparative case research on innovation diffusion. At a general level, we find that the number of cases and industry characteristics are important factors to take into account when making claims about newly developed theory from comparative case research. The focus of our analysis is the extent to which researchers' assumptions about case dependence match their own methodological requirements. Taking innovation diffusion as canonical example where social linkages are a leading source of case dependence , we argue and show that the neglect of the degree of case dependence in theory building is problematic. As we elaborate below, first, it posits threats to internal validity by undercutting within-and cross-case comparisons used to support theoretical propositions. Second, it hampers transparency in research about whether the degree of case dependence is a critical factor for the generalization or transferability of newly developed theory. Our contributions concern the design and reporting of comparative case studies. Whether this study's findings based on a general computational model of innovation diffusion are relevant to other topics in the social sciences presents a fruitful direction for future research. The central issue about the researchers' awareness of the degree of case dependence, of varied types, remains a valid concern for those interested in theory building using comparative case study. --- Innovation Diffusion: Why and When Does Case Dependence Matter? One of our article's main contributions is a clarification-through a novel use of an ABM-about whether case dependence affects new theory about innovation diffusion against commonly held standards of internal validity in social sciences. Extending discussions in social sciences , we show that the degree of case dependence may undercut the internal validity of findings about innovation diffusion. Failure to acknowledge the degree of case dependence might lead to overstating the empirical evidence in support of or against a newly developed theoretical proposition. The issue at stake is one of the validity of claims rather than a prescription aiming at restricting the diversity of ways in which theory is built from multiple cases. We heed the advice from past research that building theory from cases benefits greatly from embracing research approaches . We augment the literature by examining under which conditions the degree of case dependence, when unattended by the researcher, undercuts internal validity in innovation diffusion studies. We show that specific features of the comparative case design call for great attention by the researcher. Specifically, when the researcher addresses a limited number of cases , the researcher's assumption of low case dependence is preferred to build theory about innovation diffusion regardless of the range of interactions among the cases. However, the researcher's assumption of high case dependence is preferred when the range of interactions is high and research addresses a relatively large number of cases. The influence of case dependence for understanding innovation diffusion is particularly salient in industries where organizations are known to be highly interconnected. Therefore, and making a direct link to our simulation , researchers might learn about the implications of case dependence by, for example, taking advantage of in-depth interviews to identify the basis of the individual's attitude for innovation in case firms, examples of imitation by organizations in the environment, and organizations operating in a proximity radius to the case firm. --- Implications for Comparative Case Research and Its Variants in Social Sciences We put forward an original proposition that the issues of case dependence are better understood as a researcher's bounded rationality problem in complex organizational phenomena ). Researchers' bounded rationality is not a problem in its own right. Rather, the concern is the possibility that the researcher's bounded rationality about case dependence cases distorts her or his own assessment of the cross-case evidence in support of a theoretical proposition. It involves a mismatch between the reported versus actual states of innovation diffusion. We identify two archetypes of comparative case research based on past research: the complete research design and the ego research design. The complete research design denotes a high degree of case dependence, while the ego research design entails a low degree of case dependence. The archetype of comparative case research used by the researcher should be made explicit in future studies. Beyond adding new terms, we argue that the discussion of the degree of case dependence is informative about issues of internal validity and external validity. The exact implications of the researcher's bounded rationality vary according to the variants of comparative case research. Figure 7 shows when researchers' bounded rationality about the degree of case dependence posits a threat to validity in inductive research. Our study specifically focuses on the mismatch between using a complete case design while evoking an experiment-like analysis of across cases to support new theory propositions. An experiment-like approach has been described as following computational reasoning in social research . However, the discovery of data patterns though eliminative logic might be problematic because possible duplicates across cases are overlooked by the researcher . These duplicates stem from meaningful cross-case dependence. By meaningful, we mean social linkages that have a bearing on the research question and the subsequent answer. However, as shown in Figure 7, the authors have also suggested that "we don't discover theory, we create it!" . This approach to inductive research celebrates the role of the researcher in theory building. The warrants of internal validity are based on researchers' ability to uncover novel findings, as opposed to "experimental" comparisons between cases Note: This study focuses on the mismatch between the use of a complete case design while evoking a computational logic to build theory about innovation diffusion . Under this approach to theory building, the researchers' bounded rationality about the degree of case dependence may posit a threat to validity to newly developed theory about innovation diffusion and, more broadly, organizational phenomenon closely related to diffusion processes . We use ideal types to enhance the communication of the key insights. Our representation is illustrative. Inductive research remains a process of discovery where creativity and transparency are essential. . Inductive research is viewed as a cooking without a recipe in process of discovery in which creativity is preferred . Still, showing the chain of evidence is desirable in qualitative analysis so that the reader can fully appreciate the quality of the data and conclusions put forward . Our study has implications for external validity / transferability of findings. The degree of case dependence often entails a boundary condition that, if acknowledged, can actually strengthen the newly built theory. As shown in our review of the literature, the readers of comparative case research often have difficulty in accessing information about whether low versus high degree of case dependence influences the extent to which findings being reported can be transferred to other settings. Clarity about the constructs and their relationships, and the boundary conditions are essential features of a theoretical contribution . Our study further extends the current debate on transparency in comparative case research . We caution against practices of declaration. That is, the use of statements such as "we follow the procedures in X paper" is no substitute for an explanation of the analytical procedures that the research followed to draw conclusions from comparative case analysis. Furthermore, we present guidelines for dealing with the degree of case dependence across the main stages of building theory in comparative research design. Our guidelines furnish existing best practices in qualitative research with specific suggestions to reflect the findings of our computational simulation . When not blindly applied, these guidelines do not preclude creativity in building theory from cases or neglect the role of the researcher's interpretation during the research process. In fact, researchers' cognitive processes during theory building encompass the discovery of new theoretical insights that would otherwise be overlooked . The values of veracity and transparency have long been at the core of comparative case research in social sciences. --- Limitations and Further Research As with any study, ours has limitations. Theory building follows many processes that cannot be fully captured in a computational simulation. However, our novel use of such a simulation was advantageous to allow a "virtual experiment" . A virtual experiment was needed to manipulate the degree of case dependence, using a few generic mechanisms of innovation diffusion, which is not possible in an empirical setting in terms of qualitative research. More generally, we envisage opportunities to use agent-based simulations in innovative ways to study important questions in sociological methods and research. We examined a specific variant of the comparative case design: the multiple-case study. Our approach was desirable because the role of crosscase analysis, where cases are treated as "experiments," has been articulated in past research . Thus, our suggestions apply primarily to multiple-case study research that uses case comparison as a prime basis for deduction. Future research should explore the issue of case dependence for other variants of this research design. The impact of case dependence might depend on whether researchers use comparative case studies for theory building versus theory elaboration purposes . We also call for future studies that examine researchers' common practices in the research process, thus adding insights into the richness of practices of scientific knowledge production in comparative case design. While analyzing the extent to which ego and com network structures affect diffusion processes, we found that SNA provides a wealth of information about characteristics of the networks linking and surrounding the case organizations. Future research may further explore these differences, with particular attention to the mechanisms of diffusion other than those based on the threshold rule and affiliation rule. In doing so, future research will extend the debate about social linkages-one we started to address-as well as other sources of case dependence in the context of comparative case research. Furthermore, we envisioned opportunities for research that specifies types of innovation and empirical settings . Our simulation treats cases firms are homogeneous entities, so the exact importance of case dependence might vary according to, for example, the type of innovation and the sector of each firm. There are bountiful opportunities to extend our research about case dependence and theory building. --- Conclusion Researchers often justify the use of multiple cases based on comparisons between "experiments" that, in turn, enhance internal and external validity of the new theory. However, researchers face limited ability to identify optimally, or even satisfactorily, the degree of case dependence in complex empirical settings. Drawing on research on innovation diffusion, we created a computational simulation to learn about the implications of the degree of case dependence in building new theory. Our experimental results show that the misfit between meaningful social linkages across cases and researchers' assumptions about such linkages undercuts theory. Such concern is more salient when research uses a few cases to develop theory about highly interconnected settings. Then, as illustrated in our introductory quote, many cases that appear to be distinct might just be duplicates of the same original. The central issue is not whether Galton's challenge is relevant, but when it is relevant and how it should be accounted when building theory from cases in social sciences. and anthropology . However, these discussions are not concerned with the implications for theory building or guidance for future researchers, all of which are issues we deal with in this article. 5. We are grateful to an anonymous reviewer for suggesting this label to better specify the purpose of our simulation model. 6. We focused on the following management journals: Academy of Management Journal, Administrative Science Quarterly, Journal of Management, Journal of Management Studies, Management Science, Organization Science, Organization Studies, and Strategic Management Journal . For leading journals in innovation, we selected Research Policy, Technovation, and Journal of Product Innovation Management. 7. In a review of case study articles in the field of operations management, Barratt, Choi, and Li found that "sufficient details in research design, data collection, and data analysis were missing" in many articles. 8. We use the literature on innovation to specify the simulation parameters. An examination of whether the findings from a particular study hold true for different degrees of case dependence is outside of the scope of this article. 9. The computer code for the simulation is available from https://www.comses.net/ codebases/5131/releases/2.0.0/. 10. Our study focuses on comparative case research using a small number of cases . For the analysis of more cases , qualitative comparative analysis is a suitable method . 11. These checks fall into what most agent-based modelers call "calibration" as it "allows one to set model parameters to certain values that are chosen according to prevailing theoretical and/or empirical evidence" . These analyses are available in the Online Supplementary Material. --- Declaration of Conflicting Interests The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. --- Supplemental Material Supplemental material for this article is available online. --- Notes 1. An exception is the Special Issue on "Using qualitative evidence to inform the specification of agent-based models" edited by published by Edmonds and published in the Journal of Artificial Societies and Social Simulation. 2. We are interested in case dependence. If cases are independent, then case duplicates are not a concern because "each case serves as a distinct experiment that stands on its own as an analytic unit" . 3. For the sake of clarity, we distinguish "cases" from "observations." Unlike cases, observations refer to data points. Analyses of observations tend to follow deductive reasoning and are distinct from inductive reasoning, which is the focus of our discussion . 4. Prior discussions about case dependence have also occurred under the label of "Galton's problem" in the context of cross-national studies in political studies --- Author Biographies Nuno

Researchers increasingly take advantage of the comparative case design to build theory, but the degree of case dependence is occasionally discussed and theorized. We suggest that the comparative case study design might be subject to an often underappreciated threat-dependence across casesunder certain conditions. Using research on innovation diffusion as an illustration, we explore the role of social linkages across cases when building theory through comparison and contrast between cases. We develop an agent-based simulation, grounded by comparative case research about innovation diffusion, as novel way to study the implications of case dependence in theory building using multiple-case study research. Our simulation results suggest that the degree of case dependence has a nontrivial bearing on innovation diffusion experienced by case entities, specifically when the researcher draws a few case entities operating in a highly interconnected industry. Under these conditions, overlooking the degree of case dependence might weaken newly built theory against commonly held standards of

The population in America is changing. According to the US Census Bureau, the Hispanic population has grown by 43% from 2000 to 2010, and the United States continues to become more ethnically diverse. 1 However, the diversity occurring in the United States is not changing as rapidly in the medical field. The Association of American Medical Colleges defines people underrepresented in medicine as those belonging to racial and ethnic populations that are underrepresented in the medical profession relative to their numbers in the general population. 2 The number of underrepresented minorities in medical schools is not increasing, although it is in the overall population. 3 There are fewer African American men in medical school today than there were in 1978. 4 It is important to have a diverse pediatric academic medicine population to address the diverse needs of the pediatric population. 5 An academic clinician is defined as a fulltime university faculty member who conducts teaching and/or research along with his or her clinical work. 6 There has been an increase in the last half century of URM and women faculty at US medical schools. 7 The number of faculty who are URMs has increased at US medical schools from 6.8% in 2000 to 8% in 2010. 8 However, URM and women faculty are not being promoted or are leaving their faculty positions at higher rates compared with their white male counterparts. 9,10 As America diversifies, the medical field should respond accordingly to address these changes. According to the American Academy of Pediatrics, "…the provision of culturally effective care and enhancing the diversity of the pediatric workforce represent parallel and often overlapping initiatives to improve care for pediatric patients." 11 There are some programs to help increase the number of URMs in academic medicine. For example, in pediatrics, the Academic Pediatric Association operates the New Century Scholars Program for URMs, although this program is focused on residents and not medical students. 12 There have also been studies which found that mentors have a positive influence on residents in their decision to pursue academic medicine. 13 In a survey conducted by the Latino Medical Student Association, the Student National Medical Association, and the American Medical Association in 2010, 58% of the students reported an interest in academic medicine. 6 However, 40% believed that racial and ethnic minorities have a harder time succeeding in academic medicine. 6 URMs who are medical students may not identify themselves as academic medicine candidates because of a lack of URMs who are role models. Social identity is defined as how individuals within a certain social group behave and think within that group. 14 This differs from the notion of professional identity, which can be defined as the specific expertise, knowledge, or skill set that members of a certain profession possess. 15 Both definitions are important to ground this research because professional identities are socially defined, and the link between the professional and overarching social identity can provide the anchor with which an individual identifies with a specific role. It is important to understand how individuals identify themselves within the overall society and their profession. To our knowledge, there is no literature on the social and professional identities of URMs in academic pediatrics. There is, however, some research on the professional identity of minorities in other professional settings. In a study conducted at community colleges in California, it was found that being a minority faculty member can complicate the professional identity of the faculty member. 16 This notion of fitting in with the majority group can be difficult for minorities and creates an internal conflict for them. Our purpose for this study is to explore the experiences of faculty who are URMs at 2 urban medical centers, in particular, to examine how URM faculty decided to pursue academic pediatrics and to examine their social and professional identities as academic pediatricians. --- METHODS Institutional Review Board approval from Children's National Health System was obtained. For purposeful sampling to recruit participants who could contribute to our study goals, an e-mail was sent to the pediatric faculty at CNHS and Howard University College of Medicine , notifying them about a qualitative study on the experience of faculty who are URMs in academic pediatrics. Announcements were also made at CNHS division meetings and at the HUCM pediatric division meeting. All participants provided informed consent; although all had an option to voluntarily withdraw at any point in the study, none withdrew. Each participant received a $25 gift card. In 2016, the 3 focus groups that were conducted included URM faculty from CNHS and HUCM. Moderators were selected and trained by the medical education department at CNHS and were provided with the question guide . The moderators were both URM and non-URM people. Focus groups were ∼60 to 90 minutes in length. The second phase of the study was conducted in 2017 and consisted of 1-on-1 interviews in which the primary author queried URM faculty at CNHS for a more in-depth understanding of their subjective experiences in pursuing academic pediatrics and of their professional identity development. Thematic saturation was reached with the 10 interviews. The faculty were recruited via e-mail. The questions were used to ask how professional identities of URM faculty developed over time and what factors influenced URM faculty to become academic pediatricians . Each interview was ∼60 minutes in length. All focus groups and interviews were audio recorded and transcribed by a third party who signed a letter of confidentiality. Transcripts from the focus groups were hand coded, on the basis of the methodology developed by Miles et al, 17 by each author independently; consensus was reached among researchers during a series of meetings to identify key themes. Demographic and additional data from the focus groups were collected and managed by using Research Electronic Data Capture, which is hosted at CNHS. 18 The primary investigator coded the 1-on-1 interview data; members of the research team then verified codes and themes. --- RESULTS A total of 15 faculty participated in the focus groups, and 10 faculty from CNHS participated in the 1-on-1 interviews. Demographic data are displayed in Table 1. Eighteen of the participants were women, and 19 indicated African American as their race . Fifteen of the participants were fellowship trained, and a majority were trained at the assistant professor level . Seven themes were identified from the focus groups and 1-on-1 interviews . --- Theme 1: Definition of Academic Medicine Three components of being an academic pediatrician mentioned were education, research, and clinical responsibilities . A majority of the participants had an area of interest in education and advocacy . One participant stated, "I see it as three arms. There's research, education as well as administrative duties, all of which I love. I think that being in pediatrics you're a natural teacher and educator anyway." --- Theme 2: Being a URM in Academic Pediatrics The faculty felt a connection to their patients because they had similar backgrounds or through speaking the same non-English language. For HUCM, 75% of full-time faculty were URMs in 2016-2017. The percentage of URM full-time faculty for CNHS in 2016-2017 is not known. One of the Hispanic participants discussed an interaction with a Spanish-speaking patient at the hospital: …if I say, "Good morning" in Spanish or "Do you need help?" and, all of a sudden, it's like that sign of relief, "we've been looking for this clinic…"...and you end up a lot more involved in peoples' ways than you would if you didn't speak or if you didn't identify with the person. URM faculty also identified barriers in academic pediatrics for them, including the lack of other URMs in positions of leadership, being part of a low percentage of URMs practicing in academic pediatrics, and the impact of racism and implicit bias in the medical field . These experiences impacted their identities during the transition from resident to attending physician. I was the fellow on and my attending who also was an underrepresented minority, he was also black…this kid came in…and when I went in the room that's when he used the N word and said that no "N*****" was going to put a-you know touch his child and he wanted to see the senior doctor. I like left and the senior doctor, who also happened to be black, came in and was like these are your options for tonight. You don't really have anybody else and I can guarantee you that you know "Dr X [name changed]" is the most skilled person in the hospital right now to do this procedure. And the son actually The focus group demographic data were collected and managed by using Research Electronic Data Capture tools hosted at the Clinical and Translational Science Institute at Children' s National . 18 a Participants were able to select all that applied. said dad please just let "Dr X" do it… and that experience was one that was hard for me to sort-I had never been called the N word at least to my recollection, or to my face up until that point…more angry than anything, but it also empowered me that like I'm going to get this line, these lines in, because I need to show him that I'm pretty competent. I don't know why I felt like I needed to prove that to him but I definitely felt that way. Definition of academic pediatrics "I guess a more diverse portfolio than your direct clinical care such that you not only are in an academic setting, which is all I was trying to be, but that you're contributing and so-and that may be contributing through research, through education, but contributing to the academic community, adding knowledge to your field, improving practice in your field, it' s kind of like the public health equivalent I guess to your practice. It' s not just the patient in front of you but impacting your patient' s people by doing a larger scale through your academic work." Being a URM in academic pediatrics "I think when they walk into a room and see us, I think even as a child for me, when you know when you see people that look like you and the things that they're able to do…" "I feel it' s a burden mainly because it' s an honor to be at the place where I am now where I'm actually at the point where I am in a position to hire people and to influence my junior faculty and especially my trainees, but I do feel as if it is a burden in that it' s my job to make sure that people are aware of their own implicit biases and how they treat people who don't look like them. So an example of that is that in our unit we have a lot of families who are of low socioeconomic status they tend to be African-American or they tend to be underrepresented minorities, I would say more so than Caucasian and what I often hear my trainees say or I'll hear faculty say is like, 'Oh that parent, that parent doesn't care about their child, they're like never here.' And I feel like it becomes my responsibility to remind them that that family actually has a lot of burdens that doesn't allow them to be but that like every other parent they want to be here for their child but if they live in a shelter they need to get back home to their shelter or they lose their spot. If they don't have an extended family who are going to take care of all their other children they need to make sure that their other children are safe and so I feel like it' s often a burden to make sure that I provide an alternate voice for black people all the time." Advancement in academic pediatrics "I don't know that my values line up the same as what would consider me to advance in academia. What' s important to me doesn't always line up with what' s important for promotion." "I think knowing about the professor-level underrepresented minority folks we have here is a little scary because I see how hard they work. And, to be quite selfish, some days I'm just like, 'I don't think I want to work that hard, I'm tired.' I'm at a level where my family has said, 'Okay, she' s a doctor,' I sort of got to their goals and I don't really need to impress anybody else. So, really, it is a challenge because there' s already a burden to be sort of supportive of people coming up who are at levels behind me. And then, for myself to still be that pioneer, I'm just tired of the pioneer game where you're the only one in the room yet again. And, after a while, I'm sort of like, 'You know how much more energy do I have to keep playing this game?' And that really affects my decision about how much more in academia I need to go? I think I can fulfill what I want to be happy day-to-day with what I do now, potentially without ever having to go higher. So it' s a tough decision." Mentorship is important early in the career "…so there' s the top-down mentorship there' s also like the peer mentorship or peer encouragement I think is really important when we're talking about promotion you know one person finds out one thing and-or gets involved in some project and brings along their colleague or shares that information and I think that if people already have like good networks and I would think also that probably then more likely to be able to get involved in projects and programs that will lead to promotion…" "…but a Master Mentor, without question, and really opened my eyes to look at how my interests could become academic and so I think through that kind of mentorship that timing was everything for me, was afforded me at that time, it really allowed me to explore my interests and I probably was a little bit crazy with it in the beginning -you know-well I'm interested in this, this and this and this and I want to do this and oh and I need loaner payment too so maybe I should see if I can be a researcher so I can get some loaner payment." Financial burden and academic pediatrics "I'm actually straddling both worlds in that I work in private practice and in the ER here some on the academics' side teaching residents. The longer I do both, the more I recognize the financial rewards are just overwhelmingly more in private practice-lifestyle and finances are just better in private practice than on the academic side." Family influence on career "…I mean, I think in my community, my mom was one of the first…well she was the first to graduate from college in her family. So, when I was younger there was this big push on education during that time, sort of the second generation college graduates of folks of African-American descent. There is this huge push for education for sort of social-upwardly mobile, or even that want like better…that education is kind of the key or that traditionally what it has been. So, I think that foundation…I mean it was like education, education, got to get my education… so, pursuing a higher degree was probably inspired by my mother and like family and how I was raised with education being like that critical component of upward mobility. Gender and academic pediatrics "...the reason why probably there are more male providers than women providers in academic medicine is you still need to raise your family and my son and my daughter were sort of a more important priority for me so you know sometimes going you know part time in terms of 80 percent so I could be the classroom mother and do some of the field trips and be there to be the chaperones and those kinds of things, I did not want to miss and being a division chief was not on the bucket list so much as opposed to being able to say I spent some time with my family and not saying 'my wife is taking care of the kids' since I don't have a wife!" ER, emergency department. --- Theme 3: Advancement in Academic Pediatrics as a URM Despite being successful in their careers, some of the faculty felt as though they were not deserving of their accolades and acknowledged that they were experiencing imposter syndrome. 19 A majority of the participants in the study were assistant professors , and only 2 were associate professors . None of the participants were full professors. I didn't think that I deserved to be in medical school. I always wanted to be a doctor but when I got to college like my academics weren't my priority. I prioritized the extracurricular because I was told that that's how you get to the next level but I ended up doing so many extracurricular that my academics weren't solid as they could have been.… I didn't feel I was worthy basically so I always kind of came with this sort of chip on my shoulder like I need to bust my butt, I need to like work super hard because you know by the grace of God I'm in this institution.… I'm looking at my colleagues…and I'm like dang, I'm not as good as them.… You can feel from your colleagues as if you're not good enough or this whole idea of this Imposter Syndrome, but you have to realize that there's a reason you're here like you are worthy and special to be here in itself…. --- Theme 4: Mentorship Is Important Early in the Career Faculty felt mentorship early in the career was important. It was inspiring to see URM faculty in leadership positions in academic pediatrics. Faculty reflected and advised that medical students should find mentors early on in their career : "…seek out mentorship and mentorship comes in many different forms and your job is to be the best mentee ever, which means that you accept-the burden of the relationship falls on you." A key component to mentorship was having similar backgrounds. The mentor did not have to be a URM, but 1 female participant discussed having another woman as a mentor: …like meeting with some of our at least like female leadership has been helpful even if they're not underrepresented minorities.… I think within peds obviously there's the shift towards more women in the field, the leadership is still more male dominated so I think talking to some of the women at Children's has been really helpful to figure out you know just to talk to them about sort of like how people get promoted…. Theme 5: Financial Burden and Academic Pediatrics The participants discussed the burden of student loans and student loans being a deterrent for being in academic medicine. Participants discussed being paid more in private practice compared with academic pediatric medicine . It's expensive to be an academician, you don't get paid as much and underrepresented minorities, myself included, have a boatload of loans and last I checked none of this work unless you're doing 75 percent research qualifies for loan repayment and so unless you're a Ph.D. and maybe you have it that way. So I think that there's a financial barrier for those who bear the burden of their loans and push through the process is not always transparent or in an equitable way. Theme 6: Family Influence on Career Family had a major influence on why the participants became academic pediatricians . Family gave participants encouragement to enter the field of medicine. For example, one participant discussed having someone close to them suffer from HIV/AIDS: I had a lot of family and friends, and neighbors that were infected with HIV/ AIDS and dying so that was what drove me to med school. But I don't think I would have gotten anywhere near med school or college or anything were it not for my parents being like, "You kind of have to do this because we were the first ones in our family to do it and you must." Three of the participants' mothers were nurses, and 1 participant's uncle was a physician. Another participant discussed the impact of her grandfather being an African American physician in her community and her choice to become a physician: So I grew up in a small town and in that town there's probably-now there's about 60 000 people, but at the time that my grandfather practiced medicine I'm sure that it was less than that, and not only that the population of African-Americans in the town was probably less than ten percent and the one thing that resonated with me was everywhere we went, everyone said: "Are you Dr L's [name changed] granddaughter?" And he was a wonderful man and he did this for me, he did that for me. He died before I was born, but all these great stories of someone who had been gone by the time that I was there growing up for years and they still were talking about him, it just made me-I was just very intrigued that whatever he did, the passion that he had in medicine, that that was something that I wanted to do and so that was one of the reasons why I decided I wanted to be in medicine. --- Theme 7: Gender and Academic Pediatrics Participants discussed being women and minorities in academic pediatrics and how that affected their career choices. Some of the women felt a struggle with the work/life balance and also with having children and being physicians. One major theme is that pediatrics is a female-dominated field; however, most of the top positions in the field are held by men ( Family had a major influence on why the participants went into medicine. Mentorship was important, and participants advised that medical students seek out mentorship early in their career. Participants discussed not earning as much in academic pediatrics compared with private practice, and this can be a deterrent for URMs because of the financial burden of student loans. Most participants were women, and it was noted that although a majority of pediatricians are women, most pediatric chairs are men. Also, the participants with children discussed the struggle of the work/life balance and how that influenced some of their career goals in academic pediatric medicine. To our knowledge, there have not been any recent studies on the experience of URMs in academic pediatrics. A recent study suggests that some of the facilitators to becoming a URM surgeon, such as mentorship and family, are similar to those in our findings. 20 In addition to the novelty that our study offers, it also brings to the forefront voices of URM pediatricians and can inform future works. Our study was limited to pediatric faculty at 2 large medical centers in Washington, District of Columbia, which may not be transferable to other academic centers or other fields of medicine. Although we used purposeful sampling for URM faculty, recruiting through e-mails and meetings may have introduced selection bias. The lack of diversity in academic rank among our participants may have skewed our data toward the experiences of more junior faculty. Some participants may have felt a relationship to the investigator team, but we maintained anonymity throughout and feel the primary investigator was able to gain depth of responses through the individual interviews, which were discussed with the research team. --- CONCLUSIONS Although more research should be conducted on ways to enhance the URM experience and reduce barriers in academic pediatrics and other fields of medicine, mentorship and pipeline programs are key factors to help URMs in academic pediatrics. The evolution of the professional and social identities of pediatricians should also be further explored. Barriers that influence individuals, such as lack of minority leadership and promotion advancement, should be examined to help recruit more URMs in academic pediatrics. Without intentional efforts , workforce diversity will continue to lag behind our country's growing diversity, and the gulf between our patients' varied needs and what we offer will widen.

Our objective for this study was to explore the experiences of faculty in academic pediatrics who are underrepresented minorities (URMs) at 2 urban medical centers, in particular, the experiences that influenced their pursuit of academic pediatrics. METHODS: Three focus groups were conducted in 2016 with URM faculty from Howard University College of Medicine and Children's National Health System to explore how they were influenced to pursue academic pediatrics. Ten 1-on-1 interviews were also conducted in 2017 with URM faculty at Children's National Health System. Focus groups were coded and analyzed by the research team using standard qualitative methods. The 1-on-1 interviews were coded and analyzed by the primary investigator and verified by members of the research team. RESULTS: A total of 25 faculty participated in the study (15 in the focus groups and 10 in individual interviews). Eighteen of the faculty were women and 7 were men. Findings revealed that mentorship, family, and community influenced participants' career choices. Barriers for URMs in academic pediatrics included (1) lack of other URMs in leadership positions, (2) few URMs practicing academic pediatrics, and (3) the impact of racism and gender and implicit bias in the medical field. CONCLUSIONS: Mentorship and family are major influences on why URMs become academic pediatricians. Lack of URMs in leadership positions, racism, gender bias, and implicit bias are barriers for URMs in academic pediatrics. More research should be conducted on ways to enhance the experience of URMs and to reduce barriers in academia. WHAT'S KNOWN ON THIS SUBJECT: To our knowledge, there is no literature on the social and professional identities of URMs in academic pediatrics. There is some research on the professional identity of minorities in other professional settings. WHAT THIS STUDY ADDS: In this study, we examine how URM faculty from 2 urban medical centers decided to pursue academic pediatrics and their social and professional identities as academic pediatricians.

Introduction Successful diabetes care requires effective approaches to supporting behavior change of patients. Based on existing theories developed mainly in Western countries, a variety of behavioral and psychosocial interventions are implemented. Review of these interventions shows a philosophical foundation provided by ''empowerment'' of patients and a patient-centered approach enabling internal motivation to change [1]. Most theories emphasize changing patient behavior through the patient's own intention and ability [1]. The internally motivated ''Losing weight is really important to me.'' replaces the externally motivated ''My doctor wants me to lose weight.'' [2]. This concept is widely accepted and utilized by many certified diabetes educators [3]. However, external motivation may be more relevant in different cultural contexts. Recent evidence in cultural and social psychology indicates that substantial cultural differences exist in a globalized world. In different cultural contexts, people exhibit different ways of thinking, feeling, and behaving. [4]. In the past two decades, a line of research using experimental methodologies has formed a theoretical framework comparing Western cultures with Eastern cultures [5]. Western cultures are characterized by social orientation valuing ''independence'' or a model of agency that emphasizes self-direction and autonomy. In such cultures, people's internal attributes such as their own goals, desires and judgments form the predominant basis of their action [5]. This view is consistent with the existing theoretical foundation in diabetes intervention that focuses on the patients' internal motivation to health. In contrast, Eastern cultures such as China, Japan and Korea tend to place a higher value on ''interdepen-dence'' or a model of agency emphasizing harmony, relatedness, and connection with others. In such cultures, people tend to act in consideration of the expectations, desires, and needs of others as the predominant basis of their action [5]. In these cultures, patients with diabetes often may focus on the potential deleterious effects of their lifestyle change on the people around them. Patients' perception of the expectations of others may have more impact on the ability to change their lifestyle in Eastern cultures than is found in Western cultures. Indeed, results from social and cultural psychology indicate that these cultural differences lead to correspondingly divergent consequences in people's motivation and emotion. For example, North Americans are found to be more strongly motivated to maintain personal control while Japanese are found to be more strongly motivated toward relational harmony [6]. Attaining personal goals leads to enhanced well-being among European Americans while attainment of relational goals is more closely linked to enhanced well-being among Asian Americans and Japanese [7]. This emphasis on social relations is also shown in the result that East Asians' well-being is strongly predicted by social harmony, socially engaging emotions, and perceived emotional support from close others [8]. Perceived emotional support is a perception of receiving encouragement, compassion, and other forms of emotional support from the persons close to the respondent such as family members and friends. This result suggests that in an interdependent society people are more sensitive to the expectations of others. Similar results are shown in patients with diabetes. In Mexican Americans, who are characterized by a relatively higher tendency to interdependence than European Americans, social context reflected by the patients' perception that their family understands their diabetes is associated with patients' higher attention to self-care [5,9,10]. A family-centered approach in Taiwanese patients enhances patients' positive attitude toward diabetes [11]. The increasing rate of diabetes among Asians worldwide and Asian Americans calls for understanding the unique needs of diabetes care in Asian patients [12,13]. One notable qualitative study suggested that interdependence and reciprocal role responsibility additionally complicate disease management in Chinese American patients with type 2 diabetes [14]. The interdependent cultural contexts of East Asian society may be an important way to conceptualize the unique needs of Asian patients with diabetes. However, not all members of a culture internalize its values to the same degree; people within a culture vary in their relative agreement with interdependent or independent values [15]. Therefore, it is important to test for within-culture associations of individual difference in interdependence/independence values and individual variation in coping with diabetes. Psychosocial assessment is recommended in routine care of patients with diabetes; emotional well-being is associated with positive diabetes outcomes [16]. Diabetes-related distress is a psychosocial issue known to impact health outcomes: it is independently associated with self-management behaviors and perceived burden of diabetes and also predicts future glycemic control [17,18]. Interdependent social orientation in Eastern cultures might therefore play a role in other psychosocial aspects of diabetes care. Strongly held personal values of interdependence may complicate diabetes care because changes in diet and lifestyle are magnified in such settings; patients who are interdependent may be more solicitous of their potential impact on others. On the other hand, perceptions of emotional support, encouragement and compassion from people around them may be especially effective for those in interdependent cultural contexts. Such emotional support may decrease the psychological burden of diabetes care on the patients personally. In the current study, we explore differences in interdependence and perceived emotional support in relation to diabetes-related distress in two cultural backgrounds, Japan and the United States. --- Methods --- --- Procedure Kyoto University Graduate School and Faculty of Medicine, Ethics Committee and the Institutional Review Board of Christiana Care Health System approved the study protocol. Participants were recruited at the diabetes outpatient clinic of each hospital. All participants provided written informed consent prior to participation. The survey measuring diabetes-related distress, interdependence, and perceived emotional support were completed by all participants. The participants then completed the sociodemographic questions . Years from diagnosis, treatment, history of attending a diabetes patient education program and presence of diabetes complications or other comorbidities needing treatment or self-management such as hypertension, heart disease, malignant tumor, and depression, were also measured by a self-report checklist. Recent glycemic control was obtained from medical records. --- Measurements Diabetes related distress was measured using the Problem Areas in Diabetes scale , a well-validated 20-item self-report questionnaire [17]. Items are rated on a 5-point scale ranging from 0 to 4 . Summed scores are converted to a 0-100 scale by multiplying by 1.25 [19]. The PAID was translated into Japanese by Ishii et al. and the Japanese version also showed high internal consistency and validity [20]. Interdependence was measured by a well-established English and Japanese version of the Self-Construal Scale [21,22]. Participants indicated how much they agreed with 10 independent statements and 10 interdependent statements about the self . This scale has successfully measured independence and interdependence in many cultures including Japan and the United States. It distinguishes not only cultural variation but also individual variation in one culture. Measured scores are associated with psychopathological symptoms and neural activity in general population [23][24][25]. The score is the mean rating given to interdependent statements minus mean rating given to independent statements, which shows substantial internal reliability [22]. Perceived emotional support was measured by a well-established English and Japanese language version of a 16-item scale assessing the perception of receiving encouragement, compassion, and other forms of emotional support from close others. Participants were asked to think about close others and then to indicate the extent to which these close others provided each of 16 types of emotional support [8]. Self-esteem was measured by Self-Competence scale, a well-established scale in English and Japanese [26,27]. HbA1c measured in US was expressed according to National Glycohemoglobin Standardization Program and HbA1c measured in Japan was expressed as NGSP equivalent value [28]. --- Data analysis Participant characteristics and survey responses are presented as means and SD or sample size and percent. Distribution of variables was checked visually and by Shapiro-Wilk tests. Independent-sample t tests and Mann-Whitney U tests were used to explore group differences for normally distributed variables and for non-normally distributed variables, respectively. Fisher exact test was used for categorical data. Pearson's correlation coefficient was used to identify correlation among PAID, interdependence, perceived emotional support, self-esteem, and potential confounders such as sex, age and education level. The associations between PAID and interdependence and between PAID and perceived emotional support were assessed by two-way scatter plots and Pearson's correlation coefficient after adjusting for the identified confounders by regression model. To evaluate constructed factors of PAID that may have association with interdependence or perceived emotional support, the factor structure of PAID was analyzed in the Japanese data. In principal component analysis, an eigenvalue of.1.0 was used to identify the possible numbers of components. An exploratory factor analysis with promax rotation was performed. A loading level of. 0.40 was used for the items to be included in each component. The association between the identified factors of PAID and interdependence, perceived emotional support, self-esteem, sex, age, education level, HbA1c, years with diabetes, medications and complications were examined by multiple regression analyses. All analyses used Stata 11.0 . Statistical significance was set at P,0.05 . Missing data were not imputed, with the exception of a maximum of 2 missing values of PAID, which were estimated using the mean of their remaining items [29]. --- Results Eligible participants who completed the surveys were 152 in Japan and 64 in the United States. All 152 participants recruited in Japan were of Japanese ethnicity. Of these, three were excluded from analyses because of missing data, two for occupation and one for education. The longest experience living abroad reported by a Japanese patient was four years. Only about 21 participants recruited in the United States during the period of September 2011 through April 2012 were able to confirm ethnicity; fifteen were European Americans, four were African Americans, and two were Asian Indians. Two of the 64 participants in the United States were excluded because of depression. Twelve were excluded because of missing data, one for HbA1c, one for education, five for interdependence and five for self-esteem. Of the remaining 50 participants, one reported that she had lived in Puerto Rico for 36 years, two of them had lived in India for 35 and 21 years respectively, one had lived in Ireland, England and Germany for 30 years total, one in India and Canada for 25 years total, one in England for 11 years, and others had lived abroad for no more than one year. In this study, these 50 participants comprised the American patients. Finally, 149 Japanese patients and 50 American patients were included in the analyses . No significant differences were observed between the two groups in sex, age, education level, HbA1c, treatment, or diabetes education history. Japanese patients had lower BMI and slightly fewer years with diabetes than American patients. More American patients had nephropathy and neuropathy. Japanese patients had a higher score of interdependence and PAID and a lower score of perceived emotional support and selfesteem than American patients. In Pearson's correlation analysis, PAID had significant correlations with interdependence, perceived emotional support, selfesteem, sex and age in Japanese, and with self-esteem and age in Americans . Interdependence had significant correlations with perceived emotional support, self-esteem, sex, age and education level in Japanese, and with self-esteem in Americans. Perceived emotional support had significant correlations with sex and age in Japanese, and with self-esteem in Americans. Based on these results, association between PAID and interdependence was assessed after adjusting for perceived emotional support, selfesteem, sex, age and education level. The association between PAID and perceived emotional support was also assessed after adjusting for interdependence, self-esteem, sex, age and education level. Adjusted interdependence showed a weak but significant positive association with PAID in Japanese , while it did not in Americans . Adjusted perceived emotional support showed a significant negative association with PAID in Japanese , but in Americans it had no association with PAID . The two-way distributional patterns of adjusted perceived emotional support and PAID were strikingly different between Japanese and Americans. The more emotional support Japanese patients perceived, the less distress they reported. On the other hand, American patients who perceived more emotional support did not as frequently report less distress. Confining the analysis to patients without missing data of PAID did not influence the results. Confining the analysis to those who had not lived abroad more than 5 years also did not influence the results [30]. In the Japanese patients, principal component analysis identified four factors of PAID with eigenvalue of 9.25, 1.36, 1.17, and 1.06. Each factor accounted for 46.3, 6.8, 5.8 and 5.3% of the variance, respectively. Assuming 2 to 4 factors, exploratory factor analysis was performed with promax rotation. In the 2 and 4 factor solutions, each factor was not homogeneous and hard to interpret. The conceptual congruency of items supported the 3 factors solution. The first factor included 8 items with loadings from 0.40 to 0.77, and could be interpreted as negative feelings about total life with diabetes . The second consists of 8 items with loadings from 0.43 to 0.74, and could be interpreted as negative feelings about living conditions with diabetes. The third consists of 2 items with loadings of 0.75 and 0.77, and could be interpreted as negative feelings about treatment of diabetes. Cronbach's a as a measure of internal consistency for the 3 factors were 0.90, 0.84, and 0.82, respectively. The mean score of the items was calculated for each factor and used as a score of each subdimension of PAID. The score of the first subdimension, ''negative feelings about total life with diabetes'', was 1.560.9 , and ranged 0 to 4. The second, ''negative feelings about living conditions with diabetes'', was 0.860.7, and ranged 0 to 3.25. The third, ''negative feelings about treatment of diabetes'', was 1.261.0, and ranged 0 to 4. The association between the three subdimensions of PAID and interdependence and perceived emotional support was further evaluated by multiple regression analysis. As potential predictors of the identified three subdimensions, sex, age, education level, HbA1c, years with diabetes, medications and complications were considered to add to the three main predictors, which are interdependence, emotional support and self-esteem. Since age and education level were not significant predictors for all three subdimensions, they were removed from the model. ''Negative feelings about total life with diabetes'' was significantly associated with interdependence, perceived emotional support, self-esteem, sex, oral hypoglycemic agent, insulin and complications . ''Negative feelings about living conditions with diabetes'' was significantly associated with interdependence, perceived emotional support, HbA1c and insulin. ''Negative feelings about treatment of diabetes'' was significantly associated with perceived emotional support and HbA1c. --- Discussion We investigated the contribution of social orientation emphasizing harmonious relations with others to diabetes-related distress in Japanese and American patients. The results indicate that a patients' tendency to interdependence may increase diabetesrelated distress, and that a perception of encouragement and compassion from people around them may decrease the distress especially in Japanese patients living in an interdependently oriented society. In the current study, Japanese patients with higher personal values for interdependence reported higher levels of distress, and Japanese patients who perceived more emotional support reported lower levels of distress. Such cross-sectional correlations need to be cautiously interpreted. An interdependent social orientation is reflected in one's generalized pattern of thought, feeling, and action [5,15]; among patients who value harmonious relations with the people around them, diabetes self-care requires adjustments to relationships that will add additional distress to the patients. Japanese patients who are interdependent may be especially conscious of, and concerned about, the impact of their required lifestyle changes on close others. Notably, the association between perceived emotional support and diabetes-related distress observed in Japanese patients was not observed at all in American patients. Although the perceived emotional support addressed in the survey was general and not specific for diabetes, there was a significant negative association with diabetes-related distress. This suggests that patients' feelings that close others encourage and empathize with them may play an especially important role in diabetes self-care in a highly interdependent culture. This finding replicates past studies in which perceived emotional support was positively related to subjective well-being among Filipinos and Japanese, but not among Americans [8]. Japanese patients with higher interdependence were more likely to experience distress related to life with diabetes than to treatment of diabetes. This seems reasonable considering that interdependence affects total life and living conditions with diabetes but not treatment of diabetes. In addition, higher self-esteem may be effective in reducing the distress related to total life with diabetes, but the effect of self-esteem was smaller and was not related to the distress related to living conditions with diabetes and treatment of diabetes. Self-esteem measured in this study was confidence in one's ability not specific for diabetes. The result indicates that Japanese patients with high confidence in their general ability may be relatively resistant to distress related to total life with diabetes, but that their tendency to interdependence may increase the distress related to total life with diabetes. Perceived emotional support had a general positive effect in all three subdimensions of distress. In an interdependent society such as Japan, encouragement and compassion from people around them may have a wide range of effects on diabetes-related distress. Among other potential contributors to diabetes-related distress, male sex was significantly associated with a lower level of distress about total life with diabetes, and poor control was significantly associated with higher level of distress about living conditions with diabetes and treatment with diabetes. Duration of diabetes was not a significant contributor to diabetes-related distress in this study. These results accord with a previous report [18,19,29]. Patients treated by medication had a higher level of distress related to total life and living conditions with diabetes but not to treatment. This result suggests that negative feelings about treatment do not necessarily stem from the medication itself. However, the result also suggests that medication itself may nevertheless be an important factor in increasing distress related to total life and living conditions with diabetes. We also found a counterintuitive association between diabetes complications and distress. Patients without any diabetes complications showed a higher level of distress related to total life with diabetes. One possible explanation is that patients without any complications were more anxious about developing complications than were those already having one or more complications. In this subdimension analysis, we used only Japanese data because of the small sample size of American data for factor analysis. Previously reported factor structures of PAID vary from study to study, and one to four factors are identified. Our results are relatively similar to the results in Dutch and Swedish patients [29,31]. We used well-established measures of interdependence, perceived emotional support, and diabetes-related distress and a growing body of evidence [8,17,20222,26,27,32]. Our study found that emotional well-being of patients with diabetes was predicted by different variables in Japan than in the United States. Distress caused by diabetes self-care was influenced by both individual and cultural variation of interdependence. Cultural psychology takes the view that human cognitive and affective processes vary as a function of cultural environments that provide unique social contexts in which psychological processes develop and are shaped [4,32]. Although past research has shown that independence and interdependence are associated with a variety of daily behaviors in healthy people, the unique contribution of the present study is that it studies these cultural variables in a specific health context. Our work thus replicates and extends past work on cultural patterns to the diabetes context. A similar comparison may be applicable to other societies and cultures; a cultural emphasis on interdependence is also known to be a characteristic of many African cultures, Latin-American cultures, and many southern European cultures [5]. Asian-American cultures within the United States are reported to similarly emphasize interdependence, although the magnitude of this difference may be smaller than in those living in Asian countries [5]. The findings in this study provide better understanding of the differences between European-American patients with diabetes and other patients. This study focuses on the difference between two cultural contexts: Japanese patients recruited in Japan and American patients recruited in the United States. It enables us to interpret the results without the influence of acculturation or linguistic barrier for minority. An important limitation of this study is the small sample size of the American patients. Further study is required to investigate the impact of individual variation of interdependence on diabetes-related distress among American patients. This study suggests a potential link between interdependent social orientation and various outcomes of diabetes care. Interventions appropriate for interdependent social orientation are required. Family-centered approaches may be an effective option in such interdependent patients. --- The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files. --- Supporting Information

Objective: Recent evidence in cultural and social psychology suggests Eastern cultures' emphasis on harmony and connection with others and Western cultures' emphasis on self-direction and autonomy. In Eastern society, relational harmony is closely linked to people's well-being. The impact of this cultural and social orientation on diabetes-related distress was investigated. Research Design and Methods: Japanese and American patients with type 2 diabetes were surveyed by well-established questionnaire in Japan and in the United States, respectively. The association of personal values for interdependence, perceived emotional support, and the Problem Areas in Diabetes scale (PAID) were analyzed. Results: A positive correlation between interdependence and PAID (r = 0.18; P = 0.025) and a negative correlation between perceived emotional support and PAID (r = 2 0.24; P = 0.004) were observed after adjustments for other factors in Japanese data (n = 149), but not in American data (r = 0.00; P = 0.990, r = 0.02; P = 0.917, respectively, n = 50). In Japanese data, the three-factor structure of PAID (negative feelings about total life with diabetes, about living conditions with diabetes, and about treatment of diabetes) was identified, and interdependence showed significant positive correlations with the first and second factors and perceived emotional support showed significant negative correlations with all three factors of PAID. Conclusions: These results suggest that personal values for interdependence may be linked to the level of diabetes-related distress and that the distress may be relieved by perception of emotional support, especially in an interdependent cultural context.

Introduction --- R ussia's transition to a market economy in the early 1990s had profound implications for the health of its population. Life expectancy, lagging increasingly far behind the West since the 1960s, fell even further. The major contributors were deaths from cardiovascular disease, injuries and violence and the acute effects of alcohol among working age men, all linked to hazardous drinking. 1 Research in Russia and its neighbours has contributed to an understanding of the factors underlying these changes. One was a rapid increase in the supply of cheap and highly concentrated alcohol 2 especially products sold as aftershaves, solvents and medicines that, although officially not sold as beverages, were widely consumed. 3 However, hazardous drinking was a coping response to social disruption, with individual studies demonstrating the negative health consequences of unemployment , while ecological analyses identify the role of mass privatization 5 and resulting rapid labour turnover. 6 Some Russians profited greatly from the transition. They include not only the oligarchs but also an emergent urban middle class, especially those employed in services and finance. But others fared much less well 7 as the Soviet-era safety nets were torn leaving many to fall through the gaps. 8 Although the main drivers of rising inequalities were the radical economic policies pursued during transition, most existing research on mortality has focused on the role of education as a basis of inequality. 9,10 These studies reveal large differences in mortality even in the late 1970s, with the gaps especially wide for injuries and violence and causes directly related to alcohol, such as alcoholic poisoning. These gaps widened during the mid-1990s and beyond, 11,12 and while life expectancy at age 30 fell among the least educated, it increased among those with university education. 13 Education is, however, only one factor and will interact with changing labour market conditions, employment status, including markers of unstable employment , 4 and perceived social status. 14 In Russia, both low education and unemployment are strongly associated with consumption of non-beverage alcohols 15 while low education, as well as not being in a relationship, are strongly associated with death from causes associated with hazardous drinking. 16 This is consistent with research from Poland and Hungary showing unmarried men faring much worse during the 1980s. 17,18 The role of occupational class is relatively neglected in the research on inequalities in Russia, unlike in some western countries. Although state socialism was purportedly building a classless society, marked class and status divisions existed prior to the political transition. 9,19,20 While class and status are notions commonly applied in sociological studies of inequalities and labour market stratification, they are often combined into a unidimensional construct of socio-economic status in public health studies. As invoked by sociologists, class is the relationship of an individual to the system of economic production. Status, on the other hand, is a person's social standing, reflecting the prestige or honour conferred on them. Both class and status are hierarchical and are interrelated, as a person's class can influence their status. Occupational class and status have differing effects that are distinct from income and education in political attitudes and cultural consumption , [21][22][23] and a recent Swedish study found that occupational status is a strong correlate of mortality. 24 It is argued that these concepts can provide a more nuanced understanding of health inequalities. 25 Based on the existing social science literature on winners and losers in Russia's market reforms, 7,[26][27][28][29][30][31] we hypothesize that the health of low-skilled, manual workers in Soviet factories would have suffered most during rapid market reforms, because their skills were rendered less valuable in the changing market circumstances and they had greater risk of becoming redundant. 5,32,33 However, the situation may be more complicated as workers with the highest skills may have the highest expectations of reform and more stressful experiences, reflecting a gap between their expectations and actual outcomes, whereas managers and entrepreneurs who participated in growing areas of the economy would fare better. Thus, those who perceived themselves to have lower status in the new society would experience higher risks of poor health. In this article, we investigate the role of class and perceived status as bases of social inequalities in Russian mortality. --- Methods We extracted data on Russian individuals from the second wave of the Russia Longitudinal Monitoring Survey , an annual representative household panel survey, covering about 4000 households and 10 000 individuals in each round. We used data for the years 1994-2006 . The Supplementary Appendix describes the details of data collection and compares our estimates of mortality rates to those previously reported for the RLMS, finding them slightly lower than existing studies mainly because our sample includes persons who changed households. 10 Occupational class was coded using the new European Socio-Economic Classification . 34,35 This categorizes occupation, employment status and supervisory status into nine classes, ranging from manual routine workers to higher managers and professionals . The ESeC was recently used in a study of inequality in self-assessed health in Europe. 36 To account for low numbers of self-employed farmers in Russia, we merged self-employed persons who were not in agriculture and self-employed agricultural workers into one class. Following the literature on social mobility in Russia, we split managers and professionals into three different classes . 37 While an individual's class as measured by occupation could hypothetically change during the course of transition, we were interested in class as a determinant of people's experiences and opportunities in the course of transition, which could be obscured by allowing it to change over short periods of time . Thus, each person's class was coded based on his/her occupation at the time of the first observation. If the occupation of a person was missing in the round when he/she was first observed, we used the information from the next round. For people who were retired at the time of the first observation, we used their reports of what they had been doing in 1990 and, where that was missing, what they had been doing in 1985. The sample was limited to men aged 21-70 years. There is only a limited number of deaths of women in this age group, precluding us from estimating inequalities in female mortality reliably. Thus, women were excluded from all subsequent analyses. We also excluded men who lived in single-person households. Following this procedure, we were able to code class for 88% of men in our analytic sample. The remainder included people who were not employed at any point of the study . This missing group was included as a separate category in all the analyses. Income was measured as household income per capita, adjusted for inflation at the level of 1994 with the official deflator. Education was measured as a categorical variable with five levels , secondary completed, specialized secondary and university degree), according to the highest educational qualification achieved. Marital status was defined as a categorical variable with three values . The values for the ethnicity variable took the values Russian, non-Russian and no answer. Region was defined as one of the 38 primary sampling units . To measure perceived social status, we drew on three questions that assess an individual's perception of their position in the social hierarchy with regard to wealth, power and respect. These responses were scaled from 1 to 9 . In Cox models, all variables measured on interval scales , were standardized with means equal to zero and standard deviations equal to 1. We estimated class-specific crude and age-standardized mortality rates [using the indirect method of standardization, taking the RLMS analytical sample as the standard population with a standardized mortality ratio of 100]. Using the Kaplan-Meier method, we estimated the class-survivor function, which we used to estimate class-specific life expectancies at age 21 . Life expectancies were calculated as the areas under the Kaplan-Meier curve. Cox proportional hazards analysis was applied to model the relationships among class, perceived status, and mortality. In all Cox models, we set age as the analytic time variable, with entry to risk at age 21 and exit at age 70. The crucial assumption of Cox models is the proportionality of hazards. We checked this assumption after fitting the models by examining Schoenfeld residuals and did not find evidence of non-proportionality for class, education, income and perceived wealth. We also tested for potential multi-collinearity among these variables, finding that, while there were intercorrelations, these were below conventional thresholds used to indicate significant multicollinearity. As our main hypothesis focused on class, we present unadjusted models of the class-mortality association, subsequently introducing demographic controls and status, income and education variables. All models were estimated using Stata 11. Plots were created with R. --- Results Table 1 presents class-specific crude and age-standardized mortality rates . Higher and lower professionals have the lowest SMRs, followed by managers. The SMR for intermediate workers is also low, but the range of uncertainty is high due to the small number of deaths in this group. The highest mortality ratios were observed among the lower sales and service class, followed by manual supervisors and technicians and manual workers, both skilled and unskilled. Men who were not in the labour force had by far the highest SMR. Figure 1 presents the bivariate associations of class, status, education and income with mortality for men, adjusted for age. The association of mortality with class differs from analyses in other Western European setting in that there is not a clear, linear gradient. Higher and lower professionals and managers have about the same mortality hazards. Unskilled routine, skilled lower technical workers and manual supervisors and technicians also have similar mortality hazards. If we exclude those who are not in the labour force , the magnitude of the gap between best and worst category is as large for class as for education and household income. The Cox proportional hazards analysis of class and the hazard of dying are presented in table 2. As shown in Model 1 in table 2, occupational class is a significant determinant of mortality hazards, even after controlling for several background variables . Next, we added the two measures most commonly used to assess social position, education and household income per capita . Adjusting for these two factors reduced the strength of the association between class and mortality, but not substantially. An increase of household income per capita by one standard deviation was associated with a reduction of the hazard of dying by 12% . After controlling for age, year, marital status, ethnicity, region, income and occupational class, men with less than secondary and lower vocational education had on average 39% and 26% higher mortality hazards compared with men with completed secondary education, while university educated men had 8% lower mortality hazards. In Model 4, we added three perceived status variables for perceived wealth, respect and power. After controlling for class, education, household income and other variables, perceived wealth showed a statistically significant association with mortality , while the hazard ratios for perceived respect and power were smaller and failed to reach the level of statistical significance. Note that the focus in model 4 is on the hazard ratios for perceived status rather than class. As perceived status can be affected by class, this makes the interpretation of the coefficients for class in Model 4 less clear than in other reported models. 38,39 The aim of Crude and age-standardized mortality rates, 1994-2006 . Crude mortality rates were standardized using the indirect method. The RLMS male sample was taken as the standard population having an SMR of 100. Model 4 is to show that perceived status remains a statistically significant predictor of male mortality in Russia even after controlling for objective socio-economic characteristics, such as class, education and household income. After finding a robust association of occupational class with mortality, we estimated the association of class with longevity. As noted above, the mortality rates estimated with the RLMS sample were lower than those published by the Russian Statistical Office, so that the estimated life expectancies were higher than official ones. As calculated from the RLMS data, male life expectancy at age 15 was 48.3 years, compared with 45.8 years using average official estimates for the same period . To correct for this, we assumed that deaths were under reported equally in all classes in the RLMS sample and lowered estimated life expectancies by 5.2% for all classes. Figure 2 presents adjusted class-specific life expectancies at age 21 for men. The results reveal that a skilled manual worker aged 21 would expect to live 40 more years, whereas a higher professional would expect to live 48 more years. Men for whom class could not be coded had particularly low life expectancy at age 21 . Note that as life expectancies are based on sample estimates, there is larger uncertainty for the estimates for classes with a smaller number of deaths . --- Discussion Before evaluating the implications of our study for health policy, we must address its limitations. First, the RLMS is a small sample for the analysis of mortality, making it more difficult to identify a statistically significant relationship should one actually exist. Nonetheless, we identified robust associations of occupational class, perceived social status, education and income with mortality. Second, our estimates likely underestimate the effect of class, as, like in all analyses of class, it is difficult to assign a class position to persons who are economically inactive and persons with manual class backgrounds have greatest risks of leaving the labour force. Further, we assumed that deaths in the RLMS were equally under-reported in all classes. In fact, in manual classes under-reporting is probably higher . Thus, our estimates of the class gap in mortality are conservative. Men, 21-70 years old. Age was set as the analytic time variable. All model estimates with the sample of 6586 subjects, 40 046 person-years and 593 deaths. Model 1 controls for year of death, Models 2-4 control for marital status, ethnicity, year of death and primary sampling unit ; the coefficients are not reported. All models report confidence intervals adjusted for the complex survey design. Third, we excluded from the analytic sample men who lived in single-person households. However, this exclusion is unlikely to seriously bias the results, as these men account only for 4.5% of all person-years in the RLMS sample . Ideally, we would also estimate class inequality in mortality among working age men only, but unfortunately doing so would greatly constrain the power of our analysis as the number of deaths is reduced by $40% as compared with the sample of men aged 21-70 years. However, we did conduct sensitivity analysis, constraining the sample to men aged 21-60 years. None of our findings was qualitatively changed, although the confidence intervals did widen due to lower statistical power. Overall, we found evidence consistent with our hypotheses that low-skilled workers had the greatest risks of dying during the transition and that those who further perceived themselves in lower status positions had elevated risk. The general pattern of inequality in mortality in Russia is similar to Western countries, whereby manual classes have greater hazards of dying than non-manual. However, unlike in some western countries, we observed no difference in the mortality experience of male higher and lower professionals and, while this may reflect a lack of statistical power, it may also indicate that the former have fared less well during transition. This is consistent with previous studies demonstrating the disadvantaged position of Russian higher professionals during the transition. 7 On the other hand, male skilled workers had no advantage over their unskilled counterparts . As most skilled workers were employed in threatened industrial enterprises, while unskilled workers could be employed in different sectors, this may indicate that skilled manual workers in Russia were especially vulnerable during the deep industrial crisis that characterized the transition. Comparing the magnitude of our findings to Western European studies, we found that the male manual to non-manual mortality gap among middle-aged men in Russia is considerably larger. In the age group 45-59 years, the manual to non-manual mortality rate ratio is 2.29 , while in Europe it varies from 1.24 in Denmark to 1.65 in France. 40 The larger class mortality gap in Russia is likely to be explained by differential alcohol consumption that mediates the association between class and mortality, although further research is required to test this claim empirically. We found that life expectancy at age 21 is about 10 years longer for managers and professionals than for skilled manual workers. To put these levels in perspective, among manual classes, male life expectancy in Russia is similar to, for example, Ghana and Haiti, while among non-manual classes, it is similar to the average for men in Turkey or Hungary. In summary, occupational class position is a powerful determinant of inequalities in mortality among Russian men. Further research is required to understand the pathways that link occupational class and premature mortality among men, especially the important role of hazardous alcohol consumption as well as of any factors that may provide some degree of protection for Russian women. --- Introduction A s in the European Union, regions and local authorities in The Netherlands are becoming an increasingly important and administrative level in public health policymaking. Since 2003, the Dutch Public Health Act has obliged local authorities to assess municipal public health on the basis of epidemiological analysis, once every 4 years. 1,2 These assessments are carried out by 28 Regional Health Services , serving 418 municipalities and dedicated to the monitoring, protection, and promotion of public health. 3 Differences in these assessments between RHSs, for example, in the selection of topics reported, methodological issues and reporting format, make it difficult to compare the public health status of a particular municipality with the national public health status or that of other municipalities. Moreover, the RHSs face the problem of making their epidemiological knowledge more useful for both regional and local --- Supplementary data Supplementary data are available at EURPUB online. Conflicts of interest: None declared. --- Key points The gap in life expectancy at age 21 between male managers and professionals and skilled manual workers in Russia is about 10 years. Occupational class is significantly associated with mortality after adjusting for education and other confounders. The manual to non-manual mortality gap in men is considerably larger in Russia compared with Western Europe.

Background: Russia's market reforms in the early 1990s led to marked social inequalities. We analysed inequalities in risks of dying for Russian men by occupational class and perceived social status in the post-transition era. Methods: Cox proportional analysis of the hazard of dying by occupational class, education, household income and perceived social status was performed for 593 deaths that occurred between 1994 and 2006 using a representative sample of Russia's male population (n = 6586 people, 40 046 person-years). Occupational class was coded based on the European Socio-Economic Classification; social status was based on survey questionnaires about people's perceived economic, power and respect status. Results: Manual occupational class is significantly associated with greater hazards of dying among men, after adjusting for age, education and other potential confounding variables. Groups at highest risk were men who were manual workers, manual supervisors and technicians, and lower sales and service workers. Substantial gaps in life expectancy at age 21 of up to 10 years were observed between male managers and professionals and manual workers. Conclusion: Substantial inequalities in risks of dying exist by both occupational class and perceived status in Russia, with patterns by class differing from those in the west.

Class, Trust and Confessional Media in Austerity Britain The Jeremy Kyle Show first aired in the UK in 2005. Scheduled in the daytime television slot previously occupied by Trisha on ITV between 9.25 and 10.30 in the morning, although repeated across ITV's portfolio of channels at various times, day and night, it adopts a confrontational talk show format that would be familiar to a global audience; although there are some notable dissimilarities , the show is comparable to Oprah, Ricki Lake, The Jerry Springer Show, Trisha and numerous others. Rather than focusing on broad social issues -as did the earlier UK talk show, Kilroy -The Jeremy Kyle Show sees guests air their personal issues and conflicts in front of a live studio audience, often to controversial effect. Commonly, participants on the show will present their strained romantic relationships, voice accusations of infidelity, seek resolutions to disputed paternity, and relate the stresses of marriage or family life where drug use or alcohol abuse is involved. So, The Jeremy Kyle Show deals with personal issues at times of distress or in difficult interpersonal scenarios in a public setting. These are worked through on stage by the eponymous host, who, we are told, 'is not afraid to speak his mind, and believes the only way to solve a problem is to get it out in the open' , and supplemented by a back-stage team. To facilitate this "openness", the show makes use of lie detectors , DNA profiling for paternity, and drugs testing. As well as reproducing the tried-and-tested formula of the television talk show, The Jeremy Kyle Show sits within the broader medium of reality television, understood as a unmediated and voyeuristic presentation of authentic people and scenarios . If reality television is best understood as possessing a voyeuristic appeal, coupled with the authenticating mechanism of being seen , then the talk show also emphasises the process of being heard. The very premise of The Jeremy Kyle Show hinges on participants' willingness to confess domestic and emotional failures or conflicts in front of an audience, and so is indicative of relaxing attitudes towards discussing private matters in public , in what might be called a 'confessional culture' or 'confessional society' . As such, early talk shows such as Oprah were seen to play an important political role in giving airtime to previously marginalised, and unheard, groups and their everyday experiences . Not only were these narratives given a media prominence, but participants were encouraged to use talk as a kind of therapy for dealing with personal issues that were related to wider social problems . It is important to note, however, that the talk show evolved from this personal-as-political therapeutic talk to a more confrontational style, with shows such as Ricki Lake and The Jerry Springer Show marking a movement away from social injustice towards 'social shock with therapeutic overtones' . This, it has been argued, moves the format away from emancipatory visibility and towards normative discipline , as shocking dysfunctionality is brought into line by the talking therapy marshalled by the host. The sort of dysfunction that is presented on The Jeremy Kyle Show might be understood here as a moral failure to adhere to middle-class cultural norms , with such shows submitting working-class impropriety to surveillance in order that it might be simultaneously entertaining and policed . In this article it is argued that the function of The Jeremy Kyle Show is inextricably bound up in a neoliberal agenda towards rolling-back state apparatuses and expenditure, whilst simultaneously shaming those who depend on it or have refused, for whatever reason, the mantra of self-help, in what has come to be called Austerity Britain -predating the financial crisis and yet prefiguring responses to it. In response to the global recession of 2008-2009, the British Conservative-Liberal Democrat coalition government has pursued neoliberal economic policies, reducing the national deficit by cutting public spending, shrinking -perhaps even beginning to dismantle -the welfare system, and ushering in an age of austerity. In the first section it is argued that technologies of confession are used on participants in The Jeremy Kyle Show who are deemed, by virtue of class position, to be untrustworthy, undermining the very basis of therapeutic talk by deferring to scientific measurement of the body in order to derive truths. In the second section it is argued that this measurement does not simply identify bodily truths, but locates bodies that lack the discipline to contribute efficiently to the austerity agenda. As such, this measurement also determines the value of the participants, and, in closing this section, it is argued that the identification of low value individuals provides a platform for shaming that amounts to an attack on the principles of welfarism. In the concluding remarks, it is argued that austerity television, by presenting a kind of austerity realism where there are no alternatives to cuts and self-reliance, is an important area of study -even as we face our own conditions of austerity in the academy. The focus of this article, then, is on the use of technologies of confession and how they situate the show in narratives of austerity . --- Technologies of Confession As Skeggs argues, reality television equates ordinariness with working-class life and working-class people. With talk shows we have seen 'the general spread of the languages of confession and therapy' , where seemingly there is no problem that cannot be addressed by a talking therapeutic intervention , such that what emerges is 'therapy as entertainment' . The television talk show gives a platform for people to 'talk about the ways in which they make themselves and others miserable' , and since the object of the show is ordinariness, and the ordinary is equated with working-class people and their lives, it offers a spectacle of 'failed selves' whose failure, and misery, is judged from a classed position. The predominantly working-class participants on these shows are held up to middle-class standards of propriety , and found lacking -and in need of transformation . The problem is that therapeutic talk is an instrument of middle-class culture, and workingclass people are perceived to be lacking in the ability to fix "failed selves" by their own volition . As such, on television talk shows, "ordinary" participants are submitted to a 'basic therapeutic creed that we are perfectible and that identity can and perhaps ought to be shaped by wilful self-management and introspection' , despite an absence of trust that they are capable of doing so. It is not only that they are judged to be engaging in improper behaviour, but also that their inability to narrate their selves in the talk format is indicative in itself of a lack of responsibility and self-governance . Thus, the participants are seen as "lacking"; they lack propriety; they lack knowledge of how to display their selves properly; and they lack the means to engage in the corrective -self-help through talk . So, Wood and Skeggs argue that reality television operates as a kind of training course in middle-class culture, and in the case of talk shows this can be understood as training in the talking cure. However, we can observe that through the prominent and repeated use of technological interventions on The Jeremy Kyle Show, participants are subjected to self-help through talk, but instead of being trained in the culture of confession, are instead not trusted to be able to adequately participate. Hence we see the use of technologies of confession, primarily: the lie detector, the paternity test, and drug or alcohol testing. --- The Lie Detector Over half of the guests on The Jeremy Kyle Show request a lie detector . The test is used most frequently for accusations of infidelity, either at the request of a partner who suspects unfaithfulness or one who wants to prove that such suspicions are unfounded, although it is used less commonly in other contexts, for example, to resolve intra-family disputes over missing money or valuable items. In the case of infidelity, the audience usually sees Kyle reading through three set questions, the answer given by the participant, and then the judgment of the polygraph operator on the validity of this answer. These three questions are: Have you kissed anyone else whilst in a relationship with x? Have you had sexual contact with anyone else whilst in a relationship with x? Have you had sexual intercourse with anyone else whilst in a relationship with x? With the truth in the open, Kyle can then offer his own value judgments on the situation and dispense advice to the couple. However, the audience does not get to see the operation of the lie detector itself, which takes place behind the scenes, under the authority of the show's polygrapher. Whilst the show's website states that polygraph practitioners claim a high degree of accuracy for the procedure, we are also informed that this latter claim is disputed, and that the test is not admissible as evidence in UK courts of law . However, the nature of the disputation of the accuracy of the test is not clearly outlined, nor is the supposed science behind it -rather than simply the procedure -explained. Lie detectors work by measuring blood pressure, pulse, respiration and skin conductivity , and it is these measurements taken during the answer that the polygrapher analyses in order to detect deception. Put simply, the polygraph machine does not detect lies -that is an interpretation on the behalf of the operator. There are three criticisms raised by The National Academy of Sciences : that the test is unscientific; that it is unreliable; and that it is biased . The hypothesis that deception can be detected by such measurements is insufficiently evidenced, since the test yields accurate results at rates above chance but that are insufficiently high. In addition to this unsatisfactory rate of success, there is insufficient evidence that accuracy is stable across, for example, different sociodemographic groups, psychological or medical conditions, or personality types. As well as the potential for false results for certain groups or individuals, the role of the polygrapher in interpreting the results potentially biases the procedure -especially given that , they will be in contact with the participant, which may influence the interpretation of the measurements. So, despite the gloss of The Jeremy Kyle Show, this is a test considered dubious in both legal and scientific contexts. As Palmer observes, the lie detector 'is a cold and dramatic device to bring into the glossy world of daytime television'. This incongruity may in part be explained by a desire for producers of such shows to offer realism amidst accusations of fakery ; for example, in the UK, the use of actors on The Vanessa Show . In order to instil confidence in the audience, the participants are forced to pay a price: as Palmer notes, at least one of them will shamed, either as a liar or as the maker of false allegations, and this shame is apportioned by allegedly scientific means. This passage from science to shame, we should add, is facilitated by an erosion of trust in a self-help-throughtalk context. If talking through one's issues in a public forum is to be understood as of any use, then it would need to rest on the idea that through talk we can gain self-knowledge and through self-knowledge we can mend damaged selves. However, the imposition of the lie detector subordinates knowledge to truth statements, and so the therapeutic process is derailed by supplication to science. As Lyotard famously argued, there is more to knowledge than truth, and more forms of knowledge than science alone. Yet the lie detector here denies the participant the opportunity to narrate what is perceived to be their failing, and privileges scientific measurement of bodily responses over alternative ways that the guest might know themselves. By disempowering the guest, by placing legitimation of their statements in the hands of the polygrapher, the lie detector bars admission to self-knowledge and self-help through talk alone -and the whole premise of therapeutic talk on reality television collapses. Lyotard's rhetorical question, 'who will know? ' , is apposite here. For Lyotard, the problem with over-privileging scientific knowledge is not just who has access to what has been determined as true, but who gets to determine what is true and how this determination is made. In the case of The Jeremy Kyle Show, it is the polygrapher. Rather than a process of mending, there is simply the presentation of a broken self. Thus, that the participants on the show cannot be trusted to engage in therapeutic talk by virtue of their lowly class position, undermines the very foundation of self-help-through-talk since they are not allowed to narrate or heal for themselves their broken self -and leads to the public shaming of individuals judged failures by middle-class norms. --- The Paternity Test This 'obeisance [...] to the dictates of science' is continued with the paternity test, although here the science itself is far more trust-worthy. The determination of paternity is achieved through DNA profiling. This requires the collection of reference samples, such as skin cells from inside the cheek, saliva, blood or semen, from the putative parent and offspring. These samples are compared and a determination of paternity can be made to a degree of 99.9% certainty . On The Jeremy Kyle Show, uncertainty over paternity usually centres on newborns, and tests are undertaken when the mother does not know who the father is, or when the individual she identifies as the father has reason to doubt their claim to paternity. Less frequently, the paternity test is performed at the behest of an adult guest whose heredity is uncertain, for whatever reason. Palmer argues that the paternity test has a similar function to that of the lie detector. However, it ought to be recognised that whilst infidelity is once more at the centre, the moral and social issues surrounding paternity heighten the stakes. Since the DNA sampling is considered to be a near-infallible scientific corrective to heredity doubt, Jeremy Kyle, once he has the results, gains permission to shame the transgressive participants -and where there is doubt about paternity, all parties are in shame. Like polygraphs, the use of the paternity test on talk shows demonstrates 'how "the truth" is located in bodily responses and is therefore measurable, quantifiable' . Once more, the trust is in techno-science and not in the participants -or between participants. However, the impact of the paternity test on trust might be seen as of a different order of magnitude to that of the lie detector; that is, whilst both lie detector and paternity test are employed on untrustworthy subjects, the paternity test is further corrosive of trust. As Žižek observes, '[p]aternal authority is irreducibly based on faith, on trust as to the identity of the father [...] because we do not directly know who our father is, we have to take him at his word and trust him ' . That is, since it is the mother who bears the child, her identity is as close to a certainty as to make it a fact, yet there is always the possibility that the assumption of fatherhood is incorrect, given that their role is limited to conception. For Žižek, this plays an important formative function in the grounding of trust in society, with fatherhood as a foundational symbol. The paternity test, he argues, by introducing scientific certainty, displaces this symbolic function of the father; 'truth can be established through DNA analysis', cementing '[t]he hegemony of scientific discourse' and establishing an authority based on biopolitics . This, Žižek argues, is an attack on the entire edifice of trust in society, executed in the name of science. This undermining of trust ought to be considered alongside the class position of the participants and, especially, the public perception of their reproductive activities. Tyler has demonstrated the role of media presentation of working-class mothers in sustaining and furthering their vilification. In particular, she notes the perpetuation of an image of working-class mothers as sexually irresponsible, observing the demonisation of teenage mothers, women with children by multiple partners, and mothers of mixed-race children. Piecing together criticism of the working-class mother from a variety of media sources, Tyler argues that she is presented as 'the quintessential sexually excessive, single mother: an immoral, filthy, ignorant, vulgar, tasteless, working-class whore ' . Thus, the working-class mother is doubly untrustworthy: she cannot be trusted to act with sexual decorum; and, as a result, she cannot be trusted to know who the father of her children are. Tyler does not examine the public perception or media presentation of the working-class father, but, from the image she presents of working-class subjects in general, we can say that, simply by virtue of being working-class, they are similarly seen as sexually irresponsible, as well as feckless. The father, then, is also twice distrusted: they cannot be trusted to take self-responsibility over their sexual activity; and they cannot be trusted to take responsibility -understood primarily in financial terms -for the product of that activity. So, on The Jeremy Kyle Show, both parents are subjected to the paternity test as a corrective to the perceived irresponsibility that undermines faith in their determination of fatherhood. Whilst we need not agree with Žižek that the entire social edifice of trust is here under attack, we can say that this subordination to the discourse of science perpetuates the image of untrustworthy working-class subjects, and, what is more, legitimates the idea that submission to testing is the only grounding for trust in or between them. The very fact of their testing submits them to public shaming for their sexual irresponsibility; the results then allow Kyle to shame the participants for their irresponsibility as parents. --- The Drugs Test The final technology of confession to be discussed here is the drugs test. This is less common than use of the lie detector or paternity test, but it is no less indicative of the relationship between class and trust on The Jeremy Kyle Show. The drugs test is usually employed where conflict has arisen between partners or within families due to one of the participants' drug use or alcohol consumption. Samples of blood, urine, hair, sweat or saliva are taken in order to test for the presence of a specific substance or substances. Occasionally, this is coupled with a medical examination of the guest's internal organ functionally, undertaken by a doctor, in order to ascertain the impact of substance abuse. Whilst the drugs testing is undertaken on both men and women, the most striking example for the present study is its use on expectant mothers, since this speaks eloquently to the correspondence of class and trust, and the scrutiny of the body in order to assuage the lack of the latter. Balsamo has argued that surveillance of the maternal body has migrated from a clinical setting into society in general. With this, she sees a move from private responsibility for foetal health to public interference, and the normalisation of what she calls 'public pregnancies ' . For Balsamo , this shift is inseparable from media narratives of outrage against irresponsible mothers, for example, the "Crack Baby" media event of the late 1980s and early 1990s in the US, which centred around foetal damage or death resulting from mothers' use of crack cocaine. In this instance, the behaviour -and bodies -of predominantly poor women, or, 'welfare mothers ' , were seen to be in need of external disciplining. In the "Crack Babies" case, this resulted in the drugs-testing of at risk pregnant women, such that the mother's body was used against her because she was not considered a reliable source of the truth . Similarly, we can understand the drug-testing of pregnant women on The Jeremy Kyle Show as a public interference into the pregnancy of those deemed incapable of taking bodily responsibility for themselves or the foetus. This is consonant with Skeggs' argument that reality television presents 'working-class families, especially mothers, as incapable of knowing how to look after themselves and others ' , that is, as being irreparably irresponsible. This lack of trust can be extended to non-pregnant women and men alike, whose class position dictates that they lack the capacity for self-responsibility, and who cannot be trusted to engage in talk alone in order to address problems resulting from substance abuse. Once more, without the opportunity for the self-restorative nature of therapeutic talk, since scientific measurement of the body circumvents this, the effect is to shame participants' for their own irresponsibility. --- Technologies of Austerity If Jeremy Kyle is committed to resolving conflict through openness and honesty , then this openness is facilitated, and the honesty underwritten, by technologies of confession. The effect is that truth, in this context, is recast as a measurement of the bodily functions or responses of the participants, which in turn short-circuits talking as a tool for self-betterment, resulting in a display of shame. As such, The Jeremy Kyle Show ultimately works to place working-class people in the stocks . In what follows, it is argued that the technologies of confession addressed above ought to be understood as commensurate with neoliberal processes of both roll-out and roll-back, whereby regulatory incursions go hand-inhand with the withdrawal of state institutions and cuts to government expenditure . Thus, the technological production of shame on The Jeremy Kyle Show might be understood as contributing to the normalisation and reinforcement of austerity measures. --- Confession, Measure and Value Jensen and Tyler understand austerity 'as a discourse which produces accounts of waste and inefficiencies, moral conduct and lifestyle, work, worth and labour' in order to legitimise emergency measures in the wake of economic crisis, namely, 'to shrink the state, to compact and condense public spending, to become lean, pursue "efficiency" and to eliminate "waste"'. However, this sort of crash diet for the state requires careful monitoring and so we see an increasing culture of measurement and audit, the intervention of the state, emerging during the financial crisis . As Gane argues, this means that neoliberalism, within which the austerity discourse sits, 'does not simply involve the devolution of state or institutional powers but also the emergence of particular forms of governance' . Neoliberalism is not the absence of the state, but rather 'an argument for the state to be marketized to its core' , that is, to instantiate the principles of market freedom in its own exercise of power and to work tirelessly to regulate, not the market but people, such that the market can operate unhindered. Gane's argument here is built on a detailed reading of Foucault's biopolitics lectures, where the role of the state is understood as facilitating the 'general regulation of society by the market' . The harmful impact of the market, as Bauman argues, is no longer the concern of the state, which turns its attentions instead to personal safety -most notably here threats to the body, for example, by policing poor health or lifestyle choices. The result is an intensification of surveillance of people and their behaviours, to such a degree that Fisher terms it, with a rhetorical flourish, 'market Stalinism ' . Analyses of the act of confession on talk shows and reality television in general have often drawn on the work of Foucault . There is not the space here for a full exegesis, but, in a nutshell, we can say that for Foucault: confession necessitates a relationship of power, since it involves the presence of an authority who judges, forgives, punishes or consoles ; confession is inextricably linked to the body, since it uncovers the secret truth of sexuality ; and the authority during confession was originally a role taken by the Church but is now performed by the state . Palmer uses this account of confession in order to understand surveillance of the body as 'the calculated directing of human conduct, the means by which behaviour is shaped to various ends by the expertise of the many agencies of the state ' . In times of austerity, then, the state has an added impetus for regulating human behaviours that are deemed improper in so far as they require government expenditure, and in so doing can employ audits that seek to identify inefficient bodily conduct. As such, economic crisis is projected onto working-class bodies. The supposed irresponsible sexual activity of working-class people -promiscuity, for example, but especially sex without contraception -represents a cost to national resources insofar as the National Health Service might be required for the treatment of sexually transmitted infections, or state support in the form of benefits might be needed for low-or no-income families. Similarly, the consumption of drugs or alcohol represents a drain on resources in terms of treatment for addiction or related disease, whilst the policing of drug use and alcohol-related crime, including the cost of trials and incarceration, is a further expenditure. If there is a perception of working-class corporeality as 'a body beyond governance' , then, in austere times, it is seen to be a body that we can no longer afford not to attempt to govern. Policing the body, then, is intimately bound-up with reducing the national deficit -as evidenced by the use of that quintessential Jeremy Kyle Show technology of confession, the lie detector, to assess the veracity of benefits claims . Under conditions of neoliberalism, there is no longer a question of what is touched by the market -only how things are touched . Hallin has similarly argued that the market is touching parts of the social that were previously beyond its remit, as we have seen with the regulation of profligate bodies, adding that the media is the most important of the social institutions that are now subjected to neoliberal logic. However, it is perhaps how media have been touched that is most important, for present purposes, the way that reality television shows have begun to parallel neoliberal forms of governance . For example, Ouellette has argued that the use of drug testing on Judge Judy 'both duplicates and extends the surveillance of the poor and working class carried out by welfare offices, unemployment centers, and other social services ' . The use of technologies of confession on The Jeremy Kyle Show -and this show in particular has made these technologies routine in a way unprecedented on UK television screens -exemplifies the way that reality television shares in the role of the state in the measurement of human value. The lie detector, paternity test, and drug testing are all doubly efficient: they allow for the location of truth in the bodies of subjects who cannot be trusted to speak it; and they locate bodies that are perceived as inefficient in austere times. Put another way, the technologies of confession represent value: technological efficiency is value for money, whatever the human cost; human bodies can be measured for value for money towards meeting the demands of austerity; and those that represent poor value for money are marked as falling short of the values of Austerity Britain. Wilson has argued that Kyle's predecessor in the ITV talk show slot, Trisha Goddard, provided a way for people to negotiate personal difficulties away from a welfare state that was straining to the point of collapse, going so far as to classify Trisha as public service broadcasting. The Jeremy Kyle Show might be seen as the logical extension here: with the welfare system now deemed largely unaffordable and in need of dismantling , and volunteerism and private enterprise seen as part of the solution , this commercial television production can be read as a sort of perverse public service, auditing fiscally unviable bodies in order to shame them for their burden on the nation. --- Austerity Court Whilst the connection between television talk shows and the conditions of neoliberalism has been made explicit , it is through the personality of the host that these shows are differentiated -and so each different host will illuminate different and specific connections. Marsh and Bishop argue that Kyle's approach to hosting his show is almost that of a judge in a people's court -which is both perceptive and in need of further attention. On other talk shows the audience is permitted, often encouraged, to intervene in the narration of personal troubles between participants and host, applauding and booing, volunteering advice or opprobrium, to the extent that Palmer argues that 'the power of the norm is vested in the audience ' . On The Jeremy Kyle Show, however, whilst the audience claps or jeers they can find themselves rebuked by the host for doing so; Kyle rarely goes into the audience to solicit their opinions, and those that call out or try to intervene in the discussion find themselves censured. This has the effect of casting Kyle as the sole authority on stage, a judge in front of a public gallery, such that the host more closely resembles that of Judge Judy than any other talk show. Guests participate on Judge Judy in order to resolve legal matters that might otherwise go through the court system. Whilst the host, Judge Judith Sheindlin, is a real judge, Judge Judy is a television show in the style of a court of law rather than televised court proceedings, which gives her the leeway to inject her opinion at various stages, introducing her own prejudices. As such, Sheindlin is able to use the show as a platform for castigating people who have failed to govern themselves, as well as for voicing her guiding ethos, that everyone can be successful if they take responsibility for themselves . That people have fallen so short of success, to the extent that they find themselves in front of her in a mocked-up court, is "evidence" for Sheindlin that they are either foolish or stupid . This, then, obscures any other cause for the problems that participants find themselves in, such as economic inequality. This promotion of selfresponsibility -and the tongue-lashings that befall those who have not embraced it in the name of personal change -cannot be separated from Sheindlin's position as a judge. By presenting conflicts between participants as petty disputes that arise from personal failings, the show narrates them as being an unnecessary drain on the legal system; the show then acts as a litigation sluice, diverting supposedly undeserving legal subjects away from the actual courts . The overall message is that if you cannot look after yourself, you cannot expect the state and its apparatuses to help. As such, Judge Judy, by setting itself up as a private alternative to public courts, 'draws from the symbolic authority of the state to promote both the outsourcing of its governmental functions and the subjective requirements of the transition to a neoliberal society' . The show, then, advances 'a neoliberal agenda of cuts to social programs and the dismantling of the final vestiges of the welfare state' . Much the same can be observed of The Jeremy Kyle Show. In the first section it was argued that technologies of confession are used on people deemed untrustworthy by virtue of their class position, undermining the talking therapy that would allow for the healing of wounded selves and resulting in only the presentation of those wounds; and we saw above that these technologies act to audit the value of bodies in terms of their efficiency in conditions of austerity. The first of these functions provides the platform, and the second the authority, for Kyle to shame his guests. As with Sheindlin, Kyle delivers tongue-lashings, berating his guests for their personal irresponsibility, 'pointing fingers and offering scathing judgements' and delivering acerbic "common sense". Like other talk show hosts, then, Kyle is presented as a moral authority , who espouses the mantra of self-responsibility that is prevalent across reality television, a dictate especially directed at working-class people . So, after lie detectors and paternity tests have audited and revealed sexually irresponsible bodies, Kyle will admonish the male guest for their failure to use contraception with what has now become something of a catchphrase: Why didn't you put something on the end of it? He inquires into guests' employment status in order to determine their ability to provide for a child . Similarly, Kyle will ask after the employment status of guests revealed to be consuming drugs or large levels of alcohol, and add to this his most telling catchphrase: Who's paying for x? This demand for the origins of the means to afford things -including cigarettes, scratch cards, even nappies -is a rhetorical question that nonetheless Kyle demands his guests answer; the insinuation is that Kyle himself, his studio audience and everyone who makes National Insurance contributions is funding the fiscal and bodily irresponsibility of the guest in question. The shaming of guests who have turned to the welfare state to support their children, or who require benefits due to unemployment, moves The Jeremy Kyle Show away from the personal development narrative of talk shows like Oprah and more closely resembles the austerity rhetoric of those in government against welfare claimants. Indeed, it chimes with a report from the Guardian that consumption of cigarettes and alcohol is being used as a factor in determining discretionary housing benefits . It plays to stereotypes of unemployment, alcohol or drug consumption, and pregnancy as lifestyle choices for "feckless chavs" who are ripping off Britain via the welfare system , and to the idea that it is the welfare system itself that encourages such behaviours . Kyle does not just elide the connection of his guests' problems to the wider socio-economic situation, as observable in many reality television depictions of working-class people , he attacks the very principles of welfarism itself whilst eliding an economic crisis that creates the socioeconomic situation of his guests. As with Judge Judy, The Jeremy Kyle Show acts to 'construct templates that complement the privatization of public life, the collapse of the welfare state, and most important, the discourse of individual choice and personal responsibility' . But even Judge Judy presented a narrative where people could make successes for themselves if they would just stop relying on the state and assume self-responsibility; The Jeremy Kyle Show is simply about fire-fighting failure, where the narrative is not so much that everyone can be a success, but simply that everyone can and should be less of a financial burden to the state. The show, then, adopts state practices of rollout, measuring the value of individuals, in order to promote the roll-back of state expenditure and apparatuses. So it is that Kyle presents as an austere judge in a court of austerity. --- Concluding Remarks on Austerity Television As Raphael and Kleinhas and Morris demonstrate, the television talk show, court television and reality television in general, were products of austere times, appearing in the late 1980s as a cheap way to fill air time when television network budgets were straitened. The Jeremy Kyle Show is an example of how austerity television has evolved to become a technology of austerity itself. Andrejevic has argued that reality television presents capitalism as a social construct, but one that uncannily lies outside of human control. Couldry and Littler have argued that reality television -their example is The Apprentice -acts to reify the rules of neoliberalism. Austerity television, then, presents austerity measures as a social construct -performed, on The Jeremy Kyle Show, by the hostbut one whose norms and values have an unquestionable, concrete existence. This acts to promote austerity as the only option for dealing with economic crisis, such that we might talk here, drawing on Fisher's idea of capitalist realism -where capitalism is promoted as natural such that there are no apparent alternatives -of austerity realism. This is not to claim that the producers or the host intend for The Jeremy Kyle Show to function in this way , but rather that the show produces a narrative that normalises austerity measures, including fatal cuts to the welfare state. At this point, we should remember that the show has been broadcast since 2005 -before the recent economic crisis took hold. However, we might understand this temporal precedence in terms akin to those famously evoked by the neoliberal economist Milton Friedman: Only a crisis -actual or perceived -produces real change. When that crisis occurs, the actions that are taken depend on the ideas that are lying around. That, I believe, is our basic function: to develop alternatives to existing policies, to keep them alive and available until the politically impossible becomes politically inevitable . This is not say that The Jeremy Kyle Show operated as a blueprint for austerity, but that it has a longer history of narrating human value in terms of the relative efficiency of the poorest subsection of British society, in determining that value through technologies of confession, and in shaming those who do not represent value for money. That is, the ideas behind austerity were not just lying around before the economic crisis, but already legitimated through austerity television. The Jeremy Kyle Show is but one example of the austerity television genre. Benefits , but they share in the promotion of austerity realism. More research is needed into these kinds of shows. With the increasing infiltration of neoliberal market competition and measurement into the academy, it has been argued , value is placed in certain kinds of academic activity, with some research areas being seen to have more value than others -something particularly acute with the financial restrictions since the recession. 'Glossy topics' , such as reality television, may be seen as of lesser value, 'deemed to be frivolous, unimportant or perhaps even a luxury that cannot be afforded in more "austere" times' . Yet by studying something like The Jeremy Kyle Show we can understand the ways that austerity measures are reinforced, and the same submission to measurement and value that occurs in the academy legitimated, on television screens across Austerity Britain.

Research at York St John (RaY) is an institutional repository. It supports the principles of open access by making the research outputs of the University available in digital form.

INTRODUCTION The Rohingya are an Islamic ethnic minority who have resided in Myanmar's predominantly Buddhist country for generations . They are not counted among Myanmar's 135 designated ethnic groups and have been refused nationality since 1982, thereby making themselves stateless. An estimated 1 to 1.5 million Rohingya people reside in the three settlements of the North Rakhine region-Maungdaw, Buthidaung, and Rathedaung-in Myanmar. Human rights breaches against this population group are centered on discrimination and unfair treatment . In their native state, the Rohingya have endured decades of systemic oppression, deprivation, and intentional brutality. Such oppression has been pushing Rohingya women, youth, vulnerable children, and men into Bangladesh for several years, with notable increases after violent invasions in 1978, 1991-1992, and 2016. The largest and fastest refugee migration into Bangladesh, though, took place in August 2017. Ever since, at least 745,000 Rohingya people have fled to Cox's Bazar, along with more than 400,000 infants. Thousands of civilians have been set on fire, families have been separated apart and killed, and girls and women have been the victims of sexual assaults. The significant proportion of the people who were able to escape were severely traumatized after witnessing terrible brutalities . In the proximity of Cox's Bazar, Bangladesh, that currently is home to the largest refugee camp in the world, these people sought interim asylum in refugee camps. In March 2019, the Ukhiya and Teknaf Upazilas housed around 909,000 stateless Rohingya refugees. The large percentage of Rohingya refugees reside in 34 overpopulated shelters, most notably the Kutupalong-Balukhali expanded complex, which is home to 626,500 of them . The presence of Rohingyas in Cox's bazar posed a significant issue to Bangladesh's government, which, as a poor country, had to deal with a large number of refugees. Evidence says Rohingya people living in Bangladesh are involved in crime and human rights violations, posing a serious threat to the local population. Despite several approaches by Bangladesh to end the problem with a peaceful arrangement with Myanmar, they all ended up in the darkness . Bangladesh has attempted international cooperation several times, but each time has failed due to Myanmar's refusal and denial of agreement. It is therefore important to uncover the underlying causes of resulting the failure of crisis agreement and its impact in Bangladesh as a developing nation for an effective and efficient resolution of the state with global cooperation . Consequently, the objective of this study is to explore causes, consequences and remedies of Rohingya crisis in Bangladesh. Eventually, the existing literature does not adequately address the core reasons and serious outcomes of the Rohingya problem, especially focusing on Myanmar's absence of democracy and the emergence of extremism, drug smuggling, and other significant hazards. As a result, the current paper aims to identify the fundamental reasons behind the Rohingya people's exodus from their home country to Bangladesh, in addition to the resulting effects and consequences. --- METHOD RESEARCH Rohingya as an issue is a complex matter and has several variables that require different classes of methods for investigation . This paper follows an exploratory research approach focuses on the publication of earlier related works in secondary sources. To collect data, the author utilized several search engines and websites such as Google, Google Scholar, Science Direct, Research Gate, Social Science Research Network , Academia.edu, and others. The information was sourced from reputable journal articles, reports by international organizations, legitimate newspapers, and other reliable sources. Moreover, the author adhered to the qualitative research approach in the domain of social science in an appropriate approach. That is, the data in this study was descriptively evaluated and interpreted. At the same time, the content analysis technique has been used as a qualitative research instrument. In essence, this research is a qualitative study that relies on secondary data. --- RESULTS AND DISCUSSION --- Causes of Rohingya Crisis Rohingya crisis begins with a decade of violation, oppression and discrimination against Rakhaine Muslim . Their greatest migration started in August 2017 when an upsurge of cruelty erupted in Myanmar's Rakhine State, resulting millions of residents, half of whom were children, to flee to Bangladesh. The majority of Rohingya homes were ravaged by fire, thousands of people were massacred or torn apart from their households, and numerous violations of human rights were noted. Some key socio and geopolitical factors impacted the procedure at a later point. --- Documentation and State Policy Promulgated in Military Citizenship Law of 1982 was one of the final nails of the statelessness of Rohin gya people . The law established two types of citizenship: "Associate" citizenship, which was available to those whose citizen petitions were still being processed, and "Naturalized" citizenship, which was available to non-citizens who could show that they and their forefathers had resided in the nation prior to its independence. As a result, all Rohingyas who called Myanmar their residence should have been able to maintain their citizenship under the Act's Section 6 as naturalized citizens, or as associate residents . Despite the fact that most Rohingyas can locate their ancestors back to the colonial period, the great majority have been denied citizenship due to a lack of sufficient documentation. People who were qualified under the citizenship law of 1948, which granted citizenship to those whose kin had remained in the state earlier to 1948, would no more be entitled under this new regulation. They would therefore be regarded as associate citizens if they applied for nationality in 1948. To become a naturalized citizen, one must show proof of entry and residence in Burma before to the country's independence in 1948 . The Rohingyas were not affected by the Foreigner Act [Indian Act III, 1846], the Registration of Foreigners Act [Burma Act VII, 1940], or the Registration of Foreigners Rules, 1948. This was true before and after Burma became independent . --- Rise of Authoritarianism Aung San, the founder of democratic Burma, was a charismatic figure who was able to unite all of Burma's ethnic group before the country's establishment. The 1947 Panglong Agreement, signed under San's leadership, guaranteed that minority groups' desires for autonomy and federalism would be honored. After him in 1948, U Nu became the first Prime Minster of independent Burma. Several militant groups arose during his rule, effectively hijacking Burma's governmental apparatus . It's questionable whether U Nu was motivated by a desire to spread Buddhism in Burma. Nu primarily supported Buddhism in Myanmar to undermine the dominance of communists, according to a groundbreaking research by Human Rights Watch. Buddhism was not utilized as a strategic weapon against the Rohingya in any of the ways it was employed for political goals during the country's parliamentary system. At the time, the Rohingya were also not denied basic rights. They were engaged voters in every election. They were even recognized as one of Burma's indigenous ethnic groups by the parliamentary administration. All of these liberties have been increasingly curtailed by Burma's uprising authoritarian regime . --- Religious Conundrum Myanmar is home to one of the world's most diverse cultures. The majority of the population is Buddhist, accounting for around 90% of the total. The Rohingya Muslims of Myanmar's Rakhine state are an ethno-religious minority that have lived in the state for millennia . Ruled by Buddhist king the presentation of Rohingya minority was always negligible. Over time the government persecuted the innocent Rohingya people and denied them as legal ethnic group within the country. Human Rights Watch stated in 2002 that the state provided military orders mandating the destruction of unlicensed mosques. The government has converted former mosques and Islamic institutions into administrative offices. A total of 28 mosques and religious schools were attacked by mobs in 2001. Not only did state security fail to prevent the threats, but it actively contributed to the destruction. Since 1990, religious persecution of Rohingya Muslims has increased dramatically. The Burmese military has closed and destroyed Islamic schools , assaulted Rohingyas during prayer, prohibited the majority of religious exercises, and detained and tortured Islamic teachers and students . --- Impact of Crisis in Bangladesh Economic Impact State officials, humanitarian groups, medical experts, environmental officials, and law enforcement officials have ascertained that the occurrence of Rohingya refugees in Bangladesh has a severe, long-term detrimental effect on Bangladeshi society, industry, politics, and climate, particularly in the south of the country. The Arakan crisis forced people to flee their homes and seek asylum in Bangladesh, which has had a detrimental influence on Bangladesh's economy, social system, and safety . The Rohingya crisis, conferring to some Bangladeshi economists, has hindered Bangladesh's rapid economic growth in several areas. The inflow has already resulted in a spike in the price of daily necessities, a fall in salary rates, and a decrease in poverty rates in Bangladesh's southern region. In that sector, government spending has increased, excluding grants from other countries. The government's economic burden from Rohingya refugees is likely to grow if the crisis drags on . --- Social Impact Several NGO employees have expressed concern about the Rohingyas' lack of awareness. Some implementers have built very shallow tube wells, at a much lower height than is ideal, which has prompted additional concerns about the optimal location of these systems. It reflects that tubewells will eventually become obsolete since they will not be able to reach groundwater . As a result, there is likely to be a scarcity of safe drinking water. Kids in the camps either don't own shoes or refuse to wear them. As a result, family planning-related illnesses pose a significant threat. The number of Rohingya people living in the camps is expected to rise in the coming days, putting additional stress on already-scarce resources like food, employment, healthcare, and more. --- Environmental Impact A considerable number of Rohingyas from Myanmar have migrated to Teknaf's forest areas, posing a serious danger to the region's long-term forest management. The everincreasing population has an impact on local resources and the environment. Many additional challenges, such as biodiversity conservation, climate change adaptation and mitigation, and coastal resilience, are intertwined with forest resources. The 2017 Rohingya influx was a completely different kind of shock for Bangladesh. The Rohingya response strategy must include a sustainable forest management plan . Teknaf Wildlife Sanctuary covers 11,615 hectares and is home to 538 plant species and 613 animal species, including Asian elephants. Forests and trees play a variety of roles in enhancing community resilience and reducing vulnerability to climate-related pressures. As a result, the current deforestation condition is concerning for the Teknaf peninsula's coastal resiliency . --- Government Initiatives A multitude of initiatives has been set in motion to grapple with the intricate Rohingya issue in Bangladesh, a challenge that has captured global attention . These initiatives, stemming from both local and international actors, aim to address the multifaceted dimensions of the crisis. The complexity of the Rohingya issue necessitates a comprehensive and collaborative approach, and these initiatives represent a collective endeavour to alleviate the suffering and promote lasting solutions for the affected population in Bangladesh. --- Humanitarian Support The Rohingya people have received humanitarian aid from both national and international NGOs as well as the Ministry of Disaster Management and Relief of the Bangladeshi government. In order to provide aid to the Rohingya population, a Geneva pledge conference was organized by the UN, IOM, UNHCR, and OCHA, with the support of Kuwait and the EU. The conference aimed to gather funds and requested a total of US$ 434 million to address the humanitarian needs of the Rohingya people. --- Diplomatic Initiatives The United Nations High Commissioner for Refugees has called the persecution of the Rohingya in Myanmar an act of ethnic cleansing. Prime Minister Sheikh Hasina of Bangladesh has put up a five-point plan: i) Immediately and permanently, Myanmar must put an end to the violence and ethnic cleansing in the Rakhine State. ii) The UN Secretary-General ought to dispatch a Fact-Finding Mission to Myanmar without delay. iii) It is imperative to ensure the safety of all civilians, regardless of their religious or ethnic backgrounds. This could be accomplished by establishing UN-monitored "safe zones" inside Myanmar. iv) Safely and permanently reintegrate all Rohingya refugees living in Bangladesh back into their communities in Myanmar. v) The entire, definite, and instantaneous implementation of the Kofi Annan Commission Report's recommendations is required . --- Bilateral Agreement On the 24-27 October 2017, Bangladesh's home minister visited Myanmar and met with Myanmar's minister of home affairs to discuss possible ways to return the Rohingyas. Bangladesh and Myanmar engaged two pacts on security and border assistance during the visit. They promised to make tangible steps and take concrete measures to ensure the Rohingyas' safe, honourable, and secure return to Myanmar. --- Legal Status of Rohingyas The Myanmar government refers to the Rohingyas as "illegal Bengali immigrants to Myanmar." However, the Bangladeshi government disagrees with this classification and opposes Myanmar's stance on the Rohingyas' ethnic identity. Rather than referring to them as Myanmar's Rohingya refugees, Bangladesh identifies them as "forcibly displaced Myanmar residents" to avoid any confusion regarding their identity . --- CONCLUSION Bangladesh is dealing with a multi-faceted dilemma as a result of the Rohingya catastrophe. The international community must step up its efforts to provide long-term support and assistance to the Rohingyas, including providing them with basic necessities like food, water, shelter, and healthcare. Additionally, the Bangladeshi government should work with the international community to address the root causes of the crisis, such as the persecution of the Rohingyas in Myanmar. This can include diplomatic efforts to pressure the Myanmar government to stop its human rights abuses against the Rohingya population, as well as working to create safe conditions for their return. Finally, more attention needs to be given to the mental health needs of both the Rohingya refugees and the host communities in Bangladesh, who have been impacted by the influx of refugees. Overall, a number of coordinated, sustained, and comprehensive approach is needed to address the Rohingya crisis and ensure the protection and well-being of all those affected by it. In summary, the Rohingya crisis in Bangladesh remains a pressing global concern, demanding sustained attention and collaborative efforts. The initiatives undertaken thus far, spanning diplomatic, humanitarian, and legal realms, reflect a collective commitment to address the multifaceted challenges faced by the Rohingya population. As the international community strives for lasting solutions, continued collaboration and a comprehensive approach are imperative to alleviate the immediate plight and pave the way for a more secure and sustainable future for the affected communities.

Bangladesh, a nation marked by its dense population, continues to wrestle with multifaceted challenges even 52 years after gaining independence while navigating the sophistications of development. Among these challenges, the pressing issue of the Rohingya people's plight takes center stage, with various forms of oppression, including mass killings, executions, forced labor, deportation, and the denial of fundamental human rights, rendering them stateless. This study seeks to dissect the roots of the Rohingya crisis and explore its repercussions on Bangladesh. Employing an exploratory research approach, data from secondary sources were qualitatively analyzed using content analysis to discern underlying themes and patterns. The findings reveal that the Rohingya crisis is primarily driven by the lack of documentation, the rise of authoritarianism, and a complex religious landscape. Additionally, the paper identifies social, economic, and environmental challenges as direct consequences of refugee migration in Bangladesh. These insights hold significant implications for policymakers and practitioners actively involved in addressing the Rohingya crisis. Offering a comprehensive analysis, the study also proposes practical policy recommendations grounded in the specific context of the crisis, aiming for tangible applications in real-world scenarios.

Background With better treatments and earlier detection, an increasing number of people are surviving cancer, with recent estimates suggesting seven out of every 10 people diagnosed with cancer in Australia will survive 5 or more years [1]. While most people recover well after cancer treatment, many experience a range of physical, emotional, psychological, social, financial and practical challenges [2][3][4]. Some survivors are at risk of developing complications many years after finishing cancer treatments [4] and survivors are at risk of developing other cancers [4,5]. A number of sociodemographic factors have been associated with variation in outcomes for cancer survivors. Cancer survival is known to vary by socio-economic advantage/disadvantage [6][7][8][9][10][11] with this difference partly explained by variations in stage at diagnosis and treatment differences [6]. Recent evidence suggests that psychosocial factors including living alone, being unmarried and higher levels of depression can also negatively impact survival [12,13]. Identifying whether socio-economic and demographic disparities also influence Victoria M. White and Karolina Lisy contributed equally and are joint first authors. QoL outcomes for cancer survivors and the potential drivers of any differences can inform the type of survivorship care that should be directed to different cancer survivors to help ensure equitable outcomes [14]. Few studies have utilised a population-based sample to assess associations between socio-demographic factors such as socio-economic position, rurality and comorbid conditions and patient-reported outcomes, including QoL. Results in this area have been variable with associations differing by cancer type and outcomes. For instance, in a populationbased study from the UK, QoL was significantly worse for those from the most deprived areas, compared to those from the most affluent areas for breast and prostate cancer but not non-Hodgkin lymphoma or colorectal cancer [15]. While similar inverse associations between socio-economic position and QoL have been found in other studies for breast and prostate cancer [16], different patterns of association have been reported for colorectal cancer [17][18][19]. While studies examining the association between socio-economic position and QoL generally control for disease, treatment and other demographic factors, controlling for presence of comorbid conditions is variable. Multiple studies have shown a consistent inverse association between number of comorbid conditions and QoL [17,[20][21][22]. As socio-economic position is inversely related to presence of comorbidities, associations between socio-economic position, comorbidity and QoL in cancer survivors can be confounded. Understanding sources of disparities in survivorship outcomes would enable efforts to minimise variation, improve outcomes and inform models of care. The provision of information to assist survivors to manage their post treatment, health and wellbeing is a key component of quality survivorship care [23] yet many survivors report unmet information, supportive and emotional needs [24][25][26][27][28]. QoL outcomes are inversely associated with levels of unmet information and supportive care needs [23,29,30]. Several studies have identified disparities in the experience of unmet needs with younger, those having: a non-white race/ethnic background, comorbid conditions and lower education and income, reporting a greater number of unmet needs [23,30]. As cancer survivorship care should seek to reduce disparities in outcomes, this study aimed to understand whether and if so how QoL outcomes , social difficulties, employment) for cancer survivors vary by different socio-demographic factors, presence of comorbidities and information needs. Given potential confounding between socio-economic position, comorbidities and unmet information needs, we aimed to determine the direct and indirect effects of socio-economic position on QoL by exploring whether its association is mediated by comorbidity and unmet information needs. The study focuses on survivors of common cancers, three with 5-year survival rates over 90% and two with lower 5-year survival ) [31]. --- Methods --- Study design Cross-sectional survey of individuals diagnosed with five specified cancers where invited participants had a diagnosis date 1, 3 or 5 years prior to study approach. --- Subject eligibility Potential participants were identified through the populationbased Victorian Cancer Registry that maintains a list of all people with a new diagnosis of cancer in Victoria, Australia's second most populous state. Eligibility criteria were diagnosed with breast, prostate or colorectal cancer, melanoma or NHL approximately 1, 3 or 5 years prior to study approach, aged 18 years and over at diagnosis, and resided in Victoria at diagnosis. People were ineligible if they were deceased, had a previous cancer diagnosis or did not meet the cancer-specific morphological or histopathological criteria. --- Patient approach and consent Following VCR standard recruitment practices, notifying clinicians of potential participants were contacted by mail to confirm eligibility. Unless clinicians indicated otherwise, eligible people were approached by mail during the month corresponding to the anniversary of their diagnosis. Reminders were sent to non-responders after 3 weeks. Invitation letters included the study information sheet, questionnaire and reply-paid envelope. A toll-free phone information service was provided. Potential participants were advised that, consent was implied by return of the questionnaire. --- Questionnaire design and content Questionnaires developed for NHL, breast, colorectal and prostate cancer were based on UK's National Health Service experience of cancer care survey [15]. Using the format, structure and topic areas covered by the NHS surveys, a new melanoma patient-reported outcome questionnaire was developed in collaboration with an expert working group. Outcome and predictor variables were the same in each survey. --- Outcome measures HRQoL was measured by the EQ-5D-5L scale, a standardised measure of health status which consists of five QoL domains , with respondents selecting one of five response items in each domain to reflect their current status [32]. Australian-specific calculations were made with the nominal range of values being 0 to 1 [33]. The social difficulties index is a 21-item questionnaire used to assess everyday problems people affected by cancer may experience [34]. All items are responded to using a 4-point scale ranging from "0" = no difficulty to "3" = very much difficulty. Items assess difficulties across a range of activities, for example working, domestic chores, self-care, communicating and financial issues. Responses are summed to provide an index of difficulties with scores ranging between 0 and 44. Participants were asked if they were working at the time of their diagnosis and their work status at the time of completing the survey. Respondents working full/part-time or casually were classified as being in paid employment at that time point. --- Predictor variables Presence of other long-term conditions was assessed with respondents indicating if they had any of 17 listed conditions that included heart disease, diabetes, asthma and high blood pressure with an option to indicate another condition if needed . The number of conditions indicated was counted to form a comorbidity index ranging from 0 to 8 . Respondents indicated their need for further information in 10 areas with responses grouped into no needs, 1-3 needs or ≥ 4 needs. The VCR provided information on the cancer diagnosis, age, gender and residential postcode. Residential postcode was used to infer socio-economic position using the relative disadvantage scale from the socio-economic indexes for areas developed by the Australian Bureau of Statistics. This area-based indicator ranks each area in Australia according to its socio-economic disadvantage based on a number of indicators including unemployment, income, education and home-ownership in the area [35]. For this study, we used postcode level rankings of relative disadvantage which were grouped into deciles with low scores indicating greater disadvantage. --- Control variables Language spoken at home , country of birth , residential location and response to treatment were indicated by participants on the survey and controlled in analyses. --- Statistical analysis Bivariate and multivariate analyses were conducted to examine associations between the QoL outcomes for all participants and for participants aged under 60 and in paid employment when diagnosed, for the outcome of not working at the time of the survey. Key predictor variables in all analyses were cancer type, time since diagnosis, gender, age, socio-economic disadvantage, comorbidities and unmet information needs. Country of birth, residential location and response to treatment were controlled in all analyses. In multivariate regression analyses, the base category was used as the reference group for all categorical predictors. Socioeconomic disadvantage deciles and number of comorbidities were treated as continuous variables. Mediation analysis aims to identify mechanisms that may explain or clarify the relationship between an independent and dependent variable. It examines whether some of the association between the independent and dependent variable arises due to their association with another variable, with the analysis testing whether this other variable has a significant role in the causal pathway. The effect of the independent variable through the mediator is known as the indirect effect and the effect not through the mediator is known as the direct effect . In this study, mediation analyses examined whether comorbidities and unmet information needs mediated associations between socioeconomic position and the three QoL indicators. Mediation analyses were conducted using the PROCESS macro [36] for SPSS26. This macro utilises a bootstrapping procedure to conduct unstandardised multiple linear regression to provide estimates of the direct and indirect effects of the independent variable. For outcomes, EQ-5D-5L and SDI, unstandardised beta coefficients and 95% confidence intervals are estimated for the direct and indirect effects. For these analyses, if the 95% CIs do not include 0, the association is considered statistically significant. For the binary outcome , logistic regression estimated odds ratios and their 95% CIs. In these analyses, if the 95% CI includes 1, the association is not statistically significant. The percentage of the total effect accounted for by the mediator can be estimated by dividing the IE by the total effect. All mediation models controlled for variables identified as significant in regression analyses described above. --- Ethical approval Ethics approval was granted by the Cancer Council Victoria Human Research Ethics Committee . --- Results --- Survey response rates and sample characteristics The survey response rate overall was 45.3% . While response rates were slightly lower for those 5 years post diagnosis, previous work with this dataset showed no significant difference by time since diagnosis [3]. Table 1 summarises the demographic profile and self-reported comorbidities, response to treatment, information needs and symptom score of participants. There were roughly similar proportions of breast, prostate, colorectal and melanoma survivors in the sample with lowest numbers found for non-Hodgkin's lymphoma survivors . Slightly more participants were male and the majority were under 70 years of age . Time since diagnosis varied by cancer type with the proportion of participants 5 years post diagnosis greater for NHL and least for melanoma . Most respondents indicated their disease had responded fully to treatment. The average decile position on our indicator of socio-economic disadvantage did not differ across cancer types, with the average participant in the 6th decile indicating less disadvantage on average. On average, participants had just over one comorbid condition, with higher levels of comorbidity found for those with NHL and prostate cancer . Frequency of the specific health conditions are outlined in Supplementary Table 1. While most participants did not report any information needs, 28% reported 1-3 needs and 13% reported 4 or more. Women with breast cancer were most likely to report having 4 or more information needs , compared to only 5% of those with melanoma. --- Health-related quality of life Unstandardised and standardised betas from analyses that regressed total EQ-5D-5L scale on cancer type and demographics, information needs, socio-economic disadvantage and comorbidities variables are shown in Table 2. There were relatively small differences in EQ-5D-5L scores in bivariate analyses across the different demographic and disease specific factors. EQ-5D-5L scores were inversely correlated with number of comorbid conditions but positively correlated with socio-economic position . These relationships were maintained in multivariate analyses, with those 80 years or older , those with more information needs and those with a greater number of comorbid conditions more likely to have lower EQ-5D-5L scores. Residing in areas with less socio-economic disadvantage was associated with higher EQ-5D-5L scores . --- Social difficulties index Overall, average SDI scores were low , indicating that on average participants in our study did not regularly experience significant social difficulties. However, there was some variation across different groups. On average, melanoma survivors reported the lowest levels of social difficulties while breast cancer survivors reported the greatest levels . The number of comorbid conditions was positively associated with SDI scores ; mean SDI scores for those with no comorbid conditions was around half that found for those with three or more comorbid conditions . Multivariate patterns of associations were similar to those seen in the bivariate analyses. Having higher information needs increased the likelihood of experiencing higher levels of social difficulties. Comorbidities were positively associated with social difficulties in multivariate analyses, while socio-economic disadvantage was inversely associated . --- Employment Not being in paid employment at the time of the survey was assessed in a subset of the sample who were aged under 60 years and in paid employment at diagnosis . Fifteen percent of this group of participants were not working at the time of the survey. Melanoma survivors were most likely and prostate and colorectal cancer survivors least likely to be working at the time of the survey . Not working was more likely for people aged in their 50 s compared to those aged under 50 at diagnosis and for those 5 years post diagnosis compared to those 1 year post diagnosis. In multivariate analyses, respondents who were 5 years post diagnosis and those aged between 50 and 59 years were more likely to not be working at the time of the survey. While decile increase in socio-economic disadvantage reduced the odds of not being in the workforce by 12% , every additional comorbid condition increased the odds of not working by 61% . --- Country of birth --- Information needs Given the associations between information needs and QoL and social difficulties outcomes, we examined multivariate associations between information needs and disease and socio-demographic variables. Survivors of melanoma, older survivors, men and those with fewer comorbid conditions were less likely to report information needs . --- Mediation models Results from mediation analyses are shown in Table 4. Comorbidities partially mediated the effects of socio-economic disadvantage on EQ-5D-5L and SDI but did not mediate the association between socio-economic disadvantage and work status after cancer for those aged under 60 years. However, while statistically significant, the IE of socio-economic disadvantage through comorbidities for both EQ-5D-5L and SDI was, at 17% and 14% respectively, small. Unmet information needs did not mediate associations between socio-economic disadvantage for EQ-5D-5L, SDI or work status . --- Discussion While QoL outcomes were generally very good for all survivors, this study found that older participants, those with multiple comorbid conditions, those from socio-economically disadvantaged areas and those with more information needs were at risk of poorer QoL outcomes. Cancer type also influenced QoL outcomes with higher HRQoL and lower SDI found for people with melanoma and men with prostate cancer compared to women with breast cancer. While comorbidity status partially mediated the associations between socio-economic position and HRQoL and SDI, it did not mediate the association with returning to work. In addition, information needs did not mediate associations between socio-economic position and QoL outcomes. Our results suggest that survivorship care needs to be tailored to individuals given their cancer, age, comorbidity status and socio-economic position. We used the EQ-5D-5L as our indicator of HRQoL. While this brief non-cancer specific tool allows comparison to population norms, we note it does not provide a detailed assessment of the many dimensions that comprise QoL. Population norms on the EQ-5D-5L for countries similar to Australia in terms of economic development, health system and age of population, range between 0.85 and 0.90 [37]. While average scores on the EQ-5D-5L in our study were slightly lower for those who were 1 year into their survivorship care , for those three or more years post diagnosis scores were around 0.85, suggesting that for the survivors in our study, HRQoL resembled levels found in the general population. Other studies have found that when non-disease specific indicators of QoL are used, with time, cancer survivors experience a HRQoL similar to age matched controls [21,[38][39][40][41]. However, despite the similarity in global HRQoL, detriments in disease specific indicators of QoL including on measures of fatigue, cognitive function, pain and mental health wellbeing tend to persist over time [42]. Similarly mean scores on SDI were generally well below 10, the recommended cut point for indicating that problems with social activities may need further investigation [34]. Despite this, our study found that scores for survivors with three or more comorbid conditions and those with four or more unmet information needs were higher than average. The authors of the SDI suggest that a difference of three or more points on this scale indicates a clinically meaningful difference. Further investigation of survivors with multimorbidity or high information needs is needed with the aim of establishing interventions to reduce social difficulties in these groups. Our study's finding that comorbidities reduced HRQoL and increased social difficulties is consistent with a growing body of literature highlighting the role comorbidities have on QoL [21,22,43]. Chronic health conditions become more common with age. In Australia, 80% of adults over the age of 65 years have at least one chronic health condition, with arthritis , heart disease and diabetes the most common. With 72% of respondents reporting a long-term health condition, comorbidity in our sample is similar to the broader Australian population. Other studies have reported higher levels of comorbidity in cancer survivors; for instance, a recent US study reported that 90% of survivors had a comorbid condition [22], and in a German study of survivors 5 and 10 years post diagnosis, 95% reported comorbidity [20]. Unlike the current study, both included hypertension, with the US study also including high cholesterol. As most chronic health conditions have an independent negative impact on HRQoL [22], findings suggest that quality survivorship care needs to consider the other health conditions that many cancer survivors have. Our study highlighted the importance of addressing cancer survivors' unmet needs with lower QoL and higher social difficulties associated with a greater number of unmet information needs. While the cross-sectional nature of our design means we cannot determine whether unmet information needs are a cause or a consequence of lower HRQoL and higher social difficulties, the strength of the associations found suggests the importance of information to improve QoL outcomes. Similar to others [28,30], we found younger survivors reported more unmet information needs, as did those more recently diagnosed. Socio-economic position was not related to information needs and information needs did not mediate any effect of socio-economic position on our three QoL outcomes. While our results need to be replicated, they suggest that unmet information needs are common across socio-economic groups. We found that a greater number of comorbidities were associated with more unmet information needs. Our findings suggest that quality survivorship programs need to develop strategies to ensure the regular monitoring of information needs of survivors particularly those survivors with comorbidities. We found small but statistically significant associations between socio-economic disadvantage and our three QoL outcomes in multivariate analyses. This is similar to findings from the USA where survivors from lower socio-economic backgrounds have lower health-related QoL outcomes, even after adjusting for presence of comorbidities [18,22]. Given known associations between socio-economic position and comorbidity status, our study formally explored the role of comorbidities as a mediator. Our finding that comorbidities mediated some but not all the association between socioeconomic position and QoL outcomes suggests that strategies are needed to ensure the health outcomes of people from more socio-economically disadvantaged areas are maximised. --- Strengths and limitations While the study has several strengths including large sample size, good representation from different cancers and inclusion of survivors at 1, 3 and 5 years post diagnosis, several limitations should be acknowledged. The crosssectional study design means we cannot determine the direction of associations we identified. While a survivor bias is likely to be evidenced, as the difference in the responses of those 5 years and those 3 years post diagnosis was small we think this bias will be minimal. We did not assess when people were diagnosed with their comorbid condition in relation to their cancer nor did we assess the impact of this condition on respondents. Thus, whether lower QoL scores for people with comorbidities are due to the comorbidity entirely or a combination of cancer and comorbidity is not known. The associations between socio-economic position and QoL outcomes were relatively small. As our use of an area-level indicator of socio-economic position may underestimate this association, more work is needed to understand associations between socio-economic position and QoL outcomes. Our unmet information needs indictor reflected the number of needs rather than the type or degree of need experienced. The lack of nuance in this measure is a limitation. Further work exploring how different types and levels of needs influence QoL is required. Cancer survivors from areas with greater socio-economic disadvantage were underrepresented in our sample, as were survivors from a non-English speaking background. While we controlled for impact of disease response in analyses, we did not control for stage of disease at diagnosis or treatment. Finally, the low response rate must be noted and may give rise to some of the potential biases noted. While a response rate under 50% suggests that caution is needed when extrapolating findings, we note that this response rate is comparable to other patient studies in Australia using cancer registries for recruitment [44,45]. --- Conclusions Using a population-based cancer registry to identify cancer survivors at 1, 3 and 5 years post diagnosis, we found that survivors from more socio-economically disadvantaged areas, those with more comorbidities and those with more unmet information needs were consistently more likely to report poorer outcomes on three QoL indicators: healthrelated QoL, social difficulties and employment. Our results show the importance of establishing survivorship care programs that can regularly assess and address the information needs of cancer survivors. Importantly our results show the need to establish survivorship care plans that consider the overall health of the survivor including other health conditions they may have at diagnosis and those that develop subsequently. Being aware of any socio-economic barriers that may reduce access to quality care and implementing strategies to reduce these barriers may assist in achieving optimal outcomes for all cancer survivors. --- Data availability The data that support the findings of this study are available from the corresponding author on reasonable request. The data are not publicly available due to privacy and ethical restrictions. Code availability Not applicable. --- --- --- Author contribution --- Consent for publication Not applicable. --- Conflict of interest The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. --- Authors and Affiliations

Purpose To examine how socio-demographic, comorbidities and information needs influence quality of life (QoL) outcomes of survivors of breast, colorectal, or prostate cancer, non-Hodgkin lymphoma or melanoma. Methods Cross-sectional postal survey with eligible participants identified through a population-based cancer registry. QoL outcomes were assessed by EQ-5D-5L, social difficulties index (SDI) and, for those employed at diagnosis, current employment. Regression analyses explored associations between outcome variables and cancer type, age, time since diagnosis, residential location, socio-economic disadvantage, comorbidities and unmet information needs. Mediation analyses examined whether comorbidities and information needs explained relationships between outcome variables and socio-economic disadvantage. Results 2115 survivors participated. Mean EQ-5D-5L scores (mean = 0.84) were similar to population averages and SDI scores were low for the entire sample (mean = 3.80). In multivariate analyses, being aged over 80, greater socio-economic disadvantage, comorbidities and unmet information needs decreased EQ-5D-5L scores. Higher SDI scores were associated with socio-economic disadvantage, comorbidities and unmet information needs. Not being employed was associated with being aged over 50, more comorbidities and socio-economic disadvantage. Comorbidities but not information needs partially mediated the impact of socio-economic disadvantage on EQ-5D-5L and SDI accounting for 17% and 14% of the total effect of socio-economic disadvantage respectively. Neither comorbidities nor information needs mediated the association between socio-economic disadvantage and employment outcomes. Conclusions To improve quality of life, survivorship care should be better tailored to address the needs of individuals given their overall health and impact of comorbidities, their age and type of cancer and not simply time since diagnosis.

Introduction: People experiencing homelessness are disproportionally a ected by the COVID-pandemic. The realities of their daily lives have been given little consideration in the pandemic response. They are not represented in existing health information campaigns, and many are structurally excluded from digital information. The project aimed to develop inclusive COVID--information material to strengthen infection prevention and control of PEH. Material and methods: In a participatory process, PEH were involved in the planning, production, and evaluation of poster and video information material on COVID-. Various stakeholders were consulted for external supervision. Service providers all over Germany were informed about the material that could be ordered free of charge. For the evaluation, semi-structured interviews with homeless service providers and PEH were conducted, and the online views of the videos were measured. --- Results: Sixteen PEH participated actively in the project. Two COVID--information videos were launched in languages in February . Posters promoting vaccination against COVID-were produced in languages. As of May , the videos have been viewed more than , times. A total of service providers for PEH and public institutions received the posters, thereof upon request. Twelve service providers and PEH participated in the evaluation. They pointed out the lack of targeted information material for PEH. The consideration of the concerns and the diverse representation of PEH was perceived as particularly important. Most of the service providers were unable to show the videos due to technical and spatial limitations. Digital challenges for PEH, like the lack of and maintenance of a smart phone, became apparent. --- Conclusion: The cooperation of research, practice and the community were key factors for the realization of this project. Strong links to the community and the involvement of relevant stakeholders are indispensable when working --- Introduction People experiencing homelessness are disproportionally affected by the COVID-19 pandemic. Precarious living conditions on the street, in encampments and cramped shelters, limited access to hygienic supplies and prevention measures, stigmatization, marginalization from social, political and economic resources as well as exclusion from health services result in high rates of underlying health conditions and a high risk of SARS-CoV-2-infection . The prevalence among homeless individuals may be similar to that found in the general population, however, the increased risk of outbreaks with high infections rates has to be considered . Also, social determinants and pre-existing health conditions place PEH at higher risk of severe COVID-19 infection . However, studies to properly assess the outcome of SARS-CoV-2 infection in PEH are still required . Since the beginning of the COVID-19 pandemic, the need to consider the living conditions of PEH when implementing measures of infection control and prevention has been addressed by the German national working group on homelessness services . In July 2021, the Robert Koch Institute, the national public health institute in Germany, published targeted recommendations for COVID-19 in the context of homelessness that were developed together with experts from the field . Many good practice solutions were implemented locally to protect PEH during the pandemic, such as provision of adequate isolation and quarantine options addressing possible complex needs of PEH, 24/7 accommodation with single rooms, regular voluntary universal testing for SARS-CoV-2, and mobile vaccination campaigns . However, in many places, isolation, access to vaccination and testing as well as targeted information remained a challenge. This is particularly critical because it is known from past respiratory viral outbreaks that these control measures are crucial in managing epidemics or pandemics . In our opinion, PEH in Germany are to date not sufficiently addressed in the pandemic response including information campaigns. With the onset of the pandemic, digitalization was discussed more publicly than before as it affected everyone to a substantial extent. Being digitalized took on a new relevance, as it was, for example, a prerequisite for the digital EU certificate, which allowed unrestricted access to public buildings, use of transport and freedom to travel between international borders . Multiple problems faced by PEH result from their precarious socioeconomic situation which also affects the ability to maintain a digital device and to have access to internetbased services . Digital inequalities result in further social exclusion as it limits career opportunities, represents a barrier to maintaining social and service-related contacts, causes financial hardship and are a determinant of health . At the same time, digitalization can be an opportunity for better social inclusion . In a previous COVID-19 project among PEH, we identified the necessity to address language barriers, to include digital information formats, and to use participatory approaches considering homeless people's needs and life situation . This follow-up project aimed a participatory development of inclusive health communication on COVID-19 to strengthen options for IPC for PEH. We describe the development, implementation, and evaluation of targeted digital as well as non-digital health communication material . --- Materials and methods This project was conducted over a period of 11 months, from October 2020 to August 2021 in Berlin, Germany. According to the principle that in participatory projects participants should benefit directly from the research process , the perspective of PEH represented the basis for all project steps, and all PEH had decision-making power and were paid for their work effort. --- Study team and recruitment of community partners The study was initiated and supervised by two experts from the fields of medicine and public health. The study team also included a clinician, a health scientist, --- FIGURE Presentation of study team with expertise and responsibilities, including people experiencing homelessness as community partners and various stakeholders. two social workers, two student assistants and a professional communications designer. Four of the team members had longterm work experience with PEH. The PEH who participated as community partners were actively involved in the planning, production, evaluation and decision-making processes. They were the protagonists with one main actor for the videos and 12 further actors for portraits in the videos and posters. Three PEH had reservations to be filmed or photographed, and were involved in translations, sound recordings or evaluation rounds. Community partners were recruited in a social facility for PEH operated by the Berliner Stadtmission in Berlin. The facility included a 24/7 shelter, a medical outpatient clinic and a clothing store, where people could easily be approached during the day and in the waiting areas. Some members of the study team were already known to the PEH from their work at this facility, which formed the trust base to approach people directly. We showed a short film clip to the PEH in order to introduce them to the project. The communication designer created the clip especially for recruitment purposes. Inclusion criteria were age above 18 years, current or previous homelessness and written informed consent. We aimed at partnering with PEH from different age groups, genders, abilities, languages and country of origin to reflect the diverse image of PEH in Berlin and to create material that is easier to identify with. We consulted the following stakeholders for external supervision: the German national working group on homelessness services , a self-organization , the coordination group for homeless shelters in Berlin , the Robert Koch Institute as well as staff of the Berliner Stadtmission . --- Videos Scripts in German language were created for two videos with general information about COVID-19 and SARS-CoV-2 testing. Expertise of the medical specialists and social workers and the experience of the community partners were combined to gain a better understanding of the health, personal needs and coping strategies of PEH during the pandemic. The scripts were edited in a multi-step revision process. After they were shared with and adapted by all stakeholders, the main protagonist translated them to a simple and clear language. Native speakers translated the scripts into four other languages that were identified in the previous project we performed . Quality control was performed by professional translators. The main actor speaks German in the videos. Three PEH and one professional actor have dubbed the clips into Russian, Polish, Romanian and English. --- Posters Information posters with precise key messages on access to vaccination were developed and provided in a digital and printed version to support the public COVID-19 vaccination campaign for PEH that started in Berlin in March 2021. To identify main questions or concerns of PEH regarding vaccinations, we consulted PEH and staff of service providers, as well as our stakeholders. --- Participation and participatory decision making The shooting locations were chosen together with the protagonists. The aim was to find locations that would create a pleasant working environment and at the same time provide images that PEH could identify with and which would not reproduce stereotypes. For the whole process, the individual needs of the protagonists were considered. This included for example a transport service if needed, food and drinks, as well as access to barrier-free sanitary facilities at all locations. The selection and editing of the video material and photo motifs took place during several feedback rounds with community partners. The videos were watched together several times on a screen in the common room of the homeless shelter. Attending community partners and other PEH were asked for their opinion and criticism, e.g., about the content and the locations. Stakeholders who were not directly involved in the production were also asked for feedback. --- Dissemination of the information material On April 2021, a website was created in order to provide open access to the videos and posters, as well as to provide further information about the project . Furthermore, the material was disseminated through various social media channels . The printed posters were sent to all homeless service providers that were part of the Kältehilfe Network in Berlin . The German national working group on homelessness services informed service providers all over Germany through their network about the project and the offer to order the posters free of charge. --- Evaluation and data analysis All institutions that received material were listed. A telephone survey with randomly selected homeless service providers who had received the printed posters was conducted. Recruitment for the survey took place via email. After written informed consent, a semi-structured telephone interview was conducted for 15 min addressing the practical implementation and perception of the posters and the videos. A social worker interviewed PEH in two shelters of the Berliner Stadtmission to determine how the material was perceived by PEH. Participants were randomly approached during the service hours of the shelters. After written informed consent, a semi-structured interview was conducted for 20 min. The data analysis was based on written notes taken during the qualitative phone interviews and face to face interviews. A systematic qualitative content analysis was undertaken . After reviewing the data material, it was coded by an inductive procedure and summarized in categories. Furthermore, the usage of the project website with the number of video views was measured. --- Ethics This study was approved by the Ethics Committee of the Charité -Universitätsmedizin Berlin . All PEH who contributed to the implementation of the project were paid for their work. The study was explained to PEH in the preferred language, and written informed consent was provided for participation. The scope and time frame of the project were transparently communicated, as well as the possibility to withdraw participation at any point of the project without repercussions. As PEH are a particularly vulnerable group, privacy, data security and a familiar atmosphere were taken into careful consideration during the interviews. --- Results --- Participatory production Sixteen PEH participated actively in the production of the information material. All participants were experiencing homelessness at the time of the study. Ten protagonists were recruited in the 24/7 shelter, 3 in the clothing store and 3 were recruited in the streets during the shooting days. Four of the participants were women and 12 were men, aged between 25 and 75 years from 6 countries speaking 8 different languages, two of them were in a wheelchair. We produced two multilingual COVID-19 information videos under the slogan "We keep Corona off the streets". The protagonist changed some parts of the script for a better understanding. For example, regarding information about the opioid substitution programme in the quarantine facility, --- FIGURE Multilingual poster disseminated digitally and in print to support the public COVID-vaccination campaign for people experiencing homelessness . "substitution is provided" was replaced by "you can stay there, also if you consume". The 5-day shooting of the videos took place at 15 locations. The videos were produced with 13 protagonists. Six protagonists chose spots on the grounds of the Berliner Stadtmission or in the near vicinity as shooting locations, while 7 chose spots around the main railway station, a nearby park and public places. The first video clip contained general information about COVID-19 . It explained the transmission modes of SARS-CoV-2, symptoms, increased risk of infection among PEH and strategies for selfprotection . The second video clip contained details about COVID-19 testing and the proceeding after a positive test result for PEH who lack the possibility of self-isolation at home . Thus, the video talks about the possibility of isolation in quarantine accommodations that consider the needs of PEH in a sensitive way. In the second step, we designed multilingual posters with seven different motifs . With two versions, we have covered a total of nine languages identified as most relevant -1st language version: German, Polish, English, Farsi, Russian -2nd language version: German, Romanian, Bulgarian, Arabic, French Out of the 13 protagonists in the videos, three participated in the production of the posters. Two more were recruited in the 24/7 shelter. Portraits were taken on two shooting days. The main questions or uncertainties about the SARS-CoV-2 vaccination among PEH were identified through discussions with our community partners and staff of homeless service providers. These were having access to vaccination without a permanent address, documents or health insurance, as well as implications of drug use. Accordingly, the posters were designed with the slogan "You can get vaccinated. Even with no fixed address, no health insurance, no documents. Get informed in the social or health care services you know". In another poster version we included the sentence "Even if you use drugs". The posters that address drug use do not contain portraits of PEH to avoid stigmatization . As a result of the feedback rounds, the hybrid nature of the posters was extended by including a QR-code linking the poster with the project website. --- Dissemination The videos were launched on February 11, 2021 • Berlin Kältehilfe list • online search • homeless service providers • source unknown Diverse and barrier-free material was appreciated Brochures and flyers continue to be popular in respondents' workplace. Regardless of the type of materials, importance was placed on: • a simple, clear, and concise message All interview partners used the posters in locations for PEH Positive aspects that were mentioned: • appropriate for the counseling context • multilingualism • professional design • diversity-sensitive • simple and clear language, suitable for functional illiterates • accepting drug policy • opened the conversation on vaccination and enabled further counseling • initiated conversation about taboo topics • enabled PEH to inform themselves 'quietly' in their preferred language • The following aspects could be considered in the posters: • poster version for those who are undecided about vaccination • clear reasons/arguments pro vaccination • clarification about fake news and vaccination myths • more diversity for example in terms of women Posters can be placed in different locations such as: • outpatient clinic buses • shelters • consulting sites • restrooms • train stations • community welfare centers • --- Achievements Challenges Recommendations Reasons for not using the videos: • lack of time The answers were structured in categories commenting on achievements, challenges, and recommendations. In the case of multiple mentions of a specific aspect, the number of mentions is given in brackets. a All 12 participants were familiar with the posters before the interview. b Six out of 12 participants were familiar with the videos before the interview. was automatically sent out to all 91 institutions that were part of the Berlin Kältehilfe list, the others received the material upon request. The institutions included facilities for PEH such as clothing facilities, day care centers, consulting services, hygiene facilities, medical facilities and night shelters , facilities for drug users , facilities for refugees and municipalities and public facilities such as a library and public authorities . The distribution period lasted from February 2021 until May 2022. The dissemination via the social media channel of the Charité -Universitätsmedizin Berlin took place in February 2021. Between February 8, 2021 and May 31, 2022, the videos have been viewed 2,064 times via the project website. We registered peaks in the numbers of requests of the posters at the beginning of the vaccination campaigns in spring 2021 and then again in winter 2021. At that time, the booster vaccinations started, the night shelters opened for winter season, and the institutions were informed about the material by email. --- Evaluation Out of 20 homeless service providers that were invited to the evaluation, 12 agreed to participate in a telephone interview --- Positive aspects Negative aspects Recommendations • statements perceived as clear and visible • variety of languages used • people feel addressed/can identify Positive aspects on the choice of motifs: • underprivileged people are the protagonists • individual portraits • variety of motifs Motifs with PEH or urban motifs do not appeal to everyone because it can remind people of their own harsh realities Excessive text --- Unsuited for illiterates The following aspects could be considered in the posters: • also depict drug consumers • utilize city landmarks for identification • portray more women • portray disabilities more obviously • use of one language per poster is clearer The videos were appreciated and perceived as: • • concise • easy to understand People felt addressed and could identify: • with the protagonists • with the speaker • with the diversity of protagonists • due to the realistic representation "It's the reality" Suggestions to improve the videos: • voice could be livelier • to depict individuals who take drugs • to provides Turkish and Arabic translation • key messages conveyed at the beginning of the videos The answers were structured in categories commenting on positive and negative aspects as well as recommendations. In the case of multiple mentions of a specific aspect, the number of mentions is given in brackets. a Three of the 8 participants had seen the posters before. b None of the participants had seen the videos before the interview. which were conducted by two of our study team members in August 2021. At the same time, another team member interviewed 8 PEH on 2 days in front of a shelter and in a day center. All were experiencing homelessness at the time of the study. Out of the notes that were taken during the interviews, thirteen categories of different topics were identified . Not all categories were addressed by everyone. There was a general appreciation of how the information was presented in terms of acceptability and sensitivity toward the targeted population. All service providers approved that the posters were widely used for the vaccination campaigns. All 3 respondents who knew the videos considered them helpful for providing targeted information and allowing PEH to feel addressed directly. However, the streaming of the videos in the services was reported to be difficult due to lack of digital devices. Only 1 institution reported to have used the videos within the counseling context. Eight out of 12 service providers pointed out the digital challenges for PEH, e.g., owning and maintaining a mobile or smart phone, and the lacking digital infrastructure in services for PEH. It was acknowledged that digital tools and offers can be a chance to provide better health and social care for PEH, also those living in hidden homelessness. All respondents spoke about the lack of targeted information material for PEH in general. It was suggested to address a wide range of further topics with targeted digital information material. Out of 8 PEH, 3 had seen the posters in relevant facilities of homelessness services in Berlin. None had seen the videos before. Participants appreciated that PEH were protagonists of the videos and posters, but even more diversity would have been appreciated. They could identify with the material and pointed out that mentioning consuming drugs and alcohol was crucial. It was reported that the posters encouraged PEH to think about a vaccination. Precise information on vaccination offers in the respective locations would have been helpful. Both, dissemination of information through posters and videos was appreciated. It was pointed out, that visually and auditorily appealing videos could be suitable especially for illiterate people. Respondents appreciated the idea of disseminating information via videos. One person addressed the digital gap, saying that it was only useful if you had the convenience of owning a digital device. --- Discussion The participatory approach involves all levels of knowledge A key factor for the realization of the videos and posters was the set-up of an interdisciplinary team together with community partners and thus bridging the gap between research, practice and community. Of particular significance were good contacts of the study team with homeless service providers and PEH. Furthermore, it was crucial to involve various local and national stakeholders right from the beginning of the project for advice, support and distribution of the materials. The participatory approach enables active generation of knowledge together with practice and communities . Accordingly, we addressed hierarchies transparently and valued PEH's own life experiences as equivalent to knowledge from professionals within the fields of social work and public health. A challenge was to deal with poverty and precarious living conditions in a sensitive way. Poverty was neither to be tabooed nor trivialized. The goal was to show a realistic picture of the social system and living environment of PEH without reproducing stereotypes or exposing people. The image and portrayal of homelessness in times of the pandemic was therefore to be determined primarily by the community partners themselves. The main protagonist decided to participate by stating "I think I'm someone people accept". In front of the camera, most PEH placed emphasis on a proud attitude showing an active and upright posture. For the evaluation, 8 PEH and 12 service providers were interviewed. They acknowledged the sensitive implementation of the project. The consideration of the concerns and the diverse presentation of PEH was perceived as particularly important. In this manuscript we have chosen the term "people experiencing homelessness" because it presents homelessness as one aspect among others and is less generalizing than the term "homeless people". --- Participation and its limitations To enable participation according to the needs of the community partners, work conditions were defined and arranged together. For example, the main protagonist's condition for participation was the assurance of permanent access to barrier-free sanitary facilities. He knew from personal experience that this would be a main challenge, especially during the pandemic when sanitary facilities were even less available for PEH. This highlights one of the fundamental problems that people living in the streets face on a daily basis. For some people there were barriers that made participation difficult or impossible such as hidden homelessness, illegalization, and precarious employment. We have tried to enable safe participation for them as well. For some, a solution was to participate by dubbing the videos, translating the scripts or by attending the feedback sessions. Others chose to not participate altogether. Our impression was that it has been easier for people to participate if there was pre-existing contact with project staff or other PEH who had already taken part. In regard to the evaluation, we found it easier to find interview partners in the day center which provided a safe setting for the interview compared to the recruitment in front of the shelter where the interviews were performed outdoors. It is crucial to be aware of and respect the right to non-participation. We did not collect socio demographic data from the PEH who participated in the production and evaluation of the materials. All were experiencing homelessness at the time of the study. Some had shelter available. Similar to varying infection risks, differences in the participation and responses between sheltered and unsheltered individuals are possible, but this could be not further analyzed in this study . --- Impact of the information material Despite the mainly positive feedback in the evaluation, the impact of the videos could only be verified to a limited extent. Most of the homeless service providers were unable to show the videos on their premises, due to technical and spatial limitations. The videos and posters were widely distributed via social media and email lists. It remains unclear to what extent they reached the actual target group. However, considering the important role that social media has in shaping people's state of information and attitudes toward public health interventions such as vaccination campaigns, efforts should be increased to utilize these means of communication . The posters seem to have been particularly useful for the vaccination campaign according to the feedback of service providers and the number of poster-orders throughout Germany. Addressing the specific questions and concerns of PEH-e.g., having access to vaccination without a permanent address, documents or health insurance, as well as implications of drug use-may be one approach to increase vaccination coverage . The type of institutions that ordered the posters reflect the broad range of services that address the complex needs of PEH . Among them were services targeting people without official documents , people living in poverty, people using drugs, as well as services exclusively provided for women. The way how people are approached as well as the information that is provided must be tailored to PEH's situation and needs. The videos and posters of this project demonstrate a step into this direction. A concise point that was emphasized by staff in homeless service providers is that no information material can replace personal contact and face-to-face conversation. Relationships to social workers and medical staff remain essential-especially when people are in precarious situations, access to information is difficult and social contacts are limited. --- Limitations of the information material In the evaluation, it was critically mentioned that the material provided only general information without any details on how to get access to vaccinations. As the provision of tests and vaccinations were locally organized in various different ways, specific information could not be integrated into the material. Some institutions therefore added distinct information for the local context directly on the posters. The material intended to encourage testing and vaccination. One of the statements was that both is possible despite alcohol, other substance usage, and without official documents. Although this information was in line with official regulations, we could not guarantee appropriate implementation. In fact, reports from homeless service providers and PEH showed that people lacking identification documents had difficulties accessing public COVID-19 test and vaccination centers . The material also stated that people with positive test results would be cared for according to their needs. However, it was frequently reported to us that quarantine and isolation capacities for PEH were insufficient, and that substitution or medical care was only partially offered. --- Benefit and challenges of digital inclusion among PEH During the pandemic, interpersonal contact became limited and digitalization crucial. Most of the population has been able to continuously access updated information about the pandemic, whereas many PEH were excluded from this information flow due to technical and socio-economic reasons that make it difficult to acquire or maintain digital devices. Even with an internet-enabled device, access to information remains difficult with a lack of free Wi-Fi and a limited access to electricity. When fighting a pandemic, it is important to reach all people in society equally, albeit through different channels and in different languages. To ensure this, digitalization has to be facilitated for social and health institutions as well as for PEH . The digital gap also had to be considered in the planning and implementation of our project. Our community partners without mobile-or smartphones were visited personally by our study team for recruitment and further collaboration. People who were temporarily sheltered in a 24/7 facility were easier to locate than people living on the streets. Creative and individual ways had to be identified to stay in contact with all community partners, regardless of their housing or digital situation. This shows the daily challenges of PEH without digital access, that also affects the interaction with friends and relatives, employers, social and medical service providers as well as with public institutions. However, it has to be noted that PEH are not per se excluded from digitalization. A considerable number of PEH manage to purchase and maintain a mobile phone or smartphone and find ways of charging it and accessing internet. --- Progressive health communication There was a wide range of institutions that ordered the elaborated posters, including a library, facilities for people who use drugs as well as different facilities for specific populations. This demonstrates that PEH can and need to be reached in different ways that consider their complex needs and situations including hidden homelessness. Particularly with a group in which digitalization is highly variable, hybrid material enables analog and digital use simultaneously. For personal contact with medical professionals or social workers, the integration of digital formats such as online counseling should also be considered. --- Recommendations Valuable lessons were learned through the project, that can help to strengthen participation of PEH and to consider their perspectives in health communication strategies . Engaging marginalized populations as community experts and partners in the planning, implementation and evaluation of projects is crucial to adequately grasp their situation and needs. Strong links to the community, trust and the involvement of relevant stakeholders are indispensable when working with PEH. Progressive health communication in terms of hybrid information material considers the heterogeneity of the target group, identifies the specific needs and addresses them concretely but sensitively, without stereotypes. The material should be readily accessible, multilingual with a simple and clear message, addressing taboo topics while being designed in a discrimination-sensitive way. In this project people were addressed in their various spoken languages, including minority languages and a clear and simple language for functional illiterates. The experiences and feedback of peers and communities must be included in the development of information material. Access to mobile and digital devices positively impacts the daily lives and health of PEH. Sustained use of the devices requires access to electricity and Wi-Fi as well as safe storage facilities. Social institutions should receive financial and professional support to develop a digital infrastructure and have it available for its users. Exclusion from information, on the other hand, is a major factor in the structural marginalization of PEH. Closing the digital gap can be a contribution to counteracting social and health inequalities. --- Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. --- Ethics statement The studies involving human participants were reviewed and approved by the Ethics Committee of the Charité -Universitätsmedizin Berlin . The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual for the publication of any potentially identifiable images or data included in this article. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. --- Supplementary material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh. 2022.1042677/full#supplementary-material --- SUPPLEMENTARY VIDEO General information about COVID-. --- SUPPLEMENTARY VIDEO COVID-: testing and quarantine .

with PEH. Exclusion from digital information is an increasingly important component of the structural marginalization of PEH. Digital inclusion for PEH and service providers can help to counteract social and health inequalities. The lessons learned through this project can contribute to strengthen participation of PEH and to consider their perspectives in future health communication strategies.

Background Rates of injection drug use have grown in parallel with the US opioid epidemic since 2000 [9,12]. IDU is a growing public health concern given the associated risk of the transmission of blood-borne diseases, such as human immunodeficiency virus , hepatitis C virus , and of skin and soft tissue infections [9], [24]. Common IDU-related SSTIs include skin abscesses, cellulitis, osteomyelitis, and infective endocarditis [6]. Bacteria commonly found in skin flora, commonly Staphylococcus aureus, may contribute to IDU-related SSTIs when introduced into the subcutaneous space during injection [1,12]. In people who inject drugs with particular constitutional vulnerabilities , these potential pathogens, as well as others, may also seed deeper tissues leading to infection nidus, abscess, and/or septic emboli. PWID have demonstrated higher rates of S. aureus colonization in skin flora than the general population, that has been posited to potentially predispose them to a greater risk of cellulitis and cutaneous abscess [1]. While S. aureus is by far most commonly identified pathogen, identified in > 75% of SSTIs [3], other microorganisms include streptococci, gram-negative bacilli, and fungi [16]. Additionally, a study by Harris et al. found the overuse of acidifiers, such as citric acid, in preparing substances to be associated with increased odds of prolonged SSTI and inflammation . Venous damage as a result of the use of acidifiers can lead to increased risk of IDU-related SSTI development . Infectious diseases among people who use drugs are associated with a variety of social determinants and behavioral factors . Infections associated with IDU disproportionately affect non-Hispanic white Americans experiencing economic distress and populations living in rural and suburban areas [9], [24]. Sociocultural factors surrounding injection drug use place females at a higher risk of developing SSTIs, such as relying on other individuals to prepare their drug, increased potential for shared injection equipment; and greater difficulty finding veins [4,14,26,29,32]. Individuals recently experiencing homelessness are also more likely to report a history of SSTI, potentially relating to injecting in unsafe or unsanitary environments [4,14,26]. Behavioral factors, such as high risk injection practices, are commonly associated with SSTIs among PWID [26], Galea and Vlahov . Using and sharing contaminated single-use syringes, needles, and ancillary equipment places individuals at higher risk of spreading infectious agents; prior HIV and HCV outbreaks throughout the USA have been closely associated with the sharing of such equipment [14,23,24], Galea and Vlahov 2002). In addition to non-sterile injection equipment, previous results suggest water sources for drug dilution also play a role in the transmission of bloodborne diseases and bacterial infections [22]. A lack of proper skin cleaning technique, such as using alcohol wipes, also confers higher SSTI risk [20]. The type of injection has also been shown as a major risk factor associated with abscess formation, and 'skin popping' and 'muscling' are highly associated with SSTI development [15,25,32]. Similarly, individuals injecting in their hands, groin, or legs are also at higher risk of developing an infection (Hope et al. [15]. Regardless of the location of injection, repeated injection attempts at the same anatomical site increases overall risk of tissue and venous damage and bacterial inoculation [29]. Several studies have demonstrated the varying risks of injecting different illicit substances. Injecting a mixture of heroin and cocaine, otherwise known as 'speedball, ' puts individuals at a higher risk of developing abscesses; this is thought to be due to the synergistic effects of local irritation and tissue vasoconstriction [15,29], Galea and Vlahov 2002;Lloyd-Smith et al. 2008). The use of methamphetamine is a growing concern throughout the USA, with nearly sixfold increases in the presence of methamphetamine found in urine drug samples [33]. Additionally, methamphetamine use has increased since 2015 among people reporting any opioid use, with threefold increases being seen among people reporting heroin use [30]. Chronic methamphetamine use is associated with a higher risk of contracting infections, and individuals who use methamphetamine suffer from altered judgment and reduced inhibitions, potentially provoking them to partake in risky behaviors and thus increasing their overall risk of acquiring bacterial microbes and other opportunistic infections [28]. Hospitalizations to treat IDU-related SSTIs have risen in conjunction with increasing incidence of such infections [18]. Younger populations are at increased risk of hospitalization due to IDU-related infections [18], Nenninger et al. [21]. Between 1998 and 2001 in the USA, Takahashi et al. [31] estimated that there were 106,126 hospitalizations for IDU-related SSTIs, costing an estimated $193 million annually, and these patients were often uninsured or covered by publicly funded insurance programs, such as Medicare or Medicaid [31]. Contrarily, individuals who frequently develop IDU-related SSTIs have been shown to use a variety of self-treatment methods in lieu of seeking professional medical treatment, potentially due to their confidence in self-diagnosis and treatment due to the commonality of such infections [19]. The prevalence of opioid prescriptions and related overdose deaths has increased over the past several years throughout the USA, with rural areas being disproportionately affected; particularly, the Midwest has seen dramatic increases in opioid-related morbidities and mortality rates [8]. In Wisconsin, rural counties facing economic distress report some of the highest opioid overdose rates [27]. In addition to overdose, rural areas may be disproportionately affected by IDU-related SSTI hospitalizations because access to harm reduction services is often limited in these areas. The lack of resources could be preventing PWID from easily accessing safe injection supplies and subsequently increasing their risk of infection [21]. Limited studies have been conducted to understand injection drug use among rural-dwelling individuals. Similarly, infection risk and SSTI treatment practices among this population have yet to be explored in-depth. Given the limited resources and barriers in accessing treatment within these communities, it is important to understand their current practices and needs in order to promote more targeted interventions to decrease the risk of SSTIs. Research among rural counties is merited to determine which unsafe injection practices and substances are closely correlated to history of SSTI and how harm reduction strategies could target this population to more effectively reduce rates of IDU-related SSTIs. This study aims to: estimate unsafe injection practices currently used among PWID residing in rural areas, particularly among those with a history of IDU-related SSTIs; assess treatments utilized for IDU-related SSTIs among rural residents; and gather data that could guide future interventions and educational materials aimed at reducing the overall risk of IDU-related SSTIs. --- Methods --- Institutional review This study involved human participants and was reviewed by the Institutional Review Board of the University of Wisconsin-Madison. Written informed consent to participate in this study was provided by each participant. --- Survey development This study is part of a larger study assessing unmet health care needs among rural-dwelling PWID. The one-time, cross-sectional survey used audio computer-assisted selfinterview software to assist in ease of administration. Between March and May 2019, a total of 13 questions specific to SSTIs were developed and added to the overall survey, making the survey a total of 211 items. The 211-item survey, which included the SSTIspecific questions, was then administered to participants. These questions were adapted from Phillips and Stein's [25] Bacterial Infections Risk Scale for Injectors , a 17-item scale aimed at capturing a variety of behavioral practices closely associated with increased risk of SSTIs among PWID [25]. Similarly, the SSTI-specific questions created and adapted for this study focused on behavioral practices related to IDU, including skin-cleaning methods and water sources used for drug dilution. Additionally, information surrounding injection type and anatomical location of injection was collected. SSTI prevalence was estimated based on if individuals reported developing an infection within the year prior to responding to the survey. If a respondent cited SSTIs within the year prior to response, treatment methods used for their most recent SSTI were collected. Pre-existing questions within the overall survey collected additional data surrounding demographics, syringe and needle sharing behaviors, hospitalizations related to SSTIs, and primary drug of choice. Demographics information for survey respondents included: sex, age, race, education level, insurance status, and history of experiencing homelessness. All participants resided in rural communities. --- Recruitment The larger study recruited a total of 998 participants through Vivent Health, a statewide organization offering a large, multi-site syringe exchange program. This provided an opportunity to survey a geographically disperse population of PWID in rural communities across Northern Wisconsin. Participants were recruited using respondent-driven sampling [13]. Using a respondentdriven sampling approach allowed for the recruitment of individuals who may not utilize the syringe service program and may be at higher risk of adverse outcomes associated with injecting drugs. The larger survey was administered from September 2017 to July 2019 at six different Vivent Health locations in Brown, Douglas, Eau Claire, La Crosse, Outagamie, and Marathon counties in Wisconsin. The revised survey containing the SSTI-specific questions was administered between May and July 2019 and received a total of 80 responses from all six locations. While the larger study sought to enroll a specific number of participants, the SSTI questions were added midway through data collection for the larger study and served as exploratory data. Therefore, a specific sample size was not sought after for responses to the SSTI-related questions. Eligible respondents were 15 years or older, injected drugs within the 30 days prior to survey response, and resided in a rural community. Eligible participants were directed to their local Vivent Health location to simultaneously fill out the ROI survey and receive rapid HIV, HCV, and syphilis testing. All respondents were compensated $20 in cash for completing the survey and an additional $10 for each peer referred to the study. --- Statistical analysis Logistic regression models were used to estimate the odds of SSTI based on demographics and injection practices, for categorical variables. Chi-square tests were used for assessing overall associations of SSTI with specific skin cleaning methods and water sources. Mann-Whitney-Wilcoxon tests were used to assess associations with SSTI history and ordered variables. Significance was assessed at the alpha = 0.10 level. SSTI history was defined as a reported SSTI within a year prior to survey response. Analysis was conducted using R version 3.6.3. --- Results In total, 80 responses to the SSTI-specific questions were collected and analyzed. Shown in Table 1, respondents were primarily Caucasian males between the ages 30 and 39 who had received a high school diploma or GED , received health insurance through Medicaid , and were unstably housed within the 6 months prior to survey response . Respondents most commonly reported methamphetamine as their primary drug of choice, while 25% of respondents reported heroin. Of the 80 respondents, 62 individuals reported no history of skin and soft tissue infections within the year prior to survey response, while 18 individuals reported a history of SSTI during this same time frame. Females were more likely to report of history of SSTI in comparison with males . Shown in Table 2, respondents reported using a variety of sterile skin cleaning practices prior to injecting, including alcohol pads , soap and water , hydrogen peroxide , or Purell . However, respondents also reported using tap water , saliva , and never cleaning their skin prior to injecting . Participants reported using sterile skin cleaning practices were no less likely to report a history of SSTIs in comparison with those who did not . Respondents citing less frequent hand washing prior to injection had no increased odds of reporting a history of SSTIs . Water sources for drug dilution bore a significant relationship with SSTIs, and individuals using purified sources of watersuch as water purchased from a store or sterile water obtained from syringe exchange programs-were more likely to report a history of SSTIs in comparison with those who used non-purified sources of water-such as tap water from either their home or a public bathroom sink . No respondents reported the use of nonsterile water from a public or private toilet. Regarding injection site Table 3, individuals reporting 'skin popping' and 'muscling' were significantly more likely to report a history of SSTI in comparison with those injecting intravenously . Individuals reported injecting into various anatomical locations, including cubita fossa , forearm , hand , and upper arm . The anatomical location of injection was not significantly correlated with a history of SSTI . Individuals able to inject on their first attempt more frequently were significantly less likely to report a history of SSTI . Frequency of injection-whether daily, weekly, or monthly-did not exhibit significance in relation to SSTI prevalence . Individuals reporting a history of SSTI within the year prior to survey response reported a variety of treatment methods .At-home remedies for SSTIs included warm compresses , the use of over-the-counter medications , and selfdrainage of the abscess outside of a clinical setting . However, individuals also sought out a range of professional medical care for the treatment of their SSTIs. Professional medical resources included going to an emergency room or urgent care , using prescribed medications , having either a procedure --- Discussion Primary findings of the described study indicate that participants reporting a history of SSTIs in the prior year were likely to partake in a variety of unsafe injection practices. While skin cleaning methods prior to injection did not exhibit significant correlations with a history of IDU-related SSTIs, the type of water sources used for drug dilution did. Females are up to three times more likely to report a history of SSTIs in comparison with males, leading us to believe that there may be sociocultural and geographic factors in the rural setting contributing to such findings. Although the anatomical location of injection did not bear a significant correlation to SSTI prevalence, the type of injection-such as 'skin-popping' , 'muscling' , or intravenous-demonstrated significance; individuals reporting subcutaneous and intramuscular injection practices were significantly more likely to report a history of SSTIs compared to those injecting intravenously. Additionally, individuals able to regularly inject on their first attempt were significantly less likely to report a history of SSTIs. Lastly, individuals reporting a history of IDU-related SSTIs within the year prior to survey response cited several treatment methods, including both at-home remedies and professional medical care. A number of studies have investigated and found similar trends of unsafe injection practices correlating to higher risk of SSTIs among PWID; however, many of these studies are in the context of urban or suburban populations, rather than in rural settings. Similar to our study, several others investigating urban samples have reported higher incidence of IDU-related SSTIs among females [4,14,29]. Tuchman [32] found that urban females were more likely to initiate their injection drug use after being influenced by other females and commonly relied on others to inject for them, citing inability to inject themselves and difficulty finding an injection site as primary factors driving their decision [32]. Such circumstances increase their risk of utilizing unsafe injection practices and may lend to the higher rates of SSTIs among females who inject drugs. In the context of rural populations, the gender composition of and interactions among social networks may differ in comparison with urban settings. Cultural norms among rural populations are also different than those in urban areas, potentially serving as additional factors lending to higher rates of SSTIs among females in comparison with males in rural settings. Site of injection has also been cited as a common factor associated with IDU-related SSTIs. Similar to our findings, Smith et al. [29], Phillips et al. [26], and Murphy et al. [20] revealed 'skin-popping' as a common risk factor associated with IDUrelated SSTIs in urban populations [20,26,29]. Our findings among PWID in rural areas are consistent with their urban peers, suggesting that both populations use higherrisk sites via either 'skin-popping' or 'muscling' for injection practices. It should be kept in mind that the extent to which overall duration of injection use, and related health vulnerabilities, might be a contributing, or even driving, factor for this finding is a remaining question in the context of interpreting results of this cross-sectional study . Those citing a history of SSTIs reported the use of a variety of both professional medical treatments and at-home remedies. These findings are consistent with existing literature surrounding urban populations. Monteiro et al. [19] described an urban population of PWID often utilizing emergency medical services to treat SSTIs, while also reporting the use of at-home abscess drainage by non-medical personnel [19]. Although findings are similar between urban and rural populations, individuals residing in rural areas often experience a variety of health-related disparities that may serve as barriers to accessing proper medical care [5]. Among other factors, obstacles in accessing adequate health care in rural areas and stigma attached to seeking care for an injectionrelated health issue may contribute to the use of at-home remedies for the treatment of IDU-related SSTIs rather than seeking medical treatment. Similar to our findings, Wright et al. found no significant relationship between skin cleaning prior to injection and SSTI history among a larger urban population, insinuating that skin cleaning may not serve as a strong protective factor against infection . However, given that our results present skin cleaning methods used at the time of survey response, rather than the skin cleaning methods used at the time of infection, these results may be indicative of changes in skin cleaning behaviors among PWID following infection in order to help prevent the development of further SSTIs. Contrarily, several studies suggest that skin cleaning methods are a significant predictor within urban populations. Smith et al. [29] and Murphy et al. [20] both present the use of alcohol for skin cleaning prior to injection as a significant protective factor against IDU-related SSTIs in urban populations [20,29]. Similarly, Phillips and Stein [25] cited infrequent skin cleaning prior to injection as a common risk factor associated with IDUrelated SSTIs among an urban population [25]. Urban and rural populations may be impacted differently by skin cleaning practices despite partaking in similar techniques. Rural populations often experience several barriers when accessing syringe exchange programs [5]. Such barriers may place rural populations at a disadvantage for obtaining skin cleaning supplies, such as alcohol wipes, in comparison with individuals residing in urban areas. This may contribute to the differences in skin cleaning significance in the context of IDU-related SSTIs between urban and rural populations who inject drugs. However, based on the contradictory findings between several studies, further research is needed to determine the role of skin cleaning methods and their effectiveness in preventing the development of SSTIs. Contrary to our results, several studies have found that the anatomical location of injection is significantly correlated with risk of infection. Although our results suggest that individuals injecting into their hands and upper arms were slightly more likely to report a history of SSTI in comparison with individuals injecting into their cubital fossa, there was no significant correlation found. Conversely, past literature surrounding urban samples demonstrates that injection into either upper and lower extremities or the groin region corresponds with history of infection [6,14]. Additionally, the majority of our sample cited the use of methamphetamine as their primary drug of choice. Although our study found no significant correlation between primary drug of choice and history of SSTI, both Murphy et al. [20] and Phillips and Stein [25] reported the use of 'speedball, ' or heroin mixed with cocaine, to be commonly associated with IDU-related SSTIs in urban populations [20,25]. Access to and preference of substances may differ between urban and rural populations, potentially lending to the differences between infections related to the substance injected. Lastly, our study presents water source for drug dilution-whether sterile or unsterile-as a significant risk factor associated with reporting a history of SSTIs among this rural population. Although significance is not reported, Phillips and Stein's [25] findings suggest that PWID in urban populations also use a combination of sterile and non-sterile water sources to inject their drugs [25]. Likewise, despite respondents not reporting the use of public or private toilet water among this sample, Harris et al. report the use of toilet and puddle water or water alternatives, such as cola or lemon juice, used for injecting, which may further lend to risk of SSTI . Such results indicate that both rural and urban populations partake in risk behaviors associated with water sources and both populations are at risk of developing SSTIs related to their water sources and usage. Further investigation is required to determine the specific relationship between SSTI risk and water source, because although unsterile water sources lend to overall risk of SSTI, individuals with a history of SSTI may also be more likely to seek out sterile water sources to help prevent related infections. Overall, higher-risk injection practices were common among participants reporting a history of SSTIs living in rural areas. These results are similar to comparative studies conducted among urban and suburban populations. These findings suggest that educational materials targeting PWID not in treatment should encompass a variety of injection behaviors-including 'skin-popping' or 'muscling' ; proper skin cleaning practices; and the use of sterile water sources for drug dilution. Future studies should aim to understand social determinants and cultural factors, such as social networks, present in rural populations which may influence risky injection practices and the general barriers of safer injection practices to prevent SSTIs. With females being up to three times more likely to report a history of infection, future studies should investigate risk factors that are unique to females who inject drugs in rural communities. Given the increasing prevalence of methamphetamine, future research should also focus on the differences in injection practices between people who use methamphetamine as their primary drug of choice versus heroin or cocaine. Urban hospitals and clinical practices witness high rates of emergency department visits and hospitalizations related to infectious comorbidities associated with injection drug use [31]. Our results suggest that rural populations partake in similar risky behaviors associated with developing bacterial infections and seek out similar medical and non-medical treatments for SSTIs as urban populations. However, rural-dwelling individuals may experience more frequent barriers when attempting to access proper health care services [34]. This places rural populations at a high risk of experiencing life-threatening sequelae. Therefore, rural hospitals should consider partnering with outreach organizations and/or pharmacies to target PWID with better wound education, safer injection practices, and education regarding when it is safe to treat their infections at home versus when to seek medical treatment. By introducing such proactive intervention and prevention methods in both the clinical and public health setting, hospitals and clinical practices will potentially see decreases in emergency department visits and hospitalizations associated with infectious comorbidities related to injection drug use. --- Public health implications The opioid epidemic is one of the most pressing public health crises of our time. Co-morbidities associated with injection drug use, namely skin and soft tissue infections, place a large financial and care burden on public health agencies and clinical settings [31]. Syringe exchange programs have the potential to serve as a cornerstone in preventing and minimizing the effects of SSTIs as they relate to IDU. Given their expansive coverage across networks of people who inject drugs, syringe exchange programs should consider implementing or expanding existing early prevention and education materials specifically targeting risky injection practices as they relate to SSTIs. Additionally, these public health programs should consider providing resources on proper treatment methods and medical resources for individuals who have already developed an infection. Such prevention materials are one of the first lines of defense in combatting infectious co-morbidities associated with injection drug use and have the potential to make a widespread impact on both populations of people who inject drugs and the notable workload placed on hospitals to care for such patients. --- Limitations This study should be considered in the context of several limitations: the data presented was collected using a one-time cross-sectional approach, potentially introducing recall bias; the prevalence of skin and soft tissue infections among survey respondents was estimated based self-reported infections within the year prior to survey response and could be skewed; the sample size was relatively small, with all respondents being clients of Vivent Health's statewide syringe exchange program and could impact power; and the use of respondent driven sampling may underestimate the variability within populations because of the tendency of participants recruiting others with similar characteristics. --- --- Abbreviations ACASI: Audio computer-assisted self-interview; HCV: Hepatitis C virus; HIV: Human immunodeficiency virus; IDU: Injection drug use; PWID: People who inject drugs; SSTI: Skin and soft tissue infection. --- --- Competing interests The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. • fast, convenient online submission • thorough peer review by experienced researchers in your field --- • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year --- • At BMC, research is always in progress. --- Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: ---

Introduction: Skin and soft tissue infections (SSTIs) are among the leading causes of morbidity and mortality for people who inject drugs (PWID). Studies demonstrate that certain injection practices correlate with SSTI incidence among PWID. The opioid epidemic in the USA has particularly affected rural communities, where access to prevention and treatment presents unique challenges. This study aims to estimate unsafe injection practices among ruraldwelling PWID; assess treatments utilized for injection related SSTIs; and gather data to help reduce the overall risk of injection-related SSTIs. Methods: Thirteen questions specific to SSTIs and injection practices were added to a larger study assessing unmet health care needs among PWID and were administered at six syringe exchange programs in rural Wisconsin between May and July 2019. SSTI history prevalence was estimated based on infections reported within one-year prior of response and was compared to self-reported demographics and injection practices. Results: Eighty responses were collected and analyzed. Respondents were white (77.5%), males (60%), between the ages 30 and 39 (42.5%), and have a high school diploma or GED (38.75%). The majority of respondents (77.5%) reported no history of SSTI within the year prior to survey response. Females were over three times more likely to report SSTI history (OR = 3.07, p = 0.038) compared to males. Water sources for drug dilution (p = 0.093) and frequency of injecting on first attempt (p = 0.037), but not proper skin cleaning method (p = 0.378), were significantly associated with a history of SSTI. Injecting into skin (p = 0.038) or muscle (p = 0.001) was significantly associated with a history of SSTI. Injection into veins was not significantly associated with SSTI (p = 0.333).Higher-risk injection practices were common among participants reporting a history of SSTIs in this rural sample. Studies exploring socio-demographic factors influencing risky injection practices and general barriers to safer injection practices to prevent SSTIs are warranted. Dissemination of education materials targeting SSTI prevention and intervention among PWID not in treatment is warranted.

INTRODUCTION The Code of Children and Adolescents of Ecuador in its article 26 ratifies the right to a dignified life, with free multiple basic services, among which access to quality education for children and young people without exclusion of some type of disability, ethnicity, or social class. However, this type of right is affected when due to different social problems, such as poverty, the continuity of children and young people in the educational system is conditioned, because the priorities of the parents are focused on the economy of the home, leaving education in the background . One of the main breaches is known as school dropout, a problem that has been rooted for many years in our country. Portoviejo Canton is not exempt from a school dropout, according to data from the Ministry of Education of Ecuador, in the 2019-2020 school year "three percent of the enrolled student population dropped out." The coordination identified among the most predominant causes of school dropout is the low economic level of dropout families and, secondly, intra-family conflicts. This problem is increased when addressing the school dropout of students with special educational needs , due to the scarcity of qualitative studies that identify the causes of dropout of this vulnerable group of students. The present work addresses this social problem from a qualitative perspective in which it is intended to identify the main causes of school dropout of a group of students with SEN associated and not associated with disability and answer the following questions: What are the reasons that the representatives of students with SEN associated with a disability must leave the Educational Institution? Are the reasons for dropping out of school different between students with SEN and students who do not have SEN? Will the preparation of teachers influence issues such as methodological adaptations in the educational abandonment of students with SEN? In this sense, interviews were conducted with directors of six Educational Units of the Portoviejo canton, whose registration records have students with SEN who recently abandoned their studies. In addition, the reasons for dropping out of school were analyzed through case studies applied to dropout students. Thanks to the application of the techniques and instruments, it is possible to contrast the information presented by the directors of the Educational Units involved and the representatives of the students who justified their reason for dropping out of school. --- LITERATURE REVIEW School dropout is also present in students with SEN and when a bibliographic search of a national approach is carried out about this problem, the results obtained are scarce, which reveals the little importance that is usually given to this problem. The vulnerable group is present in the educational community. However, some national and foreign research that addresses this problem appropriately is highlighted below. In the first place, the investigation of Rojas stands out, which, through the application of case studies, seeks to know the family factors that influence the school dropout of students with SEN in Peru. The author concludes in his research that the family factors that influence school dropout were intrafamily problems, associated with the psychological strength that parents must have when accepting their child according to their personality, since parents prefer to withdraw to his representative of the Educational Unit to prevent affectations such as bullying, marginalization, among others. During the same year, the Ecuadorian researcher Fuentes elaborated on an inquiry called the school desertion of students with disabilities from the special education institutions of the Esmeraldas canton, study period 2011-2012. Which, after finishing its qualitative work, concluded that the most important factors of school dropout of students with SEN are intra-family problems since there are conflicts when discussing the presence or absence in education, considering it unimportant. Calderón highlighted in his research the decrease in school dropout in rural areas, due to the opening of educational centers in areas of difficult access. However, it also calls for better teacher preparation. The author considers that the teacher should be a person who lives in a rural area since he knows the deficiencies of his students, especially the limitations that homes with SEN students have. Years later, the Ecuadorian researcher Yánez proposed an investigation entitled "Students with disabilities and school dropout in the Faculty of Philosophy, Letters and Education Sciences of the Central University of Ecuador". Even though his study sample was university students, his objective is striking, since he intended to verify whether a disability is a reason for academic desertion. Concluding that having a disability is not a reason for academic dropout, however, external factors associated with poorly accessible facilities, and lack of teacher preparation, among others, influence. Finally, the Ecuadorian author Tubilla carried out an interesting investigation entitled "Public management of inclusive education and school dropout of students with disabilities in public educational institutions, UGEL 03, 2019". In it, he analyzed the changes in Ecuadorian public management regarding the inclusion of students with SEN in the public education system, to know if these changes affected the school dropout of students with disabilities. Concluding that the budgetary allocation to the institutions for their equipment in learning resources for students with SEN is insufficient, as well as the public management in inclusive education brought therefore school desertion of students with disabilities because their management has not been efficiently addressed. Addressed the investigations, it is possible to identify factors of school dropout of students with SEN that have in common with the previously mentioned works. In the first place, intra-family problems due to parents' acceptance of the differences their children have with others, possible integration problems, and bullying, these intra-family conflicts affect school dropout. Second, educational institutions sometimes do not have adequate physical conditions and proper teaching staff preparation to achieve significant student learning. And, thirdly, the socioeconomic factor present in Ecuadorian families affects the continuity of education. These factors are related to an investigation carried out by Romero and Hernández who classify the causes of educational dropout into three types: a) those that focus on the student's perspective, that is, those that are related to abilities, motivation, among others, b) those that focus on social, economic and cultural factors, c) those that focus on the educational system, that is, those that are related to teaching programs, pedagogy and teacher strategies, among others. others. --- METHODOLOGY Multiple case studies were applied to 11 dropout students with SEN. The case studies to be applied are of descriptive characteristics because it is intended to identify the main causes of school dropout. The application of case studies finds its justification in that, as Sampieri, Fernández, and Baptista point out, this approach is particularly appropriate for testing hypotheses and developing theories about social phenomena. For the development of the research, six Educational Units of the Portoviejo canton participated, represented by the director and/or the general secretary of the institution. Who, after agreeing to participate in the investigation, interacted in a semi-structured interview. The selection of the Educational Units is due to different aspects that are detailed below: The selected Educational Units have more than five students with SEN, whether associated or not associated with disability, of which at least 2 students dropped out of school. The SEN that students have will be both associated with disability and not associated with disability Table 1. Selected educational units. The present investigation applied the semi-structured interview as a data collection technique due to the freedom that the interviewer must introduce the necessary questions to obtain precise information on the desired topics. The interviewer must keep in mind that the answers expressed are related to the proposed topic. The preparation and selection of the items that made up the interview were subject to the contribution of the research present in the theoretical framework, in general, the various researchers highlight three groups of potential factors in the decision to abandon, and these factors will be explained in the following chapters. Subsequently, the questions were analyzed and validated by the tutor of this research. The interviews were conducted in person and online through the ZOOM platform, the selection of the medium to conduct the interview was subject to the availability of time by the director or the staff in charge of the institution. The interview was recorded regardless of the physical medium in which it was conducted, and its structure consisted of two parts. The first part of the interview presents generic questions, which sought to know the inclusion of students with SEN in a general way, including the preparation of teachers and access and infrastructure. The second part of the interview aims to know certain characteristics of each student who left the institution. These questions are classified into the personality of the student, the cooperation and participation of the representatives with the institution and finally, they are questioned about the delivery of the document justifying the school dropout. --- Table 2. Description of categories and codes --- RESULTS AND DISCUSSION After analyzing the different realities of the sample in relation to school dropout, three main factors were identified as causes of dropout, these factors do not differ with the type of SEN, whether it is associated with disability or not, that is, the reasons for Identified dropouts are present in our sample both in students with SEN associated with disability and students without SEN not associated with disability. Figure 1 summarizes the main causes of abandonment in our study sample. 2011) "There is a positive relationship between household income and the probability that the adolescent attends school and does not work." Mayancela "The risk of school dropout increases with the presence of situations such as economic and intra-family problems". Cevallos and Macias "The socioeconomic factor is more prevalent than the cultural factor as a reason for dropping out of school." Barbecho "Personal and socioeconomic reasons are the main causes of school dropout." In our research sample, this factor was also present as a justification for student one, in which the representative made this decision because he was "unemployed due to a pandemic" . What made it difficult to acquire the necessary computer resources for their represented to be part of online learning, in this case of abandonment, the lack of employment directly affected the ability to acquire the necessary computer supplies for virtual education, therefore, despite belonging to a public institution, whose education is free, the socioeconomic condition prevailed over free education. A similar reality was experienced by the representative of student four, whom he mentioned this dropout factor due to the "closure of his business due to restrictions due to the pandemic" . In this sense, the economic condition influenced the decision to drop out since the student belonged to a private institution, however, even though the representative did not provide documentation formalizing the specific reason for dropout and considering that education in our country is mandatory for 12 years, the approach to school dropout changes from being permanent to temporary. Another case in which there is evidence of school dropout due to the economic factor has as its protagonist the representative of student five, who delivered the document requesting a change of institution to a public entity. However, the student's data is not part of the students enrolled in public or private institutions in the Portoviejo canton and considering that the student had debts with the institution, the dropout is temporary. Finally, the case of student seven has the economic condition as the cause of abandonment, because the representative had delays in the monthly payment of a private institution, so in this case, it can be seen how the economic situation prevails over the educational quality that the student could receive. Curricular adaptations: This factor refers to the level of preparation of the teaching staff regarding curricular adaptations for students with SEN. The dropout factor is highlighted by Navarro et al., "Curricular adaptations are the instrument to achieve the individualization of teaching". Ayala "the curricular adaptations recommended by the Ministry of Education are suitable". Vera "The implementation of curricular adaptations improves the academic performance of students with SEN". This dropout factor was justified by the representatives of student three "lack of teaching commitment" . Student six has "virtual learning difficulty due to his disability" . Student eight "lack of teacher preparation" . And finally, student 11 "lack of teacher preparation" . It should be noted that, in the cases described, the supporting documents were delivered, the four cases indicated a lack of preparation on the part of the teacher regarding teaching strategies or methodologies, regardless of the type of institution in which the students were part. Within educational institutions, the preparation of the teacher is not the only factor of school dropout, the infrastructure adapted to students with SEN is a dropout factor identified in the academy, as highlighted by Mendoza "There are high numbers of children with disabilities who cannot access education, mainly due to the lack of adequate infrastructure", however, in our study sample the adequacy of adequate infrastructure that facilitates the mobility of students with SEN was not considered a determining factor in the school dropout of students, due to the fact that public and private educational institutions use different strategies such as the accompaniment of the student through a tutor teacher to guide the student during their stay in the institution. Family: A priori it can be interpreted that the economic condition is a characteristic immersed in the family, however, the results obtained from the study sample suggest perceiving the economic condition as an independent dropout factor. The family factor has various components, among which are intrafamily problems, single-parent families, and lack of responsibility on the part of parents, among others. The dropout factor was analyzed in previous research, the same ones found in the theoretical framework, mentioning Martínez and Álvarez "The participation of student representatives with the institution is important to guarantee an optimal educational environment". Espinoza et al., "Families that are large, single-parent and have a low socioeconomic level are directly related to school dropout." Rojas The family factors that influence school dropout were intra-family problems". Sources . "The most important factors of school dropout of students with SEN are family problems." Guanoluisa and Llango "Parents tend to detach themselves from academic responsibilities." In our sample, this dropout factor was the justification by the representative of student two, indicating "lack of time" . This type of justification is associated with the lack of academic responsibility that parents can present when it comes to a student with SEN, in this case, the student has a hearing disability. Ignorance about the teaching processes and pedagogical actions applied at home, trigger frustration for parents, the same ones who usually entrust the growth and intellectual development of their client to the Educational Institution. However, this was frustrated due to the new modality of virtual studies in which the representative acquires a leading role in the consolidation of his son's learning. The cause of educational abandonment of student ten, clearly evidences a disengagement from academic responsibilities, given that the student suddenly interrupted his studies. On the other hand, the student's representatives have not requested a change of institution and, furthermore, they did not have any debts with the entity, so school dropouts due to socioeconomic factors are ruled out. In addition, it can be said that the student was a model student, even though he had autism, his academic performance was remarkable, also ruling out educational abandonment due to the school environment. In this sense, the educational disengagement of a student with SEN totally adapted to the institution can cause a negative impact on the student's personality, because he is being denied his right to a quality education regardless of the duration of this educational abandonment. External factor: The presence of COVID-19 in Ecuador since February 29, 2020, was a challenge for a country that in educational matters did not have adequate knowledge to function virtually in the field. The confinement and the non-attendance of many activities, including education, exposed the deficiencies of the educational system, in addition to increasing the economic gaps of a predominantly middle-class population. In the opinion of this author, this external factor potentiated the reasons for dropping out of school exposed in the investigation. The economic condition as a factor of abandonment was strengthened with the confinements, due to the layoffs suffered by the representatives that justified this reason. Curricular adaptations as a dropout factor apparently evidenced the lack of training of teaching staff in the educational institutions of our sample regarding the methodology of online education for students with SEN. During confinement, the group of students with associated SEN and no disability was particularly affected, due to the restriction of the only place that allows them to develop socially, the school. Figure three shows the presence of COVID-19 as an external factor present in the reasons for dropping out of school discussed above. --- CONCLUSION AND RECOMMENDATION The reasons for dropping out of school do not differ in relation to the presence or not of associated SEN and not to a disability, because the reasons are part of external factors that affect the decision to drop out. Three dropout factors present in the selected sample were identified: a) the socioeconomic factor, b) curricular adaptations, c) those related to the family. The development of this research generates a positive impact on two aspects. From the academic point of view, this study provides recent information on the educational dropout of students with SEN in the city of Portoviejo, providing bases that allow future research focused on this vulnerable group. On the other hand, socially, it constitutes the first independent qualitative research that descriptively identifies the causes of school dropout. People with SEN associated with or not with a disability acquire several limitations during their growth that prevents them from functioning in the society of which they are a part. The impediment usually starts from the home, by not considering beneficial activities for this vulnerable group. This author is convinced of breaking down the barriers that prevent people with SEN associated or not with disabilities from interacting with society by providing equal opportunities that are a right. --- CONFLICT OF INTEREST The author declare that there is no conflict of interest. --- REFERENCES 1) Abril, E., Román, R., Cubillas, M., y Moreno, I. . ¿Deserción o autoexclusión? Un análisis de las causas de abandono escolar en estudiantes de educación media superior en Sonora, México. Revista electrónica de investigación educativa, 10, 1-16. http://www.scielo.org.mx/scielo.php?script=sci_arttext&pid=S160740412008000100007&lng=es&tlng=es.

Educational abandonment has become for many years a social problem that affects Latin American students. Consequently, several nations, including Ecuador, focus their efforts on mitigating this problem. The research about the educational abandonment of students with special needs aims to identify why 11 students did not continue studying. The sample was part of six public and private schools in Portoviejo. The school was selected due to the high number of students with special needs enrolled in the last school year. For achieving this qualitatively oriented study, semi-structured interviews were applied as a data collection technique, and the case study research method was implemented. Among the results found in the study sample, three factors of educational abandonment stand out. One of them was the socioeconomic situation, which takes precedence over the educational quality that the student could receive regardless of the type of educational institution of which he is a part. The other one was the educational environment stands out, which mainly alludes to the teacher's preparation regarding the methodological adaptations that students with disabilities require to consolidate their learning. Finally, the family factor emphasizes the lack of responsibility of parents towards their children, such as limiting their right to education. This type of research is especially relevant due to the little literature about this vulnerable group, which will allow future explorations to be carried out that seek to enforce their rights.

Introduction Human trafficking is a major problem for human society and is present everywhere in almost every country and society, somewhere more widespread and somewhere less, in one country it appears in one form and in another in another form. Even more difficult and challenging is the sexual trafficking of minors. The involvement of minors in trafficking carries and then brings many challenges, problems and requires the undertaking of many measures, programs to solve them. Sexual trafficking of children is present everywhere and even in our country and this is talked about very little or not at all our time. Human trafficking is a difficult crime to detect and prevent. Human trafficking includes the trafficking of human beings for the purpose of commercial sexual activities as well as forced labor. These crimes are occurring all over the world. Research shows that organized crime, prostitution, massage parlors and brothels are closely related to the crime of human trafficking. Government corruption and international criminal organizations contribute significantly to this crime, and financial gain is usually the primary motive. Definitions of "human trafficking" include the use of force for the purposes of labor or sex. A major report published by the Anti-Trafficking in Persons Unit of the United Nations Office on Drugs and Crime29 . The Global Program Against Human Trafficking specifically defines human trafficking as30 : Recruiting, transporting, transferring, harboring or receiving persons, by threat or use of force, or other forms of coercion, kidnapping, fraud, abuse of power or a position of vulnerability or giving or receiving payments or benefits for obtain the consent of a person who has control over another person, for the purpose of exploitation. Exploitation includes, at a minimum, exploitation of the prostitution of others, or other forms of sexual exploitation, forced labor or services, slavery or similar practices with slavery, servitude or removal of organs. --- What does Human Trafficking Involve? Human trafficking refers to "the illegal trafficking of human beings through the abduction, use or manipulation of force, deception, or sale for the purpose of sexual exploitation or forced labor"31 . This form of trafficking generally includes sex trafficking and labor trafficking, where victims are distributed through brute force, psychological coercion or abuse for sexual or labor manipulation. In particular, the child is treated as a commercial object and the return may be financial or in kind. For example, a child involved in prostitution may be compensated monetarily or sexual acts may be exchanged for basic needs such as food or shelter. This exchange of sexual acts for basic needs is also known as survival sex. --- Victims of human trafficking Close examination or investigation of victims of human or sexual trafficking is very difficult, however, various authors and researchers have managed to carry out such studies. Obtaining reliable information has proven to be as difficult as finding an agreed definition, as noted above. The reasons for this are various. For example, individuals who may be considered victims of trafficking are already in vulnerable positions and often fear the consequences of explaining their situations to the authorities. Furthermore, the individual who may be trafficked, is likely to believe that he/she has had some involvement in accordance with the traffickers' instructions and may be reluctant to self-identify as a victim or provide relevant information about offenders. 32 In research, there are common needs seen in most victims of human trafficking. These include, "emotional and moral support, legal assistance, safe housing, high school diploma or general education diploma assistance, identification documentation, job training, resume and job search assistance, medical and dental appointments, cell phone assistance, childcare, transportation, safety planning, and clothing and food assistance" . However, this is not all-inclusive, and new needs may arise with each client. Victims' needs may increase every day, e.x: -Emotional support -Emotional and mental health needs may be the most critical of all, as they are most debilitating to their daily lives. All literature has highlighted that many of these victims experience Post Traumatic Stress Disorder , depression, anxiety, self-loathing, detachment, despair, and difficulties with interpersonal and intimate relationships. In addition, victims have been known to suffer from other mood disorders including panic attacks, obsessive compulsive disorder, fear and hopelessness about the future -Physical health -Physical health needs are of great concern to victims after trafficking, because it can also affect their daily life activities. Children who are involved in sex trafficking may be more prone to physical illness due to their immature physical systems. -Support from other victims -It is necessary for victims of human trafficking to have contact with others who have also been victims of human trafficking. Often, victims are unwilling to accept the trauma and exploitation they have experienced. As a result, many survivors are reluctant to seek treatment. However, when working with others who have had similar experiences, many individuals feel more at ease 33 . -Family Support -Family involvement in helping victims post-trafficking is an important need, as their family members can be a great support system for them as well. -Proximity of services -While there are some services available, they are not many and they are often spread over large areas. This forces customers to have to travel long distances and without their own means of transport this can mean long journeys on public transport. -Reintegration -Many of these individuals may lack the skills necessary to find a job and achieve economic stability after escaping their trafficking. In many cases, these clients do not receive the full range of integration services they need, such as assistance in finding and maintaining employment, finding affordable and longterm housing, or assistance with citizenship status. --- Results This chapter presents the results of the study, which were generated, of course, by the process of analyzing data collected through interviews. In the development of this chapter, all the research questions raised in the introduction of this study are answered. For the purposes of this study, 40 children were questioned, of which 15 were girls and 25 were boys. Their ages ranged from 7 to 12 years old. The interviewees stated that they were in the situation of forced labor for 3 to 5 years, specifically: • 20 of them had more than 2 years of work; • 15 of them 1 year; • 4 of them more than 4 Regarding their education, they said that none of them had completed a full education, but had completed only a few years separated, thus: • 5 of them had 1-3 years of primary education; • 22 of them had 3-5 years of primary education; • 13 of them had 5-7 years of education Regarding ethnicity, they stated that 35 of them belonged to the Roma community and 4 of them belonged to the Egyptian community. In terms of income or profit from work, they said that: • 27 of them had very little profit; • 3 of them not at all; • 10 of them enough When asked who they work for, they said that they mostly worked for others, specifically 28 of them and only 12 worked for their parents. Regarding the question of whether they have made any attempt to leave the situation they are in, only 5 of them said that they had tried without success and 35 of them said that they had never made an attempt to leave and this due to pressure, coercion or lack of hope for a better future: "....you can't do anything because we are being protected all the time..." "...if we think we're going to leave, we'll be beaten, no matter where we go..." In the closing, they were asked if they have ever been stopped by any representative of the state bodies or any police officer, knowing that they are always found on the road and there are also traffic police or others right next to them. They categorically stated that they were never stopped or questioned! --- Conclusions Human trafficking in general, without being banned in the categories or forms of its manifestation, constitutes a serious problem for human rights and for the international community as long as it is a phenomenon that extends beyond the borders of a country. Human trafficking is mainly more problematic and very painful when it extends to children, children are victims of trafficking in various forms, such as sex trafficking or forced labor. For the latter, for the purposes of this paper, a field study was conducted to see the situation closely, and from this study, what requires a solution and raises the alarm in my personal vision is the silence of state employees in front of small children who see day, at every hour on the streets of Tirana or Durrës asking for alms, offering services such as washing car windows, etc., and no one comes to ask them. In this way, we are all co-authors of this crime. And as long as no one tries to free them, they will continue to be in this state because, as they themselves have expressed, it is impossible for them to try to leave forced labor because they are subjected to physical, psychological violence, pressure and are always under guard. Moving on to the general plane of human trafficking, it must be said that there are significant unmet needs of trafficking victims, and a lack of evidence-based research to focus on the main issues related to this type of trauma. However, there is a good research base from which to begin an assessment of these needs. There is also a variety of services available, but not nearly enough to facilitate "healing" for a larger number of survivors of human trafficking. This leaves room for further research, particularly with survivors themselves. In general, the services that are most effective in dealing with this population are health, physical support and reintegration services. Further research will identify more of these needs, as many develop each day depending on the individual. Conducting further research will determine ways to better implement strategies to serve victims and survivors of human trafficking. In conclusion, I would add that it is necessary to train the appropriate authorities in the recognition of trafficking, as well as in cooperation and communication between police forces and relevant human service organizations. Better, well-defined laws are also needed to assist law enforcement agencies in locating traffickers. Having the best interests of victims in primary consideration is essential to help prevent trafficking.

Research on human trafficking has attracted increasing attention since the early 1980s and is extremely important as it highlights the problems and through it identifies the appropriate ways that can provide solutions. The object of this paper is the treatment of trafficking in human beings and in particular will focus on the sexual trafficking of minors. This topic is very current not only in Albania but also in the international arena (due to it's nature) and requires the commitment of all states and international organizations to combat human trafficking and protect victims of trafficking as part of the field of human rights. Therefore, this topic is covered in section I of this conference. The research questions of this paper are: what does trafficking in human beings involve, what are the needs of the victims, what is the exploitation and sexual abuse of children, what are the consequences that trafficking leaves on children and why children are recruited. Numerous studies and studies have concluded that trafficking represents "a denial of the individual's rights to freedom, integrity, security and freedom of movement" Victims of human trafficking need, among other things, emotional support , family, health care and especially for reintegration into society. On the other hand, child sexual exploitation is a form of human trafficking and consists of adult sexual abuse and monetary or in-kind reward for the child or a third person or persons. and it is precisely the commercial element that separates sexual exploitation from other sexual crimes.Juveniles are recruited more easily because small victims are the most optimal, useful and lower cost, they are easy to be guided by traffickers and slave owners and to adapt to the new situation. Current studies show that some effects of Sex trafficking is sexually transmitted diseases, diseases and other infections, physical harm, substance abuse and malnutrition.

INTRODUCTION The severe acute respiratory syndrome coronavirus 2 pandemic has put a spotlight on long-standing racial/ethnic health inequities that have existed in the USA for centuries. Black and Latino populations have among the highest age-adjusted rates of COVID-19 infection and hospitalizations. Moreover, non-Latino Black and Latino people are 1.5-2.5 times as likely to die from the illness compared with White and Asian people [1]. Experts attribute the excess disease burden to the increased likelihood of exposure to COVID-19 , increased vulnerability to severe health consequences of the infection due to long-standing structural forces contributing to disparities in comorbidities, and limited access to healthcare. Regardless, many experts concur that the "return to normal life" will not be possible until a prophylactic vaccine for COVID-19 is widely available [2,3]. High rates of vaccine uptake will be necessary to ensure that a large proportion of the population becomes immune , thereby limiting community transmission of the virus from person to person [4]. Estimates vary but it is believed that between 70% and 80% of the population would have to be vaccinated to achieve herd immunity [5,6]. Given the critical role that a vaccine would play in controlling the pandemic, it is essential to understand potential uptake of the vaccine and factors associated with acceptance or resistance to vaccination, particularly among populations that have been disproportionately impacted by the pandemic. This is particularly important given the current context of declining vaccine confidence [7], mistrust of science [8,9], and extreme political polarization, which has shaped COVID-19 behaviors and beliefs in the USA. As the nation launches a massive vaccine distribution effort, effective vaccine campaign messages, and public health strategies need to be developed. The purpose of this study was to assess intention to obtain a COVID-19 among a diverse sample of American women and to examine differences in vaccine intentions across racial/ethnic groups. We assess vaccine intentions among women, as they often serve as gatekeepers for medical care in families. --- METHODS Data for this analysis are from a larger study of human papillomavirus vaccine knowledge, attitudes, and intentions among a convenience sample women ages 27-45 administered between April 13, 2020 and June 8, 2020. In that study, we conducted an online survey with a Qualtrics panel of respondents . Quotas were instituted to overrepresent women of diverse racial/ethnic groups to produce a sample that was 25% non-Latina Black , 25% Chinese, and 25% Latina. We elected to oversample Chinese women, as opposed to other Asian subgroups, because of rampant anti-Chinese sentiment in the USA which may impact vaccine intentions. The survey included an item to assess COVID-19 vaccine intentions, our primary outcome. Respondents were asked: "If there were a vaccine to prevent coronavirus, would you get it?" with response options of "Yes," "No," or "Don't Know/ Unsure." For those who responded that they would not get the vaccine or were unsure, we asked, "Why not?" and respondents were able to enter free text responses. Questions assessing prior testing for and diagnosis of COVID-19 were items developed by the National Institutes of Health [10]. For perceptions of vaccine safety and efficacy, we asked respondents their level of agreement with the following statements: "Most vaccines are safe" and "Most vaccines are effective" with 5-point Likert type response options . To assess trusted sources of vaccine information, we asked respondents to indicate their level of trust in a variety of sources with response options on a Likert scale . The primary predictor of interest, race/ethnicity, was categorized as non-Latina White , NLB, Latina, Chinese, and multiple races. Additional sociodemographic characteristics, including age , income , education , employment , and insurance status , were assessed using items from the Behavioral Risk Factor Surveillance System [11]. --- Analysis Descriptive statistics were produced for all women in the sample and stratified by race/ethnicity. Chisquared tests were used to assess crude associations between women who did, did not, and were unsure about their intention to be vaccinated by sociodemographic and COVID-19-specific variables. Statistical significance was considered at the p < .05 level. An unadjusted logistic regression model was run with COVID-19 vaccine intention as the outcome and self-identified race/ethnicity as the main predictor. Sequential adjusted logistic regression models were run to assess the relationship between self-identified race/ethnicity and COVID-19 vaccine intention adjusted for age , marital status, income, education, employment, and insurance , belief in safety and efficacy of vaccines , and trust in healthcare professionals . Model 6 additionally adjusted Model 5 for belief in safety and efficacy of vaccines. We used this sequential process to develop models to ensure adjustment for grouped variables associated with socioeconomic status, health behaviors, and healthcare experiences. The data analysis for this paper was generated using SAS software, Version 9.4 of the SAS System for Windows . --- RESULTS --- Characteristics of study sample A total of 396 individuals participated in the study. Approximately one-quarter were NLW , and one-quarter NLB , with slightly fewer Latina women and few women of multiple races . Most had at least some college education, were employed , and had health insurance . In total, 56.8% reported that they would get a COVID-19 vaccine and 25.5% were unsure or did not know if they would get it. Results from the "free text" response to the question of why the respondent did not want to be vaccinated or were unsure showed that many had concerns about potential side effects , believed that there was insufficient time for vaccine testing , had concerns about lack of vaccine efficacy , or were distrustful of vaccines more generally . --- Bivariate results There were significant differences in vaccine intentions by race/ethnicity. Chinese women were the most likely to report that they would be vaccinated , followed by women who were NLW , multiple races Latina , and NLB women reporting an intention to be vaccinated . More Latina women reported that they were not going to get the vaccine than any other group . Among those who reported that they would be vaccinated, a greater percentage were insured , had higher levels of education , and expressed the belief that vaccines were safe and effective . More women with high income reported that they would get the vaccine than women with low incomes . Among those who highly trusted information from healthcare professionals, the majority were willing to be vaccinated . Similar patterns were observed for those who trusted information from public health and government agencies . More individuals who were not intending to get the vaccine relied on information from social media, although this was of marginal significance . --- Multivariate results In the logistic regression model adjusted for sociodemographic characteristics , NLB women had lower odds of reporting the intention to be vaccinated than women who identified as NLW . Latina women were more likely to report that they would not be vaccinated when compared with NLW women . In Model 4 , the AOR for NLB versus NLW women was no longer statistically significant. However, in Model 5, which included all of the variables in Model 3 plus trust in vaccine information healthcare professionals, the AOR for NLB versus NLW women was again significant . In Model 6, which includes all of the control variables plus belief in safety/efficacy and trust in healthcare providers), differences between NHW women and NHB women are no longer statistically significant and estimates are virtually unchanged from Model 4 . --- DISCUSSION We found that nearly 57% of women in the study said they would get vaccinated if a COVID-19 vaccine were available. More than a quarter of the sample remained unsure about their intention to get a vaccine. Those with lower levels of income, education, and those who were uninsured were less likely to report the intention to be vaccinated. NLB women were significantly less likely to report that they would be vaccinated than NLW women, after adjusting for sociodemographic, COVID-19-specific covariates, and trust in information about vaccination from healthcare professionals. However, after adjusting for belief in vaccine safety and efficacy, racial/ethnic differences were no longer statistically significant. Notably, none of the Chinese women in our sample said they would decline vaccination; most said they would accept the vaccine with the remainder being undecided or unsure about vaccination. Our findings regarding intention to be vaccinated are consistent with a national poll conducted by Tufts University/IPSOS [12] in May/June that found that 57% of NLW respondents reported they would get a COVID-19 vaccine compared with 48% of NLB respondents. Similarly, a May 2020 study from the Yale Institute for Global Health found that 68% of NLW respondents would be willing to get a COVID-19 vaccine compared with 40% of NLB respondents [13]. That same study found that NLB Americans reported lower COVID-19 vaccine acceptance and influenza vaccine uptake than nearly all other racial groups [14]. More recent national polls have found that about 60%-71% of the U.S. population intends to be vaccinated [15,16], but that those with lower levels of income and education, and NLB populations are less likely to report that they would be vaccinated [15][16][17][18]. Several published studies have also found that beliefs about vaccine safety and efficacy are associated with vaccine intentions [13,19,20]. Concern about vaccines and "vaccine hesitancy" is not a new problem. In 2019, the WHO declared vaccine hesitancy as a top threat to global health [7]. Growing anti-vaccination sentiment has been stoked over the past decades by misinformation about potential vaccine side effects and long-term health consequences, perceived nefarious motives of pharmaceutical companies, and conspiracy theories about the source of various infectious agents [8,9]. These types of messages have been found to spread more rapidly and effectively than legitimate public health messaging delivered through common social media platforms [14]. The COVID-19 pandemic has created a perfect storm to fuel these concerns: it is a novel virus, mistrust of government is high, and there is intense political polarization in this country [21][22][23]. Furthermore, the rapid vaccine development and testing process has led to perceptions that shortcuts may be taken [24]. Before discussing implications of our findings for practice and research, we acknowledge study limitations. First, we conducted a convenience sample among a Qualtrics panel, so care is needed when generalizing findings [25]. While electronically recruited panels are understood not to be completely representative of the U.S. population, they are equivalently representative as traditional recruitment approaches [25]. In our sample, 82.6% had a high school education or higher, compared with 87.7% for the U.S. population, and 48.9% of our sample had income under $50,000, compared with 42.1% for the U.S. population [15]. These comparisons suggest that the impact of misrepresentation arising from the panel is likely minimal. However, those willing to complete research studies are likely to be more receptive to health interventions than the general public [26] and data for this analysis were drawn from a larger study about HPV vaccination. We purposely oversampled diverse racial/ethnic groups and sexual and gender minorities but did not apply population weights to assess the needs of these underrepresented and undersurveyed groups. Our sample was limited to individuals who identified as women ages 27-45 years, therefore information about men and those in other age groups is needed, since we cannot assume that our findings apply to other groups. We included only one question on vaccine intentions and did not assess intentions over a specific timeframe, as is often done in "stages of change" models. A more nuanced understanding of vaccine intentions could be generated by assessing additional sociocultural and political influences on perceptions of the COVID-19 vaccine, and qualitative methods could be especially helpful in this process. However, study strengths include timeliness of findings and having the ability to compare across different racial/ethnic groups for whom there has been inadequate representation in prior research. Our findings point to the importance of mounting a national COVID-19 vaccine communication plan that includes a community-based participatory action research approach to ensure a trusted and culturally informed messaging strategy specifically tailored to Black and Latino populations, as appropriate. Interventions should address multiple levels of the socioecological model, including individual, interpersonal, community, and structural levels of influence [27]. A health communications campaign targeting individual behavior should emphasize that vaccine safety and efficacy are not being sacrificed by the expedited timeframe for development, and that the vaccine has been rigorously tested among diverse population groups. At the interpersonal and community levels, our findings suggest that trusted information sources should take the lead in disseminating vaccine information. While our data cannot speak to the impact of normative beliefs and altruistic motivations for vaccination, these types of messages may be particularly relevant in the context of the pandemic. Additionally, research has found that individuals and groups can help in combatting misinformation disseminated through social media by checking the accuracy of a message before sharing it. Recent data show that subtle messaging encouraging people to consider the accuracy of information before sharing it decreases the likelihood of sharing misinformation [28]. Additionally, social corrections to incorrect information has been found to be equivalently effective [29]. Several social media platforms have explored blocking or flagging of inaccurate or untrue claims, although some are concerned that this constitutes censorship [30]. A strong physician recommendation has been associated with uptake of other vaccines [31] and women in our study who trusted providers as purveyors of accurate information had higher intention to vaccinate. As such, we need to ensure that providers are skilled in communicating about this particular vaccine. Efforts to increase vaccine uptake have found that strong, "presumptive" messaging, which assumes that a patient will be vaccinated, is more effective than conversational language [32]. System interventions, such as reminders to providers in electronic medical records, have been found effective for other vaccines and should be instituted for COVID-19 [33]. To reduce barriers associated with access to care and cost/lack of insurance, the COVID-19 vaccine should be free and/ or covered by insurance without preauthorization. Vaccination should also be widely available with flexible hours in convenient settings, such as provider offices, urgent care facilities, federally funded health centers, school health programs, and pharmacies. At the policy level, federal or state mandates for vaccination for employment or school enrollment has been a highly effective strategy [34,35] and should be considered, although these strategies have backfired in some instances [35,36]. Our findings show that it will also be essential to develop a COVID-19 vaccine plan that addresses the concerns about vaccine safety and efficacy of NLB communities. Any efforts to reach this population will require significant efforts to build trust and confidence that the vaccine is safe and effective. It is essential that historical atrocities such as the Tuskegee Syphilis Study be acknowledged as they have created justifiable mistrust in medicine and in government [37]. Without acknowledgment, it may be easy to "blame the victim" for being "misinformed" and ignore the fact concerns are justified [37]. We also need to acknowledge that the COVID-19 vaccine will be rolled out in the context of recent protests against the killing of unarmed Black people, which has highlighted deeply entrenched racism that devalues Black lives [38]. With these events in mind, vaccination campaigns should be developed in partnership with Black communities and disseminated in a culturally and racially affirming manner. Trusted and racially/ethnically diverse community leaders should deliver clear and consistent messaging, stress that the vaccine is safe and effective, especially since the difference in vaccine intentions in NLB women compared with NLW women was nonsignificant once the belief in vaccine and safety was added to multivariable models. Ensuring that there is sufficient racial/ethnic diversity in vaccine trials may also help to allay concerns that study findings do not apply to all racial/ethnic groups. Additional approaches may include recruiting community health educators from intended audiences to do outreach and education, working with faith-based organizations, and hiring healthcare providers reflective of the population to be on the "front lines" of vaccine delivery. In addition, we should explore ways to engage the services of popular and diverse social media "influencers," an approach that has shown early promise with other health issues [39,40]. In addition to health messaging, we must prioritize vaccine distribution to the communities that have been disproportionately impacted by the pandemic. Some have raised concerns about how decisions about distribution of the vaccine will be made, since there are insufficient doses to cover the entire population at this time [41]. Moreover, we must not ignore the ongoing need to address underlying structural forces that give rise to health inequities and poorer health among racial/ethnic minorities which increase the risk of adverse consequences of COVID-19. Addressing limited access to healthcare and mitigating other social determinants of health that have produced these inequities should remain a priority. Without these efforts, the vaccine may further exacerbate inequities in COVID-19 incidence and mortality among these populations. Research on COVID-19 vaccine acceptability is in its infancy. Additional studies are needed to further understand how the vaccine will be received and perceived among nationally representative samples and we strongly recommend oversampling of racial/ethnic minority population subgroups in these studies. Additionally, recent COVID-19 studies have found that men are more likely to report they would get a vaccine when compared with women [13] and persons with liberal political views had the strongest vaccine intentions followed by moderates, and then conservatives. It is important for future studies to gain a deeper understanding on how demographic variables influence COVID-19 vaccine attitudes and intentions specifically. While there is much to be learned and leveraged from past successful and less successful vaccine campaigns , the current and unique sociopolitical context during which the COVID-19 pandemic is unfolding will require the development novel strategies to mitigate mistrust, increase uptake of the vaccine, and reduce cost and access barriers to healthcare. --- Conflict of Interest: None declared. Authors' Contributions: JDA conceived and designed the study, drafted the initial manuscript. HF made substantial contributions to the design of the work, interpretation of the data, drafted sections of the manuscript, and approved the final submission. NA performed and interpreted the analyses, drafted sections, and approved the final submission. RR drafted sections of the manuscript and approved the final submission. Ethical Approval: All study procedures were approved by the Institutional Review Board at Tufts University, Medford, MA. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All study participants provided informed consent. Informed Consent: Informed consent was obtained from all individual participants included in the study. --- Welfare of Animals: This article does not contain any studies with animals. Study Registration: This study was not formally registered. Analytic Plan Preregistration: The analysis plan was not formally preregistered. Data availability: Deidentified data from this study are not available in an a public archive. Deidentified data from this study will be made available by emailing the corresponding author. Analytic code availability: Analytic code used to conduct the analyses presented in this study are not available in a public archive. They may be available by emailing the corresponding author. Materials availability: Materials used to conduct the study are not publically available.

Widespread uptake of the COVID-19 vaccine is critical to halt the pandemic. At present, little is known about factors that will affect vaccine uptake, especially among diverse racial/ ethnic communities that have experienced the highest burden of COVID. We administered an online survey to a Qualtrics respondent panel of women ages 27-45 years (N = 396) to assess vaccine intentions and attitudes, and trusted vaccine information sources. 56.8% intended to be vaccinated and 25.5% were unsure. In bivariate analyses, a greater percentage of non-Latina White (NLW) and Chinese women reported that they would be vaccinated, compared with Latina and non-Latina Black (NLB) women (p < 0.001). Those who were uninsured, unemployed and those with lower incomes were less likely to say that they would be vaccinated. In analyses stratified by race/ethnicity, NLB women remained significantly less likely to report that they would be vaccinated compared with NLW women (adjusted odds ratio: 0.47; 95% confidence interval: 0.23, 0.94), controlling for age, marital status, income, education, employment, and insurance status. When analyses were additionally controlled for beliefs in vaccine safety and efficacy, racial/ethnic differences were no longer significant (adjusted odds ratio: 0.64; 95% confidence interval: 0.31, 1.34). Given that NLB women were less likely to report the intention to be vaccinated, targeted efforts will be needed to promote vaccine uptake. It will be critical to emphasize that the vaccine is safe and effective; this message may be best delivered by trusted community members.

INTRODUCTION A lthough use of effective contraception is growing in many parts of the world, only about 17% of women of reproductive age in sub-Saharan Africa use a modern contraceptive. 1 Low levels of usage, however, do not indicate a lack of interest in family planning among women and their male partners; while fertility desires in many countries in sub-Saharan Africa are high, so too is the demand to both space and limit births, even among young women. 2 Contraceptive use, as measured by the contraceptive prevalence rate , in Ghana has not changed significantly since a rapid increase from 12.9% in 1988 to 25.2% in 2003. 3 The most recent Ghana Demographic and Health Survey showed a CPR in 2014 of 26.7%. Although the National Population Policy has an explicit objective "to ensure accessibility to, and affordability of, family planning means and services for all couples and individuals to enable them regulate their fertility," 4 the total fertility rate in 2014 was 4.2 births per woman, a slight increase from the 2008 figure of 4.0 births per woman. 5 This is in contrast to the marked decline in the fertility rate observed between the mid-1980s and the 1990s. Unmet need for contraception is the percentage of women of reproductive age who want to stop or postpone childbearing but who report that they are not using a method to prevent pregnancy, 6 and is used as an indicator of the gap between the demand for contraceptives and contraceptive use. Current unmet need for contraception is high in Ghana at about 30% of married women and 42% of sexually active unmarried women. 5 This suggests potential high interest and demand for family planning among women and their partners. Encouraging new adopters of contraceptive methods is important, but so too is understanding consistency of use among existing users. 7 In 14 of 15 low-income countries, the majority of unplanned and unwanted births were the result of either contraceptive failure or discontinuation of a method for reasons other than a desire to get pregnant. 8 Women's expectations of and experiences with side effects may lead to satisfaction with their method and continuation with using it or dissatisfaction and discontinuation of their method. 9 Further, recent evidence suggests that the experience of side effects is increasingly a reason for why women have an unmet need for contraception. 10 In Ghana specifically, intolerance of side effects is often an underlying reason for method discontinuation 11,12 and is cited as the reason many women avoid initiating contraception. 13 In recent years, health concerns and the experience of or concerns about side effects are increasingly driving non-use of contraception by Ghanaian women; the percentage of women who mentioned these as their reason for non-use increased from 14% in 1988 to 43% in 2008. 14 This could, in part, be due to the heavy reliance on oral contraceptives and injectables, 5 methods for which women report high levels of displeasure with side effects. While there is a large body of evidence based on clinical trials detailing the possible physical symptoms and side effects that may be caused by using a particular contraceptive method, ranging from bleeding changes, headaches, breast tenderness, and weight change with hormonal methods to increased bleeding with the copper IUD, [15][16][17] how users feel about these side effects is inherently subjective. In many settings, including Ghana, side effects and health concerns or fears are conflated. 13 Further, Ghanaian women's unfavorable attitudes toward contraceptive methods, a key driver in use or non-use, originate from fears regarding the safety of these methods and intolerance of menstrual side effects, rather than from social or moral objections. 14 Menstrual disruption is often considered to be a clinically benign side effect and is therefore sometimes minimized or dismissed by health personnel. 18 However, many women have a low tolerance for menstrual changes, and thus side effects, and in particular bleeding changes, are a key reason for discontinuation of contraceptive methods. 19 Understanding individuals' expectations of and tolerance for side effects is an important part of counseling and a means to ensuring satisfaction with the chosen method, and ultimately to ensuring consistent use of contraceptive methods. In this study of Ghanaian family planning adopters, we aimed to describe method characteristics women find desirable and side effects women report would be untenable; describe the side effects women were counseled to expect from the method they adopted; and characterize to what extent women chose methods that matched their desires for acceptable vs. intolerable side effects and other method characteristics. --- METHODS --- Setting and Participant Recruitment We conducted this cross-sectional study in urban family planning clinics of 2 teaching hospitals and 3 district hospitals in Kumasi and Accra, Ghana. Data were collected between June 1, 2015, and August 31, 2015. All women attending these clinics for family planning counseling and method choice were approached by a member of the study team and informed about the study to gauge their interest and eligibility for inclusion. Inclusion criteria consisted of being over the age of 18, intending to adopt a new method of family planning, and being able to converse in either English or Twi or Ga . If the women met these inclusion criteria and agreed to participate, they were taken through a comprehensive verbal consent process. Consenting participants were interviewed both before and after their family planning counseling. This study used a convenience sample; all women who attended the clinic --- Survey Instrument The survey was developed by the authors, an international study team with experience in family planning in Ghana. Questions were developed based on literature and expert opinion. The questionnaire was pretested for clarity and flow among women in the study clinics who met the inclusion criteria before the beginning of data collection. Revisions to the questionnaire were made based on this pilot testing. Questionnaire items included previous use of contraceptive methods and reasons for discontinuation, as well as what method of contraception women wanted to adopt, what method characteristics they desired, which side effects they would find bothersome, and which would be intolerable, causing them to discontinue the use of a contraceptive. After their counseling session, women were asked whether they were leaving with a method and, if they were, whether they were counseled to expect any side effects. If they had been counseled about side effects, they were asked what side effects they were counseled to expect. These answers were collected as free response and grouped. Multiple responses were maintained. --- Survey Administration Interviews were conducted both before and after women's family planning counseling session in a private room near the family planning clinics. Only the woman and the research assistant were in the room where the interview took place, and no identifying information was collected. Phone numbers were used to link the pre-and postcounseling surveys. All data were collected on a Google tablet computer using Qualtrics software and results could not be seen once the form was completed. Questionnaires were interviewadministered by trained research assistants. --- Data Analysis To determine women's preferences, participants were asked 3 sets of questions. First, women were asked a variety of questions about method characteristics, such as, "I would prefer a method that I take every day." The answers were recorded on a 5-point Likert scale from strongly disagree to strongly agree. We then created a dichotomous variable to represent yes/no agreement to the statement by grouping "strongly agree" and "agree" to indicate agreement with the statement, and "neither agree nor disagree," "strongly disagree," and "disagree" to indicate disagreement. In the second set of questions to determine preferences, questions about side effects, such as, "I would not like a method if it stopped me from bleeding," were asked also on a 5-point Likert scale. Similarly, "strongly agree" and "agree" were grouped, as were "neither agree nor disagree," "strongly disagree," and "disagree." Finally, participants were asked whether the experience of a variety of side effects would be intolerable enough to cause them to stop using the method. The answers to these questions were collected as a dichotomous variable and included such statements as, "Increased bleeding would cause me to stop using the method." In the post-counseling survey, participants were asked about the method they had adopted and also if they were counseled to expect side effects. Those who answered they were counseled to expect side effects were asked which side effects they were expecting. The side effects they were expecting, as well as those they reported as being so intolerable as to cause them to stop using the method, were compared with their adopted methods. Those who adopted methods that are known to cause such side effects were determined to have adopted a method that was not concordant with their preferences. For example, if a woman indicated decreased bleeding would be bothersome enough to cause her to stop using her method, but she adopted the injectable or the implant, she was determined to have adopted a method discordant with her preference. Women who reported they were counseled to expect side effects that are in fact shown in the literature to be caused by the method they adopted were determined to have been counseled appropriately to expect side effects common with the chosen method. For example, a woman who adopted an IUD and reported she was counseled to expect increased bleeding was determined to have been counseled to expect a side effect common with her chosen method. These comparisons were done using crosstabs in SPSS version 22. Data are presented as descriptive statistics. --- Ethical Review All study materials and methods were reviewed and approved by the Ghana Health Service Ethical Review Committee and the University of Michigan Institutional Review Board. --- RESULTS --- Background Characteristics A total of 414 women completed the precounseling survey, and 411 completed the post-counseling survey. Of the original 414 participants, 336 left with a method and were matched between the 2 surveys . A total of 55 participants did not leave with a method and an additional 23 participants from the pre-counseling survey could not be matched with the post-counseling survey. Thus, our analytical sample consisted of the 336 participants who had complete records. Participants were generally well distributed across sociodemographic variables; the mean age of the total sample of 414 women who completed the pre-counseling survey was 29.3 years , and 248 participants were married . Of the 411 women who completed the post-counseling survey, 337 left their counseling appointment with a method . The primary reasons women gave for not leaving with a method were needing to wait for a pregnancy test , the clinic not having the supplies/equipment/providers necessary , having high blood pressure , and needing to consult with the husband . Almost half of the sample had previously used some form of contraception, with the injectable being the most commonly used method , followed by the pill . --- Women's Stated Preferences The majority of the full sample of participants agreed or strongly agreed that they would prefer a method that protects them for years . A substantial number of women agreed or strongly agreed that they would prefer a method that they take every few months or that lasts forever . Only 67 women agreed or strongly agreed they would prefer a method that they had to take every day. A majority was not opposed to using a method that requires a visit to the facility to stop using it. Participants had generally unfavorable attitudes toward bleeding changes caused by contraceptives. For example, a majority reported that increased bleeding , irregular bleeding , and amenorrhea would be intolerable enough for them to stop using the method . Decreased bleeding seemed to be more tolerable to women as only 94 women reported it would be intolerable enough to stop using the method. --- Women's Method Choices The majority of women who left with a method and were matched with the precounseling survey chose a long-acting reversible contraceptive ; 52 women chose the IUD and 135 women chose the implant . A substantial number chose injectables. The remaining chose either sterilization or short-acting methods . --- Counseling About Side Effects Of the 336 participants who adopted a method of contraception, 218 reported they were counseled to expect side effects. Table 3 presents data on the percentage of women who reported being counseled to expect certain side effects by the method adopted; the side effects presented in the table are those side effects that are known to commonly occur with the method. Among the 52 women who adopted the IUD, only 16 reported they were counseled to expect an increase in bleeding . Among the implant adopters, 28 reported they were counseled to expect irregular bleeding, while 14 and 37 reported they were counseled to expect decreased bleeding and no menses, respectively. For those using injectables, 46 reported they were counseled to expect their menses to stop. --- Consistency Between Women's Method Choice and Stated Preferences Figure 2 shows the concordance between women's method choice and their stated preferences at the pre-counseling survey with regard to desired duration of effectiveness and intolerable side effects . The intolerable side effects included in the figure are those bleeding side effects that are commonly expected with use of that particular method; for example, increased bleeding for IUDs. Figure 2 counseled to expect bleeding changes with their chosen method and the proportion reporting they were not counseled to expect bleeding changes. Among IUD adopters, 73.1% stated in their pre-counseling survey that experiencing increased bleeding would cause them to stop using the method, suggesting discordance with their chosen method because copper IUDs have been shown to increase bleeding among many users. Similarly, 70.4% of implant adopters stated at the pre-counseling survey that irregular, decreased, or no bleeding would cause them to stop using the method and 72.2% of injectable users stated that decreased or no bleeding would cause them to stop using the method, suggesting discordance since these are commonly experienced side effects with use of these methods. The majority of IUD, implant, and injectable adopters reported they were not counseled to expect bleeding changes. Bleeding changes are in fact expected to occur with all 3 of these methods. There seemed to be greater concordance between women's choice of method and their stated desired duration of effectiveness: at least 87% of the adopters indicated a desired duration of effectiveness that lined up with the duration of effectiveness of their chosen method . --- DISCUSSION This study, conducted in urban areas of Accra and Kumasi, Ghana, sought to characterize women's contraceptive preferences and to examine whether their adopted method was consistent with their stated preferences. We found that the vast majority of women attending family planning clinics had a method in mind before the family planning counseling session, and many had strong preferences for and against method-specific qualities or side effects. It seems that most women received the method that they had been planning to use, and most women adopted a method that appeared consistent with some of their stated preferences, most notably duration of effectiveness. However, the majority of women left the clinics with a method that is known to cause the side effects they had characterized as intolerableones that would cause them to stop using the method. Furthermore, most women reported they were not counseled to expect these commonly occurring side effects. Improving access to and use of modern contraception is an important strategy to both reduce maternal mortality and assist women to meet their fertility goals. Side effects are a key reason both for not using contraception and for discontinuing use of contraceptives. 10,14 Women are diverse in preferences and providers need to guide them to the method that best suits their preferences. Our results show that while there is generally a good match with women's preferences regarding desired duration of effectiveness, many women are leaving with a method well known to cause a side effect they characterized as intolerable. It is not clear from these data if women were well informed about potential side effects; many did not report being educated on these side effects. Adopting a patient-centered model of care has been demonstrated to increase consistent contraceptive method use in high-resource settings, 20 and has been tested on a limited basis in lowresource settings. 21 More specifically, providers need to tailor information for women so that they know what side effects are common and not harmful and what things might cause temporary or longer-term discomfort, compared with potential complications to look for. It appears that not all women in our study were provided with clear information about the side effects they should expect with the method they were given. Others too have reported that women are sometimes poorly informed about side effects, 2 and this could lead to dissatisfaction and discontinuation. A recent analysis of the 2008 Ghana Demographic and Health Survey showed that 71% of IUD adopters discontinued the use of this method, mainly due to side effects and health concerns. 22 Discussion of method side effects appears to not have changed much over the past decades. In 1995, Bongaarts and Bruce 23 found that fewer than 50% of new adopters were counseled about their method's side effects and about 35% of women had discussions about how to manage these side effects should they arise. 23 More recently, in Niger in 2012, only 40% of women reported being informed of possible side effects of methods. 24 Some providers may be reluctant to discuss potential side effects for fear of the "nocebo" effect, 25 the phenomenon where women experience side effects after being told to expect them because of the power of suggestion. However, evidence suggests that the discussion of negative side effects is not detrimental to method adoption and may be beneficial to method adherence. [26][27][28] In fact, patients report that the reluctance of providers to discuss side effects makes them distrustful of counseling. 20 Providers need training and support in goodquality counseling skills. The counseling session is an opportunity for providers to help align clients' desires and preferences with method characteristics so that clients receive a method that is likely to suit their needs, and thus one that they are more likely to continue using. Above all, it is the responsibility of providers to help women make an informed family planning choice by giving them information about the methods available to them, the characteristics of those methods, including side effects, and how to correctly use their chosen method. Given this information, some women may decide to sacrifice some of their preferences when those preferences do not align perfectly with available methods . While supply-side barriers have long been determined to be factors limiting contraceptive uptake and continuation, demand-side barriers are increasingly being determined to be important as well. 14 Once a woman overcomes the many potential barriers to adopting a method, the provider can help ensure her satisfaction with the chosen method by reassuring her that side effects are common with use and not harmful, which may improve her adherence to and continued use of the method. If women do not have good family planning experiences, this may erode their trust of the health care provider or health center and influence members of their social network to avoid using contraceptive methods. Clearly, more research is needed to fully understand the patient-provider interaction in this particular study setting. Beyond the individual patient-provider relationship, it is possible that mass media campaigns aimed at reducing misinformation and myths about common side effects could help women to be more tolerant of innocuous side effects. The ability of women to use family planning is, of course, a larger societal issue, and many societal and cultural barriers reduce demand for family planning. 29 The most frequently cited side effect that women said would cause them to stop using the method was amenorrhea, followed by increased bleeding. This finding supports research conducted in the 1970s, which suggested that amenorrhea was unacceptable to many women, 30 although no African women were included in that study. In a qualitative study of women from a range of countries from the 1990s, women were "dismayed" by the possibility of a contraceptive method causing amenorrhea. 31 However, in more recent work, Glasier and colleagues 32 found that Nigerian women were generally willing to accept a method of contraception that induces amenorrhea, even though a majority said monthly bleeding was important to them. This could be indicative of a shift in the acceptability of a cessation in menses, although most women in our sample were not eager to adopt a method with such an effect. Education on the non-contraceptive health benefits of reduced blood loss associated with hormonal contraceptives, and reassurance that current global evidence-based guidance advises this is not harmful, 15,33 could make some clients adopt methods with such a side effect. --- Limitations This study is not without limitations. The relatively small sample size, as well as the fact that all study sites were located in urban areas, reduces the ability of these findings to be generalized. Further, these women were all seeking services at family planning clinics at district hospitals or tertiary care centers and thus might have been more interested in LARCs than women who seek contraceptive services from other locations. Qualitative investigation is important to develop clearer and deeper understanding of the concerns women have about contraceptives and what they expect from service providers. It is possible that counseling from the provider during the session changed women's perceptions toward side effects, and so their stated intolerance of a side effect before their visit changed post-counseling. These data would not capture a woman's change in preferences due to counseling. --- CONCLUSION This study interviewed women before and after their contraceptive counseling session to determine to what extent women were receiving methods that matched their stated desires. Women had clear preferences regarding several contraceptive characteristics, including duration of effectiveness and side effects. It is important for providers to understand women's individual preferences and needs and how these preferences may interact with each other to guide women to make informed family planning choices. While women generally adopted methods that matched their stated desires related to duration of effectiveness, the adopted methods were less well matched to women's preferences regarding potential menstrual side effects. Future work could investigate whether the experience of side effects, both expected and unexpected, is associated with contraceptive adherence and continuation. --- Competing Interests: None declared.

Women's method choice largely matched their stated desired duration of effectiveness but not their desires to avoid certain side effects. While most women reported they were counseled about side effects, many fewer reported being specifically counseled about common menstrual side effects with their chosen method, including side effects the women said would cause them to stop using the method.