PrEP and the syndemic of substance use, violence, and HIV among female and male sex workers: a qualitative study in Kisumu, Kenya

Abstract Introduction Female and male sex workers experience heightened vulnerability to HIV and other health harms that are compounded by substance use, physical and sexual violence, and limited access to health services. In Kisumu, Kenya, where sex work is widespread and substance use is a growing public health concern, offering pre‐exposure prophylaxis (PrEP) for HIV prevention could help curtail the HIV epidemic. Our study examines “syndemics,” or mutually reinforcing epidemics of substance use, violence and HIV, in relation to PrEP acceptability and feasibility among female and male sex workers in Kenya, one of the first African countries to approve PrEP for HIV prevention. Methods From 2016 to 2017, sex workers in Kisumu reporting recent alcohol or drug use and experiences of violence participated in qualitative interviews on HIV risk and perspectives on health service needs, including PrEP programming. Content analysis identified themes relating to PrEP knowledge, acceptability, access challenges and delivery preferences. Results Among 45 female and 28 male sex workers, median age was 28 and 25 respectively. All participants reported past‐month alcohol use and 91% of women and 82% of men reported past‐month drug use. Violence was pervasive, with most women and men reporting past‐year physical (96% women, 86% men) and sexual (93% women, 79% men) violence. Concerning PrEP, interviews revealed: (1) low PrEP knowledge, especially among women; (2) high PrEP acceptability and perceived need, particularly within syndemic contexts of substance use and violence; and (3) preferences for accessible, non‐stigmatizing PrEP delivery initiatives designed with input from sex workers. Conclusions Through a syndemic lens, substance use and violence interact to increase HIV vulnerability and perceived need for PrEP among female and male sex workers in Kisumu. Although interest in PrEP was high, most sex workers in our sample, particularly women, were not benefiting from it. Syndemic substance use and violence experienced by sex workers posed important barriers to PrEP access for sex workers. Increasing PrEP access for sex workers will require addressing substance use and violence through integrated programming.


| INTRODUCTION
Globally, sex workers experience increased risk of HIV, sexually transmitted infections (STI), and other adverse health outcomes, yet often have limited access to vital health services [1]. Female sex workers have an estimated 13 times higher risk of HIV acquisition than the general population of women [2]. Male sex workers have over 20 times higher HIV prevalence than the general male population [3] but have received much less research and programmatic attention than female sex workers [4]. Most male sex workers exchange sex with other men, but HIV prevention interventions targeting "men who have sex with men" (MSM) may not reach those who do not identify as gay or as "sex workers" [5,6]. In sub-Saharan Africa, where the majority of the HIV epidemic is concentrated, sex work among women and men is criminalized, stigmatized and often driven by poverty [7][8][9]. Thus, female and male sex workers require HIV programming that is responsive to their complex lived experiences.
Further compounding sex workers' vulnerability to HIV infection, sexual exchange is often negotiated in alcoholserving establishments, where many female and male sex workers engage in alcohol and drug use that can exacerbate their risks for HIV and other health harms [10]. Research from Kenya, which has an entrenched HIV epidemic, has linked substance use with unprotected sex [11] and physical and sexual violence among female sex workers [12,13]. Emerging research has also identified a high prevalence of substance use and violence victimization among male sex workers in Kenya [14,15], as well as multiple forms of stigma that limit their use of existing HIV prevention services [16].
Recognizing the need for new strategies to fight its persistent HIV epidemic, in 2017, Kenya became one of the first African countries to make antiretroviral pre-exposure prophylaxis (PrEP) available for HIV prevention among "key populations, " including sex workers, MSM and people who inject drugs. Although these categories could include high-risk sex workers, to date, PrEP service delivery has yet to address the specific contexts of co-occurring alcohol and drug use and violence victimization experienced in sex worker populations. Given that social and behavioural factors significantly influence PrEP utilization [17,18], the success of PrEP for HIV prevention in sex worker populations affected by substance use and violence will require understanding their attitudes on, experiences with and preferences for PrEP service delivery.
This qualitative study focused on "syndemics, " or mutually reinforcing epidemics of substance use, violence, and HIV among female and male sex workers in Kisumu County, western Kenya. Syndemic theory refers to the clustering of two or more health conditions. This perspective highlights how synergistic interactions among psychosocial and biological conditions underlie patterns of disease clustering within local political, economic and social contexts, helping to identify needs for comprehensive healthcare services for those affected [19,20].
Sex work is criminalized in Kenya, but its widespread practice in Kisumu reflects economic inequalities and limited opportunities. While HIV prevalence in Kisumu County reaches 18.7% (vs. 5.6% nationally) [21], it is disproportionately concentrated among key populations. As of 2008, the latest estimate available, there were nearly 1700 female sex workers in Kisumu, among whom HIV prevalence reached 56.5% [11]. While no similar estimates are available for male sex workers, recent studies with MSM in Kisumu have documented widespread substance use, mental health conditions, and physical and sexual violence victimization that could exacerbate HIV risk [22]. Drawing from syndemic theory, we explored how substance use, violence and HIV risk shape PrEP acceptability, access and intervention needs among sex workers in Kisumu.

| Study design and population
From 2016 to 2017, to recruit sex workers, peer educators from local HIV prevention organizations conducted outreach and targeted sampling at sex work "hotspots" [23]. Snowball sampling later involved asking enrolled participants to refer other sex workers to the study [24]. We obtained oral consent to screen for eligibility, which included being ≥18 years old and reporting: past-month trading of sex for money, alcohol, drugs or other material items; past-month "problematic" alcohol use (binge drinking of ≥5 drinks or being drunk "most" or "all" of the time when drinking) or any injection or non-injection drug use; and past-year physical or sexual violence from intimate partners, clients or police. We used purposive sampling methods to ensure that our sample was diverse in age and substance use [25], which necessitated additional recruitment of women. Eligible participants provided written consent. Ethics review boards at The Ohio State University and Maseno University in Kisumu, Kenya, approved all study protocols, which were translated into Kiswahili and Dholuo.

| Data collection
Trained, multilingual interviewers conducted semi-structured interviews lasting up to 90 minutes in private offices in participants' language of choice (English, Swahili, Dholuo). Interviewers assured confidentiality and no names or identifiers were recorded. Semi-structured interviews explored sex work, substance use, experiences of violence and perspectives on health needs (e.g. "How did you become involved in sex work? Have there been times when you have experienced violence?"). Regarding PrEP, we asked if participants had ever heard of it, and if not, provided a basic description. We then asked, "What do you think about PrEP for HIV prevention for women [men] who do sex work? Is this something you would be interested in? Why [not]?" We then explored personal interest in PrEP and preferred service delivery locations and providers. Finally, we noted that PrEP is effective and promotes adherence when taken around the same time every day and asked, "What do you think about this? How easy or difficult would it be for you to take PrEP every day?" Interviewers referred interested individuals to organizations offering PrEP for more information, and followed up as necessary. Interviews were audio recorded, transcribed verbatim and translated into English for text analysis following a structured protocol [26]. We continued interviewing until we reached saturation [27], the point at which we determined as a team that we were repeatedly hearing similar information across interviews. We then decided that our sample size was sufficient to answer our main study questions (i.e. additional interviews would not yield important new insights) around health needs in syndemic contexts.

| Data analysis
Codebook development involved several steps [28]. First, the research team read through selected transcripts and independently generated lists of initial themes based on the interview guide content (e.g. PrEP knowledge) [29]. The team met to discuss, operationalize and organize themes into a set of codes. One research assistant applied final codes to all transcripts using MAXQDA. The PI assessed code application for consistency and met regularly with the RA to discuss and clarify any issues. Content analysis allowed us to identify the most common experiences regarding PrEP knowledge, use, acceptability and preferences for service delivery within contexts of syndemic substance use and violence. We adopted a constant comparative method [30] to assess similarities and differences across women's and men's experiences. Findings are illustrated with quotes and pseudonyms are used to protect confidentiality.

| Sample characteristics
Among 45 women and 28 men (n = 73 sex workers), women were slightly older than men (median 28 vs. 25 years;  Table 1). Attesting to the considerable mobility of both groups, most had lived outside of Kisumu (91% of women, 89% of men). All participants reported past-month alcohol use, with 56% of women and 71% of men reporting "always" or "often" drinking to the point of becoming drunk. More women than men reported ever injecting drugs (49% vs. 36%) and recently using heroin (47% vs. 32%). Other pastmonth drug use was common, as 91% of women and 82% of men used drugs, including bhang (marijuana; 80% women, 75% men), miraa (khat; 27% women, 61% men), and non-prescribed medications (18% women, 43% men). Violence victimization was pervasive, with the majority of participants reporting past-year physical (96% women, 86% men) and sexual (93% women, 79% men) violence. HIV prevalence was high, with 31% of both women and men self-reporting HIVpositive status.
Through qualitative interviews, we identified the following key themes: (1) differences between women and men in PrEP knowledge and experience; (2) high acceptability and perceived need for PrEP within syndemic contexts of sex work, substance use and violence; and (3) preferences for greater sex worker involvement in PrEP service delivery.

| PrEP knowledge and use
Stark differences emerged between women and men in PrEP knowledge and experience. Despite several major PrEP clinical trials and demonstration projects having been conducted in Kenya (including in the western region) [31], only 20% (n = 9) of women had even heard of PrEP. Many women confused PrEP with post-exposure prophylaxis (PEP), which they were familiar with due to the frequency of rape and sexual violence and the availability of PEP in local clinics. After the interviewer explained PrEP, women expressed strong interest in trying it and asked numerous questions, including where to obtain it. Esther, 27, one of the few women who had previously heard of PrEP, described being turned away when trying to access it, illustrating the discretion of healthcare providers in determining PrEP "eligibility:" He [healthcare worker] was asking about the high risk of being infected. I told him, if I am a sex worker, definitely I am at high risk, even though I use condoms. . .one day it can burst. If I get PrEP, I am safe. But because I was using condoms, I was not eligible for PrEP; that is what he told me.
In contrast, most men (86%, n = 24) knew about PrEP. Of the seven HIV-negative men who had experience using PrEP, four were using it at the time of the interview; all four held favourable views of it. While describing PrEP as "awesome, " two mentioned having to overcome unpleasant side effects: [Taking PrEP] is hard but it has benefits and challenges for a few weeks because. . .. It's like you are tired and drowsy and you have a bloated stomach, but [the side effects] end.
Side effects caused two men in our sample to discontinue their PrEP use. Two HIV-positive men were involved with local organizations as "PrEP ambassadors, " helping educate other men about PrEP and "mobilize" them into services. They also reported that side effects caused several of their peers to stop using PrEP. The other individual in our study who discontinued PrEP feared that because he received PrEP at a clinic that also provided HIV care, people would assume him to be HIV-positive and taking ARVs.

| Acceptability and perceived need for PrEP in syndemic contexts of sex work
Both women and men described the multifaceted social, legal and health hazards of sex work. Overall, narratives revealed high acceptability of and perceived need for PrEP due to syndemic conditions wherein substance use and violence exacerbated HIV risk. Given the high HIV prevalence in Kisumu, most participants expressed concern at not knowing clients' HIV status. Furthermore, many participants had clients who refused to use condoms despite not disclosing their status.
The pervasiveness of heavy alcohol use within sex work contexts further increased sex workers' perceived need for PrEP. Women and men described frequently drinking alcohol with clients in bars and clubs prior to having sex, which sometimes led them to "black out" and "forget" about condoms or agree to have condomless sex for higher pay. Combined with In such cases in which sex workers became "so drunk they lose their senses, " PrEP would provide an extra level of protection from HIV.
Moreover, women discussed how the likelihood of sexual violence increased in contexts of sex work and heavy drinking, including forced condomless sex: It [sex work] is a job that can cause your death. Someone can stab me with a knife the moment I leave here. Or someone can have sex with me without a condom, or someone can rape me. They can come as a group of three people, break the door and all the three people can rape me.-Sera, age 21 Our risks come in that the condom might burst. Maybe I can be raped somewhere by someone whose status I don't know. It [PrEP] will therefore protect us.-Rose, age 27 Although sexual violence was described more frequently by women, men also viewed PrEP as a valuable prevention tool in contexts of alcohol use and violence victimization: [PrEP] is good [because] there are some clients who can rape you, and alcohol is stronger than some sex workers. After he has taken so much alcohol and he is drunk, when he goes to the room he will not remember to use condom[s], so I think it is good for sex workers.-Steven, age 23 Overall, perceived acceptability of and need for PrEP was high among female and male sex workers.

| Preferences for PrEP delivery
Given the syndemic health risks experienced by sex workers in Kisumu, maximizing the HIV prevention potential of PrEP will require understanding their preferences for PrEP delivery. Across interviews, participants stressed the importance of accessing PrEP from non-judgmental providers within service delivery settings that were "friendly" to sex workers. Participants did not want to be discriminated against due to their sexual orientation, substance use, or engagement in sex work. While some women described feeling comfortable interacting with non-judgmental clinical providers in health centres, others explained that fellow sex workers would be more appropriate sources of PrEP information, referrals and services because they would understand the multifaceted risks of sex work and could converse comfortably about PrEP in the context of those risks: It should be a fellow sex worker. We can tell her everything that we are going through. If you look at us, not all of us are fools. There are those who have reached the university level. They can be trained a little and later come to do such work. We want our freedom. What I mean by freedom is a place where I will enter and talk to them the way I want.-Rose, age 27 Given their understanding of the contextual realities of sex work, fellow sex workers could also help women address potential challenges with PrEP adherence: If you want to help people then it is good that you [employ] people who are like [us] because they have been in the field [practicing sex work] and they understand our lives. So it is good that you employ those who have done the same job [sex work] to dispense the drug for them . . . And even the counselors should also be sex workers.-Nancy, age 31 As noted above, some men in our sample already had experience discussing PrEP with their peers as "PrEP ambassadors. " These men stressed the importance of PrEP provision involving healthcare workers who were comfortable serving males who had male sex partners. When asked about who should deliver PrEP, Johnson, age 32, responded: Make sure they are people who are not homophobic, that they are people who are friendly to gay people. Because if they are homophobic, they are going to make remarks that will make people be angry. And then people will not be accessing that kind of service from that place.
Some men also specified that providers should be comfortable discussing same-sex sexual behaviours and sex work in non-judgmental ways. As Samuel, age 26, said, providers unfamiliar with same-sex behaviours should not provide PrEP to male sex workers because they would "not know anything about sex work. " Importantly, women and men emphasized that PrEP providers should understand sex workers' specific syndemic vulnerabilities as described above. Participants provided specific suggestions for improving PrEP uptake among sex workers. First, due to the concentration of sex work and drinking establishments catering to sex work, PrEP should be provided through convenient, confidential spaces in these areas. Second, PrEP services should be sufficiently flexible and involve creative strategies to accommodate the high mobility of many sex workers (e.g. mobile PrEP services). Finally, due to the myriad health concerns of female and male sex workers with substance use and violence victimization, efforts to deliver PrEP should be holistic and integrated with other health services. According to Jackson, age 30: We need facilities that work 24 hours a day. Even if you are raped, you can easily rush to the facility. I also said that we needed people to fight for our rights. . .Economic empowerment, security training, HIV prevention sensitization, training on behaviour change and risk reduction. [Sex workers] also need training on drugs and substance abuse. You need to create employment.

| DISCUSSION
In the context of an ongoing HIV epidemic in Kenya, national guidelines recommending PrEP for HIV prevention among key populations carry enormous potential to curb the epidemic. In Kenya, where multiple PrEP projects are ongoing, an estimated 53,000 to 54,000 individuals are currently prescribed PrEP [32]. In regions shouldering the burden of the HIV epidemic like Kisumu, understanding how syndemic conditions of substance use and violence increase sex workers' vulnerability to HIV and ability to adopt new prevention technologies like PrEP will be critical [33]. Viewing this situation through a syndemic lens, which conceptualizes how multiple health conditions synergistically reinforce each other [21,22], can help address sex workers' health and social needs within local contexts [34]. We provide several recommendations for improving the impact of PrEP for HIV prevention among women and men engaged in sex work.
First, we found differences in PrEP knowledge and use between women and men. While men had specific concerns about side effects and being viewed as HIV-positive if taking PrEP, women lacked even basic access to PrEP information and services. This is particularly concerning because young Kenyan women have the highest HIV incidence nationally [19], and sex work and syndemic vulnerabilities stemming from substance use and violence further increase HIV risk. Since collecting our data, a local sex worker organization has responded to women's exclusion by conducting PrEP outreach with sex workers and launching radio media campaigns to increase PrEP awareness and reduce stigma around its use (unpublished fieldnotes). In coordination with ongoing efforts by the Kenya Ministry of Health, additional organizations in Kisumu now offer PrEP to expand access to multiple populations, including additional safe spaces for MSM and men who engage in sex work. As PrEP continues to be scaled up across Kenya, continued involvement of sex workers in raising awareness and reducing stigma will be critical to its success.
Second, we found high acceptability and perceived need for PrEP within contexts of sex work, substance use and violence. Approaching questions of PrEP access and delivery from this syndemic perspective helps emphasize the interconnectedness of HIV and other health and social concerns and highlights the need for more comprehensive and integrated services for sex workers. As our study found, entrenched substance use-particularly heavy drinking-and co-occurring physical and sexual violence and underlying poverty all interact to increase HIV risk and PrEP need [35]. Furthermore, increasingly prevalent drug use in this region of Kenya may compound the social stigma already surrounding sex work, resulting in further marginalization of sex workers [36]. We recommend that facilities offering PrEP link with the growing number of psychosocial health services in Kisumu to address these complex issues, including organizations that focus on gender-based violence, addiction and mental health issues (unpublished fieldnotes). A communitybased "technical working group" that joins all parties together, including sex workers, could collaboratively brainstorm around PrEP challenges and share resources, as this model has been helpful in coordinating efforts towards other dimensions of the local HIV epidemic. Our findings reflect the need for biomedical prevention tools to be delivered with consideration of sex workers' "real world" vulnerabilities [37].
Finally, as found elsewhere [38], sex workers in our study emphasized the need for non-judgmental PrEP service delivery. While women and men worried about stigma against sex work, men were particularly concerned about homophobia and provider discomfort with same-sex behaviours. We suggest that stigma is another important factor in creating syndemic vulnerability, which has also been identified as major barrier to PrEP uptake across marginalized populations [39]. To ensure community ownership of new HIV prevention technologies, sex workers experiencing syndemic substance use and violence should be involved in PrEP intervention development and implementation [40]. Innovative PrEP intervention strategies could include training peer sex workers in PrEP outreach (e.g. "PrEP ambassadors"). Well-trained sex workers who are actively taking PrEP and aware of the syndemic risks experienced by fellow sex workers could effectively speak to overcoming challenges, including side effects, alcohol use and concerns about stigma, as well as provide peer-based adherence support. Importantly, all of the above efforts would benefit from broader efforts to change the structural conditions that criminalize sex work and render substance use and violence so pervasive in the first place [1].
Our study has limitations. Our sample was purposively selected from a high-risk population and may over-represent levels of drug and alcohol use among sex workers. We also collected data during a time of rapid changes in PrEP policy, and our results reflect only the very early stages of dissemination. We provided very general information about PrEP; future, in-depth research is needed to explore attitudes regarding initial clinical screening, side effects, ongoing STI screening, sexual and reproductive health needs, and potential social implications of using PrEP. Nevertheless, our study provides rare insight into the experiences of female and male sex workers who report substance use and violence who to date have largely been excluded from PrEP research and programmes.

| CONCLUSIONS
Given the interlinked nature of substance use, violence and HIV vulnerability among sex workers in Kenya, our findings suggest that successful implementation of HIV prevention technologies for vulnerable populations can be strengthened using a syndemic perspective [18, 38,41]. PrEP research and programming efforts should engage sex workers who use alcohol and drugs who have not yet been included in research or outreach efforts [33]. As Kenya continues to scale up PrEP delivery as part of comprehensive national HIV programming, soliciting the perspectives of key populations experiencing syndemic HIV vulnerability will be critical.