TU’WASHINDI NA PREP: WORKING WITH YOUNG WOMEN AND SERVICE PROVIDERS TO DESIGN AN INTERVENTION FOR PREP UPTAKE AND ADHERENCE IN THE CONTEXT OF GENDER-BASED VIOLENCE

Abstract

HIV pre-exposure prophylaxis (PrEP) reduces HIV acquisition among adolescent girls and young women (AGYW). Existing evidence suggests that uptake and adherence are low among AGYW and that relationship factors such as gender-based violence (GBV) are important barriers. Through a community-based participatory research (CBPR) process, a youth advisory board (YAB), service providers (SP), and a study team developed the Tu’Washindi na PrEP intervention to support AGYW PrEP use in the context of GBV. The YAB also guided the formative research and interpretation of results. The authors pretested the intervention with SP, AGYW and their partners, and community change agents, and then developed guides for AGYW support clubs, community-based male sensitization sessions, and couples-based events that included formulation of story lines for dramatized PrEP negotiation and information dissemination skills. Stakeholder engagement led to an intervention responsive to AGYW’s needs for PrEP support in the context of their relationships, which was evaluated through a 6-month pilot community randomized controlled trial.

Adolescent girls and young women (AGYW) who live in a context of heightened gender inequality and high risk of gender-based violence (GBV) represent a large subpopulation who are uniquely vulnerable to HIV infection. There is strong evidence that the experience or fear of GBV in sexual relationships is associated with having limited relationship power (Pulerwitz, Mathur, & Woznica, 2018; Rigby & Johnson, 2017) and increased HIV incidence (Bello et al., 2019; Coker, 2007; Decker et al., 2009; Dunkle & Decker, 2013; Jewkes, Dunkle, Nduna, & Shai, 2010; Kouyoumdjian et al., 2013; Pulerwitz et al., 2018; Zablotska et al., 2009). GBV increases HIV risk both directly, because of physical trauma from forced and unprotected sex (Dunkle et al., 2004; McLean, Roberts, White, & Paul, 2011), and indirectly, because of psychosocial consequences from the loss of relationship power (Coker, 2007; Dunkle & Decker, 2013; Pulerwitz et al., 2018), diminished ability to negotiate use of HIV prevention strategies (Kacanek et al., 2013; Swan & O’Connell, 2012), and high-risk behavior resulting from decreased self-worth, social isolation, and poor self-efficacy (Brown et al., 2014; T. A. Roberts, Auinger, & Klein, 2005; Salazar et al., 2005). With 47% of Kenyan women ages 20–24 reporting past experience of physical/sexual violence (DREAMS Partnership, 2020), developing HIV prevention approaches that target GBV is critical to the goal of achieving epidemic control.

While pre-exposure prophylaxis (PrEP) offers an opportunity to reduce HIV incidence among AGYW and was initially envisioned as “female-controlled,” it has been reconceptualized as “female initiated” given evidence suggesting that GBV and gender inequality introduce barriers to uptake of and adherence to this method (Lanham et al., 2014; Leddy, Weiss, Yam, & Pulerwitz, 2019; Montgomery et al., 2015; S. T. Roberts et al., 2016). These barriers occur at multiple levels. At the individual level, poor self-efficacy hinders seeking and persisting with health-promoting behaviors (Brown et al., 2014; Hatcher et al., 2016; Woollett & Hatcher, 2016). While many women in PrEP trials express a desire for disclosure and have disclosed PrEP use to their male partners (Lanham et al., 2014), for others, dynamics at the partner level, such as poor communication, low decision-making power in inequitable relationships, and fear of negative or even violent reactions, limit disclosure about PrEP use, thereby decreasing support for the use of this method (Corneli et al., 2015; Lanham et al., 2014; Montgomery et al., 2015; Stadler, Delany-Moretlwe, Palanee, & Rees, 2014). These influences are further compounded by a lack of PrEP awareness, inequitable norms, and stigma associated with taking PrEP, which contributes to partner and community opposition to female PrEP use (Hartmann et al., 2016, 2019a, 2019b, 2019c; Stangl et al., 2019; Velloza et al., 2020). A multilevel intervention to address these barriers is therefore essential to ensure that AGYW subjected to inequality and GBV can benefit from available HIV prevention options.

Despite advances in research on PrEP uptake and adherence, the lack of youth-designed interventions to increase PrEP use in the context of gender inequality and GBV among AGYW represents a critical gap in the field. In a 2019 commentary, Mannell and coauthors called attention to the reasons for the failure of the behavioral interventions targeting violence and HIV, citing the lack of meaningful involvement of young women and a focus on individual-level risk factors. They called for a process of codevelopment that involves giving young people an opportunity to analyze their own lives and identify intervention models that resonate with their realities (Mannell et al., 2019). This call echoes lessons from youth participatory action research, which highlight the increased relevance and transformative power of youth involvement (Oliver, Collin, Burns, & Nicholas, 2006; Wallerstein & Duran, 2008).

In response to the evidence gaps around youth-designed, multilevel interventions addressing the co-occurrence of GBV and HIV, we designed the Tu’Washindi na PrEP intervention. This article details the participatory process used to develop and refine the locally relevant multilevel intervention, and it also describes the final intervention model designed to address GBV and relationship power among AGYW and thereby increase PrEP uptake and adherence.

METHOD

STUDY SETTING

Tu’Washindi was conducted in Siaya County, which is located in the former Nyanza Province in western Kenya, along the shores of Lake Victoria. As of 2019, the county had a population of 993,183 (Kenya National Bureau of Statistics, 2019). The county is home primarily to peri-urban, fishing, and rural communities, with subsistence agriculture and fishing being the main economic activities (Republic of Kenya, 2015). It has the second highest HIV incidence in Kenya (2.5% per year), as well as the highest prevalence of GBV in the country, with 22% of women aged 15–49 having reported sexual violence and 56% having reported physical violence at least once since age 15 (Kenya National Bureau of Statistics & ICF International, 2014).

INTERVENTION DESIGN METHODS

The Tu’Washindi intervention is guided by social cognitive theory (Bandura, 1994) in combination with a socio-ecological framework for PrEP introduction (Mathur, Pilgrim, & Pulerwitz, 2016). The original model, based on a review of the literature, consisted of three components aimed at addressing barriers to PrEP use at the partnership, clinic, and community levels: (1) clinic-based counseling for AGYW on relationship dynamics, PrEP disclosure, and GBV safety planning; (2) the integration of GBV screening into PrEP delivery, including training and support materials for PrEP providers; and (3) the use of support clubs to address community barriers such as stigma and social isolation, as well as self-efficacy for PrEP use. Video vignettes were proposed to be incorporated into support club activities to present a role model PrEP use and improve outcome expectations for AGYW in violent relationships who wanted to use PrEP.

To encourage sustainability of the intervention, the study and intervention were embedded within the DREAMS initiative, a public–private partnership that is implementing a comprehensive package of programs to reduce HIV risk among AGYW (PEPFAR, 2015). In Siaya County, more than 48,000 AGYW have been enrolled into the program. Key activities include school- and community-based HIV and GBV prevention; referrals for postviolence care; reenrollment of girls in schools; and provision of PrEP, HIV counseling and testing (HCT), condoms and contraception, and social asset building through “Safe Spaces.” Safe Spaces are community locations where groups of AGYW convene weekly for sexual/reproductive health and life skills education led by a slightly older female “Mentor” who has completed secondary education and is rooted in the community. There are more than 500 Safe Spaces in the county, and these serve as the coordinating point for many other DREAMS activities, including clinical services delivered by mobile teams of clinicians and counselors. The DREAMS clinical team has been delivering daily oral PrEP with combination tenofovir and emtricitabine (TDF/FTC) through the Safe Spaces since 2017, and about 4,000 confirmed HIV-negative AGYW, identified as “high risk” according to the 2016 Guidelines on Use of Antiretroviral Drugs for Treating and Preventing HIV Infections in Kenya and with no recent or current illness consistent with acute HIV infection (Ministry of Health & National AIDS and STI Control Program, 2016), have initiated PrEP to date. In addition, DREAMS “Change Agents” lead community HIV- and GBV-prevention initiatives locally, particularly to men, through community sensitization around these issues according to existing participatory learning curricula (e.g., SASA! and SPHERE) (Saul et al., 2018).

The study design included a plan to vet the original intervention model through formative research and youth and service provider engagement. To do this, we conducted a series of activities, including qualitative formative research with AGYW and DREAMS service providers (SP), an intervention design workshop with key stakeholders, and pretesting of intervention materials. Advising all activities was a Youth Advisory Board (YAB) consisting of 14 AGYW selected to represent diverse geographic areas in the county and a range of ages (i.e., 18–24). Members were selected with the input of DREAMS staff to identify those AGYW who would be engaged throughout the intervention development process and who represent a range of experiences with PrEP, GBV, geographic areas, and prior research experience. They were approached by the research team and invited to participate. Seven of the originally identified 21 YAB members were dismissed due to inconsistent attendance.

Formative Research.

Our formative research included in-depth interviews (IDIs) with DREAMS participants, stratified by GBV history and PrEP use, and service providers of a variety of cadres of staff involved in DREAMS implementation. All participants were selected across four DREAMS Safe Spaces currently providing PrEP, which were chosen to represent peri-urban, rural, and fishing communities, and a range of experiences in the uptake of DREAMS services, including PrEP and postviolence care. DREAMS participants were aged 15–24 years old, enrolled in DREAMS, and HIV-negative by self-report, but at high risk for acquiring HIV and thus eligible for PrEP, as defined by the 2016 Guidelines on Use of Antiretroviral Drugs for Treating and Preventing HIV Infections in Kenya (Ministry of Health & National AIDS and STI Control Program, 2016). A combination of DREAMS programmatic data and provider recommendations were used to purposively select DREAMS participants for this phase of research. Using programmatic data, we identified AGYW who did and did not have prior reports of GBV, as well as those who were and were not on PrEP. From this list, we asked clinicians to recommend young women who may offer the most insight about relationship barriers and facilitators to PrEP use. Service providers were at least 18 years old and involved in providing DREAMS services to AGYW attending selected Safe Spaces. The research study was reviewed and approved by the Maseno University Ethics Review Committee in Kisumu, Kenya, and all participants provided informed consent, or parental consent and assent, prior to participation.

All interviews were conducted using a semistructured guide by trained research assistants in private locations, and audio-recorded with the permission of participants. IDIs with AGYW used a vignette storytelling technique to explore gender roles, violence, and use of PrEP and other HIV prevention methods, first through the construction of a story about couples in their community, followed by reflections on their own feelings and experiences. Images and a drawing activity were also used to explore participant perspectives on potential interventions. Research assistants were trained to respond appropriately if participants became distressed discussing issues of violence, including to provide previously vetted referrals to support organizations and counselors. Interviews with service providers covered their perceptions of and experiences working with AGYW, including what AGYW share with providers about their relationships and violence, as well as feedback on proposed intervention components.

Immediately following each interview, the research assistant would complete a debriefing report while listening to the interview recording, which served as a summary of the primary interview content. These reports underwent a rapid analysis process to inform intervention development. They were structurally coded within Dedoose and summarized in analysis memos stratified by participant type (e.g., AGYW with GBV history and no PrEP use, AGYW with GBV history and PrEP use) and topic (e.g., personal experience with DREAMS). Audio files of each IDI were later transcribed and translated into English and thematically coded. The socio-ecological model was used to stratify identified themes by level of influence, particularly as they related to GBV and barriers to PrEP use. Intervention ideas discussed by participants were also mapped according to the identified barriers they could address using an Excel workbook.

Intervention Design Workshop.

A preworkshop was held with a larger group of DREAMS providers (e.g., clinicians, counselors, other technical leadership) to share formative research results, seek alignment with provider experiences, and share the plan for the intervention design workshop. A 2-day intervention design workshop was then held with select DREAMS service providers, of all ages, and the YAB, where formative research outcomes were presented by research staff and discussed, evidence-based intervention models were presented, and intervention models were prototyped and voted on.

The first day of the workshop included numerous breakout activities designed to engage participants’ input on both the interpretation of formative results and the design of the intervention concept and activities. Formative research findings were presented in the framework of a story, in line with the design of the AGYW IDI guides. Breakout groups were then formed for participants to reflect on the resonance of the findings to their lives following a structured activity guide and group presentations, with reports back to the larger group. Prior to the workshop, the research team selected potential intervention models (e.g., couples counseling, male involvement, provider training, screening and counseling) based on an evidence review and designed to align with challenges and barriers identified through the formative research. The goals and structure of each model were presented to the group after the breakout session, along with evidence-based relevant examples, and questions were posed to the attendees to connect the ideas back to the local context and discuss feasibility. Following these presentations, breakout groups were formed for participants to flesh out potential adaptations to models they identified as promising, as well as add new approaches. Following completion of the exercise, each participant voted on the top three ideas that could be feasibly integrated into the existing DREAMS structure.

The second day of the workshop focused exclusively on intervention design using the top three ideas selected during the vote from the first day. Participants were divided into three groups and spent the day building out components of the intervention approaches. The day concluded with presentations from each group, followed by discussion. Following the workshop, meeting notes were reviewed and discussed by the study investigative team, who then began the development of the chosen intervention model and relevant materials, including a manual, handouts, and training materials and slides.

Pretesting Intervention Materials.

The final step in intervention development included pretesting of materials, which were designed by the research team following the workshop. This process included convening the YAB and DREAMS staff separately to review the intervention plan and draft materials, followed by the convening of four sessions with four purposively selected stakeholder groups to review specific intervention component materials: (1) DREAMS Change Agents, (2) male partners of DREAMS participants, (3) couples, and (4) DREAMS participants. All meetings were facilitated using a guide, which outlined the steps and core intervention content to be discussed, as well as key questions to guide refinement of tools and implementation.

RESULTS

FORMATIVE RESEARCH RESULTS

Twenty-four IDIs were conducted with AGYW and 12 with DREAMS service providers. Half of the AGYW interviewed had reported GBV to DREAMS, and 15 were using PrEP. Service providers consisted of Safe Space mentors (n = 4), clinicians (n = 4), and other providers (e.g., counselors, community educators, and technical advisors; n = 4).

Key themes related to barriers to PrEP use that were considered for intervention development are presented according to level of the socio-ecological model.

At the individual level, we found that AGYW felt a sense of limited knowledge, confidence, and power to prevent HIV in their relationships, including an inability to negotiate PrEP use. They expressed a lack of knowledge of or access to PrEP and fear of side effects, as well as a sense of blame and hopelessness that prevented them from changing the individual or relationship factors that put them at risk. These were in part due to norms around male control, which required men to give permission or consent for women’s health-seeking behaviors.

The following excerpt is a discussion of the vignette featuring a married couple called Grace and Harrison:

Interviewer: What aspect of Grace’s relationship makes her to worry of getting HIV now that she is HIV negative?

Respondent: What makes her to worry is… She thinks that because Harrison is positive, he might force her to have sex since she is his wife, and Harrison would not want to use a condom. This makes her have fear of getting HIV.

I: Okay. And … what can she do to protect herself from getting HIV?

R: If Harrison was understanding, Grace could talk to him to agree to use a condom, but I know most men do not agree.

I: Why?

R: They say, “Do you know that you are my wife and not a girlfriend?” They do not agree.” (AGYW, No GBV/no PrEP, age 23)

At the partnership level, men were reported to control relationship and household decisions. One critical challenge cited by AGYW was also limited existing knowledge of PrEP among men and resulting challenges in introducing a new prevention method into relationships. Women feared that their partners would be violent if they asked to use PrEP based on their economic dependence on men and other past experiences attempting to address men’s multiple partnerships. Partners were also described as holding misperceptions about PrEP, believing it to be antiretroviral treatment (ARVs), that they implied mistrust in the relationships or that they meant the female partner was having multiple partners. These misperceptions heightened fears around a violent backlash in response to PrEP, regardless of past history of intimate partner violence (IPV).

On my side, when I tell my partner that I use PrEP and he does not agree, I will use it secretly. Because these drugs are in bottles and to them, they feel they are ARVs. And because of that, he keeps beating me all the time and because of that I am forced to use them secretly in hiding. (AGYW, No GBV/PrEP, age 23)

Increasing men’s understanding of PrEP was therefore seen as critical to simultaneously reduce GBV and increase PrEP use. Out of fear of raising the topic themselves and limited communication confidence, women suggested that providers or male community members engage men to address harmful norms and misinformation. On the other hand, some participants also suggested a need for ways in which AGYW could use PrEP without disclosure to male partners, such as changing the product formulation or packaging or educating them on storing and using PrEP discretely.

So that group is where she can be told that “I use PrEP but my husband doesn’t know and I hide it in such and such a place,” so that she can also take PrEP and keep it the way so-and-so does, so that she can use it without the husband knowing that she uses PrEP. (AGYW, GBV/No PrEP, age 23)

At the service provision level, AGYW expressed a sense that providers blamed them for violence and placed the burden on them to educate their partners about PrEP. Provider accounts of interactions with AGYW occasionally confirmed these reports. Confidentiality, lack of discussion of possible covert PrEP use, and limited provision of other referral resources also informed AGYW’s perceptions of inadequate provider support.

Those who give us education on PrEP should be open to share information and may not wish to ask a question but doesn’t find it easy because there is no privacy and she might be judged negatively. (AGYW, GBV/no PrEP, age 22)

Finally, at the community level, factors such as poverty, unintended pregnancy, school dropout, male partner alcohol use, inequitable gender norms, and stigma around PrEP all limited AGYW’s ability to control their own health and well-being, further reinforcing AGYW’s reports of low self-efficacy and a desire for support to overcome numerous barriers.

I: Okay, now I would like to know ahhh … how their relationships affect their health. Like it is putting them at risk of getting HIV. Do you think these girls that are married or the girls who have boyfriends … is it the kind of the relationships they are in that put them at risk of HIV?

R: Yes … Because you’ll find … maybe, let’s say she is in school, she’s caned on the legs when in school and the parent cannot even give her 20 shillings to board a motorbike. If she gets a motorbike rider who offers her lift [ride] in the evening, they’ll start dating with that person. You see and she will be at risk of getting HIV and AIDs. (SP, DREAMS Mentor)

In addition to barriers, AGYW discussed and suggested numerous intervention approaches. They suggested models such as couples counseling, economic incentives, and a DREAMS program for boys to address limited male knowledge. Table 1 presents these ideas according to the barrier to PrEP use they may address. On the basis of the formative findings as well as evidence from the literature and practical considerations, the research team selected six intervention approaches to carry forward to the intervention development workshop: clinic-based screening and counseling, couples counseling, male involvement approaches, support clubs, vignettes, and changes to PrEP packaging and delivery.

TABLE 1.

Intervention Ideas According to Barriers to PrEP Addressed

	Intervention Components From Proposal						Additional Strategies Identified in IDIs
Barriers to PrEP	Individual counseling	Role-plays	Adherence support clubs	Video vignettes	Enhanced training	Routine GBV screening	Economic incentives	Peer advocacy	Couples counseling	Print media, radio theatre	Engage chiefs community leaders	DREAM Boys	Hotline GBV reporting	Parent education	Couples testing	Education through churches
Individual level
Knowledge re: HIV risk perception, PrEP as prevention tool	X		X	X			X		X	X	X			X
Self-efficacy to use PrEP	X	X	X	X					X	X
Partner level
Partnership power/control dynamics	X								X			X			X
PrEP disclosure	X	X	X						X	X		X	X	X	X
Safety planning	X	X
Restrictions on Safe Space participation & PrEP use by partners									X						X
Men controlling decision-making: shift toward joint decision							X		X						X
Economic control (employment & financial resources)							X
Community level
Stigma			X							X	X
Social isolation			X
Responses to IPV (e.g., chiefs, other family members)											X					X
Need for PrEP sensitization in community								X	X	X	X					X
Clinic level
Inconsistent screening for GBV in HC setting						X
Need to address GBV-specific challenges to PrEP uptake & use	X				X				X		X
Lack of PrEP knowledge among Safe Space mentors					X
Mentor capacity to address multiple types of violence					X
Linkages to trained/sensitized providers						X			X				X
Confidentiality (perceived concerns) in accessing HC/by nurses					X

Note. Italicized thematic barriers/need were identified specifically in in-depth interviews (IDIs) that extended those delineated in the proposal. Shaded boxes were intervention strategies discussed in IDIs.

INTERVENTION DESIGN WORKSHOP OUTCOMES

Thirteen DREAMS service providers participated in the preworkshop and five DREAMS services providers and 14 YAB members participated in the 2-day workshop. The primary outcome of the workshop was the identification of three areas of focus for the intervention from among the six intervention models presented. The selected models included (1) the use of support clubs to create a supportive environment and build unity among AGYW for PrEP use, and male engagement through (2) community sensitization targeting low knowledge and misperceptions of PrEP among men, and (3) couples’ education on PrEP to further garner male support. The use of vignettes to reinforce messages was also maintained as an intervention tool that would be woven into both the support clubs and couples’ education, rather than as a stand-alone intervention component. Given findings related to concerns around provider confidentiality and judgment, a clinic-based model of support was not included in the final selection. In addition, it was determined that the DREAMS program could address provider sensitization outside the context of this intervention, and that any challenges identified in the formative research needed to be addressed with all providers through additional training rather for a subset through intervention activities. A summary of the main considerations for each model that arose during the workshop is presented here.

Support Clubs.

Creation of in-person support groups where all AGYW eligible for and interested in PrEP could meet regularly was identified as important to supporting their PrEP use. In addition, the groups identified various aspects critical to their design, structure, content, and implementation.

Forming the basis of the groups, six main goals for the support clubs were identified: educate (i.e., increase knowledge of PrEP), support each other emotionally, problem solve, create unity among the participants, build confidence, and refer for other needed support such as in the case of IPV. Building on the desire for confidentiality, which led participants away from a clinic-based model we had initially envisaged, participants suggested ground rules designed around the need to protect confidentiality, build trust and support, and encourage participants to share their experiences openly with each other. These included things like listening to one another, being considerate of others’ feelings, and keeping everything confidential, along with an opportunity at the start of the support clubs in each setting to add other relevant rules.

In addition, both the structure and the content of support clubs were discussed. It was determined that a mixture of structured and unstructured content would most effectively allow for engagement, fun, and knowledge building. Participants preferred meetings being held regularly within the existing Safe Space framework and led by mentees rather than exclusively by mentors, with whom some participants occasionally had difficult relationships. However, the added support of clinicians and/or counselors to provide “professional” guidance on certain topics was also desired. Participants identified several key topics of interest, including what PrEP is, myths and misperceptions about PrEP, how to disclose PrEP use to a partner, how to use PrEP without a partner’s knowledge, and how to address violence. Finally, the AGYW felt it was important to allow the groups to be open to girls who were and were not experiencing violence, given the formative findings around the role of fear of violence as a barrier to PrEP use, as well as to those who were and were not actively using PrEP, such that nonusers could learn from current users.

Finally, the use of vignettes to reinforce education provided in the support clubs was felt to be an important intervention mechanism. Attendees workshopped five storylines that could be presented as dramas, given limited technological capacity at sites to show videos. These would address barriers faced by AGYW in their PrEP use depending on their circumstance (e.g., AGYW in polygamous relationships, AGYW who had to engage their partner’s support). In addition, it was proposed that these dramas could be outlined in advance, but the details would be further developed by support club participants as an intervention activity.

Male Engagement Through Community Sensitization and Couples’ Events.

The need and desire to engage men in the community and male partners about HIV prevention came out strongly from the workshop activities. Participants felt that men needed more information about PrEP, but that AGYW did not have enough status in their relationships to provide the information themselves—it needed to come from another man in the community. At the same time, AGYW felt it was important for men to know that AGYW and their partners were receiving the same messages about PrEP to maintain trust. The attendees proposed that this could happen through two primary mechanisms: (1) through sensitization of men in the community, and (2) through events explicitly designed to engage couples connected to the DREAMS program (i.e., DREAMS participants and their male partners). Attendees felt an approach that included values clarification and messaging connecting positive family values to PrEP use across both activities would be successful in engaging men in supporting PrEP. However, they also identified a need to address basic information on PrEP such as side effects, availability, and myths rooted in inequitable gender norms that PrEP promotes promiscuity, represents a lack of relationship trust, and may inhibit fertility, all of which are used to rationalize GBV and stigmatize PrEP use, particularly among women. These messages were meant to complement existing messages around GBV prevention from SASA!

For the community sensitization, participants proposed an iterative approach of engaging with existing community structures, such as local chiefs and group leaders, in order to identify and be allowed access to meetings of men, such as boda-boda (motorbike) driver associations and other groups of men who were deemed both high risk and common partners for AGYW. Once group meetings were identified, Change Agents along with DREAMS clinicians would present key messages on PrEP according to topics outlined above during a short presentation followed by a question-and-answer session. It was envisioned that numerous sessions would be held over a period of 3 months leading up to the couples’ event to ensure that greater numbers of men were reached by messages and that men were adequately primed to attend an event with their female partners.

As noted above, the couples’ event (i.e., “Buddy Day”) was recommended as a mechanism to demonstrate that men and women were receiving the same messages on PrEP, which would increase men’s trust in the messages and consequently PrEP. Different models were discussed; however, the context of a health fair was identified as a feasible and compatible setting to deliver information on this topic. Methods of encouraging attendance by couples were discussed, and suggestions included delivery of formal invitations to male partners of DREAMS participants, providing incentives only to individuals attending as a couple, and offering other services besides PrEP and HIV testing that might entice individuals to attend (e.g., blood pressure measurement). To incorporate “edutainment” into the event, as well participants’ desire for exchange across support clubs from different DREAMS safe spaces, it was proposed that the dramas developed by support clubs could be presented at the Buddy Day as a means of sparking reflection and discussion on PrEP.

Collating these three areas of intervention focus, along with feedback on specific implementation approaches, the research team outlined an intervention involving support clubs and male engagement through community sensitization and a Buddy Day for couples. Figure 1 lists both the originally proposed intervention components along with revised components and the key stakeholder-driven insights that led to these revisions.

FIGURE 1.

Proposed and revised intervention components.

Pretesting Results and Intervention Refinement.

Finally, 25 individuals were engaged in pretesting across four group activities. These included five Change Agents, two male partners of DREAMS girls, three DREAMS girls and their partners together as couples, and 12 DREAMS participants. Change Agents (community leaders conducting sensitization on HIV and GBV) were included at this stage because they were not included in the intervention development workshop and would play a critical role in delivering newly conceptualized male engagement activities. Pretesting with Change Agents included presentation and role-play of community sensitization messages, which they further tested with men in a follow-on session. Couples participated in a mock Buddy Day and review of information, education, and communication materials, and DREAMS participants reviewed the support club session plan and materials. These sessions primarily led to implementation refinements along the lines of where to conduct sensitization events (e.g., Chief meetings), the types of incentives to offer couples to attend Buddy Day, clarification on where to access PrEP, and the integration of the dramas into Buddy Days via cross-site presentations by support clubs. The implementation of the dramas across sites was designed to protect the confidentiality of support club participants. Table 2 provides further detail on the final intervention model, which was designed to be implemented over the course of 6 months.

TABLE 2.

Final Tu’Washindi na PrEP Intervention Model

Activity	AGYW Support Clubs with drama	Community sensitization of men	Buddy Day (i.e., couple’s sensitization)
Goal	Build unity and emotional support among girls, as well as sharing knowledge and self-efficacy around PrEP use, disclosure, adherence, etc.	Build knowledge about and support for PrEP and address myths and misperceptions among men in the community and male partners of DREAMs participants	Provide PrEP sensitization to couples as a unit to facilitate safer discussion and disclosure
Location	Safe Space venues	Community meetings and gathering spaces for men with venues identified during community mapping	Community venue (e.g., churches or community halls)
Leadership	Mentors/peer mentees with DREAMS clinicians and counselors leading some activities	Change Agents	DREAMS clinicians, counselors, and Change Agents
Participants	AGYW eligible for and interested in or currently using PrEP	Men in the community ideally reaching male partners of DREAMS participants	Focus on DREAMS couples through invitation and incentivization, but open to all community members
Format	Participatory sessions with a mixture of structured and unstructured content, including a game and check-in activity to build community; recap of the previous session; a new topic related to PrEP and relationships; drama activities for role modeling; and a closing activity to reflect on lessons learned and build confidence and support	A short (<30 minute) presentation followed by Q&A, delivered once per week over a 10-week period leading up to Buddy Day	Single-session health fair
Content	8-session curricula covering: 1. PrEP information, 2. PrEP disclosure, 3. undisclosed PrEP use, 4. healthy relationships, 5. GBV and PrEP use, 6. unstructured session (topics of participants’ choice), 7. unstructured session, and 8. future goals.	PrEP sensitization including factual information about PrEP (e.g., safety, effectiveness, and eligibility), discussion of common myths and misperceptions, and messages encouraging men to support their partners’ PrEP use (e.g., “PrEP use can reflect a commitment to the relationship or couple. A woman’s decision to use PrEP indicates her commitment to staying HIV negative for her own health and that of her family.”)	PrEP sensitization, following messages used in the community sensitization, as well as the performance of a drama developed by Support Club members, and other health services
Timing	Throughout the 6-month intervention period	Conducted over the first 3 months of the intervention period	In Month 3 of the intervention period

DISCUSSION

The Tu’Washindi na PrEP intervention was designed through a research-based participatory process, which led to meaningful changes from the originally proposed model, and ultimately to an intervention responsive to AGYW needs for PrEP support in the context of gender inequity and GBV. The final intervention addressed challenges to PrEP use faced by AGYW at multiple levels of the socio-ecological model, including stigma at the community level, limited male knowledge and partner opposition to PrEP at the relationship level, and a limited sense of power at the individual level, among others. Incorporating community sensitization targeting men, couples’ events to raise awareness around PrEP, and support clubs for AGYW using and eligible for PrEP, the intervention aimed to build peer, partner, and communal support for AGYW PrEP use in Siaya County, Kenya.

Evidence suggests that participatory approaches can increase the relevance and ultimately acceptability of intervention models. This approach leads to interventions that are more contextually grounded in a community’s needs and offers an opportunity to build community and youth capacity to address them (Oliver et al., 2006; Wallerstein & Duran, 2008). Youth are less often included in participatory research (Jacquez, Vaughn, & Wagner, 2013) and at least two recent reviews of behavioral interventions for HIV prevention have cited this as a reason for limited success (Gibbs, 2016; Mannell et al., 2019). In the Tu’Washindi intervention development process, youth and other stakeholders identified needed revisions that would ground Tu’Washindi in their specific needs and contexts. These revisions ranged from large changes in design such as the incorporation of male sensitization at the community level, as well as smaller changes such as the translation of video vignettes into culturally relevant dramas structured around several common relationship settings and challenges for young women. Both repeated engagements, from study inception through to results discussion and dissemination, and methodologies that allow for creative engagement such as storytelling should be considered in future participatory intervention development work.

The multilevel model of Tu’Washindi fills a gap in the literature and, to our knowledge, it is the first AGYW PrEP support intervention that aims to specifically address the relationships between GBV and PrEP use. Partner opposition, low relationship power, and GBV are increasingly recognized as inhibiting the uptake and use of novel HIV prevention methods such as PrEP (Cabral et al., 2018; Lanham et al., 2014; Montgomery et al., 2015, 2019; Palanee-Phillips et al., 2018; S. T. Roberts et al., 2016, 2020; Rousseau-Jemwa et al., 2018), and our research identified both fear of violence and the actual experience of violence to be barriers to AGYW PrEP uptake. By adjusting the intervention model to offer support clubs to all AGYW regardless of previous experience of GBV, addressing harmful norms among men that connect PrEP with promiscuity and mistrust, and offering skill-building to AGYW to use PrEP safely either through nondisclosure or safe disclosure to partners, Tu’Washindi aimed to minimize both concerns about and actual experience of violence as a result of PrEP. It may also serve as a model for other PrEP delivery programs, particularly given growing recognition of the need to address GBV associated with women’s PrEP use as evidenced by new PEPFAR requirements for routine GBV screening in PrEP delivery programs (PEPFAR, 2020).

There are several limitations worth noting about this work. These include the potential selection bias in our formative research and workshop participants, as well as the small numbers of men engaged in intervention pretesting, which may have limited the representativeness of the input we received. For example, women at the highest risk for HIV may not be engaged in DREAMS and therefore not selected for the formative research or YAB participation. Even among the YAB it is possible that the removal of members due to inadequate participation may have limited the diversity of perspectives received on effective strategies to engage and retain AGYW in the intervention. Likewise, the small number of men engaged in pretesting messages may have hindered our understanding of effective messages to address men’s concerns about PrEP. Despite these limitations, the multiple phases of stakeholder engagement and iterative and participatory development process lend strength to the intervention design, which is being tested further.

Our evaluation of the final Tu’Washindi na PrEP intervention in Siaya county, Kenya, through an RCT pilot was completed in late 2019 and will provide critical data on the feasibility and acceptability of the intervention among participants and implementers. This evidence, along with preliminary evidence for the intervention’s impact on GBV reduction and PrEP uptake and adherence, will aid in determining the effectiveness of our participatory approach to intervention design as well as offer a potential model for other settings where AGYW PrEP use is hindered by gender inequity and GBV. Results of the pilot study are anticipated in mid-2021.

CONCLUSION

Involving stakeholders, particularly youth, in designing health-related interventions through an evidence-based process can result in meaningful contributions to design outcomes. The Tu’Washindi na PrEP intervention was contextually grounded and responsive to youth’s needs as a result of several phases of youth and health care provider engagement in the design process, from their contributions to formative research interpretation to pretesting of the final model and messages. The multilevel intervention model addressing men, couples, and AGYW was designed to respond to barriers to PrEP use resulting in gender inequity and GBV, issues found to be critical to the spread of HIV and limited use of PrEP.