Adapting an evidence-based gender, livelihoods, and HIV prevention intervention with street-connected young people in Eldoret, Kenya

Abstract

Despite being highly vulnerable to acquiring HIV, no effective evidence-based interventions (EBI) exist for street-connected young people (SCY) in low- and middle-income countries (LMICs). Therefore, this paper describes the research process of adapting an existing EBI in Eldoret, Kenya using a modified ADAPT-ITT model with a young key population. From May to August 2018 we adapted the combined Stepping Stones and Creating Futures interventions. We used community-based participatory methods, focus group discussions, and working groups with four Peer Facilitators and 24 SCY aged 16 to 24 years. At the inception of this project, a matched-savings program was integrated into the intervention to further address structural drivers of HIV. Numerous adaptations came forth through the participatory process. Engaging SCY in the adaptation process ensured the program was responsive to their needs, relevant to the street context, and respected their right to participate in the research process.

Introduction

The number of new HIV infections in sub-Saharan Africa continues to decline. Nevertheless, young people living on the continent still account for a large proportion of new cases of HIV, with young women aged 15-24 disproportionately acquiring HIV (UNAIDS, 2017). Consequently, continued efforts to prevent HIV acquisition among young key populations are essential using combined biomedical, behavioural and structural interventions (Pettifor et al., 2015). Street-connected young people (SCY), for whom the streets play a central role in their lives and identities (UNHCR, 2017), can be classified as a young key population given their substance use, precarious housing, experiences of violence, and engagement in selling sex (Woan, Lin, & Auerswald, 2013).

Driven by structural and social inequities and their need to survive on the street, SCY in sub-Saharan Africa are highly vulnerable to acquiring HIV. In sub-Saharan Africa, SCY participate in transactional sex (Winston et al., 2015; Woan et al., 2013), experience sexual and gender-based violence (Kudrati, Plummer, & Yousif, 2008; Wachira et al., 2015), have multiple concurrent partners (Anarfi, 1997; Embleton et al., 2015), engage in sex under the influence of drugs and alcohol (Embleton, Ayuku, Atwoli, Vreeman, & Braitstein, 2012), and report inconsistent condom use (Kayembe et al., 2008; Winston et al., 2015).

Research in Kenya suggests that SCY are contracting sexually transmitted infections (Embleton, Wachira, et al., 2016; Kaime-Atterhög, Lindmark, Persson, & Ahlberg, 2007; Winston et al., 2015), and may have an HIV prevalence that exceeds that of other young people in the country, with street-connected young women disproportionately acquiring HIV (Braitstein et al., 2019; Goldblatt et al., 2015; Shah et al., 2018; Winston et al., 2015). This heightened prevalence may be due to engaging in transactional sex, gender inequities, and experiencing sexual and gender-based violence in the street subculture (Embleton, Wachira, et al., 2016; Embleton et al., 2018; Sorber et al., 2014; Wachira et al., 2015; Winston et al., 2015). Street-connected young women engage in transactional sex with young men connected to the streets for survival, and also with individuals outside of the street subculture for economic benefit (Embleton, Wachira, et al., 2016; Embleton et al., 2015, 2018). Likewise, street-connected young men engage in transactional sex with both street-connected and non street-connected individuals (Embleton et al., 2015). However, street-connected young men’s reported engagement in transactional sex is very limited in comparison to street-connected young women’s (Sorber et al., 2014; Winston et al., 2015). SCY in Kenya are economically marginalized typically earning less than 100 Kenyan Shillings (Ksh) (~$1.27 CAD) per day (Sorber et al., 2014). Street-connected young men engage in a range of income generating activities and earn more per day in comparison to street-connected young women whom primarily rely on begging and transactional sex for survival (Embleton et al., 2015, 2018; Sorber et al., 2014). Importantly, SCY’s primary sexual partners and relationships are with each other. Sexual relationships are engaged in for economic benefit, desire, procreation, and for informal ‘marriages’ in the street subculture (Embleton et al., 2015, 2018; Wachira et al., 2016). Moreover, the street subculture propagates harmful social and gender norms, which promote substantial gender inequities and is the source of much of the sexual and gender-based violence experienced by street-connected young women (Embleton et al., 2018; Sorber et al., 2014; Wachira et al., 2015). Therefore, including both street-connected young women and men in interventions to reduce gender inequities and sexual risk practices while improving livelihoods, is critical to addressing SCY’s HIV vulnerability in this context. Changing deeply engrained social and gender norms, such as those present in the street subculture (Embleton et al., 2018) is challenging, however using strengths-based approaches from within a community by building on a community’s strengths and examining harmful practices can catalyse change (CUSP, 2017).

To date, there is a dearth of evidence on effective interventions to reduce sexual risk practices among SCY (Naranbhai, Abdool Karim, & Meyer-Weitz, 2011), and very few interventions have been tested and evaluated with this highly marginalized population in low- and middle-income countries (LMICs) (Berckmans, Velasco, Tapia, & Loots, 2012; Coren et al., 2016; Dybicz, 2005). As no specific HIV prevention interventions have been rigorously tested and evaluated for SCY in LMICs (Berckmans et al., 2012; Coren et al., 2016; Naranbhai et al., 2011), adapting an existing evidence-based intervention (EBI) with and for SCY may be a viable approach to fill this gap. A number of systematic reviews have been conducted to identify and assess effective HIV prevention interventions for young people in sub-Saharan Africa (Maticka-Tyndale & Brouillard-Coylea, 2006; Michielsen et al., 2010; Napierala Mavedzenge, Luecke, & Ross, 2014), many of which may be suitable, acceptable, and effective for SCY. Many of these interventions use combination approaches to address structural and social drivers of HIV acquisition. Given the structural and social inequities impacting SCY’s vulnerability to HIV acquisition, combination and multifaceted interventions are particularly suitable for this population. To our knowledge, no one has adapted an existing EBI with SCY in sub-Saharan Africa. With the lack of effective interventions for this population in LMICs, determining the feasibility of and providing a model for adapting existing interventions is an important step in increasing the number of effective interventions for this marginalized population.

Adaptation is defined as the process of modifying or altering an EBI for a new context to reduce mismatches, without conflicting with or negating its core elements, thereby maintaining intervention fidelity (Card, Solomon, & Cunningham, 2011; Gordon, Welbourn, & Trust., 2017; Wingood & DiClemente, 2008). Adaptations may involve modifying program content or the method of program delivery (Castro, Barrera, & Martinez, 2004), and should take into account the local context to ensure content and delivery are culturally appropriate. The term context broadly encompasses multiple features (e.g. geographical, political, social, or cultural) that may interact with the intervention to produce variation in implementation processes or outcomes (Craig et al., 2018). Few models and theoretical frameworks exist for adapting interventions; however, the eight-step ADAPT-ITT model (Wingood & DiClemente, 2008) has been used successfully for adapting HIV interventions with adolescents (Latham et al., 2010).

Given the paucity of interventions that have been rigorously adapted, piloted, and evaluated in LMICs with and for SCY, this paper describes the participatory research process of adapting the evidence-based combined Stepping Stones and Creating Future interventions (Alvarado et al., 2017; Jewkes et al., 2014; Jewkes, Nduna, & Jama, 2010; Misselhorn, Jama Shai, Mushinga, Washington, & Mbathe, 2013) with SCY in a new setting in Eldoret, Kenya using a modified ADAPT-ITT model. We describe the use of a rights-based participatory adaptation process focusing on Steps 3 (Administration) through to Step 7 (Integration) of the ADAPT-ITT model.

Materials and Methods

Study Design:

From May 2017 to January 2018 a two-phase mixed methods study was used to adapt, pilot, and evaluate a gender, livelihoods, and HIV prevention intervention for SCY in Eldoret, Kenya. In the first phase, we adapted the combined Stepping Stones and Creating Futures interventions using a modified ADAPT-ITT model based on the following 7-step process: 1. Assessment, 2. Decision, 3. Administration, 4. Production, 5. Topical Experts, 6. Integration, 7. Testing. This modified model omitted ‘Training’ as Step 7 and integrated training into Step 3 (Administration). Step 1 (Assessment) and Step 2 (Decision) occurred between 2013-2016 as part of other studies.

From 2013 to 2015 our research team identified needs in relation to HIV, gender equality, and livelihoods among SCY in Eldoret, Kenya (Step 1. Assessment) (Embleton, Wachira, et al., 2016; Embleton et al., 2015; Shah et al., 2018; Sorber et al., 2014; Wachira et al., 2015). Subsequently, we conducted a scoping review using Arksey and O’Malley’s five-stage framework (Arksey & O’Malley, 2005) to identify HIV interventions for high-risk youth. The combined Stepping Stones (South Africa 3^rd Edition) and Creating Futures (Jewkes et al., 2014) program was identified as an intervention our research team sought to adapt, pilot, and evaluate to determine if it would be suitable, feasible, acceptable, and potentially effective in our setting (Step 2). Stepping Stones and Creating Futures were selected because they address identified social and structural drivers of SCY’s HIV acquisition and identified needs (Step 1).

From May 2017 to August 2017 we completed step 3 (Administration) through 6 (Integration) to adapt the intervention. Throughout this process, we used community-based participatory methods with four Peer Facilitators and 24 street-connected young people aged 16 to 24 years, to adapt the intervention for the context through a series of community meetings, focus group discussions (FGDs), and small working groups. In our study, context referred to the social, cultural, economic and political circumstances in which street-connected young people live and work and the financial circumstances of the research project.

Intervention Description:

Stepping Stones is a behavioural intervention that was developed in Uganda (Gordon et al., 2017; Welbourn, 1995). The original curriculum consists of 13 participatory sessions that aim to improve sexual health and promote greater gender equity in relationships among men and women. Topics covered over the 13 three-hour sessions include: reflecting on love, sexual joys, body mapping, menstruation, contraception and conception, sexual problems, unwanted pregnancy, HIV, STIs, safe sex, gender-based violence, motivations for sexual behaviour, dealing with grief and loss, and communication skills (Welbourn, 1995). The training program has since been adapted for use in different settings and has been used with children, adolescents, and adults (Gordon et al., 2017; Holden et al., 2018). The Stepping Stones program is designed to be participatory and uses critical reflection, role-playing scenarios, and drama, which draw on the everyday reality of participants’ lives (Gordon et al., 2017; Welbourn, 1995), and therefore works towards shifting social norms from within a community through critical reflection and building on strengths (CUSP, 2017). The program has been tested in multiple settings and shown to reduce transactional sex, intimate partner violence, and multiple partnerships, increase HIV knowledge and condom use, and improve gender equity; however, no study has proven its effectiveness in reducing HIV incidence (Alvarado et al., 2017; Paine et al., 2010; Skevington, Sovetkina, & Gillison, 2013). Stepping Stones has also been effective at diffusing knowledge beyond participants into the community (Paine et al., 2010). Potential for diffusion of knowledge in our setting is particularly important given the substantial gender inequities and sexual and gender-based violence in the street subculture (Embleton et al., 2015; Wachira et al., 2015).

Creating Futures is a structural intervention designed to build on Stepping Stones, based on sustainable livelihoods theory (Jewkes et al., 2014; Misselhorn, Mushinga, Jama Shai, & Washington, 2014). The livelihood-strengthening curriculum consists of 11 peer facilitated single gender sessions in groups of 20, focusing on five types of capital: financial, natural, human, physical, and social. The sessions cover: securing and keeping jobs, budgeting, saving, debt, social resources for livelihood, coping with crises, incoming generating activities, setting goals and building basic business principles. The goal of the program is to empower young people to find pathways out of poverty and vulnerability (Jewkes et al., 2014; Misselhorn et al., 2014). In South Africa, the combined Stepping Stones and Creating Futures intervention was delivered over 12 weeks with out-of-school youth living in informal settlements, in twice weekly 3-hour single-sex sessions, by trained facilitators of a similar age (Jewkes et al., 2014). Participants in the combined program saw a significant increase in their earnings, had improved gender attitudes, and women (aged 18 to 30) experienced a significant reduction in intimate partner violence (Jewkes et al., 2014).

Prior to adaptation we obtained the South African program materials for Stepping Stones (Jewkes et al., 2010) and Creating Futures (Misselhorn et al., 2013) and integrated a group-led matched-savings program conditional on attendance to further address structural drivers of HIV and gender inequity in our context. It was proposed that participants would form savings groups at the outset of the intervention to promote participation and retention in the program. As a group they would decide whether they would contribute between 50 to 100 Ksh per person (~ $0.63-$1.27 CAD) on a weekly basis to their savings. If all group members were in attendance at the weekly intervention sessions, the group’s savings would be matched. In order to maintain fidelity of the intervention, we identified the core elements of the program that resulted in its effectiveness and that should remain in our adapted EBI. For Stepping Stones these included program principles including Freirian Critical Pedagogy (Jewkes et al., 2006), the sequential structure, dosage and duration (Gordon et al., 2017) and for Creating Futures, sustainable livelihoods theory (Misselhorn et al., 2014).

Study Setting:

This study occurred in Eldoret, Kenya, located in Uasin Gishu (UG) County. In 2010, UG County had approximately 894,179 individuals from 202,291 households, of whom 41.5% were aged 14 years or less. Approximately 51.3% of the UG County population lives below the Kenyan poverty line. Eldoret town has a population of 289,389. It is home to Moi University (MU), Moi Teaching and Referral Hospital (MTRH), and the Academic Model Providing Access to Healthcare (AMPATH) program, which is a partnership between MTRH, MU, and a consortium of universities from North America including the University of Toronto. AMPATH began as an HIV care and treatment program, and with support from PEPFAR currently has over 80,000 HIV infected patients in care across western Kenya. Recently AMPATH and MTRH established the Rafiki Centre for Excellence in Adolescent Health in Eldoret, which became the study site for this project.

Ethics approval and consent to participate:

This study received ethics approval from the University of Toronto Research Ethics Board and Moi University / Moi Teaching and Referral Hospital Institutional Research and Ethics Committee. We received a waiver of parental consent for minors and participants provided written consent or assent (or a fingerprint for those unable to write) for their participation in this research project with a specially trained social worker. The study received approval from the UG County Children’s Coordinator to occur. Participants did not receive compensation for their participation in the adaptation activities. Tea and chapati were provided to participants in focus group discussions and working groups.

Study Participants:

SCY were eligible to participate in the adaptation process if they were: 1) aged 16-24 years, 2) had spent a portion or majority of their time on the streets for the past 6 months, and 3) were not enrolled in or attending school.

Recruitment and Enrolment:

As part of the adaptation process, we conducted a series of community meetings with the street community in different barracks (primary locations in which SCY congregate in the town) and six other residential locations around Eldoret to discuss the proposed intervention and seek their input. In each location, SCY nominated a representative to engage in FGDs and other participatory activities. It was explained that their elected representative would communicate ideas and concerns with the study team regarding the proposed program. This system of nominating a representative from each location ensured that a diverse group of SCY across the city were involved in the intervention adaptation process. Those that were eligible and indicated their willingness to participate were invited to the adolescent-friendly clinic at AMPATH for enrolment and to undergo the assent or consent process with a specially trained social worker. In total we recruited 24 SCY in age and sex stratified groups (young men aged 16-19 n=6 and 20-24 n=6, young women aged 16-19 n=6, and 20-24 n=6). The median age of participants was 19.5 years (IQR: 17-23). The majority (79%) of participants had some level of primary education, 17% had some secondary, and 4 % had no formal education.

Data Collection:

Data to inform adaptation was collected through field notes and FGDs. The principal investigator (LE) maintained notes throughout the adaptation process. We held four age and sex stratified FGDs in a private room at the Rafiki Centre for Excellence in Adolescent Health. FGDs took an average of 60 minutes, were audio-recorded, and conducted by Peer Facilitators of the same gender in Swahili and Sheng (a dialect used on the streets). Peer Facilitators gave participants a presentation about the intervention. Participants were then asked a series of questions regarding what components of the intervention were acceptable and appropriate, those that were not, what they liked and did not like, what they would add to the intervention or take away, what they would like to change, suitable days and times of the week for attendance, program location, and how they would like the matched-savings groups to function. Participants were also given the opportunity to provide any additional feedback or ideas. Participants from FGDs sessions were invited to further adapt the intervention based on their initial feedback in working groups.

Data Analysis:

Audio-recorded data collected through FGDs was transcribed into Swahili and translated into English. Transcribed and translated data was imported into NVIVO software for analysis. Data collected during FGDs was analysed using thematic analysis (Braun & Clarke, 2006), to identify patterns and key concepts in relation to the appropriateness, acceptability, and other themes that emerged in relation to the proposed HIV prevention program components.

Results

The modified ADAPT-ITT model framework that guided the community-based participatory adaptation process with SCY can be found in Appendix I. It outlines the original ADAPT-ITT model methodologies and our modified approach with examples. Figure 1 shows the timeline, participants, activities, and outcomes used in the modified ADAPT-IT process with SCY.

Figure 1:

Timeline, participants, activities, and outcomes used in the modified ADAPT-IT process with SCY

Administration:

In May 2017, to adapt and facilitate the proposed program, we hired four young people (two young women and two young men aged 19-23 years) who had been or currently were connected to the streets to become Peer Facilitators. The PI trained Peer Facilitators over the course of 4 weeks. The Peer Facilitators were the starting point for community-based participatory research approaches to adaptation, given their lived experience, extensive knowledge of the local street context, and on-going connection to the street community.

As part of the Stepping Stones curriculum, it is recommended that when training facilitators they first experience the full program. We started as a team by experiencing each session of Stepping Stones and Creating Futures, over the course of two weeks. Through experiencing the program session by session the Peer Facilitators provided ideas and suggested both program content and delivery adaptations. All ideas and adaptations were documented by writing notes directly on the original manual materials and in the research team journal. During this initial process the following key issues and proposed adaptations arose including: low literacy levels and identifying a suitable dialogic method for intervention delivery; ensuring services, laws and products matched the political context in Kenya; the integration of drug and alcohol content; creating stories, names, role plays and imagery that reflected the social, cultural, and economic context on the streets in Kenya; modifying the Creating Futures curriculum to be relevant to the livelihood needs of SCY; and creating one comprehensive curriculum with a new title.

One major program delivery concern was adapting the curriculum extensively for low literacy levels among SCY. The program was adapted to be primarily dialogical. We omitted written exercises, homework, and journals. We integrated an Indigenous Talking Circle approach for program delivery to foster dialogue, understanding, and active listening. Typically, an object of importance is held by the person speaking, and is circulated clockwise around the Circle, person-by-person. In our setting we used a beaded ‘Maasai Stick’ or ‘Rungu’, which signifies respect and leadership in Kenyan culture. This methodology suited the Freirean Critical Pedagogy theoretical underpinnings of the Stepping Stones and Creating Futures (Gibbs, Jewkes, Sikweyiya, & Willan, 2014; Jewkes et al., 2006).

We had to make substantial alterations to the Creating Futures curriculum in relation to program content. This included: creating two new character stories that related to SCY and their social, economic, cultural, and political contexts on the streets in Kenya; omitting journals and homework sections that required writing; creating a resume; attending job interviews; and content related to scholarships and post-secondary education. We altered the curriculum to be dialogic using the same Talking Circle approach and to focus on livelihood opportunities and income generating activities. The majority of SCY in our setting have limited education, and are not in the position to apply for formal employment or scholarships to further their education. Given the need for survival on the streets, the program was altered to encourage participants to work towards an income generating activity goal in conjunction with the matched-savings groups that were structured to commence at the outset of the Stepping Stones program.

Finally, the Peer Facilitators suggested combining the program into one comprehensive manual with a new title for the program that utilized Sheng and terms that SCY would identify with. The program was renamed Stepping Stones ya Mshefa & Kujijenga Kimaisha (Stepping Stones for Street Youth & Build Your Life Up). Mshefa is a Swahili slang word that means a hustler (one who works hard to survive). It is a label used by SCY themselves as an identity. Kujijenga Kimaisha translates to ‘build yourself and your life up’, and represents the Creating Futures program.

As part of their training, Peer Facilitators hosted 4 mock facilitation sessions with guest participants who included Peer Navigators (Shah et al., 2018), youth from the community and healthcare providers. During these mock facilitation sessions, we tested our Talking Circle approach and it was well received among guest participants. Following the mock facilitation sessions, the Peer Facilitators suggested that a mixed language manual would be most useful and make the content more relevant to the streets. We therefore translated titles, discussion questions, and other key text into Swahili/Sheng, as Sheng is primarily used on the streets.

Production:

Drawing on documentation from our training and mock facilitation sessions we produced a first draft of our adapted EBI while maintaining program fidelity and core components. As each session was produced in draft format, the Peer Facilitator team reviewed the adapted sessions for content that reflected the streets and translated titles, discussion questions, and scenarios into a language that SCY use and comprehend by using a mix of English, Sheng, and Swahili. In an iterative process, this was then integrated into the manual as we produced a first draft of our adapted program.

Topical Experts & Integration:

To engage topical experts, we took a rights-based participatory approach, where SCY in our setting became our topical experts over the course of two months (UNHCR, 2017). We conducted a series of community meetings in the different barracks in town and in six locations around Eldoret. From these meetings we elicited preliminary ideas and responses about the program. Generally, the response was positive, but SCY expressed many concerns regarding the matched-savings program; particularly in regards to trust with their finances and disbelief that funds would be matched. In community meetings, SCY voiced concerns that they would not be able to travel to the adolescent-friendly clinic two times per week without support for transportation. As a team we decided it was feasible to propose providing 40 Ksh to participants each time they attended a program session based on the cost of public transport and within the financial context of the research.

Following community meetings, we hosted four FGDs with 24 participants. The Peer Facilitators created a visual and detailed presentation about the program that they presented to participants (Figure 2). Key themes from the FGDs included: program acceptability, match-savings program alterations, location of program, days and times of the week to attend, using their existing ‘talents/skills’, and a desire to focus on livelihoods to change their circumstances.

Figure 2:

Visual presentation used in FGDs to discuss the proposed program and components

Program acceptability:

In general, participants thought the program was acceptable for SCY and did not have any components they wanted to add or take away. The program delivery twice per week and concept of matched-savings were well received amongst participants as one young woman demonstrates: ‘I liked where we come for those sessions twice per week, then we have our savings doubled. Double, double save.’ (Female, 16-19)

Despite positive support and belief that the program could have a positive impact on SCY, participants expressed some scepticism regarding the implementation of the program as conveyed by one young man:

‘In my opinion this program is fine. And if it is true what is written there, you know you have written, but it hasn’t been done yet. Therefore if it is true...because as I see it lots of boys will come. And not only that, but this could potentially end homelessness on the street. If we will succeed, but if it is just propaganda and talking...’ (Male, 20-24)

In general, feedback regarding the program content of both Stepping Stones and Creating Futures was positive, as one male participant states:

‘Let me start. Tubonge (Let’s Communicate), there I don’t see any problem. We are supposed to look deeper and direct each other. We need to know both sexes. 100%. That is perfect. We need to know about sex and love. That is perfect too.’ (Male, 16-19)

However, one group of female participants 16-19 years debated the importance of also including health and drug and alcohol use content. Overall, all groups came to a consensus that the program delivery and content was suitable and would be accepted by SCY in our setting. One exception to this was with the structure of the matched-savings program.

Matched-Savings Program Issues:

Many issues regarding the proposed matched-savings program came forth in community meetings and focus groups. SCY expressed a desire to contribute a larger sum of money on a weekly basis as stated by one male aged 16-19: ‘On Friday I will bring 100 bob. I cannot give 25. That is little.’ Participants also questioned the facilitators regarding the frequency of contributions: ‘I would like to ask this. Why not let us give twice a week?’ (Male, 20-24). Participants explained that twice-weekly contributions would give them a safe place to keep their money over the course of the week, as affirmed by one young man:

‘Let me ask you something. There is somewhere you haven’t gotten me. From what I have understood you have said we should give on Friday, but if I am willing, parking gives me money and instead of spending it, I’d rather keep it somewhere. We can be giving it to you on Monday and Friday.’ (Male, 16-19)

All participants discussed that they could not rely on their peers to attend the sessions and that if their savings were matched on group attendance this could create conflict among participants. SCY expressed a desire to have their savings matched based on individual attendance. Participants also discussed distrust about leaving their money with the investigator as one young female aged 20-24 stated: ‘What if she runs away to her country?’.

In response to these issues raised, we changed the mechanism of matching from conditional on full group attendance to individual attendance to decrease the chances of conflict. Second, we increased the amount and frequency participants could contribute to the matched-savings program. We decided that it was feasible for participants to bring money to each session as long as over the course of the week it did not exceed 200 Ksh. Based on SCY’s reported daily earnings, we concluded this would not induce increased sexual risk practices (Sorber et al., 2014). Lastly, to address issues of trust, we explained why and how the study would be keeping their savings safe, and referred participants to the study social worker to reassure them.

Program Location and Times:

Given the stigmatization SCY face, issues regarding where the program would take place arose, as well as days and times of the week that each group would attend. Participants expressed different opinions about where the program should occur, but all were in agreement that they wanted a dedicated private space to attend the weekly sessions. A young woman aged 20-24 suggested: ‘Even put up a tent. That will be fine. Don’t just call us to the clinic. You see we don’t all come?’. As a team we agreed to set up a dedicated tent outside of the adolescent-friendly clinic that would be used solely for the program.

We reviewed the days and times of the week put forth by each age and sex stratified group in the FGDs and came up with a proposed schedule that aimed to meet each group’s interest. Young men in both age strata came to the consensus to have their sessions in the morning hours to ensure they would leave the program with enough time to generate income in the afternoons, as stated by one young man aged 20-24: ‘That is good. At least one leaves here and goes to look for money.’ Whereas, young women said they would prefer afternoon sessions, as many have domestic chores and childcare they complete in the mornings as stated by young women aged 20-24: ‘If you had a child, what time would you reach here? You can’t.’

Talents & Livelihoods:

Finally, in response to a desire to focus on livelihoods and use existing talents and skills, we reassured groups that the Kujijenga Kimaisha program was dedicated to topics that would assist young people in establishing income generating activities, setting goals, and savings. In addition, the team decided that we could invite SCY who were talented artists to adapt the imagery in the original program documents for the streets version. We hosted an illustration competition for the program manual covers. Prizes were given for first, second, and third place, and the winning illustrations were featured as the manual covers (Figure 3A and 3B). Subsequently, we invited all illustrators back to illustrate the complete manual. SCY were compensated 50 Ksh for each illustration they created for the manual. This resulted in creating a comprehensive manual with imagery they strongly identified with. This process was particularly effective as they had ownership over the manual, were credited as illustrators, and they felt proud of their talents being used.

Figure 3A and 3B:

Manual cover illustrations by SCY

Finally, we hosted four working groups to discuss changes to the program based on all of the feedback provided by SCY. Participants responded that they were very satisfied with how we listened to their ideas, altered the matched-savings program, and harnessed the skills of their peers to illustrate the manual. They were content with the proposed location of the program, provision of transport support, and the program schedule. The representatives in the working groups said they had no further issues to address with the proposed program. We then finalized a second draft of the adapted manual (Appendix II shows the final table of contents).

Testing:

From September 2017 to January 2018, we piloted and evaluated the adapted program with 80 SCY. We randomly selected 80 participants from a list of eligible SCY who indicated their interest in participation into age and sex stratified groups (16-19 years and 20-24 years), to attend biweekly sessions over 14 weeks. We measured changes in short-term outcomes using a mixed-methods pre- and post-test study design. Outcomes measured included: HIV-knowledge, gender equitable attitudes, condom use self-efficacy, sexual practices and economic status. Results are expected later in 2019.

Discussion

This paper describes the application of a model for adapting EBIs for SCY in LMICs. Importantly, this paper demonstrates the feasibility of adapting EBIs for this young key population. Taking a rights-based participatory approach to adaptation and using an established adaptation framework ensured that we maintained program fidelity while being responsive to the key population’s concerns and their right to participate fully in the research process. Our process ensured the program was suitable and relevant to the local social, cultural, and economic contexts of the streets, resulting in program acceptability in our setting. Our results highlight the acceptability of the combined Stepping Stones and Creating Futures programs, and its potential for use with SCY in other LMICs. Furthermore, our findings demonstrate the acceptability and suitability of the innovative matched-savings program to give SCY an avenue for savings when formal banking systems are inaccessible.

Poverty is a driving factor for young people finding themselves in street situations in LMICs (Embleton, Lee, Gunn, Ayuku, & Braitstein, 2016), and the need to earn money once on the streets results in a significant number of young women in our setting engaging in transactional and survival sex (Embleton, Wachira, et al., 2016; Embleton et al., 2015; Winston et al., 2015). The focus on savings and building livelihoods was an important program component for SCY in our setting. Interventions for SCY focused on livelihoods are an approach that have not received significant attention in the academic literature, and as Berckmans’ et al (2012) suggests, a sustainable livelihoods approach may be a viable and important avenue for interventions with SCY (Berckmans et al., 2012). Moreover, this approach aligns with what Dybicz (2005) terms secondary prevention, aimed at ensuring SCY can safely transition into adulthood and increase their ability to secure income on the street (Dybicz, 2005). Our adaptation results suggest that combining a livelihoods approach with other programmatic goals, such as HIV prevention, is feasible, suitable, and highly acceptable with this population in our context.

Given the complexity of SCY’s circumstances in LMICs, no one intervention will meet all of their needs, but involving them in the process of identifying health priorities, in intervention development, implementation, and evaluation processes may increase the effectiveness and sustainability of any program (Gordon et al., 2017; UNHCR, 2017). Service provision in the form of drop-in centres, healthcare facilities, shelters, and child welfare and protection, are critical in upholding SCY’s rights (UNHCR, 2017), and improving their health and well-being. As Coren (2013) found in high-income settings, both time limited therapeutic programs and standard services such as shelters and drop-in centres resulted in favourable changes in outcomes for participants (Coren et al., 2016). Yet, EBIs for specific health issues, such as HIV prevention, are vital given their vulnerability to acquiring HIV (Naranbhai et al., 2011). Overall, the body of literature reviewing interventions for SCY points to insignificant evidence and a need for more research on effective interventions (Berckmans et al., 2012; Coren et al., 2016; Dybicz, 2005; Naranbhai et al., 2011). In lieu of creating new interventions to increase the number of robust and potentially effective interventions for SCY, our adaptation process demonstrates the feasibility and suitability of adapting an existing EBI. Others seeking to pilot or implement programs for this population can use our adaptation process as a model to adapt existing EBIs they can then test in other settings.

This study is with limitations. Our adaptation process occurred in one geographic location in western Kenya where the research team has a long-standing relationship with the study population, and therefore it is likely this impacted the feasibility of the adaptation process. However, in alignment with ethical research practices and a rights-based approach with SCY, researchers or organizations seeking to use this model can most likely do so effectively if they build trusting relationships with the street community prior to implementation and use community-based participatory methods.

Our rigorous adaptation process using community-based participatory methods and a modified ADAPT-ITT model demonstrates that it is feasible to adapt existing EBIs for SCY in LMICs. Engaging SCY throughout adaptation of the intervention ensured the adapted curriculum was responsive to their needs, relevant to their circumstances, and the local social, cultural, and economic context on the streets. The adaptation methods respected their right to participate in the research process and led to a high level of program acceptability in our setting. This adaptation model may be an avenue to start address the gap in designing and identifying effective interventions for SCY in LMICs.