Implementation of Anyaka Makwiri: A Multicomponent Mentoring Program for Adolescent Girls and Young Women in Uganda

This article describes the development and implementation of the Anyaka Makwiri program and summarizes results from the qualitative assessment of participant experiences. Anyaka Makwiri is a multicomponent mentoring program developed for adolescent girls and young women (AGYW) ages 15 to 24 in Gulu, Uganda. The comprehensive program consisted of a curriculum covering sexual and reproductive health (SRH), financial capabilities, soft skills, and gender-based violence and gender equality; activities designed to improve participants’ social connectedness; optional onsite testing for sexually transmitted infections (STIs), HIV, and pregnancy along with STI treatment; group-based savings; and links to SRH services, including contraceptive and gender-based violence services. The program was implemented over a 6-month period and reached 490 AGYW. Findings are derived from routine program-monitoring data including administrative records, de-identified service statistics, and baseline surveys. In addition, this article summarizes some of the key findings from qualitative interviews with both mentors and AGYW participants, conducted at the conclusion of the program.  Participants generally had a favorable view of the mentoring program, particularly in terms of the curriculum topics they were exposed to, and mentors were also positive about their experiences. Despite the program’s many successes there were some implementation challenges, the most prominent being intermittent participant attendance due to a variety of difficulties. The lessons learned from the implementation of Anyaka Makwiri provide valuable insights for the design and implementation of multicomponent mentoring programs for AGYW.


Background
In Northern Uganda, myriad social and ecological factors compound the vulnerability to negative health and social outcomes for adolescent girls and young women (AGYW) ages 10 to 24 (Amone-P'Olak & Ovuga, 2017; Karamagi et al., 2018;Malamba et al., 2016;Schlecht et al., 2013). These factors include recovery from war-time disruption to the formal economy, health, and education systems; poverty; sexual violence; early and forced marriage; and orphan status.
Estimates of the HIV prevalence among women ages 15 to 49 living in this region are as high as 10% (Karamagi et al., 2018;Malamba et al., 2016). AGYW ages 15 to 19 living in Northern Uganda are less likely to use contraception than their counterparts in every other region of the country, and approximately 24% have already begun childbearing (Kabagenyi et al., 2016;Uganda Bureau of Statistics & ICF, 2017).
Countering the multifaceted vulnerability of AGYW requires a holistic approach that seeks to build protective factors needed to circumvent poor health and social outcomes. Establishing safe spaces for AGYW is a recommended protection and empowerment strategy in crisis settings, and group-based mentoring programs, delivered in a safe and supportive environment, are an evidence-based approach for building AGYW's protective assets (Plourde et al., 2017;United Nations Population Fund, 2016). The application of a positive youth development (PYD) framework in youth programming has demonstrated positive impacts on health and social outcomes in high-income countries and is a promising approach for programs in low-and middle-income countries. PYD seeks to address youth health and development from a holistic asset-building approach by intervening to build individual skills, assets, agency, and competency as well as the enabling environment (Catalano et al., 2019). It is within this context that YouthPower Action developed the Anyaka Makwiri (Smart Girl) program for AGYW ages 15 to 24 in Gulu, Uganda. A description of how program components align with key features of PYD is provided later in this manuscript.
To develop this intervention, YouthPower Action applied a three-phase process consisting of (a) a systematic review of peer and grey literature, (b) AGYW engagement workshops and, (c) a scan and assessment of evidence-based curricula. The systematic literature review (Phase 1) sought to identify mentoring interventions that demonstrated effectiveness in improving sexual and reproductive health (SRH), reducing HIV risk and infection, and building AGYW's protective assets. We define protective assets as the "skills, resources, and social and economic capital" AGYW need to reach their full potential (Plourde et al, 2017, p. 132). Group-based interventions, while less common in the literature, showed more impact for reducing AGYW's sexual and reproductive health risks than one-on-one programs. Group-based mentoring programs demonstrated promise in improving reproductive health outcomes, academic achievement, financial behavior, and social networks, as well as reducing experiences of violence (Plourde et al., 2017). The results of the review showed that the most successful mentoring programs incorporated additional components beyond just mentoring. Successful programs also included curriculum-based education and access to safe, social spaces outside of the home where participants could develop and strengthen their peer network (Plourde et al., 2017). Detailed results of the literature review, including an analysis of group-based versus one-on-one mentoring models are described in Plourde et al. (2017).
Using principles of and tools for girl-centered design and PYD, authors conducted AGYW engagement workshops (Phase 2) to better understand the most prominent issues faced by AGYW, the topics they were interested in learning about, perceptions related to safety in the community, and recommendations for mentor selection and program development (Austrian & Ghati, 2010;Coalition for Adolescent Girls, 2015;Girl Effect, 2013). Participants shared multiple challenges faced by AGYW in their community including persistent gender inequality, domestic violence, sexual violence, early marriage, early pregnancy, lack of work, and school drop-out.
Workshop participants suggested that the program should provide group discussions and information about relationships, gender-based violence and gender equality, SRH, communication, and financial capabilities.
Rather than dedicate resources to developing new educational materials when so many highquality resources exist, our team assessed evidenced-based curricula that covered soft skills, SRH, and financial literacy to identify content that met information needs (Phase 3). Materials were assessed to ensure that they were substantive, accessible, appropriate, evaluated, and designed for our target age range. Selected materials also aligned with UNESCO's International Technical Guidance on Sexuality Education, evidence-based recommendations for girls' financial education, and YouthPower Action's recommended key soft skills for cross-sectoral youth outcomes (Gates et al., 2016;Morcos & Sebstad, 2011;UNESCO, 2009).
Informed by our findings from the intervention development process, YouthPower Action developed a multicomponent group-based mentoring intervention integrating key features of PYD programs. These include skill-building activities, youth engagement and contribution, promotion of healthy relationships and bonding, belonging and membership, safe spaces, and access to age-appropriate and youth-friendly services (see Figure 1; Hinson et al., 2016). The mentor intervention consists of a curriculum covering SRH, financial capabilities, soft skills, and gender-based violence and gender equality; activities designed to improve participants' social connectedness; optional onsite testing for STIs, HIV, and pregnancy along with STI treatment; group-based savings; and links to SRH services, including contraceptive and gender-based violence services. Each mentor session begins with dedicated unstructured time, called a "sisterhood circle," for mentors to check in with participants about what is going on in their lives, any challenges they may be facing, or exciting news they may want to share. Icebreakers and games are also integrated into each session to provide time for participants to engage with each other, strengthen friendships, and have fun. Finally, each session ends with a journal prompt to encourage participants to reflect on the content provided that week. The savings component was developed to encourage participants to set and achieve personal savings goals.
Group chairs, secretaries, and treasurers were assigned to account for all money coming in and going out via group and personal ledgers. Participants could not access loans but could access their own savings at any time if needed-even if they did not yet reach their personal savings goal. The intervention was delivered to groups of 20 to 25 AGYW weekly for 26 weeks, 2 to 3 hours per weekly session. To ensure a small mentor to mentee ratio, each AGYW group was led by four mentors.

Figure 1. Program Alignment With Key Features of Positive Youth Development Programs
The program was implemented in close collaboration with the District of Gulu. Before implementation, district entry meetings were conducted with the district health officer and district health educator, and with subcounty community development officers, local council committee members, village health teams, cultural and religious leaders, and elders and opinion leaders. AGYW who participated in the intervention were recruited from participants with whom our local partner, The AIDS Support Organization (TASO), had previously worked through the

Methods
We conducted surveys at baseline and endline to test the feasibility and efficacy of the intervention. The results of the quantitative (survey) analysis are reported elsewhere. In addition, de-identified group-level data were collected by TASO service providers who, at two points in time, administered HIV, STI, and pregnancy tests; offered STI treatment; and provided referrals. The first round of testing was conducted at the start of the intervention, and the second was conducted after the 6-month intervention period. Partner program records were analyzed to assess uptake among participants of contraceptive methods and screenings for human papillomavirus (HPV), pelvic inflammatory disease, and cervical cancer. Implementation tracking tools were developed to assess short-term intervention outcomes such as number of sessions held by each mentor group, number of participants per session, and amount of money saved through savings group activities. Technical progress reports were developed by TASO quarterly.
Finally, qualitative interviews with both mentors and AGYW participants were conducted at the conclusion of the program. Interviewers requested mentors' and participants' feedback and perspectives about the mentoring program (i.e., which aspects of the mentoring intervention they liked, any program challenges they identified). Data collection was carried out by a local research firm contracted for the study. Data collectors trained in qualitative techniques conducted face-to-face, in-depth interviews with 13 mentors and 13 AGYW participants at endline. One AGYW participant and one mentor were randomly selected from each of the 13 groups in the mentoring study arm. All interviews were conducted in a private location that was convenient to the participant, and were conducted either in English or Acholi, depending on the participant's preference. Participants gave either informed assent (for those ages 15 to 17) or consent (for those ages 18 to 24) at baseline and provided contact information so they could be reached again to coordinate the endline surveys and in-depth interviews. Participants were reconsented before completing endline, in-depth interviews. We received a waiver of parental permission for minors ages 15 to 17 because not all parents/guardians were aware that their daughters participated in the DREAMS program. All audio recordings of in-depth interviews with AGYW participants and mentors were transcribed and translated from Acholi into English.
Transcripts were analyzed using qualitative thematic analysis.

Results
Between April and December 2017, the Anyaka Makwiri program trained 88 mentors and reached a total of 490 AGYW organized into 22 groups. Intervention participants were on average 20 years old, Catholic, and had at least a primary school education; approximately 64% lived in urban areas. Most participants reported previous sexual intercourse (90-91%), had at least one child, and were married (64.3%). Mentor recruitment and selection criteria stipulated that mentors must always be female and near in age (18-35) to promote relatability and comfort discussing sensitive topics as well as to protect participants' safety. Mentors were also required to reside in the same communities as the AGYW participants, have some level of high school education, be available both for the training and for 2 hours a week over a 6-month period, and be committed to improving AGYW's access to the skills and resources they need to empower themselves.
Over the 6-month intervention period, the Anyaka Makwiri program provided more than 1,000 elective pregnancy, HIV, and STI tests. In collaboration with a local partner, the program provided participants with approximately 200 elective HPV, pelvic inflammatory disease, and cervical cancer screenings as well as access to contraceptive methods. Participants who tested positive for STIs received on-site treatment, those who tested positive for HIV and/or pregnancy were referred to our implementing partner's clinic for prenatal and/or comprehensive HIV care, including access to prevention of mother-to-child transmission programs when appropriate. Through the program's savings group component, participants saved a total of 9.2 million UGX (around 25,000 USD). Some AGYW organized income-generating activities on their own, such as farming, rearing livestock, and selling food and/or drinks, to increase savings.
Many savings groups continued to meet beyond the project period. In addition, participants identified, led, and engaged in community service projects such as water source clean-up, fetching water for a health facility, erecting a hand cleaning tap (tippy tap), helping elderly neighbors, and performing a skit on HIV prevention.
Qualitative interviews with both mentors and AGYW participants sought to explore their experiences with the Anyaka Makwiri program. Participants generally had a favorable view of the mentoring program, particularly in terms of the curriculum topics they were exposed to. For example, the most frequently mentioned topic participants found helpful was HIV/AIDS and STI prevention. Other topics considered helpful were savings, family planning, and gender-based We also asked AGYW about some of the challenges they faced in participating in the mentoring program. Participants often mentioned difficulties in attending because of illness, either themselves or children and relatives, and having household chores. In a few instances, other barriers to participation cited were rain/weather, interference with school hours, and husband or partner opposition.
Overall, mentors were positive about their experiences with the Anyaka Makwiri mentoring program. To further understand mentors' perspectives, we asked them about perceived program successes. The most important program success, according to mentors, was that AGYW learned the importance of savings and were saving money. For example, two mentors specifically mentioned that some AGYW had started their own small businesses because of savings activities. In addition, other frequently mentioned perceived successes were that participants knew how to protect themselves against HIV and unplanned pregnancies and that AGYW knew their HIV status. A few mentors also mentioned that participants practiced better hygiene, were better communicators, and were more confident. As one 18-year-old mentor described: Some girls come and tell me that being in the group is changing their lives.
Because they say sometimes, they pass through frustrations but what we are teaching them gives them the emotional intelligence to bear with hard situations.
And they also say that at least now they are able to save money and now they know that they have to plan for their future, they have to invest in something.
In parallel to the question posed to participants about challenges, we asked mentors what they perceived as challenges in delivering the mentoring program. The two main challenges that most mentors cited were the difficulty in translating the English language curriculum into Acholi, the local language, as well as the irregular participant attendance. A 22-year-old mentor explained: What was challenging about facilitation is about the training manual, when we were trained, these things were passed over briefly and yet there are a number of things found in the manual that you need to teach. And you know most of the things are in English and for you and the group to understand, you need to consult another person so that you can understand and translate it in Acholi.
That is what I found challenging and also, we were required to teach yet our learners had no writing materials, you know it's not easy for learners to just listen without writing and they find it difficult to grasp everything by head.

Implementation Challenges
Despite the many program successes, there were implementation challenges. Approximately 1 month into the intervention's implementation, a stock-out of STI test kits occurred, interrupting our ability to provide this service to participants for several weeks. According to the 2018 country progress report submitted to UNAIDS by the government of Uganda, test kit stock-outs at district level facilities and storage centers are frequent despite national progress in procurement procedures and supply chain management (DREAMS, 2019). Once supply chain issues were resolved, and regular testing was resumed, the team noticed lower uptake for second-round testing among participants than for the first round. In response, messaging about the importance of retesting was provided and the need for comprehensive HIV education was reaffirmed.
Flooding and government construction led to the displacement of participants and interrupted the weekly meeting schedule of some groups. Program participants with immediate shelter and financial needs because of displacement were referred to local organizations providing emergency services. Once displaced groups reassembled, many began to meet biweekly to account for lost time. However, after they had caught up to the intervention schedule, some continued to meet multiple times per week resulting in a shorter overall duration of the intervention. Other groups decided to split each session into two weekly sessions-with one focused on curriculum content and the other focused solely on savings. While the groups' interest in meeting multiple times per week demonstrated the participants' enthusiasm for the intervention, it did lead to exposure differences that may have affected evaluation results (to be published in a forthcoming manuscript).
While recruitment criteria for mentors included their availability for the initial training and for 2 hours a week over the 6-month intervention period, a small number (eight) did resign before program completion. Reasons for mentors' early resignation included relocation, job opportunities, marriage and divorce, and meeting-time conflicts. Because the program was designed for implementation by multiple mentors per group, programming was not disrupted when mentors resigned. Mentors from groups that were in close geographical proximity to groups that lost a mentor helped to fill these vacancies as they occurred. Due to the migrant nature of our target beneficiary population, and seasonal changes (such as rainy season, farming season, and school schedules), overall group fluidity was an added complexity. The implementation of strong monitoring tools helped program managers to keep track of participants, but poor record-keeping by some mentors made this impractical in some cases.

Discussion
Mentors are the backbone of the Anyaka Makwiri program. Mentors are not only responsible for convening weekly meetings and delivering session content; they also act as role models, foster a safe space for learning and building healthy friendships, promote AGYW's self-esteem, help AGYW deal with life's challenges and provide advice, advocate for AGYW's rights, and help AGYW identify available resources and services to support their healthy transition into adulthood. Careful mentor selection and intensive training are important activities that can affect a program's success. Despite the intensive 5-day training provided to mentors, qualitative data suggest that mentors wanted even more training. This suggests a need for longer training periods and emphasizes the importance of ongoing, supportive supervision. The appointment of multiple mentors per group not only allows for a deeper connection between participants and mentors because of the small mentor-to-mentee ratio, but also helps to ensure that programming is not disrupted in the case of mentor resignation. Anyaka Makwiri mentors were provided with a stipend-providing this, or another incentive such as career training, helps to legitimize the mentor role, recognizes the contribution of mentor volunteers, and may help with retention. Training substitute or standby mentors may be another strategy to quickly replace any vacancies that open during the program period. Careful group monitoring can help to ensure fidelity to program design and track mentor and mentee attendance.
It is important to consider seasonal changes and their impact on participant attendance when identifying the most appropriate timeline for implementation, though given the length of the program there is likely to be some seasonal impact on attendance in many country contexts.
Close follow-up with mentees who have missed several sessions allows mentors to refer participants to appropriate community services if needed. It is important for mentors to understand that the reason to follow up with mentees who do not attend sessions is to ensure that their needs are met and to provide support, not to force program attendance. Strong local partnerships are critical to establishing an effective referral network. In the case of Anyaka Makwiri, partnerships with local partners allowed us to broaden the range of available SRH services for participants, provide support for emergency housing needs, and respond to participants' experience of gender-based violence.
When identifying meeting locations, the team not only considered the safety of the location itself, but also the surrounding areas and the path to the session's location. Because many of the communities we worked within were slum communities, some of the meeting places were temporary structures and were impacted by construction and flooding. Finding permanent, safe, locations nearby where participants live and can safely travel may not always be feasible.
However, community leaders can help to identify and ensure safe spaces and advocate for the use of more permanent structures such as churches or schools for meetings.
The Anyaka Makwiri program was novel in that it incorporated a concerted focus on training mentors not only to implement the curriculum, but also on approaches to create a supportive environment for the development of AGYW. In addition, the program integrated on-site access to health services. The integration of health service delivery through mentor groups is not common among other similar mentoring programs (Plourde et al., 2017). To date, few comprehensive tools to support the implementation of group-based mentoring programs are available for open access. However, literature on the successful scale-up of adolescent and youth SRH interventions notes that the availability of implementation tools and guidance documents often facilitates the transition of program activities to new user organizations (Uganda AIDS Commission, 2018). Thus, to support the scale-up of this approach through other large or community-based programs, the authors developed an implementation toolkit.

Conclusion
Addressing the multifaceted vulnerability of AGYW requires a holistic approach that empowers AGYW and includes strategies that influence AGYW's future reproductive health and educational, financial, and social outcomes. Given the complex nature of multicomponent programs, there is much to consider in ensuring the successful implementation of such an approach. The Anyaka Makwiri experience provides valuable insights for the design and implementation of multicomponent mentoring programs.