Reaching Adolescent Girls and Young Women Through Girl-Friendly Drug Vendors

Building on our team's formative work using human-centered design methods and behavioral economics theories, we have designed an intervention (the "Queen Club") to develop AGYW-friendly drug shops where we will introduce HIVST alongside contraceptives and linkages to care. In a 4-month randomized pilot study, we will compare mediating outcomes (i.e., AGYW visits, distribution of HIVST and contraception, referral to SRH services) between 10 intervention and 10 comparison ADDOs, using data from a time-location survey of customers and records for HIVST and contraception distribution and referrals for sexual and reproductive health (SRH) services.

ADDOs will be randomly selected from four wards in the Shinyanga municipality. We will use stratified randomization by ward to assign 20 ADDOs 1:1 into the two study groups. Regardless of intervention arm, we will conduct the following activities in all ADDOs: in-person training on HIVST provision, monthly HIVST supply with half of kits explicitly earmarked for free provision to AGYW customers, an HIV care referral plan, and shop record tracking. The shop records will collect data on AGYW patronage, HIVST distribution, contraception distribution, and health facility referrals for SRH services. Monthly reviews of inventory records by the local team will include review of these records to determine AGYW patronage (for any service) and the volume and types of contraception and HIVST kits distributed to AGYW.

ADDOs in the intervention arm will be trained to implement the "Queen Club" intervention. The Queen Club intervention was developed through a year-long human-centered design process with AGYW and other stakeholders and is informed by behavioral economic theory and game-design. It is intended to partially mitigate to the physical, economic, and social constraints faced by AGYW in Shinyanga, as revealed in the formative research, and motivate them to discreetly seek HIVST and contraception at local drug shops, while in a context of fun.

In brief, the Queen Club is a loyalty program through which AGYW can earn surprise mystery gifts through repeated purchases at participating ADDOs. This is intended to build loyalty and trust with an ADDO shop, and also build the customer base for the ADDO owner. The back of the Queen Club loyalty card includes symbols, selected by AGYW, that represent oral contraception, HIVST, emergency contraception, condoms, and pregnancy tests. With the symbol portion of the card, AGYW can discreetly point to their desired product and receive it for free without questioning.

In a one-day training, we will review contraceptive methods and counseling techniques with intervention ADDO owners and staff using training materials adapted from the Ministry of Health, Community Development, Gender, Elderly, and Children. At every intervention ADDO, we will provide a countertop display with contraceptive methods and an HIVST kit for demonstration (if desired by any customer), as well as a tablet computer pre-loaded with videos about various contraceptive products and the Oraquick HIVST kit instructional video in Kiswahili.

The primary outcomes are mediators of the intervention's potential effect on preventing HIV and unintended pregnancy. In this phase, our goal is to measure these mediating, "process" outcomes and maximize learning for a future effectiveness and sustainability study using ADDOs as a platform to deliver HIVST and other HIV prevention and SRH services. We will collect data on the following mediating outcomes from the 10 intervention and 10 comparison ADDOs:

• AGYW patronage: number and proportion of visits by AGYW.
• Contraceptive distribution: number and types of contraceptives (i.e., condoms, oral contraception, emergency contraception) distributed to AGYW.
• Health facility referrals: number of referrals for AGYW made for SRH services, including HIV testing or treatment.
• HIVST kit uptake: number and proportion of kits distributed to AGYW.

We will assess AGYW patronage with a time-location survey of ADDO customers (randomly selected 3-hour intervals) at baseline (during the month prior to intervention) and during the final two months of the intervention (12 episodes per ADDO total) in which trained researchers will discreetly document the number of customers, their sex, and approximate age. Monthly reviews of inventory and referral records will show the volume and types of SRH products distributed to customers, including HIVST and contraception distributed to AGYW, and the number of SRH referrals made, the number and proportion for AGYW, and for which services.

We will compare AGYW patronage, referrals to health facilities, and contraception transactions between intervention and comparison ADDOs over the 4-month pilot period using: (1) statistical tests of means (t-tests) and medians (Wilcoxon rank-sum test); (2) a Poisson regression model (for counts of AGYW visits) to estimate rate ratios and confidence intervals controlling for baseline levels of AGYW patronage using a "difference-in-differences" approach; and (3) chi-squared tests for independence to qualitatively assess differences between groups for all mediating outcomes. Results will be used to understand the potential effectiveness of AGYW-friendly ADDO environments on AGYW visits and demand generation for HIV prevention and SRH services, a basis for power calculations to determine the sample size needed for the future study of effectiveness and sustainability.