Description

In Aim 1, the investigators will conduct a cluster RCT in Kisumu, Kenya to determine the effect of nurse-navigators on PrEP initiation, persistence, and adherence among AGYW seeking contraception within a pharmacy-based PrEP delivery model. The study will randomize 20 retail pharmacies offering PrEP (10 pharmacies per randomization arm) and will enroll 1900 AGYW seeking contraception. All participants will be enrolled following purchase of a contraceptive method, offered PrEP (daily oral PrEP or the DPV-VR) and followed for 10 months. The study will quantify and compare PrEP initiation, persistence, and adherence at the pharmacy-level between randomization arms, in addition to several secondary and exploratory outcomes.

Pharmacy selection for cluster RCT (Aim 1): The investigators conducted a landscape analysis of all pharmacies in Kisumu, Kenya as part of the team's ongoing pharmacy-based PrEP delivery studies. In this analysis, the investigators gathered information on the types of customers who attend pharmacies, the types of services offered (e.g., HIV testing) and products purchased (e.g., contraceptive methods) at pharmacies, and availability of a separate private consultation room. The investigators will select 20 pharmacies based on whether they are willing to provide PrEP per national guidelines, have a separate consultation room that can be used for HIV testing and PrEP counseling, and provide the full range of contraceptives to AGYW clients (including EC, OCP, injectables, implants, and condoms).

Randomization (Aim 1): All the pharmacies will receive standard pharmacy-based PrEP delivery and be randomized to either receive (n=10 pharmacies) or not receive (n=10 pharmacies) a nurse-navigator to assist with PrEP delivery. To ensure balance between randomization arms in terms of key site characteristics, sites will be categorized on AGYW volume (i.e., monthly number of purchases made by AGYW clients) and distribution frequency of contraceptive methods purchased (based on pharmacy log data) and restricted randomization will be used for site (cluster) allocation to intervention and control arms. Specifically, all possible randomizations that evenly distribute sites on these specified factors (AGYW client volume and distribution of contraceptive methods) into 2 study arms will be generated, and one combination will be selected using a random number generator. Randomization and allocation will be performed by a biostatistician from the UW CFAR Biometrics Core, who has no knowledge of sites other than the variables included in the restricted randomization process, and the Study Statistician, who received an MPH in Biostatistics from UW, will supervise the randomization process. Once assigned, the randomization allocation will be unblinded. Since the randomization will occur at the pharmacy level, it is impossible to blind study team members or participants to the randomization assignments. However, procedures to minimize the influence of the unblinded nature of this study on outcomes will be implemented. Ongoing data monitoring will not include information about study endpoints disaggregated by site or study arm. Only the study statistician will review data on study endpoints by study arm or facility.