Pre-exposure prophylaxis for HIV prevention (PrEP) among young women at high risk for HIV in South Africa: longitudinal patterns of use and strategies to improve persistence
Embargo until
Date
2022-06-21
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Background. Adolescent girls and young women (AGYW) and female sex workers (FSW) make up a disproportionate share of HIV infections in South Africa. Oral pre-exposure prophylaxis for HIV prevention (PrEP) can prevent new infections, but low levels of persistence limit its real-world impact. We aimed to characterize patterns of PrEP use, including initiation, discontinuation, and cycling, and to identify implementation strategies that may improve PrEP persistence.
Methods. We worked with a South African non-profit, TB HIV Care, to create a cohort of AGYW and FSW eligible for PrEP using routinely collected data. First, we described persistence on PrEP among FSW in eThekwini using a discrete time-to-event survival analysis. Discontinuation was defined as a composite outcome based on 1) two months of not returning for PrEP, 2) client discontinuation, or 3) provider discontinuation. Second, we used discrete time-to-event survival analyses and group-based trajectory modeling to describe the longitudinal patterns of PrEP initiation, discontinuation, and re-initiation across both AGYW and FSW. Eligibility for PrEP initiation was defined as testing negative for HIV. Re-initiation was defined as a new PrEP initiation within 12 months of an initial PrEP prescription, following a 3-month gap in PrEP. Third, we used an interrupted time-series design to evaluate the impact of PrEP delivery strategies, including clinical mentoring for providers, SMS PrEP refill reminders and support texts, case management, and a loyalty rewards program, on 1-month PrEP persistence.
Results. Persistence at one month for AGYW (0.38, 95%CI [0.37-0.38]) and FSW (0.41, 95%CI [0.40-0.42]) was low. PrEP cycling was common, with close to half of FSW (22% “Early Cycling”; 21% “Ongoing Cycling”) and one third of AGYW (34% “Ongoing Cycling”) experiencing some form of cycling during the year following initiation. SMS support and refill reminders and provider training had a positive impact on 1-month persistence among FSW.
Conclusions. PrEP-delivery programs have focused a large portion of available resources on promoting uptake; however, low PrEP persistence and inadequate understanding of cycling will undermine PrEP as a prevention tool. SMS support, refill reminders, and provider training show promise for improving immediate PrEP persistence, but strategies to augment these are needed for sustained PrEP use.
Description
Keywords
HIV, PrEP, Female sex workers, Young women, AGYW, Young women, South Africa, Survival analysis, Interrupted time series, Quasi-experimental design