Dynamics of Pre-Exposure (PrEP) Eligibility Because of Waxing and Waning of HIV Risk in Rakai, Uganda

Abstract

Background: We conducted a retrospective population-based study to describe longitudinal patterns of prevalence, incidence, discontinuation, resumption, and durability of substantial HIV risk behaviors (SHR) for pre-exposure prophylaxis (PrEP) eligibility. Methods: The study was conducted among HIV-negative study participants aged 15–49 years who participated in survey rounds of the Rakai Community Cohort Study between August 2011 and June 2018. Substantial HIV risk was defined based on the Uganda national PrEP eligibility as reporting sexual intercourse with >1 partner of unknown HIV status, nonmarital sex without a condom, having genital ulcers, or having transactional sex. Resumption of SHR meant resuming of SHR after stopping SHR, whereas persistence of SHR meant SHR on >1 consecutive visit. We used generalized estimation equations with log-binomial regression models and robust variance to estimate survey-specific prevalence ratios; Generalized estimation equations with modified Poisson regression models and robust variance to estimate incidence ratios for incidence, discontinuation, and resumption of PrEP eligibility. Findings: Incidence of PrEP eligibility increased from 11.4/100 person-years (pys) in the first intersurvey period to 13.9/100 pys (adjusted incidence rate ratios = 1.28; 95%CI = 1.10–1.30) and declined to 12.6/100 pys (adjusted incidence rate ratios = 1.06; 95%CI = 0.98–1.15) in the second and third intersurvey periods, respectively. Discontinuation rates of SHR for PrEP eligibility were stable (ranging 34.9/100 pys–37.3/100 pys; P = 0.207), whereas resumption reduced from 25.0/100 pys to 14.5/100 pys (P < 0.001). PrEP eligibility episodes lasted a median time of 20 months (IQR = 10–51). Interpretation: Pre-exposure prophylaxis use should be tailored to the dynamic nature of PrEP eligibility. Preventive-effective adherence should be adopted for assessment of attrition in PrEP programs.