Description of study characteristics.

Successful implementation of oral pre-exposure prophylaxis (PrEP) requires retention in PrEP care among those with an increased likelihood of HIV. This study aimed to estimate the proportion of retention on PrEP and extent to which variability in PrEP retention is associated with population-, program-, and study-specific characteristics among men who have sex with men (MSM) and transgender persons (TGP). We performed a systematic review and meta-analysis examining PrEP studies and conference abstracts retrieved from the PubMed and Ovid online databases, capturing demonstration projects or observational studies published from January 1, 2010 to March 24, 2021. We included 84 studies (totaling 90 study “cohorts” analyzed) that reported on the retention of oral PrEP for HIV and included predominantly MSM and TGP. The proportion of retained participants was obtained from each study and used to estimate the cumulative probability of being retained on PrEP and the loss of PrEP retention rate over time via a random-effects meta-analysis survival model. We examined sources of heterogeneity by including study-level covariates in this model. The pooled cumulative probability of PrEP retention was 77.0%, 64.7%, 48.5%, and 24.1% at 6, 12, 24, and 60 months, respectively. Loss of PrEP retention rates were significantly (p