A taxonomy of pragmatic measures of HIV pre-exposure prophylaxis (PrEP) Use: application in a PrEP-using cohort in Chicago, IL.

OBJECTIVES As delivery of PrEP becomes an HIV prevention priority in the U.S., standard, pragmatic measures of PrEP use are needed to compare and evaluate prevention implementation programs. By using readily available electronic health record data (EHR), we describe and compare measures of persistence and retention. DESIGN Retrospective cohort METHODS:: Using EHR prescription data for patients at a large urban Federally Qualified Health Center from 2015 to 2019, we calculated measures of persistence and retention and compared them to pharmacy claims data, PrEP biomarkers, and HIV outcomes. RESULTS Total PrEP time was 19.8 months on average. During this period, average adherence by medication prescription ratio (MRxR) was 89%; 77% of patients had an MRxR ≥85% and 90% have an MRxR ≥57%. Over the first six months, average proportion of day covered (PDC) ≥85% was 53% and PDC ≥57% was 57%. Prescription fill rates, based on claims data from a pharmacy partner, ranged from 45% to 60%. Using tenofovir-diphosphate as the gold standard, PDC had high sensitivity (97%) but low specificity (≤13%). As a measure of retention, over the first six months, 59% of patients had quarterly HIV tests. CONCLUSION Total PrEP time is useful measure of overall persistence, while PDC can assess persistence and adherence at a specific time point. Adherence by PDC is more conservative compared to MRxR; both will overestimate true adherence. Retention in care can be measured by quarterly HIV tests. Using consistent terminology and reporting timepoints and adherence thresholds will help reporting and comparing PrEP delivery programs.