Persistence on Oral HIV Pre-Exposure Prophylaxis Among Key Populations at a Centre in Buenos Aires: A Survival Analysis of a Real-World Cohort.

INTRODUCTION

Although HIV pre-exposure prophylaxis (PrEP) has been rapidly scaled up in Argentina to reach key populations with high HIV burden, evidence on real-world persistence is lacking. Understanding who remains engaged in care, and why others discontinue, is essential to ensure the effectiveness of PrEP. This study evaluated oral PrEP persistence and identified demographic and behavioural factors associated with discontinuation in a cohort of individuals initiating PrEP in Buenos Aires.

METHODS

We performed a retrospective cohort study including all users initiating daily or event-driven oral PrEP at our centre between September 2021 and July 2025. The cohort consisted of transgender women (TGW), men who have sex with men (MSM) and cisgender women engaged in sex work (cSW), in line with national PrEP guidelines. Persistence on PrEP was defined as the time to first discontinuation (TFD), either self-reported or estimated as the date when the last dispensed PrEP pill would have been exhausted if taken daily. Categories for discontinuation included participant-reported causes and loss to follow-up (LTFU). TFD was evaluated using survival analysis. Cox proportional hazards models were fitted to examine associations between TFD and sociodemographic and behavioural factors.

RESULTS

Overall, 970 individuals initiated PrEP: 69% MSM, 24% TGW and 6% cSW. Median age was 31 years (IQR 26-36). During follow-up, 447 participants (46%) discontinued PrEP. Persistence declined steadily over time: 91% at 1 month, 81% at 6 months, 72% at 12 months, 58% at 24 months and 43% at 36 months. Main categories for discontinuation were LTFU (44%), user decision or perceived low risk (46% combined) and adverse events (10%). In the multivariable model, younger age, identifying as TGW or cSW, cocaine use in the last month before baseline and the absence of basal/follow-up syphilis or other bacterial sexually transmitted infections were associated with PrEP discontinuation.

CONCLUSIONS

Nearly half of the cohort discontinued PrEP over 3 years, with differences across key populations. These findings highlight the need for differentiated community-tailored PrEP delivery models that address structural and behavioural barriers to improve long-term PrEP engagement among key populations in Argentina and wider Latin America.