Background Sexually transmitted disease clinics are well-suited to deliver PEP; few provide full courses (28 days) due to cost. Since 2016, NYC Sexual Health Clinic (SHC) staff have provided full-course PEP to eligible HIV-negative patients, and thereafter attempt linkage to PrEP. We examined HIV incidence and patterns of PrEP/PEP use among a PEP cohort.

Methods We matched men-who-have-sex-with-men (MSM) who received PEP at SHC (09/2016-05/2017) to the citywide HIV surveillance registry to identify new HIV diagnoses between last PEP event during this interval and 06/30/2018. We calculated HIV incidence with time-at-risk starting 28 days after PEP provision; periods of additional PEP supplied at SHC during follow-up were excluded from time-at-risk. We examined subsequent PrEP and repeat PEP use during follow-up. For MSM without PrEP/repeat PEP, we calculated the number-needed-to-treat (NNT) with PrEP for one year to prevent 1 HIV infection (assumed 73% efficacy with 90% adherence).

Results Eleven HIV diagnoses occurred among 520 MSM with 652 person-years (PY) follow-up; HIV incidence was 1.7/100PY (95%CI: 0.8–3.0), and highest among Hispanic MSM (2.6/100PY) and MSM aged<30 years (2.2/100PY). Median time to HIV diagnosis was 178 days (range 41–410). During follow-up, 18% (94/520) received PEP again (range 1-4 times), with 58% of PEP-repeaters (55/94) also given PrEP at some point. Of 460 MSM with SHC visits during follow-up, 202 were linked to or self-reported PrEP use, 200 were not linked to PrEP, and 58 declined navigation for PrEP. Of the 202 who linked to/self-reported PrEP use, 31 (15%) had subsequent PEP event(s) at SHC. Eight of 279 MSM without evidence of PEP/PrEP during follow-up were diagnosed with HIV (incidence=2.3/100PY); NNT=61.

Conclusion Despite PrEP availability, fewer than half of PEP patients took up PrEP and a substantial proportion took PEP repeatedly; research to elucidate underlying reasons for PrEP and PEP use patterns (insurance/cost, convenience, self-perceived risk) is warranted.

Disclosure No significant relationships.