Background: Risk compensation with decreasing condom use was reported during open-label extension of the ANRS-Ipergay trial, but no risk disinhibition appeared with a constant number of sexual partners or intercourses over time. We investigated sexual behavior changes during the ANRS-Prevenir study follow-up.
Methods: ANRS-Prevenir is an ongoing PrEP study launched in May-2017 in the Ile-de-France region. Quarterly online self-questionnaires collected participants'' sexual behavior. Analysis used data until October-2018. Sexual behavior was assessed by the median number of sexual partners (previous 3 months); PrEP and condom use (systematic or not, previous 3 months); and condomless sex at most recent anal intercourse (CMRAI). Generalized estimation equation models were estimated distinguishing by PrEP use scheme at each visit (daily versus on-demand); and PrEP status at enrolment (already on PrEP versus starting).
Results: Among 2143 participants enrolled in 26 hospitals, 2035 (95%) completed questionnaires. PrEP use scheme and PrEP status information was available for 2004 participants with 5402 questionnaires (visits). Participants using daily PrEP (DPrEP) represented 47.1% of visits. Multivariable estimations showed that participants using DPrEP had 1.5 more sexual partners than those using on-demand PrEP (O-DPrEP) (95%CI[1.36-1.64]). DPrEP users were 27% more likely to use PrEP systematically with sexual partners than O-DPrEP users. Systematic PrEP use probability increased 71% per year (p.y.) for both groups (1.71, 95%CI[1.54-1.90]). Concerning CMRAI, no difference existed between participants using DPrEP and those using O-DPrEP (1.06, 95%CI[0.99-1.18]), CMRAI probability increased 20% p.y. for both groups (1.20, 95%CI [1.14-1.27]). Regarding PrEP status, 44.1% of participants started PrEP at enrolment (PrEP-beginners), representing 40.1% of visits. Participants already on PrEP at enrolment (PrEP-experienced) had 1.3 more sexual partners than PrEP-beginners (95%CI[1.17-1.45]). Systematic PrEP use with sexual partners was 2.6 times higher for PrEP-experienced than for PrEP-beginners (95%CI[2.30-2.86]). Systematic PrEP use probability increased 3.5 times p.y. for PrEP-beginners (95%CI[3.05-3.94]), and 1.23 times p.y. for PrEP-experienced (95%CI[1.14-1.32]). CMRAI was more likely for PrEP-experienced (1.25, 95%CI[1.18-1.34]) than for PrEP-beginners; probability increased 11% p.y. for PrEP-experienced (1.11, 95%CI[1.06, 1.16]) and 32% p.y. for PrEP-beginners (1.32, 95%CI[1.27, 1.43]).
Conclusions: Risk compensation existed especially for participants initiating PrEP. However, PrEP roll-out does not seem to promote sexual behavior disinhibition.