Background Limited data are available to assess the potential for increased sexual risk-taking by persons using antiretroviral pre-exposure prophylaxis (PrEP) for HIV prevention. In July 2011, the Partners PrEP Study, a randomised trial of daily oral tenofovir and emtricitabine/tenofovir PrEP among HIV-uninfected members of African heterosexual HIV serodiscordant couples, demonstrated efficacy of PrEP for HIV prevention and use of placebo was discontinued in the trial. Follow-up of study participants on active PrEP through December 2012 provided an opportunity to evaluate risk behaviour on PrEP after efficacy was announced.

Methods Among participants assigned to active PrEP pre-unblinding who continued follow-up after the placebo arm was stopped, we used zero-inflated negative binomial regression with robust standard errors and adjusted for baseline sexual behaviour, age, gender, and secular changes to compare the frequency of unprotected sex up to 12 months before versus after knowledge of PrEP efficacy.

Results We analysed 54,876 person-months (33,254 pre- versus 21,622 post-unblinding) from 3024 HIV-uninfected subjects (64% male). On average, the observed frequency of unprotected sex with the HIV-infected study partner was 58 per 100 person-months pre-unblinding versus 53 per 100 person-months post-unblinding, reflecting no immediate change or change in trend over time following unblinding (p = 0.734 and 0.264, respectively). The annual average total number of unprotected sex post-unblinding was 6 acts versus 5 that would have been expected in the counterfactual situation had unblinding not occurred. There was no significant increase in diagnoses of incident sexually transmitted infections or pregnancy post- versus pre-unblinding (p > 0.05).

Conclusion The transition from a blinded, placebo-controlled trial to all participants aware they were receiving active PrEP in the Partners PrEP Study provided a “natural experiment” to evaluate behavioural risk compensation. PrEP, provided as part of a comprehensive prevention package, may not result in substantial changes in risk-taking sexual behaviour in HIV serodiscordant couples.