Abstract Body

Antiretroviral therapy (ART) is highly efficacious in preventing HIV in randomized trials of volunteer serodiscordant couples with full serostatus awareness. However, evidence of the effectiveness of ART in preventing HIV in population-based samples, where serostatus may not be known to the individual or their partner, is lacking.  Here, we assess the effect of ART on HIV acquisition in a population-based cohort in rural KwaZulu-Natal, South Africa.  In this cohort, couples are linked via surveillance rather than via clinics, allowing for estimation of the effect of ART in a “real-world” setting where individuals are not necessarily aware of their HIV status or the status of their partner. 

All HIV-uninfected individuals present between January 2005 and December 2013 (n=17,016) with at least two HIV tests were included, regardless of whether or not they had a cohabitating partner.  Individuals with cohabitating partners were categorized as having an HIV-uninfected partner, an HIV-infected partner who was not on ART, or an HIV-infected partner who was on ART.  ART status of HIV-infected partners was determined via public-sector ART clinic data.  Interval-censored time-varying parametric proportional hazards regression was used to assess how the partner’s ART status affected HIV acquisition risk.

Of the 17,016 individuals, 1,846 had an HIV-uninfected and 196 had an HIV-infected partner over the follow-up period, of whom 76 initiated ART during follow-up. HIV incidence was 0.3 per 100 person-years among individuals with an HIV-uninfected partner (95% confidence interval [CI] 0.2-0.5), compared to 5.6 per 100 person-years (95% CI 3.5-8.4) among individuals with an HIV-infected partner who was not on ART, and 1.4 per 100 person-years (95% CI 0.4-3.5) among individuals with an HIV-infected partner who was on ART.  In an adjusted model, HIV-infected partner’s use of ART was associated with a 77% decrease in HIV acquisition risk amongst serodiscordant couples (aHR=0.23, 95% CI 0.07-0.80).  

ART initiation was associated with substantially reduced HIV incidence, but less than has been seen in more controlled settings.  Achieving effective population control of HIV incidence, and thus elimination of the epidemic, may not be possible only through ART provision to seropositive persons. A combination of additional interventions is likely to be necessary.