The population-level impact of antiretroviral treatment on HIV incidence depends on the clustering of ART patients in sexual networks

Stephane Helleringer, Johns Hopkins University
Wim Delva, South African Centre for Epidemiological Modelling and Analysis (SACEMA)

Antiretroviral treatment (ART) reduces HIV transmission, but its impact on HIV incidence may be offset by changes in sexual networks during the course of ART scale-up. Concerns about “risk compensation” (i.e., increases in individual risk behaviors due to perceptions that HIV is no longer lethal or transmissible) have largely been dispelled in empirical studies. But other emerging patterns of sexual networking may modify the effects of ART on HIV incidence. We used a mathematical model to show that ART homophily - a tendency among ART patients to preferentially form new relationships with other ART patients – can lead to the clustering of ART patients in sexual networks. These effects of these clusters on the effectiveness of ART for HIV prevention, depending on the level of HIV prevalence, coverage of, and adherence to ART. ART homophily should be measured empirically and incorporated into models used to plan and evaluate ART programs.