What level of risk compensation would offset the preventive effect of early antiretroviral therapy? Simulations from the TEMPRANO trial

Abstract

Whether risk compensation could offset the preventive effect of early initiation of antiretroviral therapy (ART) on human immunodeficiency virus (HIV) transmission remains unknown. Using virological and behavioral data collected 12 months after inclusion in the TEMPRANO randomized trial of early ART (Abidjan, Côte d'Ivoire, 2009-2012), we estimated the risk of HIV transmission and compared it between the intervention (early ART; n = 490) and control (deferred ART; n = 467) groups. We then simulated increases in various sexual risk behaviors in the intervention group and estimated the resulting preventive effect. On the basis of reported values of sexual behaviors, we estimated that early ART had an 89% (95% confidence interval: 81, 95) preventive effect on the cumulative risk of HIV transmission over a 1-month period. This preventive effect remained significant for a wide range of parameter combinations and was offset (i.e., nonsignificant) only for dramatic increases in different sexual behaviors simulated simultaneously. For example, when considering a 2-fold increase in serodiscordance and the frequency of sexual intercourse together with a 33% decrease in condom use, the resulting preventive effect was 47% (95% confidence interval:-3, 74). An important reduction of HIV transmission may thus be expected from the scale-up of early ART, even in the context of behavioral change.