the potential impact of changes in risk behavior

The potential impact of changes in risk behavior is one important determinant of the possible effect of ART in preventing transmission within a population. A model of ART use among MSM in a resource-rich setting [6] indicated that if treatment rates are high, transmission is likely to be significantly reduced; however, the rate of infection will be strongly influenced by changes in risk behavior, such that transmission could rise despite the increase in number treated. In a meta-analysis reviewing results of 25 studies of the impact on sexual risk behavior of ART and beliefs regarding transmissibility [35], being on ART or achieving an undetectable viral load was not associated with increased risky sexual behavior. However, beliefs about HIV transmission and reduced concerns about unsafe sex were associated with a greater likelihood of risky behavior.

Modeling of infectiousness of HIV-infected persons at different stages of disease suggests that persons with high viral loads associated with acute infection and end-stage disease are highly infectious; however, those with asymptomatic infection will contribute most to transmission within a population because of the longer duration of this stage [36]. This suggests that early identification and therapy of infected persons may be an important aspect of the effectiveness of ART as a prevention tool. In a model (designed to reflect South African transmission dynamics) of a program to control heterosexual transmission which included universal voluntary HIV testing and immediate initiation of therapy in those found to be infected [7], incidence of new HIV cases was reduced to less than one per 1000 persons per year within a decade, and estimated deaths occurring through 2050 were estimated to be halved when compared to the current approach of waiting to start ART until a specific CD4 count is reached.