Date of Award


Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)

First Advisor

Paltiel, A. David


While there is no magic bullet that can completely prevent HIV transmission and halt the HIV/AIDS pandemic, governments and policy makers have an array of partially-effective HIV prevention programs from which to choose, including the use of existing interventions (e.g. education and condom use) and technologies under development (e.g. microbicides and vaccines). Complex decisions regarding if, when, and how to implement various programs with partial efficacy must be made, often in the absence of data on program outcomes. Additionally, the quantitative tradeoff between program-related decreases in HIV transmission and the effects of risk behavior change is often unknown. The potential for changes in risk behavior to influence outcomes at the individual and population levels raises significant operational and ethical concerns regarding the magnitude of program benefits.Epidemiological data collection of sexual risk-taking behavior and mathematical modeling of population HIV transmission dynamics are used here in a multidisciplinary approach to examine the implications of risk behavior change surrounding the implementation of HIV prevention programs with less than 100% efficacy. This topic is explored within a South African context---specifically, the urban township of Soweto, which has a generalized, predominantly heterosexual HIV epidemic representative of populations throughout sub-Saharan Africa.The subject of risk behavior change associated with prevention program implementation arises most frequently in discussions regarding HIV vaccine development and use. The first and second papers contain analyses of self-reported data on current and anticipated sexual risk-taking behavior from adults undergoing screening and enrollment into HIV vaccine trials. The third paper describes a mathematical model to simulate the dynamics of heterosexual HIV transmission, including gender differences in the negotiation of safe sexual practices, in these populations. The potential impact of future HIV vaccination programs is considered, focusing on the interactions between vaccine efficacy and risk behavior change. The analysis is extended in the fourth paper by adapting the framework developed for HIV vaccines, likely not available for at least a decade, to consider adult male circumcision---another potential, partially---effective HIV prevention program for which clinical trial efficacy data have recently been released and for which the technology is available immediately.

Open Access

This Article is Open Access