The question driving the proposed research is simple: even though knowledge about the sexual transmission of HIV is widespread in sub-Saharan Africa, why do certain groups of heterosexual men continue to engage in risky sexual behavior, thereby putting themselves and others at risk? Despite two decades of research and more than 17,000 peer reviewedarticles on HIV/AIDS knowledge, attitudes, and practices, we are still struggling to answer this question. This proposal aims to understand the social and contextual factors that facilitate and trigger HIV risk behavior among heterosexual men in Ghana and Tanzania. To answer these questions, we will examine the social and environmental contexts that occur within the three-hour window prior to a sexual event (e.g. being away from home, drinking alcohol, discussing HIV status with partner) and the sexual event itself. We will also look at the impact of respondent characteristics, gender dynamics, and longer-term contextual factors on risk behavior. The proposed study will be carried out primarily among men at high risk for HIV, ages 18-49, in high transmission areas of two African cities with high HIV prevalence - Tema (Ghana) and Mbeya (Tanzania). Select groups of women will also be included in the study. Our approach is novel and will extend models currently used for behavioral interventions among men in Africa. It combines two theoretical,perspectives - Gender Role Theory andThe Social-Ecological Modelof Health - and focuses on identifying risk triggers. We will look at several levels of variables, and how they interact to produce risk behavior, including the formation of high-risk sexual networks. The study employs a mixed-method design that integrates participant observation, focus groups, in-depth interviews and a structured survey. The survey will employ a local network design, and ask men about their last three different sexual partners. Such a design will provide us with estimates of concurrency,patterns of mixing, and the nature and extent of epidemiological bridging. Survey data will primarily be analyzed using structural equation modeling, a technique that can account for complex interactions among variables. Qualitative and quantitative data will be explicitly linked and triangulated to enhance validity of findings and increase explanatory power. This proposal has several innovative aspects, including its focus on the three-hour window preceding risky sex, its emphasis on gender scripts by HIV risk region, the inclusion of complex structural equation modeling of HIV risk, and an explicit mixed-methods design. By embracing a multi-level and mixed-method approach, combined with an emphasis on proximate behavioral triggers, the proposed research will identify those forces that impact risk behavior most and lead to concrete and actionable intervention points for high-risk men and their partners.
| Fleming, Paul J; Mulawa, Marta; Burke, Holly et al. (2015) The role of relationship types on condom use among urban men with concurrent partners in Ghana and Tanzania. AIDS Care 27:466-72 |
| Burke, Holly McClain; Fleming, Paul J; Guest, Greg (2014) Assessment of the psychometric properties of HIV knowledge items across five countries. AIDS Educ Prev 26:577-87 |
