Purpose: Reducing concurrent sexual partnerships is an important goal for HIV prevention in sub-Saharan Africa. The impact of concurrency-related changes in sexual behavior on the spread of HIV is unknown.
Method: Using HIV and sexual partnership data from national surveys in Swaziland, Tanzania, Uganda, and Zambia, we developed a network model to compare the effects of four behavior changes that might occur with concurrency reduction campaigns: (1) changes in partnership patterns from multiple concurrent partners to serial monogamy; (2) partnership reduction among those who have three or more concurrent partners (high risk); (3) partnership reduction among individuals with two concurrent partners; (4) partnership reduction among all individuals with concurrent partners. We estimated HIV incidence and prevalence of a program that is 10% effective in a population of adults aged 15-49 over a 10-year period.
Result: The most effective type of behavior change is partnership reduction among high-risk individuals: compared to no behavior change, HIV prevalence was lower by an average of 1.7% in the four study countries after ten years, and new HIV infections decreased by an average of 12.0%. Partnership reduction among all individuals and moderate-risk individuals, as well as increased monogamy, yielded slightly less benefit, reducing HIV prevalence, respectively, by .93%, .82%, and .87%, and new HIV infections by 9.2%, 8.6%, and 9.0%. The impact of different types of behavior change depended on the country and percentage change in behavior.
Conclusion: Reduced concurrency among high-risk individuals is the most effective target for reducing HIV incidence and prevalence, but concurrency reduction in other risk groups yields nearly as much benefit, so broader concurrency reduction programs are nearly as effective. However, even complete elimination of multiple concurrent sexual partnerships cannot stop HIV transmission, so complementary HIV prevention programs are essential. 
Candidate for the Lee B. Lusted Student Prize Competition