Concurrency and HIV spread in sub-Saharan countries: biases from self-reported data and their implications for HIV prevention

Stephane Helleringer, Columbia University
Linda Kalilani-Phiri, University of Malawi
Jemima A. Frimpong, University of Pennsylvania

Concurrent partnerships (MCP) have been described as the “key driver” of generalized HIV epidemics. However, MCP are not more common in regions of Africa with high HIV prevalence than in regions with concentrated epidemics. We test the hypothesis that under-reporting of MCP during surveys may confound the association between concurrency and HIV transmission. We conducted a sexual network study on Likoma Island (Malawi). Population-level sexual networks were constructed by linking reports of sexual partnerships to village rosters, and we considered that a sexual relationship took place when one of the two partners reported it. The prevalence of MCP increased significantly according to network data. The largest increases were seen among young women (ages 18–24) and older men (ages 25 and above). Several risk factors for HIV transmission were more common in MCP omitted by self-reports. The impact of concurrency on HIV transmission may be larger than initially thought.