HIV Transmission Networks among Men Who Have Sex with Men in the United States: New Insights from Dynamic Demographic Network Models

Steven M. Goodreau, University of Washington
Nicole Carnegie, New York University
Eric Vittinghoff, University of California, San Francisco
Susan Buchbinder, University of California, San Francisco

An emerging paradigm in public health seeks to tailor multiple interventions for cost-effectiveness, using epidemic modeling to identify potential areas for interaction. However, the complexities of combination interventions require new methods for multiple reasons. We report on a modeling framework developed to estimate and simulate HIV transmission among men who have sex with men. We incorporate numerous forms of demographic, relational, behavioral, and biological heterogeneity, parameterized from large-scale surveys of MSM in the U.S. We rely on the ERGM framework for networks, with two novel extensions (Krivitsky 2009). Initial results suggest that 33% of infections occur within main partnerships, far less than the 68% estimated in a recent paper (Sullivan et al. 2009). Our estimate for the proportion of infections originating with diagnosed, untreated men is high (59%). We conclude by discussing implications of our early results, and upcoming applications to questions of combination HIV interventions for MSM.