Use of agent-based models to inform HIV prevention for men who have sex with men

Patrick Sullivan, Ph.D.
Emory University

ABSTRACT: 
Agent-based models allow a virtual “laboratory” to evaluate counterfactual scenarios, evaluate the potential impact of combinations of HIV prevention, and benchmark levels of prevention services required to achieve population-level impacts of programs. We will review agent-based models of HIV prevention approaches for MSM in 5 countries, and describe how the findings of these models have been used to design combination prevention programs and guide programmatic goal setting. The results of agent-based models in different countries support the role of smart combinations of prevention interventions (e.g., PrEP, testing, treatment) and of ancillary services to increase the impact of specific modalities (e.g., adherence promotion for PrEP).