A variety of biomedical, behavioral, and structural interventions are now available to control HIV epidemics among people who inject drugs (PWID). Since no intervention eliminates transmission entirely, current prevention paradigms focus on combination approaches to reduce, and eventually eliminate, new HIV infections among PWID. However, bringing an entire suite of interventions to scale, particularly during early stage HIV epidemics, is neither efficient nor feasible in many contexts. One possible solution to this challenge is the use of mathematical modeling to determine which and when specific interventions should be implemented in any given context. Unfortunately, mathematical models have been limited in their ability to account for evolving risk behavior, dynamic sexual/drug-using network structures, and diverse epidemic contexts. These factors influence (in crucial ways) intervention effectiveness and have substantially limited the capacity for mathematical modeling to inform targeted and more effective HIV prevention programs. While the implementation of highly adaptive prevention strategies holds potential for the elimination of HIV among PWID, studies to inform these approaches have thus remained in their infancy. I propose a pioneering research program that will integrate behavioral, epidemiological, and network data within a robust modeling platform to determine how HIV prevention strategies should be optimized for different epidemic phases and contexts. To achieve this objective, I will create a series of agent-based models representing artificial societies of PWID in settings across North America and internationally. These models will be used to reproduce entire epidemic trajectories, incorporating an unprecedented diversity of risk behavior, network structures, and intervention availability across contexts. The virtual epidemics will then be interrogated with hypothetical combination prevention strategies. By testing interventions in silico, I will discover at what point in an epidemic course specific intervention(s)-opioid substitution therapy, needle and syringe programs, pre- exposure prophylaxis, and treatment as prevention-based approaches (among others)-should be implemented and brought to scale to most effectively curtail HIV transmission. Moreover, by modeling interventions across epidemic contexts, I will determine which prevention strategies are robust to significant differences in ris behavior patterns and network structures. This project thus represents an unparalleled scientific effort in which high-resolution microsimulations will be used to inform tailored, community-specific responses to HIV transmission among people who inject drugs.
This project will determine how combination HIV prevention programs should be prioritized to reduce and eventually eliminate HIV transmission among people who inject drugs (PWID). A novel approach known as agent-based modeling will be used to inform how HIV prevention strategies for PWID should be tailored to specific epidemics and contexts. The results of this project will guide the implementation of highly effective HIV prevention strategies that are more efficient, better targeted, and more responsive to the needs of substance-using communities.
Showing the most recent 10 out of 19 publications
