Estimating Potential Effectiveness of Varied Public Health Intervention Programs Using Agent-Based Modeling Techniques

Limited funds, growing population needs, and the requirement to quantify results are placing greater strains on public health programs. Because individual interventions have varying levels of success, public health programs, like those currently being used to battle the HIV/AIDS epidemic in developing countries, currently employ a mix of approaches. In order to do so, program funds must be allocated for each intervention type, often without the benefit of knowing which intervention, or mixture of interventions, will provide the greatest social impact and economic benefit to the population. This period of trial and error can lead to wasted funds, increased disease exposure, higher rates of infection in the population, and greater morbidity and mortality. Using dynamic agent-based models allows public health practitioners and policy makers to examine various intervention assumptions and configurations, simulate varying levels of population participation for interventions, explore the outcomes, and examine costs and funding restraints. This can help implementation planners arrive more quickly at a potential optimized program. The outcomes from the simulation can also suggest the levels of measures to be used to validate the expected program effectiveness.

 This poster illustrates the steps used to create a dynamic systems model for an HIV/AIDS intervention program and preliminary results of examining participant benefits and program costs. The accompanying simulation is a working model that allows the end user to adjust and test variables to match the expected environment and program parameters. The simulation’s results are expressed in measures that can be transferred directly to observed results collected from the interventions and programs. These measures can then be fed back into the model to refine it and improve future results.