The dynamics of HIV spread: a computer simulation model.

Mathematical modeling of the AIDS pandemic has been limited by the difficulty of satisfactorily representing the marked behavioral heterogeneity that characterizes the various populations at risk. We propose an approach which models the spread of infection as a discrete-event simulation using SIMSCRIPT, a powerful simulation language. The program developed provides sufficient flexibility to adequately represent and study a wide range of risk-group dynamics. Using this tool we have verified the May-Anderson prediction relating contact rate heterogeneity to the rate of HIV spread. We have also been able to assess the sensitivity of the model to the particular choice of distribution for contact rates, disease stage durations, and intercontact intervals. It is thought that this approach will permit the empirical testing of hypotheses which do not lend themselves to a purely mathematical treatment.