A methodological study on fitting a nonlinear stochastic model of the AIDS epidemic in Philadelphia

A nonlinear stochastic model accommodating heterogeneous risk behavior and recruitment was fit to Philadelphia public health data adjusted for delays in reporting. The methodological study that was performed resulted in a finding of significant clinical importance. Namely, that the variable time from infection with HIV to seroconversion may be longer than reported in the literature. The finding challenges our understanding of the progression of HIV disease and has profound public health implications. Although the model and the software incorporated several risk categories, such as heterosexual males and females among others, those examined in this paper are confined to white male homosexual/bisexual intravenous drug users and white male homosexual/bisexual nonintravenous drug users. The results of this paper demonstrate that it is possible to fit a rather complex stochastic model to public health data, using computer intensive methods, and thus, more completely reflect the diversity of human behavior represented by the defined risk groups of the data.