Some general stochastic models for the spread of AIDS and some simulation results (Q2277192)

A stochastic model is formulated for the spread of AIDS epidemic. This model involves l (l\(\geq 1)\) populations (representing the different risk groups) and n categories (susceptible, latent and infected persons, AIDS cases). The infected persons are subdivided into k stages, and the AIDS cases are subdivided into m stages. Transition rates between the categories and from stage to stage, as well as migration and immigration rates are defined. It is assumed that the number of suceptible persons is a deterministic function of time. This general model is simplified to yield a Markov process. Using the Kolmogorov forward equations, a first order partial differential equation for the probability generating function of this process is derived. This equation is used to derive two systems of differential equations: one for the expected number of persons in each stage, and one for the variances and covariances of the number of persons in each stage. For the special case of \(l=4\), \(m=1\) and \(k=1\), 10 or 20, and assuming some constant values for the transition and migration rates (not allowing for immigration), numerical results are shown for the solutions of both of these systems of differential equations. The effects of the initial number of infected persons, the effect of a decreasing sexual contact rate and the effect of an increasing number of infected stages are discussed on the basis of numerical evaluations.