The chain multinomial models of the HIV epidemiology in homosexual populations (Q1324270)

The author has developed several stochastic models that take into account the dynamic of HIV epidemics [see e.g. the author and \textit{R. H. Byers}, Math. Biosci. 113, No. 1, 115-143 (1993; Zbl 0767.92027)]. In this paper he considers a `chain multinomial model' to develop the probability of HIV infection for susceptible people, infective people and AIDS cases at any time. He derives the means, variances and covariances. His results suggest that the infection duration of infective people and sexual activity levels have significant impact on the projection of HIV epidemics.