Selective Contact within Structured Mixing with an Application to HIV Transmission Risk from Oral and Anal Sex

Abstract

This paper presents a mathematical model of heterogeneous mixing which is designed to explore the effects of partner selection on the transmission patterns of sexually transmitted diseases. The model generates a symmetric matrix of the rate of sexual contacts between individuals categorized into discrete population subgroups. Sexual pairing results from a process with three separate stages: 1) social contact, 2) acceptance as a potential sex partner, and 3) initiation of sex in response to the needs and standards of each partner. This model can define the contact pattern within any single activity group of the structured mixing model presented by Jacquez et al (1989) in this volume. In such a formulation, the rate of new sexual partnership formation is determined by the number of available partners as well as by sexual needs and moral standards and can vary throughout the course of an epidemic. This approach provides a framework for exploring the potential consequences on infectious disease transmission of different sociological factors while at the same time providing a framework for testing specific sociological theory. We present an example of an epidemic in a male homosexual population where all sexual partnerships are short term. The parameters relating the acceptance of one group by another as potential sexual partners vary by the relative sexual activity rates of individuals and by their relative preference for oral or anal sex. Per sex act transmission probabilities for anal sex were made four times as high as oral sex. Proportional mixing results in a corresponding fourfold difference in risks of oral and anal sex. A moderate degree of selective mixing, however, generates a thirty two fold difference in the risk of oral and anal sex during the stage of the epidemic corresponding to that in which epidemiologic studies have been used to assess the relative risks of oral and anal sex.