Can bad models suggest good policies? Sexual mixing and the AIDS epidemic

Publisher Website
Google Scholar
Abstract
The epidemiology of HIV transmission, and hence AIDS, depends upon phenomena such as sexual mixing patterns, about which very little is known. Mathematical models of sexually transmitted HIV assume random mixing among susceptible and infected individuals, across differing levels of sexual activity. Given that this representation of sexual mixing is probably wrong (in the sense of descriptive accuracy), should such models be used to formulate and/or evaluate AIDS intervention strategies? This paper presents an example where random mixing, though wrong descriptively, identifies a useful policy. In the process, the concept of worst case sexual mixing is introduced. Arguments for involving simple models of the AIDS epidemic in policy analysis conclude the paper.