The research proposed in this application has as its primary goals the elaboration of a mathematical and statistical methodology for the estimation of the mixing social/sexual structures in the development of estimators of the mid- (5 to 10 years) and long term (10-30) seroincidence and seroprevalance of the human immunodeficiency virus. The methodology will be developed through computer simulations based on a mixing theory that we have recently constructed from first principles. In these simulation, populations with predetermined structures will respond to instruments (developed for this purpose and from national behavioral studies) designed to measure the distribution of sexual partners of an individual, as well as the sexual contact structure among subpopulations. Families of prior mixing parametric distributions constructed from our mixing theory will be used in conjunction with empirical Bayes; statistical methodology to improve on our estimators of sexual mixing. Finally, this research project integrates for the first time uniquely tailored, mathematically well-understood models for the dynamics of HIV, empirical Bayes; statistical methodology and data on sexually transmitted diseases (gonorrhea and syphilis) to generate mid- and long- term estimators of the seroincidence and seroprevalence of HIV among specific subpopulations.
|Busenberg, S; Castillo-Chavez, C (1991) A general solution of the problem of mixing of subpopulations and its application to risk- and age-structured epidemic models for the spread of AIDS. IMA J Math Appl Med Biol 8:1-29|