ANALYSIS OF INFECTIOUS-DISEASE DATA FROM PARTNER STUDIES WITH UNKNOWNSOURCE OF INFECTION

Partner studies are useful for estimating the transmission probabiliti es of infectious diseases. However, it is often not known which partne r was the source of the infection (the index case). The objective of t his paper is to develop statistical methods for analyzing partner stud ies when it is uncertain which partners acquired the infection from so urces outside the partnership. The approach involves simultaneously mo delling the probability of acquisition of infection from outside the p artnership, and the probability of transmission within the partnership as a function of covariates. An EM algorithm is presented. Efficiency and simulation results are given in some special situations involving heterosexual transmission studies. In heterosexual partner studies, t he methods depend crucially on the availability of a covariate that pr ovides information about which partner was the likely source of infect ion.