A MARKOV MODEL FOR HIV DISEASE PROGRESSION INCLUDING THE EFFECT OF HIV DIAGNOSIS AND TREATMENT - APPLICATION TO AIDS PREDICTION IN ENGLAND AND WALES

Back-calculation is a widely used method to estimate HIV incidence rat es, and is commonly based on times of AIDS diagnosis. Following up ear lier work, we extend this method to also incorporate knowledge of time s of HIV diagnosis (first positive test). This is achieved through the use of a Markov model which describes the progress of an HIV infected person through various stages, and which allows causal connections be tween events to be explicitly modelled. Estimation is based on maximum likelihood, the likelihood being calculated within a discretized vers ion of the Markov model. The effect of sampling uncertainty and model uncertainty (sensitivity) is evaluated simultaneously by means of a co mbined bootstrap and simulation procedure. At each replication we resa mple both the data and the model (from a set of possible models descri bed by randomizing one or more parameters). For instance, uncertain kn owledge about the incubation distribution affects the estimates of som e parameters, but not others. The Markov approach is applied to the pr ediction of AIDS incidence for homosexuals in England and Wales up to the year 2000. (C) 1997 by John Wiley & Sons, Ltd.