Longitudinal data from a panel of male veterans who have been examined every 3 years since 1968 and data from an experimental study of health insurance are being used to analyze the progression of tooth loss and the conditional probability of change in tooth conditions over time. Both data designate each tooth as missing or present. If present, a tooth is designated as sound, carious, or filled. If missing, the space is designated as replaced or not replaced. If replaced, the type of prosthetic replacement is recorded. These data will be analyzed to develop conditional probabilities of a tooth being lost over time. These probabilities will be conditioned on the status of the tooth at the previous examination and other putatively important explanatory variables. The panel of male veterans will allow several longitudinal examinations to be analyzed. The experimental data contains a wide array of potentially important explanatory variables which will allow the impact of economic and social variables on the probabilities to be assessed. Analyses are nearing completion. A report of the findings will be published.
