The focus of this project is development and refinement of statistical procedures for the design and analysis of cancer screening and related studies. Problems under investigation data analysis methods, assessment of case-control studies, development of procedures for monitoring cancer screening trials, and estimation of length bias and lead time. To assess the case-control design for screening evaluation, the MISCAN microsimulation model is being used to provide population data with and without screening. Case-control studies are then done in the screened populations, and the results are compared with the true effect to assess bias in the case-control approach. A method for estimating the effect of starting periodic screening at different ages was developed that uses data from a screened population to estimate the probability of diagnosis in an unscreened group. A noncompliance model was developed that makes possible estimates of cost-effectiveness adjusted for refusal to be screened. A new design was proposed to test the equivalence of digital versus analog mammography with respect to sensitivity and specificity, at a lower cost than would be possible with a standard design. Methods using the differences in case survival measured both from time of entry and time of diagnosis between screened and control groups were used to develop length bias estimators, and methods based on the catch-up time were investigated. Approaches for estimating the post lead time survival of screen detected cancer cases were devised with a focus on estimation under dependence of disease state durations. Approaches were defined for data monitoring of cancer screening trials, including methods for nonproportional hazards, adapting existing simulation programs.
