Statistical Models in Epidemiology, the Environment, and Clinical Trials

This volume contains refereed papers by participants in the two weeks on Clinical Trials and one week on Epidemiology and the Environment held as part of the six weeks workshop on Statistics in the Health Sciences Applications at the Institute for Mathematics and its Applications (IMA) in the summer of 1997. Donald Berry was in charge of the weeks on clinical trials, and Elizabeth Halloran organized the week on epidemiology and the environment. The collection includes a major contribution from Jamie Robins, Andrea Rotnitzky, and Daniel Scharfstein on sensitivity analysis for selection bias and unmeasured confounding in missing data and causal and inference models. In another paper, Jamie Robins presents a new class of causal models called marginal structural models. Alan Hubbard, Mark van der Laan, and Jamie Robins present a methodology for consistent and efficient estimation of treatment-specific survival functions in observational settings. Brian Leroux, Xingye Lei, and Norman Breslow present a new mixed model for spatial dependence for estimating disease rates in small areas. Andrew Lawson and Allan Clark demonstrate Markov Chain Monte Carlo methods for clustering in spatial epidemiology. Colin Chen, David Chock, and Sandra Winkler present a simulation study examining confounding in estimation of the epidemiologic effect of air pollution. Dalene Stangl discusses issues in the use of reference priors and Bayes factors in analyzing clinical trials. Stephen George reviews the role of surrogate endpoints in cancer clinical trials.