The general objective of the proposed research is to continue development of statistical methods for specific application to clinical trails in diabetes, digestive, kidney and other chronic diseases. In most of these clinical trails, the principal outcomes are measures of morbidity rather than simple mortality; multiple, possibly recurrent responses are recorded over time; data are partially missing; and restricted randomization procedures are employed. The Biostatistics Center of the Department of Statistics/Computer and Information Systems, The George Washington University, is currently the coordinating center for clinical trials in diabetes, endocrinologic and kidney diseases, and previously was the biostatistical support unit for a clinical trail in digestive diseases, all funded by NIDDK. Each of the proposed research topics is directly motivated by current and anticipated statistical needs of these and other clinical trials in these disease areas. Nine specific research projects are identified: (1) Group sequential analysis using the logrank test with error spending function based on patient exposure fraction; (2) Error spending functions for group sequential methods allowing early acceptance of the null hypothesis; (3) Implications of an alternative definition of the multiple comparison problem; (4) Testing and estimation under restricted randomization models; (5) Two- sample problems based on a Wilcoxon statistic; (6) Weighted least squares methods for categorical data with repeated measures and randomly missing values; (7) Optimal combination of RxC contingency tables; (8) Application of point process models to inference for recurrent event rates; (9) Joint covariate and dispersion parameter estimation and testing in quasi-likelihood regression models for discrete data. Additional research topics will be identified as work proceeds on these projects, and in response to the ongoing data analytic activities of the Coordinating Center staff for these chronic disease trials.
