The goals of this project are to continue our investigation into statistical methods to assess the effects of covariates on outcome in longitudinal studies. Such data arises in a variety of medical studies and includes survival data, competing risks data, data from multi-state models and data from repeated measures experiments. Our primary application is to bone marrow transplantation, but these techniques can be applied to a wide range of medical areas. ? ? Specific projects include the development of a regression approach, based on Jackknife pseudo values, that can be used as summary model for multi-state models; Continued development and study of alternatives to the Cox model for survival and competing risks data; the development and study of techniques for measuring explained variation which can be used to examine and compare models for predicting a patient's outcome. A study of the various techniques for comparing treatment efficacies in observational studies; continued studies of models for competing risks data including techniques for covariate adjustment of the cumulative incidence curve and methods to compare cumulative incidence curves; techniques to include covariates into models for multi-state data and continued development of methods for modeling repeated measurements taken over time as functions of covariate processes.
Showing the most recent 10 out of 74 publications
