Copula-Based Methods for Multivariate Survival Analysis

Mississippi State University, Mississippi State MS

Investigators

Abstract

Nonfatal disease-related hardship diminish quality of life. For example, cancer survivors often experience adverse health outcomes resulting from treatment or disease progression. Understanding the dependence among times to such events is crucial for designing effective care. This project will develop new statistical methods to model such dependencies and how the occurrence of these nonfatal events affects overall survival. The results of this research can inform new treatment strategies and improve the quality of life for many patients. This project will integrate the research with curriculum development and the training of graduate students and provide research opportunities for undergraduate and high school students. The project leverages copulas, popular statistical tools for modeling the dependence among event times, to address three major aims. The first aim is to develop a likelihood-based goodness-of-fit test for detecting copula misspecification, which would lead to invalid estimation. The second aim focuses on missing event times caused by censoring, such as the termination of follow-up. New procedures will be designed to impute the missing event times via inter-event dependence. The third aim is to establish a vine copula framework to model the dependence among times to multiple nonfatal events and their relationship with time to death. The theoretical aspects of this research effort will be complemented by developing open-source software packages in R for use by the research community. This project is jointly funded by the Statistics Program and the Established Program to Stimulate Competitive Research (EPSCoR). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.