Recurrent Event Models in the Presence of a Terminal Event: Comparison, Inference, and Data Analysis

• Activity Number: 72
• Type: Contributed
• Date/Time: Sunday, August 7, 2005 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Xianghua Luo*+ and Mei-Cheng Wang
• Companies: Johns Hopkins University and Johns Hopkins University
• Keywords: Latent ; rate functions ; regression models

Abstract

This paper discusses statistical implications of joint models of recurrent events and a terminal event. As an extension of the proportional hazards model for time to a single event, the proportional rate models are used commonly for modeling time to recurrent events. In the presence of an additional terminal event, various definitions of the rate function have been adopted in the proportional rate models. While these rate functions have different interpretations, the recognition of the differences is lacking theoretically and practically. Here, three types of rate functions for recurrent events are compared in conceptual and quantitative perspectives, which leads to the conclusion that incautious use of a specific rate function may lead to misleading scientific conclusions. We aim to provide guidance in choosing appropriate statistical models and interpreting the analytical results. Simulations are conducted for comparisons of the focused models. A data analysis of a clinical trail conducted by the Community Programs for Clinical Research on AIDS (CPCRA) is presented to illustrate the analytical results.