|
Activity Number:
|
133
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 4, 2003 : 12:00 PM to 1:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #301658 |
|
Title:
|
Model Selection for Clustered Recurrent Event Data Using Frailty Model
|
|
Author(s):
|
Xin Zhi*+ and Lynn E. Eberly and Patricia Grambsch
|
|
Companies:
|
University of Minnesota and University of Minnesota and University of Minnesota
|
|
Address:
|
718 4th St. SE, Minneapolis, MN, 55414,
|
|
Keywords:
|
EM algorithm ; likelihood ratio test ; nested frailty model ; recurrent event
|
|
Abstract:
|
Modeling clustered and recurrent event data may require two frailties.This leads to a need for appropriate tests for frailty model selection. We focus here on nested frailties, which arise naturally for recurrent event data collected in a multicenter clinical trial: patients within a clinic share a common frailty, while multiple events for each patient share another common frailty; the second frailty is nested in the first one. A shared gamma frailty model (one level of frailty) and a multiplicative double gamma frailty model (two levels of frailty) can be fitted to this type of data using EM-based algorithm. We carried out a simulation study to examine the performance of a likelihood ratio test in choosing between the single and double frailty model. Simulation results show that when the single frailty is the correct model, the LRT statistic is approximately distributed as 50:50 mixture of Chisq(0)+Chisq(1). The power of the test varies according to the strengths of the frailty effects as well as simulation setup. The Women's Fungal study of the Terry Beirn Community Programs for Clinical Research in AIDS (CPCRA 010) is used to illustrate the application of the LRT in practice.
|
