Proportional Hazards Model with a Mixture of Berkson and Classical Errors in Covariates and Calibration Data

• Activity Number: 367
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Jean de Dieu Tapsoba* and Ching-Yun Wang
• Companies: Fred Hutchinson Cancer Research Center and Fred Hutchinson Cancer Research Center
• Keywords: Berkson error ; calibration subsample ; classical error ; expected estimating equations ; instrumental variable ; proportional hazards

Abstract

In many biomedical or epidemiological studies, it is often of interest to assess the association between the time to an event of interest and some covariates under the Cox proportional hazards model framework. A problem commonly arising in such studies is that some covariates are subject to measurement error, which may be of classical type, Berkson type or a combination of both types. The issue of Cox regression with error-prone covariates has been well-discussed in the statistical literature, which has focused mainly on classical error so far. This paper considers Cox regression analysis when some covariates are possibly contaminated with a mixture of Berkson and classical errors. We propose two new approaches to addressing this problem. More specifically, we develop a SIMEX-based approximate method and a consistent non-parametric expected estimating equations method when two replicates of the mismeasured covariates are available along with calibration data for some subjects in a subsample only. The proposed methods place no assumption on the mixture percentage. Their finite-sample performances are assessed via simulations. They are applied to an AIDS clinical trial.