Division of Biostatistics
Seminar Series Fall 2007

Varying Coefficient Cox Model with Nonparametric Longitudinal Covariates

Jimin Ding
Department of Mathematics
Washington University in St Louis

Friday, October 5, 2007, 12:30–1:30 pm

GEMS classroom, 3rd Floor in Shriner's Building
Coffee, tea, and cookies will be provided

Abstract

In biomedical and clinic studies, time to some interesting events is usually recorded together with longitudinal covariates for each subject. The goal is not only to study the influence of covariates on survival time but also the pattern of longitudinal covariates. Joint modeling of the longitudinal and survival data has emerged an effective way to gain information for both processes and will be studied here. We propose to use the Cox model with timevarying coefficients to link longitudinal covariates and survival time and investigate the possible time-dependent survival regression effect. Considering longitudinal covariates are often measured intermittently and may contain measurement errors, we will estimate unobservable longitudinal process together with time-varying coefficients in the joint modeling setting.

This is the joint work with Prof. Jane-Ling Wang, UC Davis.