Joint modeling of longitudinal and time-to-event data in NLST
Principal Investigator
Name
Fenghai Duan
Degrees
PhD
Institution
Brown University
Position Title
Biotatistician
Email
fduan@stat.brown.edu
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
201111-0041
Initial CDAS Request Approval
Nov 29, 2011
Title
Joint modeling of longitudinal and time-to-event data in NLST
Summary
Currently, these two outcomes are mainly analyzed separately, e.g., fitting a mixed effects model for the longitudinal data and fitting a Cox proportional hazards model for survival outcome. However, these approaches do not consider the dependency between them. There are situations that a joint modeling is of more interest and more appropriate, e.g., analyzing the time-to-event outcome while taking account of the dependency and association between the longitudinal measures and the time-to-event data (i.e., conducting survival analysis with a time-dependent covariate) (Diggle et al, 2008; Brown 2009). The research in this area started in the late 90s and has been boosted in the last decade (Hogan and Laird 1997; Henderson et al. 2000; Xu and Zeger 2001; Song et al 2002; Chi et al 2006; Cai et al. 2006; Diggle et al 2008; Williamson et al 2008; Brown 2009; Subtil et al 2009; Ibrahim et al 2010). There are two excellent review articles given by Tsiatis and Davidian (2004), and Ibrahim et al (2010).
Collaborators
Constantine Gatsonis
Richard Fagerstrom
Timothy Church