|
Activity Number:
|
331
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
WNAR
|
| Abstract - #309680 |
|
Title:
|
Joint Analysis of Repeated State Transition in Activities of Daily Living and Death in Longitudinal Aging Study
|
|
Author(s):
|
Haiqun Lin*+ and Zhenchao Guo and Peter Peduzzi and Thomas Gill and Heather G. Allore
|
|
Companies:
|
Yale University and Yale University and Yale University and Yale University and Yale University
|
|
Address:
|
, , ,
|
|
Keywords:
|
aging study ; dependent censoring ; latent trait ; longitudinal data ; multistate transition
|
|
Abstract:
|
A general multistate transition model is proposed for repeated episodes of multiple states representing different health status. A semiparametric model with transition-specific nonparametric baseline intensities and state-specific latent traits is adopted. A latent trait captures individual tendency of repeated sojourn in a state and accounts for within-state-within-subject correlation. Between-state correlation is accounted for by the covariance of the latent traits. The factor loadings of a latent trait accommodate the dependence of transitions to different competing states from a same state. The method is illustrated by repeated transitions between independence and disability of activities of daily living with death as an absorbing state in a longitudinal aging study. A simulation study is performed to study the appropriateness of the asymptotic estimates of the standard errors.
|
