|
Activity Number:
|
44
- Joint Modeling of Longitudinal Data, Recurrent Events, and a Terminal Event
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, July 30, 2017 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
ENAR
|
|
Abstract #322228
|
|
|
Title:
|
Joint Modeling of Marker and Event Processes While Adjusting for Dependent Observation
|
|
Author(s):
|
Richard John Cook*
|
|
Companies:
|
University of Waterloo
|
|
Keywords:
|
censoring ;
dependent observation ;
marker processes ;
multistate models
|
|
Abstract:
|
In cohort studies interest often lies in modelling marker processes and examining the relation between marker dynamics and the occurrence of clinically important events. Frameworks for joint modelling of markers and lifetime data are critically reviewed and the assumptions regarding observation processes necessary for valid analysis are given. Inverse intensity-based weights are developed for such joint models to accommodate marker and event-dependent observation of individuals. The asymptotic biases arising from naive analysis of data acquired from a dependent inspection scheme are illustrated and presented along with supporting empirical findings demonstrating the performance of estimators from weighted analyses. Application to data from a study of inflammatory markers and joint damage in a cohort study of arthritis patients is given for illustration.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2017 program
|
