Online Program Home
My Program

Abstract Details

Activity Number: 408 - Joint Modeling of Longitudinal and Survival Data and Related Topics
Type: Contributed
Date/Time: Tuesday, July 30, 2019 : 2:00 PM to 3:50 PM
Sponsor: Lifetime Data Science Section
Abstract #306699 Presentation
Title: Bayesian Joint Models for Longitudinal and Competing Risks Data
Author(s): Allison Furgal* and Ananda Sen and Jeremy Taylor
Companies: University of Michigan Biostatistics and University of Michigan and University of Michigan
Keywords: joint model; competing risks; Bayesian; nonparametric prior
Abstract:

Joint models for longitudinal and time-to-event data are useful in situations where an association exists between a longitudinal marker and an event time. Much of the research in this area has been carried out for time-to-event data that arise from a single source of failure. Joint modeling of longitudinal and competing risks data is a subject of ongoing research. In this context, methodology have been developed almost exclusively using cause-specific hazard functions. Such modeling fails to capture the explicit effect of association between the longitudinal marker and the risk components. In this talk, we explore Bayesian joint models within a latent failure time framework. We investigate various parametric models within the log-location-family of distributions under different association structures. Flexibility in modeling is added via the incorporation of nonparametric Dirichlet process priors. We evaluate our method via simulations and illustrate the approach with an application to data from cirrhosis patients measuring both biological markers over time and time to death or transplant.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program