Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 231 - SPEED: SPAAC SESSION I
Type: Topic-Contributed
Date/Time: Wednesday, August 11, 2021 : 10:00 AM to 11:50 AM
Sponsor: Biometrics Section
Abstract #317904
Title: A Comparison of Two Approaches for Dynamic Prediction: Joint Modeling and Landmark Modeling
Author(s): Wenhao Li* and Liang Li
Companies: The University of Texas MD Anderson Cancer Center UT Health Graduate School and The University of Texas MD Anderson Cancer Center
Keywords: Dynamic prediction; Joint model of longitudinal data and survival data; Landmark analysis; Partly conditional model; Static prediction model; Survival analysis
Abstract:

The commonly used statistical approaches to dynamic prediction can be broadly classified into joint modeling of longitudinal and survival data, and landmark modeling. It is an important research question to understand which approach can produce more accurate prediction. There were few previous researches on this topic and they were conducted in the scenario where the data were simulated from the joint model. As a result, the landmark model was estimated as a working model under misspecification, which causes difficulty in interpreting the comparison. In this paper, we present the comparison under two scenarios, where the data were simulated from either the joint model or the landmark model. The latter scenario was made feasible by using a recently proposed algorithm to simulate longitudinal data under the assumption of landmark models. Our simulation results demonstrate that the relative performance of these two modeling approaches depends on the data settings and one does not always dominate the other. We also present a comparison of these two dynamic prediction approaches with static prediction, and discuss situations where dynamic prediction is likely to be advantageous.


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

Back to the full JSM 2021 program