|
Activity Number:
|
609
- New Approaches to Improving Accuracy, Precision, and Robustness of Survival Analysis
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 1, 2019 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #304516
|
Presentation
|
|
Title:
|
Predicting Events from Longitudinal Data: The Imputed Cox Model
|
|
Author(s):
|
James Troendle* and Eric Leifer and Xin Tian
|
|
Companies:
|
National Institutes of Health and National Heart,Lung and Blood Institute and National Heart, Lung and Blood Institute, National Institutes of Health
|
|
Keywords:
|
Hazard;
Imputation;
Measurement Error;
Time-Varying Covariate
|
|
Abstract:
|
Consider making predictions of future event probabilities based on the history of a repeatedly measured marker process. The general problem is to predict the probability of an event in a specified amount of time for an individual who has survived up to time s, given their marker process up to time s. This general problem is similar to that faced by a physician confronted by a subject at the current time. We suppose the prediction can be informed from a past database of similar subject's survival and marker process histories. Our approach is to use a very simple longitudinal model to impute expected marker values at various desired times. Then a Cox model is fit with time-varying covariates created from the imputed biomarker data. A simple bootstrap of the whole procedure is used to estimate the variance of the probability estimate for inference. We evaluate the proposed estimator via simulation from data generated from a standard joint model used in the literature.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2019 program
|
