|
Activity Number:
|
470
- Lifetime Risk, Competing Risk, and Recurrent Events
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Lifetime Data Science Section
|
|
Abstract #309792
|
|
|
Title:
|
A Statistical Framework for Time-to-Event Data with Exposure-Lag-Response Association and Its Application in Time-to-ICU Discharge
|
|
Author(s):
|
Yan Gao* and Sanjib Basu and Cheng Ouyang
|
|
Companies:
|
University of Illinois at Chicago and University of Illinois At Chicago and University of Illinois at Chicago
|
|
Keywords:
|
Survival analysis;
Exposure-lag-response association;
Cumulative effects;
Optimal latency;
High-dimensional variable selection;
Unobserved covariates
|
|
Abstract:
|
Most statistical methods for survival analysis assume the hazard rate at a specific time is impacted by either the baseline or the instantaneous value of covariates. But this framework is often violated when covariates demonstrate cumulative effect. For instance, workers develop cancers due to chronic exposure to toxic substances and ICU patients need accumulation of artificial nutrition to survive. Gasparrini et al. (2014, 2017) coined the term Exposure-Lag-Response Association (ELRA) for cumulative effect and extended the distributed lag models originally formulated by Almon (1965) in time series to the distributed lag non-linear models in survival data. However, little research in survival analysis with ELRA has identified the “optimal” latency, high-dimensional variable selection or asymptotic property. We propose an innovative framework to address these issues which aims to optimize data-driven time-varying windows of ELRA. We illustrate its application and comparison with existing methods in extensive simulation studies. We use our developed method to analyze time to ICU discharge data from a large cohort of heart failure patients in a premier hospital in Shanghai.
|
Authors who are presenting talks have a * after their name.
