|
Activity Number:
|
401
- Real-World Survival Data with Multiple Events: Challenges, Opportunities, and Recent Advancements
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 10, 2022 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Lifetime Data Science Section
|
|
Abstract #320500
|
|
|
Title:
|
Inference for Recurrent Event Processes Under Shape Heterogeneity
|
|
Author(s):
|
Yifei Sun* and Ying Sheng
|
|
Companies:
|
Columbia University and University of California San Fransisco
|
|
Keywords:
|
Counting process;
Dimension reduction;
Kernel smoothing;
Informative censoring
|
|
Abstract:
|
Recurrent event data are commonly encountered in longitudinal studies. Cox-type models such as the proportional rate model have been popular tools for analyzing recurrent event data. One of the key assumptions of the proportional rate model is that each covariate has a multiplicative effect on the rate function. The assumption facilitates straightforward interpretation of the covariate effects; on the other hand, it implies that covariates do not alter the shape of the conditional rate functions, which may not hold in general applications. We complement the existing literature by allowing the shape of the rate function to depend on a linear predictor via a functional single index model. We further allow the size of the rate function to depend on covariates through a pre-specified or unspecified link function. The proposed estimators are asymptotically normal at a root-n convergence rate. Simulation studies and a data example on cancer recurrence were conducted to illustrate the proposed methods.
|
Authors who are presenting talks have a * after their name.
