Activity Number:
|
280
- New Methodology Developments in Analyzing Complex Survival Data
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 9, 2022 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Lifetime Data Science Section
|
Abstract #323003
|
|
Title:
|
Partial-Linear Single-Index Semiparametric Transformation Models with Censored Data
|
Author(s):
|
Myeonggyun Lee* and Andrea B. Troxel and Mengling Liu
|
Companies:
|
New York University Grossman School of Medicine and New York University Grossman School of Medicine and New York University Grossman School of Medicine
|
Keywords:
|
B-spline smoothing;
EM algorithm;
nonparametric maximum likelihood estimation;
partial-linear single-index model;
semiparametric transformation model
|
Abstract:
|
In studies with time-to-event outcomes, multiple and correlated time-dependent covariates are commonly observed, and it is of great interest to model their joint effects by allowing a flexible functional form and to delineate their relative contributions to survival risk. A class of semiparametric transformation (ST) models offers flexible specifications of the intensity function and can be considered as a general framework to further accommodate nonlinear effects of covariates. Motivated by electronic health record data on COVID-19 patients, we propose a partial-linear single-index ST model. The proposed method reduces the dimensionality of multiple covariates and provides interpretable estimates of the covariate effects. We develop an iterative algorithm using the regression spline technique to model the nonparametric single index component for possibly nonlinear joint effects, followed by nonparametric maximum likelihood estimation of parameters. We propose a nonparametric test for examining the linearity of covariates. We present Monte Carlo simulation studies to compare the new method with the standard ST model and apply our proposed method to the COVID-19 de-identified data.
|
Authors who are presenting talks have a * after their name.