Non-parametric estimation in an illness-death model with component-wise censoring
Wednesday, July 21 at 3:00 pm
-
4:00 pm
Add to Calendar
2021-07-21 22:00:00
2021-07-21 23:00:00
Non-parametric estimation in an illness-death model with component-wise censoring
Department of Epidemiology & Biostatistics, Bioinformatics Presents:
Speaker: Anne Eaton, PhD, Assistant Professor of Biostatistics, Univ. of Minnesota
In clinical trials in serious disease settings, event-free survival is often used as the primary endpoint. But non-fatal events are often only detected at clinic visits, while time of death is known exactly. The endpoint therefore exhibits component-wise censoring. The standard method used to estimate event-free survival fails to account for component-wise censoring. We apply a kernel smoothing method in a novel way to produce a non-parametric estimator for event-free survival that accounts for component-wise censoring. We propose estimators for the probability in state and restricted mean time in state for illness-death models under component-wise censoring and derive their large-sample properties. Finally, we perform a simulation study to compare our method to existing methods.
Department Of Epidemiology & Biostatistics
Fei.Jiang@ucsf.edu
America/Los_Angeles
public
Department of Epidemiology & Biostatistics, Bioinformatics Presents:
Speaker: Anne Eaton, PhD, Assistant Professor of Biostatistics, Univ. of Minnesota
In clinical trials in serious disease settings, event-free survival is often used as the primary endpoint. But non-fatal events are often only detected at clinic visits, while time of death is known exactly. The endpoint therefore exhibits component-wise censoring. The standard method used to estimate event-free survival fails to account for component-wise censoring. We apply a kernel smoothing method in a novel way to produce a non-parametric estimator for event-free survival that accounts for component-wise censoring. We propose estimators for the probability in state and restricted mean time in state for illness-death models under component-wise censoring and derive their large-sample properties. Finally, we perform a simulation study to compare our method to existing methods.
Clinical Trials
Component-wise Censoring