|
Activity Number:
|
74
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #307334 |
|
Title:
|
Nonparametric Comparison of Two Survival Functions with Dependent Censoring via Nonparametric Multiple Imputation
|
|
Author(s):
|
Chiu-Hsieh Hsu*+ and Jeremy M. G. Taylor
|
|
Companies:
|
University of Arizona and University of Michigan
|
|
Address:
|
1515 N. Campbell, Room 2942, Tucson, AZ, 85724-5024,
|
|
Keywords:
|
dependent censoring ; multiple imputation ; logrank test
|
|
Abstract:
|
When the event time depends on censoring time, the conventional two-sample test method could produce an invalid test. We extend our previous work in survival estimation to propose a multiple imputation approach to using auxiliary variables to adjust for dependent censoring while comparing two survival functions. To conduct the imputation, we use two working PH models to define an imputing risk set. One is for the event times and the other for the censoring times. Based on the imputing risk set, a nonparametric multiple imputation method is used to impute a future event or censoring time for each censored observation. Simulation studies show that the sizes of the log-rank and Wilcoxon test constructed on the imputed datasets derived from the bootstrap samples are comparable to the nominal level in the presence of dependent censoring if either one of the two working models is correct.
|
