|
Activity Number:
|
469
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #308408 |
|
Title:
|
Power of Nonparametric Logrank and Wilcoxon Tests with Adjustment for Covariates: A Simulation Study
|
|
Author(s):
|
Honghua Jiang*+ and James Symanowski and Sofia Paul and Yongming Qu and Anothy Zagar and Shengyan Hong
|
|
Companies:
|
Eli Lilly and Company and Nevada Cancer Institute and Novartis Pharmaceuticals and Eli Lilly and Company and Eli Lilly and Company and Eli Lilly and Company
|
|
Address:
|
14421 Chariots Whisper Dr, Westfield, IN, 46074,
|
|
Keywords:
|
Time-to-event variables ; Cox proportional hazards model ; analysis of covariance
|
|
Abstract:
|
Time-to-event variables are common primary and secondary outcomes for oncology trials. Usual methods of analysis for these endpoints involve product-limit estimators of cumulative survival for treatments, logrank tests to compare treatment groups and Cox proportional hazards model with treatment and potential prognostic covariates. Adjustment for covariates reduces bias and may increase precision and power. However, the appropriateness of Cox proportional hazards model depends on parametric assumptions. One way to address this issue is to use nonparametric analysis of covariance (Tangen & Koch 1999). Simulation studies are carried out to investigate the power of nonparametric tests with adjustment of covariates vs. nonparametric tests without adjustment of covariates. A comparison between adjusted and unadjusted methods is also illustrated with an oncology clinical trial example.
|
