|
Activity Number:
|
380
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 11, 2004 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #301529 |
|
Title:
|
Cox Regression Methods for Two-stage Randomization Designs
|
|
Author(s):
|
Yuliya Lokhnygina*+ and Jeffrey Helterbrand
|
|
Companies:
|
North Carolina State University and Genentech, Inc.
|
|
Address:
|
Dept. of Statistics, Raleigh, NC, 27695,
|
|
Keywords:
|
two-stage randomization designs ; proportional hazards ; inverse probability weighting ; clinical trials
|
|
Abstract:
|
Two-stage randomization designs are becoming common in oncology and AIDS clinical trials. In these designs patients are initially randomized to an induction treatment, followed by randomization to a maintenance treatment conditional on their remission and consent to further participation in the trial. The goal of the study is usually the comparison of different combinations of induction and maintenance therapies to find the most beneficial treatment policy. Recently, Lunceford, Davidian, and Tsiatis (2002) and Wahed and Tsiatis (2003) developed some nonparametric approaches to the analysis of such designs. Our analysis methods are based on the Cox proportional hazards model. We consider a special situation where at the second stage patients can be randomized to either observation or active drug, and the primary objective is to compare the induction therapies with respect to a time-to-event endpoint. We propose reweighted versions of the usual score estimating equation and the score test in the Cox model. Large sample properties are derived and illustrated via a simulation study. Application of the proposed methods is demonstrated on the example of E4494 clinical trial.
|
