|
Activity Number:
|
256
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #304998 |
|
Title:
|
The Use of Regression Models in Adjustment of Treatment Assignment
|
|
Author(s):
|
Valerii V. Fedorov*+
|
|
Companies:
|
GlaxoSmithKline
|
|
Address:
|
1250 S Collegeville Rd, Collegeville, PA, 19426,
|
|
Keywords:
|
design of controlled clinical trials ; minimization ; biased coin designs
|
|
Abstract:
|
The optimization of treatment assignment can be viewed as an optimal design problem in the presence of uncontrolled covariates. This problem was partially addressed in optimal design theory in the early eighties but never was directly applied in the controlled clinical trials. I explore the connection of the approach with a couple of currently popular methods: "minimization" and "optimum biased coin designs" and show that all three lead to the very similar final results only for the simplest linear regression model. Otherwise patient allocations may have rather different.
|
