|
Activity Number:
|
277
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:30 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #309342 |
|
Title:
|
Five-Arm Dose Selection Trial with Adaptive Randomization
|
|
Author(s):
|
Mark F. Munsell*+ and Donald A. Berry and J. Kyle Wathen and Marcos J. de Lima
|
|
Companies:
|
The University of Texas M.D. Anderson Cancer Center and The University of Texas M.D. Anderson Cancer Center and The University of Texas M.D. Anderson Cancer Center and The University of Texas M.D. Anderson Cancer Center
|
|
Address:
|
1515 Holcombe Blvd, Unit 447, Houston, TX, 77030,
|
|
Keywords:
|
adaptive randomization ; dose-selection
|
|
Abstract:
|
We describe a dose selection trial with a placebo arm and four active arms, each a different dose of the same agent. Doses too low or too high will not be effective. The study outcome is treatment success. Patients are at high or low risk of treatment failure. The study is planned with a maximum of 150 patients. Patients are randomized to the five arms, with patients assigned to placebo with probability 0.20. We find the probability that the treatment success rate in each active arm is greater than the treatment success rate in the placebo arm, while considering risk of failure. These probabilities aid in adapting the randomization probabilities for the four active arms, making decisions whether to close accrual to one or more arms and selecting the best active arm. We will discuss operating characteristics and practical considerations of the design.
|
