|
Activity Number:
|
322
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #304217 |
|
Title:
|
Simon's Two-Stage Design for Phase II Cancer Clinical Trials with Two Distributions as Hypotheses
|
|
Author(s):
|
Junfeng Liu*+ and Yong Lin and Weichung J. Shih
|
|
Companies:
|
University of Medicine and Dentistry of New Jersey and University of Medicine and Dentistry of New Jersey and University of Medicine and Dentistry of New Jersey
|
|
Address:
|
195 Little Albany Street, Room 5539, New Brunswick, NJ, 08901,
|
|
Keywords:
|
(Incomplete) Beta distribution ; (Incomplete) Beta function ; prior predictive probability ; Simon's two-stage design ;
|
|
Abstract:
|
Clinical trials are usually designed to control two types of error rates from rejecting a more effective new drug and/or overpraising an ineffective new drug. Given the historical drug response rate (less than a hypothesized value p0) and the expected new drug response rate (greater than a value p1 (> p0)), error rates are controlled below any specified levels by recruiting enough patients. This talk discusses a scenario where hypotheses (the historical and expected new drug response rates) are two continuous distributions due to information uncertainty. For different settings, we show design results in terms of sample size and decision region. We also show some results regarding design existence from both practical and theoretical perspectives.
|
