|
Activity Number:
|
133
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #309461 |
|
Title:
|
Crossing Monitoring Boundaries and Not Stopping
|
|
Author(s):
|
Janet Wittes*+
|
|
Companies:
|
Statistics Collaborative
|
|
Address:
|
1625 Mass Ave NW, Washington, DC, 20036,
|
|
Keywords:
|
randomized trials ; sequential analysis ; DSMB
|
|
Abstract:
|
A DSMB monitoring a randomized clinical trial uses prespecified boundaries as guidelines for its actions. If the trajectory of the data crosses the "upper" boundary, the trial can stop with a declaration of "success." Although DSMBs are aware that these boundaries are only advisory, they recognize that stopping without crossing leads to an inflated Type I error rate and can therefore jeopardize the scientific credibility of the results. On the other hand, failing to recommend stopping a trial when a boundary is crossed incurs no statistical penalty. DSMBs are reluctant to continue after crossing because the boundary reflects the Board's opinion prior to the start of the trial as to the data that it would find convincing evidence of benefit. In the context of actual examples, the paper discusses reasons a DSMB may not recommend stopping a trial even after crossing a boundary for benefit.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
