|
Activity Number:
|
257
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, August 10, 2004 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #300053 |
|
Title:
|
Experiences with Interim Analyses and DSMBs--a CRO Perspective
|
|
Author(s):
|
John C. Conlon*+
|
|
Companies:
|
Covance, Inc.
|
|
Address:
|
Clinical Development Services, Princeton, NJ, 08540-6233,
|
|
Keywords:
|
O'Brien-Fleming ; Lan-DeMets ; Lan-Trost ; power reassessment ; Type I error ; interim analysis
|
|
Abstract:
|
Data management and biometrics support (including interim analyses) for DSMBs forms an integral part of life within our CRO. Challenges can be administrative as well as technical, and development of statistical models can be paramount to both. This paper illustrates some of the experiences and challenges encountered whilst planning and analyzing two large phase III oncology studies. Administrative challenges discussed include the organization of DSMB meetings at times to match required numbers of endpoints, development of DSMB communication plans and contingency planning for all outcome eventualities. Technical statistical challenges from the two studies include analysis using O'Brien Fleming implementation of Lan & DeMets procedures and the implementation of a conditional power reassessment according to Lan & Trost (1997). The latter covers not only the establishment of futility analysis criteria and the determination of the actual Type I error rate, but also the discussions with the FDA in getting the procedure approved.
|
- 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 2004 program |
