|
Activity Number:
|
115
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
WNAR
|
| Abstract - #304882 |
|
Title:
|
Statistical Considerations and Implications of Early Stopping in the JUPITER Trial
|
|
Author(s):
|
Robert J. Glynn*+ and Paul M. Ridker
|
|
Companies:
|
Brigham and Women's Hospital and Brigham and Women's Hospital
|
|
Address:
|
900 Commonwealth Avenue, Boston, MA, 02215,
|
|
Keywords:
|
clinical trial ; early stopping ; Lan-DeMets approach ; O'Brien-Fleming boundaries
|
|
Abstract:
|
Justification for the Use of Statins: an Intervention Trial Evaluating Rosuvastatin (JUPITER) was a primary prevention trial designed to continue until accrual of 520 confirmed cardiovascular events (PM Ridker et al. N Engl J Med 2008). The trial was stopped early, with a median treatment duration of 1.9 (maximum 5.0) years, based on determination by its independent data monitoring board that the interim data constituted proof of efficacy beyond a reasonable doubt that would likely influence clinicians' treatment of study subjects. We describe here the pre-trial basis for the trial's expected size, the formal statistical monitoring plan with O'Brien-Fleming stopping boundaries based on the Lan-DeMets approach, and the additional considerations by the board in its decision to terminate. We also discuss the implications of early stopping on estimation and testing of the treatment effect.
|
