Multiplicity Adjustments for Clinical Trials with Multiple Doses in Two-dose Regimens

• Activity Number: 405
• Type: Contributed
• Date/Time: Thursday, August 12, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Peng-Liang Zhao*+ and Hui Quan
• Companies: Merck & Co., Inc. and Merck & Co., Inc.
• Keywords: familywise error rate ; closed procedure ; strong control ; weak control

Abstract

Typically there is no prior information regarding which dose regimen of an active compound might be more effective or better, therefore two regimens (e.g., once-daily and twice-daily doses) can be considered in a phase II dose-range finding study. After the phase II study, if one regimen (e.g., once-daily dose) is numerically less efficacious than the other regimen, but is more convenient for patients than the other regimen and is also likely to be registrable, multiple doses in both regimens can be included in phase III confirmatory trials and be compared to the control. We discuss multiplicity adjustment procedures for two regimens and multiple doses in these trials. We consider procedures which use the dose-response relationship of each regimen to form the closed procedures and control the familywise Type I error rate in a strong sense. These procedures generally have a higher power than the traditional procedures such as the Bonferroni and the Hochberg procedures for the high doses. We also discuss procedures which control the familywise Type I error rate in a weak sense. Numerical examples will be used to illustrate the applications of our procedures.