Premium
Assessing non‐inferiority of a new treatment in a three‐arm trial in the presence of heteroscedasticity
Author(s) -
Hasler Mario,
Vonk Richardus,
Hothorn Ludwig A.
Publication year - 2007
Publication title -
statistics in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.996
H-Index - 183
eISSN - 1097-0258
pISSN - 0277-6715
DOI - 10.1002/sim.3052
Subject(s) - heteroscedasticity , homogeneous , econometrics , statistics , computer science , randomized controlled trial , medicine , mathematics , combinatorics
In this paper, we describe an adjusted method to facilitate non‐inferiority tests in a three‐arm design. While the methodology is readily available in the situation of homogeneous group variances, the adjusted method will also maintain the α‐level in the presence of heteroscedasticity. We propose explicit criteria for an optimal allocation. Depending on the pattern of heterogeneity, remarkably unbalanced designs are power optimal. We will apply the method to a randomized clinical trial and a toxicological experiment. Copyright © 2007 John Wiley & Sons, Ltd.