Two-stage designs for cross-over bioequivalence trials
Standard
Two-stage designs for cross-over bioequivalence trials. / Kieser, Meinhard; Rauch, Geraldine.
In: STAT MED, Vol. 34, No. 16, 20.07.2015, p. 2403-2416.Research output: SCORING: Contribution to journal › SCORING: Journal article › Research › peer-review
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - JOUR
T1 - Two-stage designs for cross-over bioequivalence trials
AU - Kieser, Meinhard
AU - Rauch, Geraldine
N1 - Copyright © 2015 John Wiley & Sons, Ltd.
PY - 2015/7/20
Y1 - 2015/7/20
N2 - The topic of applying two-stage designs in the field of bioequivalence studies has recently gained attention in the literature and in regulatory guidelines. While there exists some methodological research on the application of group sequential designs in bioequivalence studies, implementation of adaptive approaches has focused up to now on superiority and non-inferiority trials. Especially, no comparison of the features and performance characteristics of these designs has been performed, and therefore, the question of which design to employ in this setting remains open. In this paper, we discuss and compare 'classical' group sequential designs and three types of adaptive designs that offer the option of mid-course sample size recalculation. A comprehensive simulation study demonstrates that group sequential designs can be identified, which show power characteristics that are similar to those of the adaptive designs but require a lower average sample size. The methods are illustrated with a real bioequivalence study example.
AB - The topic of applying two-stage designs in the field of bioequivalence studies has recently gained attention in the literature and in regulatory guidelines. While there exists some methodological research on the application of group sequential designs in bioequivalence studies, implementation of adaptive approaches has focused up to now on superiority and non-inferiority trials. Especially, no comparison of the features and performance characteristics of these designs has been performed, and therefore, the question of which design to employ in this setting remains open. In this paper, we discuss and compare 'classical' group sequential designs and three types of adaptive designs that offer the option of mid-course sample size recalculation. A comprehensive simulation study demonstrates that group sequential designs can be identified, which show power characteristics that are similar to those of the adaptive designs but require a lower average sample size. The methods are illustrated with a real bioequivalence study example.
KW - Biostatistics
KW - Computer Simulation
KW - Cross-Over Studies
KW - Delayed-Action Preparations
KW - Humans
KW - Models, Statistical
KW - Randomized Controlled Trials as Topic
KW - Research Design
KW - Sample Size
KW - Theophylline
KW - Therapeutic Equivalency
KW - Journal Article
U2 - 10.1002/sim.6487
DO - 10.1002/sim.6487
M3 - SCORING: Journal article
C2 - 25809815
VL - 34
SP - 2403
EP - 2416
JO - STAT MED
JF - STAT MED
SN - 0277-6715
IS - 16
ER -