TY - JOUR
T1 - Multiple testing procedures for adaptive enrichment designs
T2 - Combining group sequential and reallocation approaches
AU - Rosenblum, Michael
AU - Qian, Tianchen
AU - Du, Yu
AU - Qiu, Huitong
AU - Fisher, Aaron
N1 - Publisher Copyright:
© 2016 The Author 2016.
PY - 2016/10/1
Y1 - 2016/10/1
N2 - Adaptive enrichment designs involve preplanned rules for modifying enrollment criteria based on accrued data in an ongoing trial. For example, enrollment of a subpopulation where there is sufficient evidence of treatment efficacy, futility, or harm could be stopped, while enrollment for the remaining subpopulations is continued. We propose a new class of multiple testing procedures tailored to adaptive enrichment designs. The procedures synthesize ideas from two general approaches. As in the modified group sequential approach, the procedures gain power by leveraging the covariance among statistics for different stages and different hypotheses. As in the alpha reallocation approach, the procedures lower rejection thresholds for the remaining null hypotheses after others have been rejected. The proposed procedures are proved to have power greater than or equal to several existing methods, and to strongly control the familywise Type I error rate when statistics are normally distributed. The methods are illustrated through simulations of a trial for a surgical intervention for stroke, involving two subpopulations.
AB - Adaptive enrichment designs involve preplanned rules for modifying enrollment criteria based on accrued data in an ongoing trial. For example, enrollment of a subpopulation where there is sufficient evidence of treatment efficacy, futility, or harm could be stopped, while enrollment for the remaining subpopulations is continued. We propose a new class of multiple testing procedures tailored to adaptive enrichment designs. The procedures synthesize ideas from two general approaches. As in the modified group sequential approach, the procedures gain power by leveraging the covariance among statistics for different stages and different hypotheses. As in the alpha reallocation approach, the procedures lower rejection thresholds for the remaining null hypotheses after others have been rejected. The proposed procedures are proved to have power greater than or equal to several existing methods, and to strongly control the familywise Type I error rate when statistics are normally distributed. The methods are illustrated through simulations of a trial for a surgical intervention for stroke, involving two subpopulations.
KW - Multiple testing procedure
KW - Treatment effect heterogeneity
UR - http://www.scopus.com/inward/record.url?scp=84995654457&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84995654457&partnerID=8YFLogxK
U2 - 10.1093/biostatistics/kxw014
DO - 10.1093/biostatistics/kxw014
M3 - Article
C2 - 27006374
AN - SCOPUS:84995654457
SN - 1465-4644
VL - 17
SP - 650
EP - 662
JO - Biostatistics
JF - Biostatistics
IS - 4
ER -