ABSTRACT
In a clinical trial with multiple coprimary endpoints, the efficacy require-
ment is defined as meeting two or more endpoints simultaneously. There-
fore, in an adaptive design of multiple endpoint trial, we have to deal with
the multiplicity due to multiple endpoints and multiple analyses at differ-
ent times. We will use mathematical induction to prove that the stopping
boundaries from group sequential stopping boundaries with a single end-
point can be use directly without inflation or deflation the type-I error. We
will also derive the formulations for conditional power and power for group
sequential design with multiple endpoints, which is an extension of the con-
ditional power for single endpoint developed by Lan and Wittes (1988). The
extension of the method to sample-size reestimation design is also provided
with illustrative examples (Cheng, 2014; Cheng, Menon, and Chang, 2014).
13.1 Introduction