Multiplicity issues for platform trials with a shared control arm

This paper provides in-depth discussion about different types of error generated in platform trials with a common control arm, and how they compare to the ones arisen from standard independent trials. We provide our views on some of the popular "myths" associated with such design, under the frequentist framework. It is found that platform trial generally performs quite well in terms of type I error rate, false discovery rate, and power. In most cases, these operating characteristics of a platform trial are comparable to or even better than running individual trials.