Abstract

Multi-arm trials allow testing of several experimental treatments against a shared control group. They provide statistical and non-statistical benefits compared to conducting separate randomized controlled trials of each treatment. They are therefore particularly useful for tumor types where there is more than one treatment available for testing. Going further, a multi-arm, multi-stage (MAMS) trial introduces interim analyses, which allow accruing data to be used to drop ineffective treatments early. MAMS trials can provide additional efficiency and ethical benefits beyond those provided by multi-arm trials without interim analyses. In this chapter, methods for the statistical design of multi-arm and MAMS are described. We examine error rates and how to control them. We also look at different types of statistical power and how to choose a sample size such that the type I error rate and power is controlled at a specified level. For MAMS designs, we examine two distinct types, group-sequential and drop-the-loser MAMS trials, with practical guidance provided for how to design them.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call