Recent developments in group-sequential designs

Wason, JMS

doi:10.1007/978-3-642-55345-5_6

Recent developments in group-sequential designs

Lookup NU author(s): Professor James Wason ORCiD

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

© 2014 Springer-Verlag Berlin Heidelberg. All rights are reserved.In a group-sequential trial, patients are recruited in groups, and their response to treatment is assessed. After each group is assessed, an interim analysis is conducted. At each interim analysis, the trial can stop for futility, stop for efficacy, or continue. The main advantage of group-sequential designs is that the expected number of patients is reduced compared to a design without interim analyses. There are infinitely many possible group-sequential designs to use, and the choice strongly affects the operating characteristics of the trial. This chapter discusses optimal and admissible group-sequential designs. Optimal designs minimise the expected sample size at some specified treatment effect; admissible designs optimise a weighted sum of trial properties of interest, such as expected sample size and maximum sample size. Methods for finding such designs are discussed, including a detailed description of an R package that implements a quick search procedure. Recent applications of group-sequential methodology to trials with multiple experimental treatments being tested against a single control treatment are also described.