Sample size re-estimation for confirmatory two-stage flexible multi-arm trial with normal outcomes

Yan Li, Guoqiao Wang, Jeff M. Szychowski

Research output: Contribution to journalArticlepeer-review


The need for efficient clinical trial designs has led to the development of flexible multi-arm designs. In this paper, we investigated sample size re-estimation (SSR) methods for fully flexible designs that allow for any unplanned treatment selection, early stopping for efficacy or futility. Even though SSR methods for two-arm trials are well developed, they are not directly applicable to certain fully flexible multi-arm trials due to the application of the closed test procedure. We derive SSR procedures based on the conditional power approach in the context of confirmatory multi-arm trials for three designs: inverse normal combination test, Fisher’s combination test and flexible group sequential design. We conduct extensive simulation studies to evaluate the performance of the three designs with and without SSR. Results show that the relative performance of the three designs depends on many factors. In practice, simulation studies are usually necessary to help determine the most appropriate design.

Original languageEnglish
Pages (from-to)157-178
Number of pages22
JournalJournal of Statistical Computation and Simulation
Issue number1
StatePublished - Jan 2 2020


  • Multi-arm
  • adaptive design
  • clinical trial
  • closed test
  • combination test
  • conditional error
  • conditional power
  • interim analysis
  • sample size re-estimation
  • treatment selection


Dive into the research topics of 'Sample size re-estimation for confirmatory two-stage flexible multi-arm trial with normal outcomes'. Together they form a unique fingerprint.

Cite this