Factorial trials involve randomising the same participants to more than one treatment (or factor). Our work focuses on how to design, analyse, and report factorial trials, as well as how to decide whether a factorial design is most appropriate. 

What are factorial trials and why are they relevant?

Factorial trials involve randomising participants more than once, for instance to treatment A or not A, and again to treatment B or not B. The main reason factorial trials are used is because they allow researchers to answer two distinct research questions in a single trial, without much (or any) increase to the sample size. They can therefore be much more efficient than doing separate trials for each comparison or running multi-arm trials.

What have we done?

Our work focuses on how factorial trials should be run to best maximise their benefits. We have:

  • Developed extensions of the CONSORT and SPIRIT reporting guidelines for factorial trials
  • Described a comprehensive strategy for defining estimands, methods of estimation, and sensitivity analyses for factorial trials
  • Produced guidance on when it is appropriate to use a factorial trial
  • Described how a multi-arm multi-stage trial design can be used within a factorial design
  • Evaluated the quality of the design, analysis, and reporting of factorial trials in the medical literature
  • Identified flaws in some common methods of analysis.

How will this make a difference?

Our work provides guidance on how factorial trials can be used in practice. Improvements in the design and analysis of factorial trials can lead to increases in efficiency and interpretability, ultimately leading to quicker identification of effective treatments.