Article Text

Download PDFPDF
Performing baseline testing in cluster randomised controlled trials
  1. Ushma Galal,
  2. Thomas R Fanshawe
  1. Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK
  1. Correspondence to Dr Thomas R Fanshawe, Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford OX1 2JD, UK; thomas.fanshawe{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

​Performing statistical tests of baseline characteristics is not usually advised for individually randomised trials. Should they be used for trials where participants are randomised in clusters?

In the design of clinical trials, the most important techniques used for reducing bias are randomisation and blinding. Randomisation introduces an element of chance to the assignment of interventions, ensuring a sound basis for any treatment differences found. The aim of blinding is to limit the occurrence of intentional and unintentional bias in the conduct and interpretation of the trial results.

The risk of such bias is present when knowledge of treatment assignments influences the recruitment and allocation of individuals to interventions. In individually randomised trials, even when blinding is not possible, randomisation is used and relied on to prevent allocation bias. Statistical testing of baseline measures between treatment groups is usually frowned on. The argument is that any apparent differences observed can only be due to …

View Full Text


  • Contributors UG and TRF identified the paper of interest. UG wrote the first draft, which TRF edited and added the verdict. Both authors approved the final version.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Commissioned; internally peer reviewed.