Can we achieve better recruitment by providing better information? Meta-analysis of ‘studies within a trial’ (SWATs) of optimised participant information sheets

Vichithranie W. Madurasinghe, Peter Bower*, Sandra Eldridge, David Collier, Jonathan Graffy, Shaun Treweek, Peter Knapp, Adwoa Parker, Jo Rick, Chris Salisbury, Mei See Man, David Torgerson, Rebecca Sheridan, Frank Sullivan, Sarah Cockayne, Charlotte Dack

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)
4 Downloads (Pure)


Background: The information given to people considering taking part in a trial needs to be easy to understand if those people are to become, and then remain, trial participants. However, there is a tension between providing comprehensive information and providing information that is comprehensible. User-testing is one method of developing better participant information, and there is evidence that user-tested information is better at informing participants about key issues relating to trials. However, it is not clear if user-testing also leads to changes in the rates of recruitment in trials, compared to standard trial information. As part of a programme of research, we embedded ‘studies within a trial’ (SWATs) across multiple ongoing trials to see if user-tested materials led to better rates of recruitment. Methods: Seven ‘host’ trials included a SWAT evaluation and randomised their participants to receive routine information sheets generated by the research teams, or information sheets optimised through user-testing. We collected data on trial recruitment and analysed the results across these trials using random effects meta-analysis, with the primary outcome defined as the proportion of participants randomised in a host trial following an invitation to take part. Results: Six SWATs (n=27,805) provided data on recruitment. Optimised participant information sheets likely result in little or no difference in recruitment rates (7.2% versus 6.8%, pooled odds ratio = 1.03, 95% CI 0.90 to 1.19, p-value = 0.63, I2 = 0%). Conclusions: Participant information sheets developed through user testing did not improve recruitment rates. The programme of work showed that co-ordinated testing of recruitment strategies using SWATs is feasible and can provide both definitive and timely evidence on the effectiveness of recruitment strategies. 

Original languageEnglish
Article number218
Number of pages8
JournalBMC medicine
Issue number1
Early online date23 Sept 2021
Publication statusPublished - 23 Sept 2021

Bibliographical note

Acknowledgements: We would like to acknowledge all those in the trial teams who supported this programme of SWATs, including our public contributors (Ailsa Donnelly and Judith Hogg). We also thank Paul Wallace (original MRC START applicant) and Elizabeth Murray (Help Diabetes principal investigator).

Funding: The authors wish to acknowledge the MRC Methodology Research Programme which funds this research (MRC grant reference: G1002325). The MRC has no role in the study design; collection, management, analysis and interpretation of data; writing of the report; or the decision to submit the report for publication.


  • Information
  • Randomised controlled trial
  • Recruitment
  • Research methodology
  • SWATs
  • User-testing


Dive into the research topics of 'Can we achieve better recruitment by providing better information? Meta-analysis of ‘studies within a trial’ (SWATs) of optimised participant information sheets'. Together they form a unique fingerprint.

Cite this