Internet Explorer is no longer supported by Microsoft. To browse the NIHR site please use a modern, secure browser like Google Chrome, Mozilla Firefox, or Microsoft Edge.

Studies within a trial (SWAT) and studies within a review (SWAR)


26 May 2022




Studies within a trial (SWAT) and studies within a review (SWAR)

Research studies play a central role in improving clinical care in the UK and are a key component of many clinical practice guidelines developed by NICE and others, but such studies can be inefficient. Reasons for study inefficiencies include the complex and unpredictable nature of some trials/studies, but more commonly, the lack of evidence to support processes and decision-making,

To increase the evidence base, we encourage researchers to embed a study evaluating alternative ways of undertaking a study (e.g. public involvement, trial recruitment, retention, data collection, evidence synthesis methodology etc) to provide evidence about how we can improve the way we perform that process. We welcome proposals that include embedded studies aimed at improving the successful completion of studies.

The findings of SWAT/SWAR are anticipated to be generalisable to other studies but may also be of specific benefit to the host trial itself, by supporting trial delivery process decision points. SWAT’s/SWAR's relating to recruitment and any other aspect of study delivery are within remit. Applicants who wish to include a SWAT/SWAR in their application should read the relevant sections of the first stage guidance notes for further details.

Supporting information

If you are considering undertaking a SWAT in your trial, the PROMETHEUS team would be most happy to provide support with setting up the SWAT. If this would be useful, please go to the PROMETHEUS website for further information

PRioRiTy I and PRioRiTy II used James Lind Alliance methodology to work with patients, carers, health professionals, researchers and methodologists to set research priorities around recruitment and retention to clinical trials.