A new computer model to select who is invited to clinical trials will ensure that participants are more diverse than usual.
People who enrol in health research studies such as cancer screening trials usually have better health than the target population (the "healthy volunteer” effect). A new paper led by Adam Brentnall (WIPH), in collaboration with colleagues from King’s College London, describes a computer algorithm developed to help target trial invitations to achieve more equitable representation. The new model tilts clinical trial participation invitations towards groups less likely to join, to ensure that enough people are included in the trial from all major societal and ethnic groups.
The model was tested in practice in recruitment to the NHS-Gallieri multi-cancer screening trial. Over 11 months ≈1.5 million invitations were sent to recruit 140,000 people, making it likely to be the fastest recruiting trial of this size ever. This method of inviting participants for clinical trials addresses the “healthy volunteer” effects and inequity that exist in many health research studies, and can be adapted for use in other trials or research studies.
Funding for this study was received from Cancer Research UK, including funding awarded to CRUK by GRAIL LLC.
Brentnall AR, Mathews C, Beare S, Ching J, Sleeth M, Sasieni P. Dynamic data-enabled stratified sampling for trial invitations with application in NHS-Galleri. Clinical Trials. 2023;0(0). doi:10.1177/17407745231167369