Optimising Trial Conduct and Generating Translational Gains with Studies Within a Trial (SWATs): A Case Study from the MEL-SELF Randomised Controlled Trial of Patient-Led Surveillance

Abstract

Randomised controlled trials (RCTs) are the gold standard for evaluating healthcare interventions, but inefficiencies in trial conduct and delays in uptake of effective interventions create waste and widen the evidence-to-practice gap. Studies Within A Trial (SWATs) are embedded sub-studies used to improve trial processes such as recruitment and retention, but their potential to support implementation is underused. Melanoma surveillance is a useful test case because patients require ongoing self-monitoring, increasingly supported by teledermatology, which requires sustained engagement and effective implementation. This thesis examined how SWATs can improve trial efficiency and inform implementation, using the MEL-SELF RCT as a case study.

Two scoping reviews found substantial inconsistency in how melanoma surveillance RCTs reported recruitment, retention, outcome completion, and adherence, with limited evaluation of strategies to improve these processes. Four SWATs were embedded in MEL-SELF. A baseline survey identified participant motivations for enrolment, including altruism, personal benefit, and trust in their clinician. A behavioural analysis using the COM-B model identified barriers to skin self-examination and image submission, including low self-efficacy, poor planning, time constraints, technical difficulties, and preference for clinician-led care, and was used to guide adherence support. A randomised SWAT found that motivation-informed reminders did not improve completion of low-complexity trial tasks or more demanding image submissions, suggesting that burden and usability were more important than reminder wording. Another SWAT comparing dermatoscopes found similar clinical usefulness across devices, but better image quality and fewer support needs with the higher-cost polarised device. Overall, this thesis shows that multiple SWATs can be embedded within a single RCT to improve trial conduct and inform implementation of complex interventions.