SMART MS fRCT Anonymised Dataset

<b>Background: </b>Cognitive difficulties are common in multiple sclerosis (MS) and are associated with poorer quality of life, reduced independence, and lower employment. Despite growing interest in cognitive rehabilitation, existing interventions often lack generalisability and accessibility. Strengthening Mental Abilities with Relational Training (SMART) is a digital intervention targeting relational skills, which are theorised to underpin broader cognitive functioning. This trial evaluated the feasibility, acceptability, and preliminary signals of efficacy of SMART for people with MS experiencing cognitive difficulties.<b>Methods: </b>We conducted a three-arm parallel-group feasibility randomised controlled trial. Participants with a diagnosis of MS and self-reported cognitive problems were recruited from three NHS sites and randomised (1:1:1) to SMART plus treatment-as-usual (TAU), Active control (Sudoku) plus TAU, or TAU alone. Primary feasibility outcomes included recruitment, retention, intervention adherence, and data completeness. Acceptability was assessed using intervention-usage data and participant feedback. Candidate outcome measures included the Perceived Deficits Questionnaire (PDQ), Symbol Digit Modalities Test (SDMT), Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), EQ-5D-5L, and ICECAP-A. Assessments took place at baseline, 3 months, and 6 months. The study was co-developed with people living with MS and their families through embedded patient and public involvement and engagement processes.<b>Results:</b> A total of 43 participants were randomised (72% of target). Retention was 84% at 3-month follow-up and 77% at 6 months. SMART adherence was high, with 93% of participants meeting the predefined criterion for training adherence. Acceptability feedback was consistently positive across arms. Data completeness for self-report and performance outcomes was acceptable (typically 60–80% across measures), though return rates for RBANS were lower due to postal logistics. Exploratory analysis suggested a moderate effect of SMART on perceived cognitive difficulties (PDQ, Cohen’s <i>d</i> = 0.85 vs Active control; 0.69 vs TAU), with confidence intervals supporting the possibility of a clinically meaningful benefit. No clear between-group signal was observed on the SDMT.<b>Conclusions: </b>Trial procedures were feasible, and the SMART intervention was acceptable to people with MS. Findings support progression to a fully powered trial, with minor refinements to recruitment and assessment processes.