Background and Purpose--Stroke care is complex, requiring input from professionals, patients and carers. Identifying and developing appropriate intervention components to meet these complex needs is difficult. The Medical Research Council (MRC) Framework for developing and evaluating ‘complex’ (nonpharmacological) interventions aims to improve intervention development. This study uses the Framework to review complex interventions in stroke care.

Methods--Systematic review with multiple search strategies (electronic databases, recent journals, gray literature) was used. The MRC Framework was used to guide the search strategy and assess study quality. ‘Complex interventions’ were defined as educational/psychosocial interventions to change knowledge, beliefs or behaviors.

Results--Sixty-seven studies were included: 39 randomized controlled trials (RCT) and 28 other designs. Complex interventions targeted healthcare professionals (17), and patients, carers and the general population (21 targeting primary or secondary prevention; 30 targeting adjustment and recovery after stroke). Compared with recovery studies, primary and secondary prevention studies were significantly less likely to have been evaluated in RCTs. Interventions evaluated in RCTs were significantly less likely to influence primary outcomes (26%) compared with other designs (44%). Theoretical grounding to support intervention choice was reported in 40 studies but only 14 were theoretically ‘well developed’; 21 RCTs listed multiple primary outcome measures, with 10 listing 5 or more. Of these only 3 reported considering statistical power before recruitment and none was sufficiently powered.

Conclusion--Few complex interventions in stroke care have been adequately developed or evaluated. This may explain failures to demonstrate efficacy. In future, greater attention is needed to theoretical development and methodological quality.