Developing and evaluating a complex intervention in stroke: using very early mobilisation as an example

Background: Complex interventions, those that incorporate multiple interacting components, are difficult to define, measure and implement. The aim of this research was to develop and evaluate the complex intervention, very early mobilisation (VEM) in acute stroke care. The clinical effectiveness and cost-effectiveness of VEM were evaluated whilst simultaneously considering the implications for future implementation. Methods: A mixed methods approach was used: systematic review, predictive modelling, observational study design, individual patient data meta-analysis, qualitative methods and economic evaluation. Statistical models to accurately predict mobility post-stroke were developed. A multicentre observational study was conducted to establish pre-implementation activity levels of acute stroke patients. Data from two completed and comparable feasibility trials were used to estimate the clinical and economic impact of VEM. A qualitative process evaluation was conducted to identify the barriers and facilitators to implementing VEM, if shown to be effective. Results: Two predictive models were developed with age and stroke type common factors to both. Pre-implementation activity levels were low. Patients who underwent VEM were 3-times more likely to be independent at 3 months than were standard care (SC) patients. The incremental cost-effectiveness ratio associated with VEM in comparison to SC indicated VEM to be potentially cost-effective from a societal perspective. Barriers and facilitators identified for each stage of the stroke pathway and a set of HCPs’ beliefs towards VEM were formulated. Conclusions: This research has adhered to current guidance provided by the Medical Research Council to develop and evaluate VEM. The clinical effectiveness and cost-effectiveness of VEM were estimated. The ongoing A Very Early Rehabilitation Trial phase III will provide definitive evidence for the effectiveness of VEM and the wider consequences for stroke care. This research has provided the support and the foundations for the development of a clear implementation strategy for VEM