Informing the design of a randomised controlled trial of an exercise-based programme for long term stroke survivors: lessons from a before-and-after case series study

BACKGROUND: To inform the design of a randomised controlled trial (RCT) of an exercise-based programme for long term stroke survivors, we conducted a mixed methods before-and-after case series with assessment at three time points. We evaluated Action for Rehabilitation from Neurological Injury (ARNI), a personalised, functionally-focussed programme. It was delivered through 24 hours of one-to-one training by an Exercise Professional (EP), plus at least 2 hours weekly unsupervised exercise, over 12- 14 weeks. Assessment was by patient-rated questionnaires addressing function, physical activity, confidence, fatigue and health-related quality of life; objective assessment of gait quality and speed; qualitative individual interviews conducted with participants. Data were collected at baseline, 3 months and 6 months. Fidelity and acceptability was assessed by participant interviews, audit of participant and EP records, and observation of training. FINDINGS: Four of six enrolled participants completed the exercise programme. Quantitative data demonstrated little change across the sample, but marked changes on some measures for some individuals. Qualitative interviews suggested that small benefits in physical outcomes could be of great psychological significance to participants. Participant-reported fatigue levels commonly increased, and non-completers said they found the programme too demanding. Most key components of the intervention were delivered, but there were several potentially important departures from intervention fidelity. DISCUSSION: The study provided data and experience that are helping to inform the design of an RCT of this intervention. It suggested the need for a broader recruitment strategy; indicated areas that could be explored in more depth in the qualitative component of the trial; and highlighted issues that should be addressed to enhance and evaluate fidelity, particularly in the preparation and monitoring of intervention providers. The experience illustrates the value of even small sample before-and-after studies in the development of trials of complex interventions.