The effectiveness of community-based rehabilitation for stroke patients who remain at home: a pilot randomized trial

Objective: To assess the effectiveness of community-based rehabilitation fo r stroke patients who were not admitted to hospital in South London. Design: Randomized controlled trial. Setting: Patients' homes in South London. Subjects: Stroke patients not admitted to hospital after a stroke. Intervention: Rehabilitation at home by rehabilitation team for up to three months or usual care. Main outcome measures: The primary outcome measure was the Barthel score. S econdary measures included the Motricity Index, Rivermead ADL, Hospital Anx iety and Depression score and Nottingham Health Profile. Results: Forty-three patients who remained at home were randomized to rehab ilitation team (23) or 'usual' care (20). The mean number of physiotherapy sessions was three (range 1-14) for the rehabilitation team group and two f or the usual care group. Patients (with a deficit) in the rehabilitation ar m of the trial were more likely to receive occupational, physical and speec h therapy than those in the control arm (p = 0.03, 0.01 and 0.008, respecti vely). For those patients actually receiving therapy, there was no evidence that the amount received differed between the groups. However, the number of patients in each of these comparisons was very small. The outcome for pa tients in the rehabilitation team arm of the trial was nonsignificantly hig her (0.05 < p < 0.2) than for those in the control arm for the areas of Not tingham Health Profile, anxiety, depression, caregiver strain and the propo rtion of patients living at home. Based on the data observed here, a trial with approximately 150 patients in each arm would be needed to have adequat e power to detect a 33% difference between intervention and control groups in these outcomes. Conclusion: Community therapy support for patients not admitted to hospital is feasible but to determine whether it is cost- or clinically effective w ould require trials of adequate size.