Objective: To compare stroke patients' reports with service providers' reports on the rehabilitation services patients receive.
Setting: Community, predominantly urban.
Subjects: Ninety-three stroke patients who had not been admitted to hospital.
Method: Between August 1994 and August 1996, 54 general practitioner practices in Nottingham (population 344 200) were requested to notify the Nottingham Community Stroke Project of all cases of stroke. Patients were visited one month after stroke and asked to report on whether they had received any hospital, social, community or voluntary services. Some departments were contacted and asked if the patients were known to them and whether they had been seen.
Results: Patients reported having little contact with hospital or social services. Comparison of patients' reports with that of the appropriate service department showed poor agreement for community care assistants (kappa = 0.18), fair agreement for social workers (kappa = 0.59), speech therapy (kappa = 0.48), hospital outpatient appointments (kappa = 0.55) and admissions to hospital (kappa = 0.48) and good agreement for social services occupational therapists (kappa = 0.64).
Conclusion: Stroke patients' and service providers' reports on rehabilitation services may not always agree. Care should be taken when interpreting information on service provision.