Putting stroke outcomes into context - Assessment of variations in the processes of care

Background: Comparative studies of patient outcomes consistently find that variations cannot be explained by case mix alone, suggesting that differences in the process of care may contribute to variations in outcome. We sought to describe local medical and social services available to patients participating in a multinational study of stroke services and outcomes. Methods: Key informant interviews were conducted with service providers in participating centres. A semi-structured interview schedule was used to describe local models of clinical care, types of community care available and the role of the family in caring for patients. Results: Data were provided by 15 centres in 14 European states. Models of clinical care include multidisciplinary and ‘geographical’ stroke units, dedicated stroke beds, neurologist-led care and physician-led care. Nurse to patient ratios range from 1:3 to 1:15. Three patterns of rehabilitation provision emerge and estimates of acute physiotherapy range from 30 to 180 min/day. In almost all centres the patient's family is expected to assume responsibility for care after discharge but there are wide variations in the levels of community support available to informal carers. Conclusions: The extent to which the structures and processes of care vary across a sample of European centres highlights the need for caution in assuming that studies which control for patient characteristics thereby control for all relevant variables. There are also implications for the implementation of international gold standards of care. Future analyses will relate our findings to 3 month and 1 year outcomes and assess their impact.