BACKGROUND AND PURPOSE: Large within-country variations have been described in stroke management and there have been a few studies of between-country variation (in the USA and the UK). We designed a study to examine stroke management across a wide range of countries representing different stages of economic development. Large variations would suggest the need to explore methods of increasing the uptake of evidence-based stroke practice. METHODS: Members of the International Clinical Epidemiology Network (INCLEN) from 14 centres in ten countries agreed to review the records of the last 50 patients admitted to hospital with a clinical diagnosis of stroke. Information on demographic variables, the clinical diagnosis of stroke type, investigations performed and treatments given and the discharge destination of the patient were recorded and sent to the coordinating centre in Australia for analysis. RESULTS: There were statistically significant between-centre differences in the proportions of patients cared for by a neurologist, staying in hospital for at least ten days and having CT or MRI scans. Significant between-centre differences were also seen for treatment, for example, the use of aspirin in non-haemorrhagic stroke varied from 11 to 79%. The variation (for all interventions studied) was no longer statistically significant when examined within strata according to availability of facilities. CONCLUSIONS: The large variation between centres in the management of stroke is largely 'explained' by the availability of resources, even for interventions that do not depend on resource availability. It will be important to develop management guidelines that reflect evidence-based practice of relevance across a range of economic settings.
BACKGROUND AND PURPOSE: Large within-country variations have been described in stroke management and there have been a few studies of between-country variation (in the USA and the UK). We designed a study to examine stroke management across a wide range of countries representing different stages of economic development. Large variations would suggest the need to explore methods of increasing the uptake of evidence-based stroke practice. METHODS: Members of the International Clinical Epidemiology Network (INCLEN) from 14 centres in ten countries agreed to review the records of the last 50 patients admitted to hospital with a clinical diagnosis of stroke. Information on demographic variables, the clinical diagnosis of stroke type, investigations performed and treatments given and the discharge destination of the patient were recorded and sent to the coordinating centre in Australia for analysis. RESULTS: There were statistically significant between-centre differences in the proportions of patients cared for by a neurologist, staying in hospital for at least ten days and having CT or MRI scans. Significant between-centre differences were also seen for treatment, for example, the use of aspirin in non-haemorrhagic stroke varied from 11 to 79%. The variation (for all interventions studied) was no longer statistically significant when examined within strata according to availability of facilities. CONCLUSIONS: The large variation between centres in the management of stroke is largely 'explained' by the availability of resources, even for interventions that do not depend on resource availability. It will be important to develop management guidelines that reflect evidence-based practice of relevance across a range of economic settings.