RESUMO
BACKGROUND AND PURPOSE: This work was undertaken to review research addressing the cost-effectiveness of stroke-related diagnostic, preventive, or therapeutic interventions. METHODS: We performed searches of MEDLINE, Excerpta Medica online, HealthSTAR, and Sciences Citation Index Expanded and examined the reference lists of the studies and reviews obtained. From these, we selected studies that reported an incremental analysis of cost per effect, in which the effect measure was life-years or quality-adjusted life-years. We abstracted data from each study using a standardized reporting form. Twenty-six articles met the eligibility criteria and were included in the review. RESULTS: The methodological quality of the articles reviewed has improved compared with previously reported. Many stroke evaluation and treatment policies may result in benefits to health that are considered worth their cost. Some interventions were considered cost-ineffective (anticoagulation in low-risk nonvalvular atrial fibrillation and surveillance with duplex ultrasound after endarterectomy). Different studies addressing the cost-effectiveness of screening asymptomatic carotid stenosis resulted in strikingly divergent conclusions, from being cost-effective to being detrimental. Other studies omitted important costs that, if included, would likely have had profound impact on their cost-effectiveness estimates. CONCLUSIONS: Given the divergent conclusions drawn from studies addressing similar questions, it may be premature to use the results of cost-effectiveness research in developing stroke policy and practice guidelines. Successful implementation of such evaluations in the care of patients with stroke will depend on further standardization of methodology and critical appraisal of reported findings.