Validation of Hospital Performance Measures of Acute Stroke Care Quality. Riksstroke, the Swedish Stroke Register

Abstract

Background: Registers are increasingly used to monitor stroke care performance. Fair benchmarking requires sufficient data quality. We have validated acute care data in Riksstroke, the Swedish Stroke Register. Methods: Completeness was assessed by comparisons with diagnoses at hospital discharge recorded in the compulsory National Patient Register and content validity by comparisons with (a) key variables identified by European stroke experts, and (b) items recorded in other European stroke care performance registers. Five test cases recorded by 67 hospitals were used to estimate inter-hospital reliability. Results: All 72 Swedish hospitals admitting acute stroke patients participated in Riksstroke. The register was estimated to cover at least 90% of acute stroke patients. It includes 18 of 22 quality indicators identified by international stroke experts and 14 of 15 indicators used by at least 2 stroke performance registers in other European countries. Inter-hospital reliability was high (≥85%) in 77 of 81 Riksstroke items. Conclusions: A nationwide stroke care register can be maintained with sufficient data quality to permit between-hospital performance benchmarking. Our experiences may serve as a model for other stroke registers while evaluating data quality.