INTERRATER RELIABILITY IN THE ASSESSMENT OF NEUROVASCULAR DISEASES

Empirical studies on neurovascular diseases demand valid and reliable diagnostic assessments. Using 13 standardized case summaries on cerebral ischemic events, interrater agreement was tested in this study. Eighty-one diagnosticians gave their diagnostic assessments on the basis of three different classification systems (ICD-9, draft of ICD-10, German classification of neurological diseases; CSGN). As a result, the interrater reliabilities showed marked differences between the three systems, the chance-corrected agreements (kappa) differed from 0.38 (ICD-9) to 0.72 (ICD-10). The data analysis demonstrated that the neurovascular diseases were only partly covered by the different diagnostic systems.