Background: Etiological subclassification of ischemic stroke has become increasingly important, as new therapeutic agents have been introduced. The aim of this study was to assess the inter-rater reliability of the TOAST classification applied in the acute setting, and further to evaluate the criterion validity of the TOAST classification in discriminating between small vessel disease and other etiologies.
Methods: From June to December 2001, 38 patients with acute ischemic stroke were included in the present study. All were classified according to the TOAST criteria by two junior registrars shortly after admission. Later, a consensus classification was made based on a comprehensive investigation programme.
Results: The inter-rater reliability between the two junior registrars was very good (kappa 0.88). The overall agreement between initial and consensus classifications was fair (kappa 0.30). The TOAST classification discriminated between small vessel disease and other subgroups in most patients (sensitivity 0.93, specificity 0.83).
Conclusion: Etiological subclassification of ischemic stroke requires extensive cerebrovascular investigation. Normally, such resources are not available at admission. Nevertheless, the TOAST classification should be used in this setting to help physicians differentiate between small vessel disease and other etiologies of acute stroke, particularly when therapies with possible harmful side effects are considered an option.