Comparison of conventional and modern methods in determining ischemic stroke etiology by general and stroke neurologists

Background/aim: This study aimed to investigate the consistency between stroke and general neurologists in subtype assignment using the Trial of ORG-10172 in Acute Stroke Treatment (TOAST) and Causative Classification of Stroke (CCS) systems.

Materials and methods: Fifty consecutive acute ischemic stroke patients admitted to the stroke unit were recruited. Patients were classified by two stroke and two general neurologists, each from different medical centers, according to TOAST followed by the CCS. Each neurologist was assessed for consistency and compliance in pairs. Concordance among all four neurologists was investigated and evaluated using the kappa (ĸ) value.

Results: The kappa (ĸ) value of diagnostic compliance between stroke neurologists was 0.61 (95% CI: 0.45–0.77) for TOAST and 0.78 (95% CI: 0.62–0.94) for CSS-5. The kappa (ĸ) value was 0.64 (95% CI: 0.48–0.80) for TOAST and 0.75 (95% CI: 0.60–0.91) for CCS-5 for general neurologists. Compliance was moderate [ĸ: 0.59 (95% CI: 0.52–0.65)] for TOAST and was strong [ĸ: 0.75 (95% CI: 0.68–0.81)] for CCS-5 for all 4 neurologists. ‘Cardioembolism’ (91.04%) had the highest compliance in both systems. The frequency of the group with ‘undetermined etiologies’ was less in the CCS (26%) compared to TOAST.

Conclusion: The CCS system improved compliance in both stroke and general neurologists compared with TOAST. This suggests that the automatic, evidence-based, easily reproducible CCS system was superior to the TOAST system.