Validation of a Polish version of the Israeli Vertebrobasilar Stroke Scale - An attempt to more accurately assess posterior Stroke

Abstract

ObjectivePosterior circulation stroke, in contrast to anterior circulation stroke, has a greater complexity and variability of clinical symptoms. This could be responsible for delayed diagnosis and treatment time and, as a consequence, worse prognosis. Certain blame in this respect can also be attributed to the clinimetric scales used to assess stroke severity, which are characterized by significantly lower accuracy than with anterior strokes. The Israeli Vertebrobasilar Stroke Scale (IVBSS) was the first attempt dedicated to posterior strokes and was devised for better measurement of clinical condition. We aimed to develop a Polish version of the IVBSS (PL-IVBSS) to assess the reliability, validity and psychometric properties of the tool to confirm its clinical utility. MethodsWe enrolled 126 posterior circulation ischemic stroke subjects. Four researchers estimated stroke severity using appropriate and widely accepted devices (the modified Rankin Scale – mRS, the National Institutes of Health Stroke Scale – NIHSS, the Barthel Index, and the Glasgow Coma Scale – GCS) and compared with the PL-IVBSS. We analyzed inter- and intrarater agreements, repeatability, concurrent and predictive validity, internal consistency, scalability and homogeneity, reflecting the psychometric features of a validated instrument. ResultsCronbach’s alpha coefficient was 0.67, and the median inter-item correlation coefficient was 0.22, indicating moderate internal consistency and insufficient homogeneity. A total of 63.6% of the individual items obtained required discriminatory power (r > 0.3), showing moderate scalability. The PL-IVBSS achieved a good coefficient of repeatability (CR = 1.21 95%CI 1.08–1.38) and narrow limits of agreement in Bland-Altman analysis, emphasizing the accuracy and high reproducibility. Excellent intraclass correlation coefficients and weighted kappa values (all >0.90) underlined the high reliability of the PL-IVBSS. Highly significant correlations with other relevant devices (all r > 0.5, p < 0.0001) highlighted the satisfactory concurrent and predictive validity of a validated clinimetric tool. ConclusionWe devised a validated version of the IVBSS, indicating the high reproducibility, repeatability and accuracy of the PL-IVBSS and confirming its clinical utility. Despite moderate psychometric properties, our findings support the need for its clinical application and widespread use in stroke units for a reliable assessment of posterior stroke severity.