Abstract Body: Background: Several semi-quantitative imaging scores exist to assess the extent of ischemic injury in basilar artery occlusions using various computed tomography (CT)-based modalities. Their inter-rater agreement has never been compared.
Methods: We conducted a retrospective multicenter cohort study of patients with basilar artery occlusions. Four imaging scores (PC-ASPECTS, CAPS, BATMAN, and PC-CTA) were assessed by two raters. Inter-rater agreement was compared using Cohen’s kappa statistic and reliability was assessed using the intraclass correlation coefficient (ICC).
Results: 98 patients were included for analysis. The CT perfusion-based CAPS score yielded the highest interrater agreement (kappa 0.64 [95%CI 0.52 – 0.75]) and highest reliability (ICC 0.82 [95%CI 0.73 – 0.91]). By comparison, CTA-based scores achieved fair agreement. The PC-ASPECTS score yielded the lowest levels of agreement overall (kappa 0.11 [95%CI 0.0061 – 0.21]) and in individual regions, with the lowest kappa values for midbrain (kappa 0.098 [95%CI 0.022 – 0.30]). Dichotomization resulted in some improvement in agreement for all approaches.
Conclusion: CT Perfusion-based imaging score provide the highest interrater agreement and reliability for assessment of ischemic injury in patients with basilar artery occlusions among common posterior circulation scoring methodologies.