P.113 Validation and standardization of cerebral vasospasm grading on CT angiography

Abstract

Background: The diagnosis of cerebral vasospasm, either by digital subtraction angiography (DSA), or now more commonly by computerized tomographic angiography (CTA) occurs in up to 70% of patients with aneurysmal subarachnoid hemorrhage (aSAH). The lack of standardization among vasospasm grading has made its clinical correlation with delayed cerebral ischemia challenging Methods: 36 of the 764 aSAH patients found on the St. Michael’s Hospital RIS database had both DSA and CTA performed, at time of admission and again between day 2 and 14 following SAH. Two blinded neuroradiologists graded all vessels for vasospasm on two separate scales, by consensus for DSA and independently for CTA Results: Comparing CTA and DSA, Grading Scale (GS)1 had the highest Spearman Correlation Coefficient (SCC): 0.691 (P<0.001) for Rater (R)1, and 0.687 (P<0.001) for R2. SCC was higher when only considering proximal vessels. Cohen’s Kappa (CK) measuring inter-rater reliability was 0.695 (P<0.001) for GS2 and 0.681 (P<0.001) for GS1. CK was higher in anterior circulation vessels, and tended to decrease with increasing vasospasm grade. Conclusions: Although either scale will provide the benefits of standardization to clinical practice and research, GS1 is recommended as it is more intuitive and provides higher SCCs, with only slightly lower CKs.