Grading of proximal internal carotid artery (ICA) stenosis by Doppler/duplex ultrasound (DUS) and computed tomographic angiography (CTA): correlation and interrater reliability in real-life practice

Doppler/duplex ultrasound (DUS) and computed tomographic angiography (CTA) are frequently applied methods to assess the degree of proximal internal carotid artery (ICA) stenoses in patients with acute ischemic stroke. This study evaluated the agreement and interrater reliability (IR) of both methods using a revised DUS grading system as well as different criteria (ECST/NASCET) under real-life conditions. CTA and DUS data of 281 proximal ICA stenoses [143 patients; 65.7 % male; age (mean (years) ± SD, range) 72.2 ± 11.1, 40–99] were retrospectively analyzed. For both methods, two independent raters estimated the degree of stenosis according to NASCET and ECST criteria. DUS raters applied revised German DUS criteria. For agreement and IR assessment, the linear weighted Kappa statistic was used. Correlation between DUS and CTA was substantial irrespective of the applied classification [weighted Kappa: 0.77 (NASCET)/0.79 (ECST)]. IR for DUS was almost perfect (weighted Kappa: 0.94) and better than for CTA [weighted Kappa: 0.78 (NASCET)/0.78 (ECST)]. In a real-life setting, CTA and DUS assessments of the degree of proximal ICA stenoses agreed substantially irrespective of the criteria applied (ECST/NASCET). For DUS, IR was better than for CTA.