Purpose To assess the sensitivity and specificity of common carotid ultrasound method for carotid near-occlusion diagnosis. Methods Five hundred forty-eight patients examined with both ultrasound and CTA within 30 days of each other were analyzed. CTA graded by near-occlusion experts was used as reference standard. Low flow velocity, unusual findings, and commonly used flow velocity parameters were analyzed. Results One hundred three near-occlusions, 272 conventional ≥50% stenosis, 162 <50% stenosis, and 11 occlusions were included. Carotid ultrasound was 22% (95%CI 14–30%; 23/103) sensitive and 99% (95%CI 99–100%; 442/445) specific for near-occlusion diagnosis. Near-occlusions overlooked on ultrasound were found misdiagnosed as occlusions (n = 13, 13%), conventional ≥50% stenosis (n = 65, 63%) and < 50% stenosis (n = 2, 2%). No velocity parameter or combination of parameters could identify the 65 near-occlusions mistaken for conventional ≥50% stenoses with >75% sensitivity and specificity. Conclusion Near-occlusion is difficult to diagnose with commonly used carotid ultrasound methods. Improved carotid ultrasound methods are needed if ultrasound is to retain its position as sole preoperative modality.

Unfortunately, the article was published online prematurely with errors. Initials were presented as numbers and not yet replaced with the actual initials. A translation legend for blinded information was needlessly included. The original article has been corrected. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.