Poor agreement in carotid artery stenosis detection by ultrasound between external offices and a vascular center

Abstract

Carotid duplex ultrasonography is the prime investigation used to grade carotid artery stenosis in clinical routine. We compared the carotid ultrasound (US) scans performed externally with our results. This retrospective study included 288 patients who had been referred to our outpatient department and initially presented with an external carotid duplex scan report indicating carotid atherosclerosis. The external scans were analyzed and compared with our scans in respect of the accuracy of identification and quantification of stenosis, the criteria used to grade stenosis and the duplex criteria used. Weighted Kappa coefficients (K) were computed to quantify the agreement between internal and external findings. The majority of the external reports had been performed by radiologists [70.8 % (n = 204)], followed by specialists of internal medicine [19.4 (n = 56)] and by neurologists [9.8 % (n = 28)]. Only slight agreement was registered between the external reports and those performed at our institution with regard to the identification of stenosis (K = 0.2 for the left and K = 0.12 for the right side). Greater agreement was observed in respect of the level of stenosis (K = 0.42 for the right and K = 0.54 for the left side). Overestimation of the level of stenosis was registered for 45 % in the left internal carotid artery (ICA) and 36 % in the right ICA; the overestimation was most pronounced for occlusions and high-grade stenoses, which is a source of great concern for decision-making. The present data indicate only a slight agreement between carotid duplex US imaging performed at medical offices and our results.