Evaluation of the interrater and intermethod agreement of the German multiparametric ultrasound criteria for the grading of internal carotid artery stenosis

Abstract

Purpose: The interdisciplinary German guidelines for the diagnosis and treatment of internal carotid artery stenosis (ICAS) recommend a multiparametric approach for the sonographic grading of extracranial ICAS. The aim of this study is to evaluate the interrater and intermethod agreement of this elaborated sonographic approach with different angiographic modalities.

Methods: Patients with extracranial ICAS were examined twice with colour-coded duplex sonography (CDS) by two experienced vascular neurologists. Each of the ten criteria and the resulting stenotic value were assessed. Grading of ICAS based on the multiparametric ultrasound criteria was compared with different angiography modalities (magnetic resonance angiography (MRA), computed tomography angiography (CTA), digital subtraction angiography (DSA)).

Results: Seventy-four consecutive patients with 91 extracranial ICAS were recruited from our stroke unit and neurovascular outpatient clinic. Interrater agreement for each single ultrasound criterion ranged from moderate to excellent (for the peak systolic velocity). Concerning the absolute stenotic value of ICAS, an excellent agreement between both ultrasound examiners with an ICC of 0.91 (range 0.87-0.94; p < 0.001) was found. In 96% of ICAS, the difference between the stenotic values was ≤ 10%. Intermethod agreements between CDS and DSA, CTA, and MRA were also good for both sonographers.

Conclusion: Strictly adhering to the multiparametric "DEGUM ultrasound criteria", we found an excellent interrater agreement and a good intermethod agreement compared with angiography for the sonographic grading of extracranial ICAS. Thus, multiparametric CDS is in particular suitable for the follow up of extracranial ICAS even when examinations are done by different sonographers.

Keywords: Carotid artery disease; DEGUM ultrasound criteria; Grading of internal carotid artery stenosis; Interrater and intermethod agreement; Ultrasound.

Fig. 1

Interrater agreement of the sonographic grading of internal carotid artery stenosis. The interrater agreement for the grading of internal carotid artery stenosis between both sonographers when applying the multiparametric “DEGUM ultrasound criteria” is visualised by a Bland and Altman diagram (a). In most cases, the difference of the total stenotic value was equal to/or less than 10% which again indicates a good agreement (b)

Fig. 2

Correlation of the sonographic stenotic value and selected ultrasound criteria of the 2nd sonographer (y-axis) with the stenotic value assessed by digital subtraction angiography by the 2nd neuroradiologist (x-axis, in %). Overall, there was a good correlation between DSA and velocity-based criteria (b, c, f, h). Both criteria that were dichotomised into present or absent (collateral flow (d) and confetti sign (g)) confirmed high-grade stenosis when positive but could not rule out a relevant stenosis when negative. The vertical and horizontal bars in d, f, g, h indicate the boundaries that are proposed to discriminate between a moderate and severe stenosis [9, 11]

Fig. 3

Correlation between intrastenotic diameter/cross-sectional area and peak systolic velocity of internal carotid artery stenosis. Correlation between the narrowest intrastenotic diameter (assessed by digital subtraction angiography (a) and CT angiography (b)), respectively, the narrowest intrastenotic cross-sectional area (assessed by CT angiography (c)) and the peak systolic intrastenotic velocity (x-axis, in cm/s) measured by colour-coded duplex sonography