Agreement between Multidetector-Row CT Angiography and Ultrasound Echo-Color Doppler in the Evaluation of Carotid Artery Stenosis

PURPOSE: Stroke is a leading cause of severe disability in the western world. A correct diagnostic procedure to stratify risk is necessary in order to rapidly plan the most efficient therapy. The purpose of this work was to evaluate the agreement between ultrasound echo-color Doppler (US-ECD) and multidetector-row CT angiography (MDCTA) in determining the degree of carotid stenosis. METHODS AND MATERIALS: From January 2004 to February 2007, 187 patients who had previously undergone both US-ECD and MDCTA were studied retrospectively. For each patient, stenosis degree was measured by applying the North American Symptomatic Carotid Endarterectomy Trial criteria. Data derived from MDCTA and US-ECD were then compared to calculate the inter-technique variability by using Cohen kappa statistics. For all plaque types (fatty, mixed and calcified), inter-technique variability was assessed. The Wilcoxon signed-rank test was used to highlight differences between the procedures, and scatterplots were also calculated. As a gold standard, reference surgical comparisons were performed in 50 patients. RESULTS: The percentage of observed agreements in the evaluation of stenosis degree was 72.19% with a kappa value of 0.659 (95% confidence interval: 0.604-0.715) and a weighted kappa of 0.789. After the evaluating stenosis degree in fatty, mixed and calcified plaques, the percentages of observed agreements were 79.31, 74.58 and 64.29%, respectively, with kappa values of 0.738, 0.691 and 0.565. CONCLUSION: A good agreement between US-ECD and MDCTA in the evaluation of stenosis degree was observed, together with a remarkable difference between the 2 techniques in the assessment both of calcified plaques and of plaques showing a predominant calcific component.