Objectives: The purpose of this study was to evaluate the reliability of carotid duplex ultrasound scanning performed by nonaccredited vascular laboratories and to assess the clinical effect on patient management.
Methods: We retrospectively reviewed concordance of findings of carotid duplex ultrasound scanning between laboratories accredited by the Intersocietal Commission for Accreditation of Vascular Laboratories and nonaccredited laboratories in 174 patients with asymptomatic disease referred to tertiary care community hospitals for surgical evaluation for carotid endarterectomy (CEA) between January 2001 and December 2002, and evaluated changes in clinical management made on the basis of repeat examinations.
Results: Concordant findings were noted in 171 of 348 arteries (49%), predominantly those with minimal or mild disease (114 arteries; 67%). Discordant findings of no clinical significance were found in 54 arteries (16%). Clinically significant discordant findings were noted in 123 arteries (35%) in 107 patients (61%). In 104 arteries (88 patients) stenosis was overestimated by the nonaccredited laboratory secondary to technical error (19 arteries), use of B-mode imaging data alone (36 arteries), and use of inappropriate velocity criteria (49 arteries). None of these patients underwent CEA. Stenosis was significantly underestimated in 19 arteries (19 patients); all of these patients underwent uncomplicated CEA.
Conclusions: Incorrect physician interpretation of data is the most common cause of error in carotid duplex ultrasound scanning performed in nonaccredited vascular laboratories. Results of carotid duplex ultrasound scanning from nonaccredited laboratories should be considered with extreme caution, and do not appear reliable in planning treatment of obstructive disease.