Abstract

BACKGROUND: Individuals with carotid stenosis enter surveillance or are considered for surgery on the basis of disease severity assessed by ultrasound. However, there is variation in the ultrasound diagnostic thresholds used to determine disease severity. Our objective was to describe this variation and its potential impact on patients. METHODS: To describe the variation in carotid ultrasound diagnostic thresholds, we examined testing protocols from 338 accredited vascular testing centers in the United States. To determine the potential impact of this variation, we applied the range of thresholds to carotid ultrasound parameters from 2 groups: a population-based sample ≥65 years of age in the Cardiovascular Health Study (n=4791), and a cohort of patients who underwent surgery for asymptomatic carotid stenosis in the Vascular Quality Initiative registry (n=28 483). RESULTS: Internal carotid artery peak systolic velocity was used by all centers to assess disease severity, with 60 distinct thresholds in use. The peak systolic velocity threshold for moderate (≥50%) stenosis ranged from 110 to 245 cm/s (median, 125; 5th and 95th percentile, 125 and 150), and the threshold for severe (≥70%) stenosis ranged from 175 to 340 cm/s (median, 230; 5th and 95th percentile, 230 and 275). In the population-based sample, the 5th percentile threshold would assign a diagnosis of moderate carotid stenosis to twice as many individuals as the 95th percentile threshold (7.9% versus 3.9%; relative risk, 2.01 [CI, 1.70-2.38]). In the surgical cohort, 1 in 10 (9.8%) patients had peak systolic velocity values that warranted the diagnosis of severe carotid stenosis at centers in the 5th percentile, but not in the 95th. CONCLUSIONS: The diagnostic threshold for carotid stenosis varies considerably. Whether or not a person is said to have moderate stenosis and enters surveillance, and whether or not they have severe stenosis and are candidates for surgery, can depend on which center performs their ultrasound.