Simplified ultrasound protocol for the exclusion of clinically significant carotid artery stenosis

Abstract

Objectives: To evaluate a simplified ultrasound protocol for the exclusion of clinically significant carotid artery stenosis for screening purposes.

Material and methods: A total of 9,493 carotid arteries in 4,748 persons underwent carotid ultrasound examination. Most subjects were 65-year-old men attending screening for abdominal aortic aneurysm. The presence of a stenosis on B-mode and/or a mosaic pattern in post-stenotic areas on colour Doppler and maximum peak systolic velocity (PSV) in the internal carotid artery (ICA) were recorded. A carotid stenosis was defined as The North American Symptomatic Carotid Endarterectomy Trial (NASCET) >20% and a significant stenosis as NASCET >50%. The kappa (κ) statistic was used to assess agreement between methods. Sensitivity, specificity, positive predictive (PPV), and negative predictive (NPV) values were calculated for the greyscale/mosaic method compared to conventional assessment by means of PSV measurement.

Results: An ICA stenosis was found in 121 (1.3%) arteries; 82 (0.9%) were graded 20%–49%, 16 (0.2%) were 50%–69%, and 23 (0.2%) were 70%–99%. Eighteen (0.2%) arteries were occluded. Overall, the greyscale/mosaic protocol showed a moderate agreement with ICA PSV measurements for the detection of carotid artery stenosis, κ = 0.455. The sensitivity, specificity, PPV, and NPV for detection of >20% ICA stenosis were 91% (95% CI 0.84–0.95), 97% (0.97–0.98), 31% (0.26–0.36), and 97% (0.97–0.97), respectively. The corresponding figures for >50% stenosis were 90% (0.83–0.95), 97% (0.97–0.98), 11% (0.08–0.15), and 100% (0.99–1.00).

Conclusion: Compared with PSV measurements, the simplified greyscale/mosaic protocol had a high negative predictive value for detection of >50% carotid stenosis, suggesting that it may be suitable as a screening method to exclude significant disease.

Keywords:

• Carotid stenosis
• screening
• ultrasound

Disclosure statement

The authors report no conflicts of interest.

Funding information

Financial support was given by FoU-enheten NU-hospital organisation, Sweden. Martin Björck and Anders Wanhainen are supported by the Swedish Research Council (Grant # K2013-64X-20406-07-3).