Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Noninvasive carotid artery tests detect stenosis

ACP J Club. 1995 July-Aug;123:17. doi:10.7326/ACPJC-1995-123-1-017

Source Citation

Blakeley DD, Oddone EZ, Hasselblad V, Simel DL, Matchar DB. Noninvasive carotid artery testing. A meta-analytic review. Ann Intern Med. 1995 Mar 1;122:360-7.



To determine the sensitivity and specificity of 6 noninvasive tests for detecting clinically important carotid artery stenosis.

Data sources

MEDLINE (1977 to 1993) was searched by using the terms oculoplethysmography, ultrasonography, digital subtraction angiography, cerebral angiography, magnetic resonance imaging, and vertebral, basilar, and carotid arteries. Bibliographies of neurology textbooks and retrieved articles were also reviewed.

Study selection

Studies were excluded if the test results were not compared with the results of conventional carotid angiography or intra-arterial digital subtraction carotid angiography, if the results did not allow for the identification of occluded arteries, or if the reference standard test results could not be entered into a contingency table according to degree of stenosis.

Data extraction

Method of enrollment, patient demographics, blinded outcome assessment, study results, and study centers were abstracted.

Study selection

568 articles were screened independently by a team of physicians and nurses. 2 pairs of physicians independently reviewed the final set of studies (104 studies), and discrepancies were resolved by discussion to arrive at the 70 studies selected. The reproducibility of the decision to exclude a tested subset of articles was 100%.

Main results

Studies included in the meta-analysis comprised data from 6406 patients (mean age 62 y, 65% men). 12 265 arterial sides were reported. {For detecting 100% stenosis, the sensitivities (95% CI) for carotid duplex ultrasonography, carotid Doppler ultrasonography, and magnetic resonance angiography were 82% (77% to 86%), 87% (79% to 93%), and 82% (74% to 88%), respectively; the specificities were all 98%. For detecting 70% stenosis, the sensitivities were 88% (81% to 92%), 86% (81% to 90%), and 90% (85% to 93%), respectively; specificities ranged from 89% to 94%}* (Table). Supraorbital Doppler ultrasonography performed as well as carotid duplex, carotid Doppler, and magnetic resonance angiography at 70% stenosis but had lower specificity at 100% stenosis. B-mode ultrasonography and oculoplethysmography had lower sensitivity and specificity at 70% and 50% stenosis.


For detecting artery occlusion and clinically important stenosis, carotid Doppler ultrasonography, carotid duplex ultrasonography, magnetic resonance angiography, and supraorbital Doppler ultrasonography have high sensitivity and specificity.

Source of funding: Agency for Health Care Policy Research.

For article reprint: Dr. D.B. Matchar, Center for Health Policy Research and Education, 125 Old Chemistry Building, Duke University, Durham, NC 27708, USA. FAX 919-684-6246.

*Information supplied by author.

Table. Noninvasive tests to detect carotid artery stenosis†

Outcomes Test Sensitivity (95% CI) Specificity +LR -LR
100% stenosis Carotid duplex ultrasonography 82% (77 to 86) 98% 41 0.1
Carotid Doppler ultrasonography 87% (79 to 93) 98% 44 0.1
Magnetic resonance angiography 82% (74 to 88) 98% 41 0.1
70% stenosis Carotid duplex ultrasonography 88% (81 to 92) - -
Carotid Doppler ultrasonography 86% (81 to 90) - -
Magnetic resonance angiography 90% (85 to 93) - -

†LRs defined in Glossary and calculated from data in article or supplied by author.
†Range 89% to 94%.


Techniques for the noninvasive evaluation of carotid artery disease are evolving so rapidly that a study such as this meta-analysis by Blakeley and colleagues is in danger of becoming outdated almost as soon as it is published. Nevertheless, this high-quality review synthesizes the current evidence in this field and yields useful information about the test properties of these diagnostic imaging procedures. At present, 2 noninvasive tests are clearly of value to clinicians. Duplex color ultrasonography combines anatomic B-mode views of the carotid bifurcation with Doppler blood flow velocity measurements and is the first test done in most patients suspected of having carotid disease. If substantial (> 70%) stenosis is not found, no further tests are usually necessary. If the results are positive, magnetic resonance angiography is useful for identifying additional vascular lesions and assessing the run-off circulation to the brain. The sensitivity and specificity of these 2 tests are so high that surgeons are increasingly willing to operate without requiring confirmation by invasive arteriography (1).

Both duplex ultrasonography and magnetic resonance angiography tend to overestimate the degree of stenosis and, therefore, cannot be depended on in cases of total occlusion of the internal carotid artery. Arteriography is required to distinguish complete occlusion (a non-operable condition) from 99% stenosis (which is operable).

The potential use of magnetic resonance angiography as a screening test is limited by its cost and by the fact that the test provides anatomic detail but not functional assessment of the carotid circulation. For these reasons, duplex ultrasonography and magnetic resonance angiography represent the current state of the art.

Arthur W. Feinberg, MD
North Shore University HospitalManhasset, New York, USA


1. Lustgarten JH, Solomon RA, Quest DO, Khanjdi AG, Mohr JP. Carotid endarterectomy after noninvasive evaluation by duplex ultrasonography and magnetic resonance angiography. Neurosurgery. 1994;34:612-9.