Jf. Reed et al., MAGNETIC-RESONANCE ANGIOGRAPHY AND DUPLEX SCANNING CONCORDANCE IN DETECTING SEVERE CAROTID STENOSIS - A METHODOLOGICAL REVIEW, Vascular surgery, 29(3), 1995, pp. 177-183
In this meta-analysis, the authors examine published studies on the co
nsistency (homogeneity) of study outcomes of magnetic resonance angiog
raphy (MRA), conventional angiography (XRA), and duplex ultrasound (DU
) scanning in assessment of the degree of carotid artery disease. The
sensitivity, specificity, and accuracy of the MRA as a noninvasive dia
gnostic test are exceptional. Given an adequate two-dimensional time-o
f-flight (2D-TOF) study, it provides a robust test for the evaluation
of carotid stenosis. The authors conclude from this meta-analysis that
MRA is at least as accurate as DU in a selected group of patients. Th
e question of superiority for all patients with suspected carotid arte
ry disease is not, however, answered by this meta-analysis. Another qu
estion not answered is the question of DU/MRA-XRA. concordance/discord
ance. If DU and MRA were concordant across all, rather than a selected
group of patients, would the clinical decision tree be in favor of el
imination of XRA? The cost for a DU scan is approximately one third of
the cost of an MRA and, from an economic point of view, one could not
argue the use of MRA over DU. However, as MRA technology and accuracy
improve and when the cost of MRA approaches that of DU, MRA may repla
ce DU as a screening method. Clearly, a consecutive series of all pati
ents suspected of having carotid artery disease needs to be evaluated
by use of the best available diagnostic technologies. A comprehensive
protocol would include an evaluation of the sources of difficulties in
conducting these tests.