Colour-coded duplex assessment alone before carotid endarterectomy

This study assessed whether carotid colour-coded duplex imaging can provide sufficient information to investigate patients before carotid endarterectomy. One hundred patients with symptomatic carotid artery disease (amaurosis fugax 24, transient ischaemic attack 38 and established stroke 38) were assessed by both colour-coded duplex imaging and arch angiography. Stenoses were graded at angiography using the linear method from the European Carotid Surgery Trial and on duplex using peak systolic velocity criteria. Angiograms were also analysed for any arch or intracerebral pathology which would alter surgical management. In defining 70-99 per cent stenoses, duplex had a sensitivity of 98 per cent, specificity 96 per cent, positive predictive value of 96 per cent, negative predictive value 98 per cent and accuracy 97 per cent compared with angiography for both symptomatic and asymptomatic carotids (chi 2 = 1.22, 1 d.f., P = 0.25). With respect to the symptomatic carotid arteries angiography gave different information from duplex in six patients (6 per cent): definition of 70-99 per cent carotid stenosis was different (four), visualization was poor on duplex (one) and an intracerebral aneurysm was identified by angiography. One patient died following arch angiography as a result of mesenteric thrombosis. Carotid duplex imaging alone would appear to be sufficient to assess most patients before carotid endarterectomy as an accurate, safe and relatively cheap investigation compared with angiography. In cases where duplex assessment does not allow full visualization of the carotid bifurcation angiography is required.