The objective of this study was to assess the diagnostic accuracy of colour flow Doppler ultrasound (CFD) and its potential to replace digital subtraction angiography (DSA) before carotid endarterectomy (CEA). All patients undergoing CFD of the carotid bifurcation in our department over a period of 1-1/2 years for whom both CFD and DSA results were available were included in the study. We evaluated the feasibility of CFD, its diagnostic accuracy and its potential to diagnose clinically significant stenosis (50%, 70% and 90% NASCET type diameter stenosis) compared with DSA. 225 carotid bifurcations in 116 patients met the criteria for evaluation (biplane arterial DSA without superimposition). Data analysis yielded the following diagnostic performance of CFD: sensitivity for a 50% stenosis 91.4% (95% confidence interval (CI) 83.3--96.2%), specificity 93.2% (95% CI 87.1--96.8%) and accuracy 92.4% (95% CI 88.4--95.4%); sensitivity for a 70% stenosis 89.2% (95% CI 81.9--94.1%), specificity 96.2% (95% CI 90.5--98.6%) and accuracy 92.4% (95% CI 88.4--95.4%). In 9 of 116 cases, carotid angiography was used to evaluate inconclusive CFD results. DSA disclosed relevant information not suspected by CFD in only 1 of the 116 cases. Thus, 91% (106/116) of the angiographies could have been dispensed with without loss of information. One major stroke occurred during diagnostic DSA. We conclude that DSA of the carotid arteries is unnecessary when CFD is unequivocal. The diagnostic gain of DSA must be counterweighted against its potential risks.
