A Critical Evaluation of Carotid Duplex Scanning in the Diagnosis of Significant Carotid Artery Occlusive Disease

Abstract

Carotid duplex (CD) scanning is the primary method of evaluating patients suspected of having extracranial carotid occlusive disease. It is incumbent on vascular laboratories (VL) to have internally validated criteria confirming overall accuracy, specificity (SP), sensitivity (SS), negative predictive value (NPV), and positive predictive value (PPV). Receiver operating characteristic (ROC) curves allow further analysis to update existing criteria. We correlated 127 internal carotid arteries studied by carotid duplex scanning and confirmatory modalities, which showed a SP of 83.3%, SS of 97.9%, NPV of 92.5%, PPV of 95%, and overall accuracy of 94.5% for >50% internal carotid artery stenosis. For >70% stenosis, SP was 88.8%, SS was 96.1%, NPV was 93.6%, PPV was 92.5%, and overall accuracy was 92.9%. ROC curves for the peak systolic velocities were used; for detecting a 50–69% stenosis range a sensitivity of 88%, specificity of 93%, NPV of 73%, and PPV of 97% were found. For detecting a 70–99% stenosis a sensitivity of 95%, specificity of 90%, NPV of 93%, and a PPV of 93% were found. All vascular laboratories must have a vigorous quality assurance program and must validate their own internal criteria or the recently promulgated consensus criteria for grading the severity of carotid stenosis by carotid duplex examination.