Abstract
A total of 50 legs in 34 consecutive patients with clinically suspected deep vein thrombosis (DVT) were evaluated both with colour-coded duplex sonography and phlebography, the tests being performed immediately following each other without the examiner knowing the result of the other procedure. Contrast phlebography was regarded as the standard for diagnosis. All the sonograms were considered to be technically adequate. The overall diagnostic accuracy, taking the leg as a whole from the ilio–femoral segment to the calf, was 92% and the sensitivity 78%. The accuracy in the femoral vein was 92%, sensitivity 64%. In the calf the accuracy was 90% and sensitivity 81%. although no attempt was made to localize or quantify the thrombus. The accuracy of the diagnosis of thrombus was not improved by the use of the colour-coded duplex facility. The main advantage of colour-coded duplex sonography over conventional duplex scanning is the easier identification of the venous system making it faster to perform. It is suggested that colour-coded sonography is a useful primary investigation for patients suspected of DVT or pulmonary embolism, but that about a third of patients will require contrast phlebography if the sonogram is negative or technically inadequate. It is emphasized that phlebography remains the most accurate method currently available for the diagnosis of DVT and using low osmolar contrast media is well tolerated and impressively safe.