Comparison of duplex ultrasonography and venography in the diagnosis of deep venous thrombosis

Sixty-five patients with suspected deep venous thrombosis (DVT) in 68 limbs were entered consecutively into a study to compare venography with duplex ultrasonography scanning. Both tests were performed on 64 limbs, venography being contraindicated in four. Overall, duplex scanning correctly identified 86 per cent of DVTs diagnosed on venography and correctly excluded 80 per cent with negative venograms. Nearly all errors arose in the diagnosis of calf DVT. In the femoral vein duplex scanning had a specificity of 100 per cent and a sensitivity of 95 per cent. In addition, duplex scanning provided data on the limb not undergoing venography. Of 55 limbs that underwent bilateral duplex scanning, five had thrombus in the femoropopliteal segment and a negative contralateral venogram. In addition, three Baker's cysts were diagnosed. Duplex scanning can be used in patients in whom venography is contraindicated and may also provide information about the contralateral limb. We regard femoropopliteal duplex scanning as sufficiently accurate that treatment can be initiated on the basis of the scan. Duplex scanning should replace venography as the standard method of diagnosing femoropopliteal DVT; radiographic studies should now be required only when the scan result is in doubt.