Magnetic resonance venography for the detection of deep venous thrombosis: comparison with contrast venography and duplex Doppler ultrasonography.

Abstract

Contrast venography is the gold standard for diagnosis in deep venous thrombosis (DVT); however, this technique is invasive and requires the use of potentially hazardous contrast agents. Although duplex Doppler ultrasonography is accurate in the evaluation of lower extremity DVT, it is less accurate in the assessment of the pelvic and intraabdominal veins. Magnetic resonance venography (MRV) has recently been developed, and our purpose was to determine whether MRV could accurately demonstrated DVT when compared with duplex scanning and contrast venography. Eighty-five patients underwent contrast venography and MRV from the inferior vena cava to the popliteal veins to rule out DVT. Thirty-three of these patients also underwent duplex scanning. Blinded readings of these studies were compared for the presence or absence and extent of venous thrombosis. DVT was documented by contrast venography in 27 (27%) venous systems. Results of MRV and contrast venography were identical in 98 (97%) of 101 venous systems, whereas results of duplex scanning and contrast venography were identical in 40 (98%) of 41 venous systems. All DVTs identified by contrast venography were detected by MRV and duplex scanning. The discrepancies were due to false-positive MRV (3) and duplex scanning (1) results. When compared with contrast venography, MRV had a sensitivity of 100%, specificity of 96%, positive predictive value of 90%, and negative predictive value of 100%. For duplex scanning the sensitivity was 100%, specificity was 96%, positive predictive value was 94%, and negative predictive value was 100%. It is concluded that MRV is an accurate noninvasive venographic technique for the detection of DVT.