Compression ultrasonography as a reliable imaging monitor in deep venous thrombosis.

Our experience at Hippokration Athens Hospital with high-resolution real-time ultrasonography of the deep veins of the lower extremities is described, drawn from the results of a prospective comparative study of 65 patients. We used a combination of sonography and contrast venography for the detection of deep venous thrombosis (DVT). Of the study group, 33 had clinically suspected DVT, 20 were asymptomatic high-risk patients, while 12 with varicosities of the leg and scheduled for saphenectomy were examined preoperatively for deep venous patency confirmation. For all patients, venography, the reference method, was performed within 12 hours of the ultrasound scan. Since 29 of the examinations were bilateral, the total number of extremities studied was 94. The sonographic criteria analyzed were the intraluminal echogenicity and the venous compressibility with the ultrasound transducer probe. In addition, the response of the common femoral vein to the Valsalva maneuver was studied. Among these three criteria, compressibility was by far the most accurate. Noncompressible abnormal veins were noted in all of the 24 extremities with proximal thrombosis and in six of the 12 with thrombosis limited to the calf veins. Thus, the sensitivity of the compressibility criterion was 100 per cent for proximal and 50 per cent for isolated calf venous thrombosis (83 per cent). All venographically patent veins were fully compressible (specificity, 100 per cent). Abnormal intraluminal echoes were found in 18 of the 36 extremities with thrombosis but not in those with negative findings on venography. Sensitivity of intraluminal echogenicity was, therefore, 50 per cent and specificity, 100 per cent. For the Valsalva criterion, the standard lower normal limit of 10 per cent was applied, leading to 40 per cent sensitivity and 93 per cent specificity rates. In comparison with venography, ultrasonography underestimated the extent of thrombosis in 60 per cent of the true-positive examinations, although never to a clinically significant degree. In conclusion, compression ultrasonography, a technique based upon the unique criterion of venous compressibility, is a highly accurate and objective noninvasive diagnostic method, and is also suitable as a screening test. We urge clinicians to support their therapeutic decisions concerning the management of DVT with it.