[Instrumental diagnosis of deep venous thrombosis. Prospective study with non-invasive methods].

The clinical diagnosis of deep vein thrombosis is unreliable and venography remains the best single method of investigation. In the last ten years several non-invasive procedures have been introduced in the vascular laboratories, sometime without adequate assessment of their reliability. We tested sensitivity and specificity of three different non-invasive procedures, namely phleboscintigraphy with 99Technetium, Doppler ultrasound technique and strain-gauge plethysmography in patients with clinically suspected deep vein thrombosis of lower limbs or pulmonary embolism. A total of 288 patients entered the study. Venography was used as the reference standard and was assessed independently, without knowledge of the results of non-invasive methods. In the first 36 patients phleboscintigraphy and Doppler ultrasound were evaluated: sensitivity and specificity of phleboscintigraphy resulted only a little more than 50%; the method was therefore considered unreliable and subsequently abandoned. Doppler ultrasound could be evaluated upon a total of 81 patients, showing an unsatisfactory sensitivity (63%), but a quite reliable specificity (86%). Best results were obtained with strain-gauge plethysmography, using maximal venous outflow and venous capacitance as diagnostic parameters. 209 patients entered this study, and a sensitivity of about 90% with a specificity of about 95% was observed. If we consider only acute proximal deep vein thrombosis, sensitivity approaches 97%, while it is only 60% in distal deep vein thrombosis.