Failure of computerized impedance plethysmography in the diagnostic management of patients with clinically suspected deep-vein thrombosis.

Before a new diagnostic modality can be introduced in clinical medicine, the validity of both a normal and abnormal test result have to be assessed prospectively in an appropriate patient group. We have evaluated the clinical validity of a new computerized impedance plethysmography (CIP) in the diagnostic management of 381 consecutive patients with clinically suspected venous thrombosis. In patients with serially normal CIP results, the diagnosis of venous thrombosis was refuted and, consequently, they were not treated with anticoagulant therapy and all were followed up for a period of 6 months to estimate the occurrence of symptomatic venous thromboembolism. The study was prematurely terminated by the safety monitoring committee because of an unacceptably high incidence of confirmed venous thromboembolism (10 patients, 3.2%; 95% confidence interval: 1.6% to 6%), including 4 episodes of fatal pulmonary embolism. In a subsequent explanatory study using ultrasonography in 29 other symptomatic patients who had at least 2 repeated normal CIP test results, the failure of CIP to detect proximal vein thrombosis was confirmed in 4 patients (14%). The reasons for this failure are probably related to the use of a modified device to measure impedance in the CIP apparatus, resulting in a lower ability to separate patients without venous thrombosis from those with the disease. We concluded that CIP is insensitive for the detection of proximal vein thrombosis and, therefore, not clinically useful in the diagnostic management of patients with suspected venous thrombosis.