Current issues of ACP Journal Club are published in Annals of Internal Medicine


Impedance plethysmography was not accurate in the detection of deep-vein thrombosis in high-risk, asymptomatic patients

ACP J Club. 1992 Mar-April;116:55. doi:10.7326/ACPJC-1992-116-2-055

Source Citation

Agnelli G, Cosmi B, Ranucci V, et al. Impedance plethysmography in the diagnosis of asymptomatic deep vein thrombosis in hip surgery. A venography-controlled study. Arch Intern Med. 1991 Nov; 151:2167-71.



To evaluate computerized impedance plethysmography (CIP) for the diagnosis of deep-vein thrombosis in asymptomatic high-risk patients.


The sensitivity, specificity, and positive and negative predictive values of CIP were calculated for consecutive patients scheduled for hip surgery and entered into 2 trials of pharmacologic deep-vein thrombosis prophylaxis.


Orthopedic department of a university-affiliated hospital in Italy.


246 patients with a negative CIP at admission to the study were entered; 206 had surgery for hip fracture (78% women; mean age, 74 years), and 40 had elective total hip replacement (63% women; mean age, 64 years). Patients received either placebo or 1 of 2 antithrombotic agents for 10 days postoperatively.

Description of test and diagnostic standard

CIP was done with a 45-s occlusion time test. If the result was positive, 4 more tests were done: for 45, 90, 45, and 90 s. The CIP was positive if all 5 tests were abnormal. CIP was done at baseline, then every 3 days during surveillance lasting until postoperative day 10 ± 1, or until a CIP was positive. When a CIP was positive, venography was done within 24 hours in the same leg and in the opposite leg as well if the first venogram was positive. If the CIP remained negative, bilateral venography was done on postoperative day 10 ± 1. Contrast venography was positive for deep-vein thrombosis in the presence of an intraluminal filling defect confirmed in 2 projections or after repeated injection of contrast medium.

Main results

All patients remained asymptomatic during surveillance. Complete surveillance and adequate venograms were obtained in 440 legs (89%); venography was inadequate in 6 legs and was not done in 46 legs (23 patients). Deep-vein thrombosis was identified in 145 legs: 81 in the operated leg and 64 in the nonoperated leg. There were 67 proximal and 78 isolated distal deep-vein thromboses. The sensitivity and specificity of CIP for overall deep-vein thrombosis were 22% and 87%, respectively, in the operated leg; 14% and 95%, respectively, in the nonoperated leg; and 19% and 91%, respectively, in both legs. {The likelihood ratio for a positive test result (LR+) for overall deep-ven thrombosis in the operated leg was 1.69 and the likelihood ratio for a negative test result (LR-) was 0.9. The LR+ in the nonoperated leg was 2.8 and the LR- was 0.9. The LR+ in both legs was 2.11 and the LR- was 0.9.}*


Computerized impedance plethysmography was not accurate in the detection of deep-vein thrombosis in high-risk, asymptomatic patients.

Source of funding: Not stated.

Address for article reprint: Dr. G. Agnelli, Istituto di Semeiotica Medica, Università di Perugia, Via Enrico dal Pozzo, 06100 Perugia, Italy.

*Numbers calculated from data in article.


This study provides several lessons. First, it reminds us that "negative" studies can be as useful as "positive" ones. Although studies showing no benefit of a test or an intervention have frequently been viewed as failures and not deserving publication, increasingly it is recognized that this information is often as important as that indicating a beneficial effect. Thus, although the premise behind this research makes sense, this study shows that the CIP (as well as the very similar conventional impedance plethysmography) is a poor test to use in surveillance for proximal deep venous thrombosis in patients after hip surgery.

Second, it reminds us that before we can rely on any test for patients in a specific clinical situation, the test must be evaluated using those patients. Because impedance plethysmography is so effective in the diagnosis of proximal deep venous thrombosis in symptomatic patients, it is easy to forget that it may not be equally effective in asymptomatic patients, even when their risk for deep venous thrombosis is high.

The most likely reason for the low sensitivity of the CIP is proximal thrombi that are not completely occlusive (1). Of the currently available noninvasive tests, only duplex ultrasonography has been used with success for screening high-risk postoperative patients for proximal deep venous thrombosis. One recent study found a sensitivity of 85.7% (12 of 14 patients with proximal deep venous thrombosis) and a specificity of 97.3% (287 of 295 patients without proximal deep venous thrombosis) (2).

John T. Philbrick, MD
University of VirginiaCharlottesville, Virginia, USA


1. Wheeler HB. Diagnosis of deep vein thrombosis. Arch Intern Med. 1991;151:2145-6.

2. Barnes RW, Nix ML, Barnes CL, et al. Perioperative asymptomatic venous thrombosis: role of duplex scanning versus venography. J Vasc Surg. 1989; 9:251-60.