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


Color Doppler ultrasound was not an accurate diagnostic test to detect proximal leg vein thrombosis in asymptomatic, high-risk patients

ACP J Club. 1993 Mar-April;118:51. doi:10.7326/ACPJC-1993-118-2-051

Source Citation

Davidson BL, Elliott CG, Lensing AW, and the RD Heparin Arthroplasty Group. Low accuracy of color Doppler ultrasound in the detection of proximal leg vein thrombosis in asymptomatic high-risk patients. Ann Intern Med. 1992 Nov 1;117:735-8.



To evaluate the accuracy of color Doppler ultrasound (CDU) in the detection of proximal deep-vein thrombosis (DVT) in asymptomatic, high-risk patients.


Independent, blinded comparison of CDU with ascending contrast venography in consecutive patients.


7 university and community hospitals in North America.


319 of 385 consecutive eligible patients > 18 years of age, who were receiving low-molecular-weight heparin or warfarin after elective unilateral hip or knee arthroplasty, had both CDU and venography. Exclusion criteria were contraindications to anticoagulants or venography; bleeding, procoagulation tendency, or a history of venous thromboembolism; recent anticoagulant or antiplatelet therapy; or refusal.

Description of test and diagnostic standard

CDU was done before venography, and both tests were assessed without knowledge of other results. CDU examined the popliteal vein with the patient in a supine position, using gentle but firm transducer probe compression. The test was considered normal if both full vein compressibility and complete filling of the vein with color were shown. Calf veins were not examined. Venography was done with 75 to 175 mL of contrast material injected into a dorsal foot vein and spot films of the calf, popliteal, thigh, and iliac vein regions. Thrombosis was defined as a constant intraluminal filling defect or abrupt cutoff in deep veins despite repeated injections. Proximal vein thrombosis was defined as any involvement of proximal veins irrespective of calf veins.

Main outcome measures

Proximal DVT and occlusive thrombosis.

Main results

Contrast venography identified 80 cases of DVT in 319 patients; 21 of these were proximal DVTs, and 4 of these 21 were occlusive thrombi. CDU identified 8 of 21 proximal DVTs and 3 of 4 occlusive proximal thrombi plus an additional 8 patients with calf DVT near the popliteal vein. CDU had low sensitivity and high specificity for detecting proximal vein thrombosis (Table). The positive predictive value of CDU was 25% {CI, 12% to 45%}*.


Color Doppler ultrasound was not an accurate diagnostic tool to detect proximal deep-vein thrombosis in asymptomatic, high-risk patients.

Source of funding: In part, Wyeth-Ayerst Research.

For article reprint: Dr. B.L. Davidson, Allegheny University Hospitals, 301 South 19th Street, Philadelphia, PA 19103, USA. FAX 215-545-6871.

*CI calculated from data in article

Table. Test features for diagnosing proximal-vein thrombosis†

Test Sensitivity (95% CI) Specificity (CI) +LR (CI) -LR (CI)
Color Doppler ultrasound 38% (18 to 62) 93% (89 to 95) 4.94 (2.42 to 9.10) 0.67 (0.44 to 0.86)

†+LR = likelihood ratio for presence of the disease if test is positive; -LR = likelihood ratio if test is negative. LRs calculated from data in article.


Impedance plethysmography or ultrasonography were effective for detecting deep-vein thrombosis

Radiocontrast venography is the gold standard for the diagnosis of DVT. Venography, however, is expensive, invasive, and involves risks including radiocontrast allergy, renal failure, and DVT itself. For these reasons, noninvasive approaches have been widely used in patients with suspected DVT.

Three of the most commonly used diagnostic tests for DVT are IPG, duplex ultrasound, and color Doppler ultrasound. IPG measures the rate of blood flow from the leg and is abnormal if thrombi occlude the femoral or iliac vein. Duplex ultrasound visualizes the veins of the thigh, and DVT is diagnosed if the vein cannot be compressed with gentle pressure. Color Doppler ultrasound combines the findings of duplex ultrasound with color Doppler, which detects the blood flow and complete filling of the vein.

The studies by Davidson and colleagues and Heijboer and colleagues emphasize the need to interpret the test characteristics of noninvasive venous studies in the context of the setting in which they are done (1). Heijboer found excellent sensitivity but only moderate specificity with both the IPG and duplex ultrasound in symptomatic hospitalized patients. The 14% false-positive rate for US was much higher than studies done in outpatients (2) and may reflect the presence of other illnesses that lead to false positives. Davidson and colleagues found the sensitivity of color Doppler ultrasound was only 36% in asymptomatic high-risk patients having elective hip or knee replacement. This test may have had a lower sensitivity because the thrombi are frequently smaller and nonocclusive in asymptomatic patients receiving antithrombotic prophylaxis. Both studies confirm that each of these tests has a lower sensitivity in detecting distal DVT.

These and other studies suggest the following conclusions and diagnostic approach to DVT: 1) Duplex ultrasound and IPG have similar test characteristics. Color Doppler ultrasound has not been shown to be superior to duplex ultrasound alone. The test that should be used most often depends on the experience and expertise of the laboratory doing the study. 2) In symptomatic patients, if the test is positive, treat for DVT unless the patient is at high risk for bleeding complications from anticoagulation. Venography is indicated to confirm the diagnosis in those patients. If the initial test is negative, do serial tests to detect the migration of the clot from the distal to the proximal venous system. Withholding anticoagulation from patients with negative serial tests has been shown to be safe (3, 4). In asymptomatic high-risk patients, it is mandatory to use prophylactic therapy to reduce the risk for DVT and pulmonary embolism (5). Noninvasive studies are insensitive in this setting, and cost-effective screening strategies are not available at this time.

Jeffrey L. Carson, MD
University of Medicine and Dentistry ofRobert Wood Johnson Medical SchoolNew Brunswick, New Jersey, USA


1. Wheeler HB. Diagnosis of deep-vein thrombosis. Using the right test for the right purpose. Arch Intern Med. 1991;151:2145-6.

2. Lensing AW, Prandoni P, Brandjes D, et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med. 1989;320:342-5.

3. Hull RD, Hirsh J, Carter CD, et al. Diagnostic efficacy of impedance plethysmography for clinically suspected deep-venous thrombosis. Ann Intern Med. 1985;102:21-8.

4. Huisman MV, Büller HR, ten Cate JW, et al. Management of clinically suspected acute venous thrombosis in outpatients with serial impedance plethysmography in a community hospital setting. Arch Intern Med. 1989;149:511-3.

5. Ciagett GP, Anderson FA, Levine MN, Salzman EW, Wheeler HB. Prevention of venous thromboembolism. Chest. 1992;102(Suppl):391S-407S.