The Diagnostic Approach to Deep Venous Thrombosis

C. Gregory Elliott, MD, Departments of Medicine and the Pulmonary Divisions of the LDS Hospital and University of Utah Health Sciences Center, Salt Lake City, Utah

Semin Respir Crit Care Med. 2000;21(6) 

In This Article

Diagnostic Tests for Deep Vein Thrombosis

Clinical Suspicion

Clinical information alone (history and physical examination) is insufficient for the diagnosis of deep vein thrombosis. Objective diagnostic testing is essential to diagnose deep vein thrombosis. However, objective tests for deep vein thrombosis are best interpreted with knowledge of the pretest probability that the patient has deep vein thrombosis.[7] Investigators have related clinical features to the pretest probability that a patient has deep vein thrombosis.[8] A scoring system ( Table 1 ) based upon risk factors, symptoms, and signs allows identification of patient subgroups who are unlikely or highly likely to have deep vein thrombosis.[9]

The principal disadvantage of clinical models that estimate clinical probability of deep vein thrombosis is the reproducibility of the models. Excellent interobserver reliability exists for one of the clinical models developed by Wells et al, [10] but extrapolation of these efforts into other settings (nonresearch) may increase errors, particularly when less experienced clinicians assess the likelihood that alternative diagnoses are responsible for lower extremity symptoms.

Contrast Venography

Contrast venography remains the reference standard for the diagnosis of deep vein thrombosis.[11] The technique is standard [12] ; however, it is important to assess the technical adequacy of venography. Failure to visualize all deep veins from the calf to the inferior vena cava can lead to missed diagnoses. The finding of a constant intraluminal filling defect on two or more views confirms the diagnosis of acute deep vein thrombosis.[12] Treatment with anticoagulants can be withheld safely when a technically adequate contrast venogram shows no evidence of deep vein thrombosis.[13]

There are several disadvantages of contrast venography. Many patients cannot undergo venography because of pain or lack of venous access. The test is invasive and thus may injure patients. Phlebitis or deep vein thrombosis can complicate the procedure. The contrast agent can cause idiosyncratic reactions such as urticaria, angioedema, bronchospasm, or cardiovascular collapse. It may also directly injure the kidney. For this reason, renal insufficiency (serum creatinine > 2.0 mg/dL) is a relative contraindication. Furthermore, venography is more expensive than its noninvasive counterparts.

Because of the many disadvantages, contrast venography is seldom the first diagnostic test when a physician suspects deep vein thrombosis. Venography finds its principal use when the diagnosis remains in question after clinical evaluation and initial noninvasive diagnostic testing. Investigators also use venography as an endpoint for clinical investigations of methods to prevent deep vein thrombosis.

Compression Ultrasonography and Venous Imaging

Compression ultrasonography combines real-time imaging of the deep veins with venous compression to diagnose deep vein thrombosis. The technique can be applied to deep veins of the upper and lower extremities.[14] Lower extremity venous imaging is performed with the patient supine and the head elevated (10-20 degrees). The leg is positioned with the knee slightly elevated and the hips slightly rotated externally. Upper extremity examinations also take place with the patient supine and the arm raised to permit access to the axilla.

The technique is not standardized; it varies from laboratory to laboratory. Some laboratories limit the examination to the common femoral vein and the popliteal vein, whereas other laboratories examine the deep veins from the inguinal region to the calf, including the calf veins. Some laboratories perform bilateral examinations routinely, even though symptoms or signs suggestive of deep vein thrombosis are present in one leg. Even the nomenclature used to identify the deep veins differs, and the use of the term superficial femoral vein (actually a deep vein) can confuse physicians.[15]

Compression ultrasonography with B-mode venous imaging is both sensitive and specific for the detection of proximal deep vein thrombosis in symptomatic patients. The sensitivity has ranged from 89 to 100% for the diagnosis of symptomatic deep vein thrombosis ( Table 2 ).[4,16,17,18,19,20] The specificity has ranged from 95 to 100%.

Compression ultrasonography has disadvantages. Morbid obesity, severe edema, casts, and other immobilization devices can limit the examination. Furthermore, the sensitivity and specificity of compression ultrasonography for the detection of proximal and calf deep vein thrombi are reduced in certain clinical settings, for example, for asymptomatic high-risk postoperative patients ( Tables 3 and Table 4 ).[18,21,22,23,24,25,26,27,28,29,30,31] The addition of color flow Doppler does not improve the sensitivity of compression ultrasonography for the detection of proximal deep vein thrombi in high-risk asymptomatic patients ( Table 5 ).[28,32,33] Compression ultrasonography has limited utility for the detection of isolated iliac vein thrombosis, [6] and it is less sensitive for the detection of isolated calf vein thrombosis in symptomatic patients ( Table 6 ).[34,35]

Technically limited studies may contribute to the reduced sensitivity. The sensitivity improves when investigators select only the compression ultrasonography examinations with complete visualization of the calf veins. For example, Rose et al [31] reported a sensitivity of 95% and a specificity of 100% for technically complete studies and a sensitivity of 30% and specificity of 70% for technically inadequate studies. However, technically limited studies occur often enough to compromise the usefulness of compression ultrasonography for evaluation of calf thrombosis.

In spite of the disadvantages, physicians commonly use compression ultrasonography as the initial diagnostic test for patients with symptoms or signs that suggest deep vein thrombosis. The safety, availability, and well-documented accuracy of this technique justify its widespread use. Furthermore, clinical outcome studies demonstrate the safety of withholding anticoagulant treatment when compression ultra-sonography tests do not detect deep vein thrombosis in symptomatic outpatients.[36,37,38] Research has refined the timing of compression ultrasonography for the management of outpatients suspected of having deep vein thrombosis. Birdwell [37] and Cogo [38] and their colleagues have shown that anticoagulants can be withheld safely when compression ultrasonography is negative on the day of presentation and again 5 to 7 days later. In the Birdwell study, 335 patients with two negative compression ultrasonography examinations were followed for 3 months. Only 2 of these patients (0.6%, 95%CI = 0.1-2.1%) experienced a venous thromboembolic event during follow-up. In the Cogo study, 9 of 1290 patients (0.7%, 95%CI = 0.3-1.2%) experienced a venous thromboembolic event on follow-up.

Combining D-dimer tests or clinical probability with a single compression ultrasonography examination also may permit safe management of outpatients suspected of having deep vein thrombosis. Bernardi et al [39] followed 598 patients who had a normal compression ultrasonography examination and a negative D-dimer test (enzyme-linked immunosorbent assay, ELISA; Instant-IA d-dimer kit, Stago, Asnieres, France). Only one patient (0.2%, 95%CI = 0.0-0.9%) had a venous thromboembolic event during 3 months of follow-up without anticoagulants. Wells et al [8] combined clinical probability with compression ultrasonography and reported that only 3 of 481 patients (0.6%, 95%CI = 0.1- 1.8%) with a low or moderate clinical probability and a negative compression ultrasonography developed venous thromboemebolism during 3 months of follow-up.

Impedance Plethysmography

Occlusive-cuff impedance plethysmography (IPG) is an established noninvasive technique for detecting proximal vein (popliteal, femoral, and iliac veins) thrombosis in patients with a first episode of clinically suspected venous thrombosis. The technique measures blood volume changes in the leg as a change in electrical resistance (impedance).[40] The IPG instrument delivers a weak constant current that passes through the calf and is detected by electrodes on the thigh. Changes in blood volume are produced by inflation of a thigh blood pressure cuff to a pressure that exceeds venous pressure but is less than arterial diastolic pressure (e.g., 50 mm Hg). When the cuff is deflated, rapid venous outflow occurs in normal legs; whereas venous emptying is much slower when venous thrombosis is present.[40] Unfortunately, other causes of impaired venous out-flow (e.g., pelvic tumor or elevated venous pressure) can cause false-positive tests.[41]

A standard protocol is essential.[42] The protocol includes positioning of the leg to avoid compression of the popliteal and femoral veins as well as standard inflation of the cuff and the use of a validated graph to plot the results.

Numerous investigations have reported the sensitivity and specificity of IPG in patients with clinically suspected deep vein thrombosis ( Table 7 ). Impedance plethysmography is both sensitive and specific for the diagnosis of proximal deep vein thrombosis in symptomatic patients when venography is the reference standard.[36,43,44,45,46,47,48] In contrast, impedance plethysmography is not sensitive for the detection of proximal deep vein thrombi in asymptomatic high-risk patients such as patients who have recently undergone hip arthroplasty or stabilization of hip fracture ( Table 8 ).[49,50,51]

Magnetic Resonance Imaging

Magnetic resonance imaging may provide a non-invasive alternative to venography. The technique can be applied to both upper and lower extremities, and it is particularly useful for the detection of pelvic vein thrombi. Investigators have suggested that magnetic resonance imaging is both sensitive and specific for the detection of proximal deep vein thrombosis in symptomatic patients ( Table 9 ).[52,53,54,55,56]

Limited data suggest that magnetic resonance imaging is less sensitive for the detection of isolated calf vein thrombosis ( Table 10 ).[54,57]

The sensitivity and specificity of magnetic resonance imaging remain poorly defined for asymptomatic high-risk postoperative patients.


D-dimer assays detect fragments produced by clot lysis. A number of different techniques (ELISA, latex, rapid ELISA) are available. The methods are quite different ( Table 11 ), and clinicians must understand the characteristics and limitations of the test that they request. For example, latex D-dimer assays are insensitive and have no role in screening for deep vein thrombosis. In addition, the clinician must know the evidence that validates the cutoff values in their clinical environment; that is, published results may not apply to other hospital laboratories. In general, D-dimer assays with high sensitivity may be useful for excluding the diagnosis of acute deep vein thrombosis, particularly when the pretest probability for the disease is low.[58,59] However, widespread use of D-dimer assays combined with low clinical scores to withhold anticoagulants awaits studies that report the safety of this approach.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.