The clinical diagnosis of deep vein thrombosis (DVT) is notoriously unreliable. Confirmation is necessary before subjecting patients to the risk of long-term anticoagulation. Ascending venography was the reference standard for the diagnosis of DVT, but it is invasive and associated with adverse effects. Consequently, venous duplex ultrasonography has become the routine initial evaluation of DVT.Ultrasonography or impedance plethysmography are sensitive and specific for the diagnosis of proximal, occlusive venous thrombosis. An abnormal result virtually confirms the diagnosis and treatment can begin. If a normal result is obtained but clinical suspicion remains high, then serial testing is indicated.Patients with recurrent DVT may have residual organised thrombus which makes non-invasive tests less useful and difficult to interpret. Serial scanning or the results of previous surveillance may help.The majority of the thrombi in asymptomatic high-risk patients are in the calf veins and are often non-occlusive. Ultrasonography has improved the detection of DVT in this group of patients and is sometimes used to check for DVT in high-risk patients following surgery.


Venous thromboembolism is an important cause of morbidity and mortality. In most (90%) cases of pulmonary embolism (PE), the thrombi originate from the deep veins of the legs. Proximal (femoro-popliteal) deep vein thrombosis (DVT) is associated with 10% risk of fatal PE and at least 50% risk of PE or recurrent DVT. Although calf vein thrombosis is associated with low risk of PE, extension into the proximal venous system occurs in 20-30% of cases. Early diagnosis and treatment of proximal DVT are essential in preventing PE, reducing the risk of recurrent DVT and minimising the long-term complications of chronic venous hypertension.

Clinical diagnosis
The clinical features of DVT are non-specific. More than half the cases of clinically suspected DVT are not confirmed by objective testing (Table 1). Furthermore, patients with significant DVT may have only minimal symptoms and signs. The diagnosis is improved by considering the number and type of risk factors present (Table 2).1

Objective tests
Objective testing (Fig. 1) is mandatory to confirm the diagnosis of DVT before subjecting the patients to anticoagulation. The usefulness and limitations of each test can be considered under 3 patient groups:

– initial episode of DVT

– recurrent DVT

screening of DVT in:

- patients with suspected pulmonary embolism

- selected postoperative patients with a previous history of DVT, PE or thrombophilia

- selected patients following major joint replacement

- selected high-risk patients (e.g. neurosurgical) who are not receiving anticoagulant prophylaxis

Ultrasonography is now the first test for DVT. It can be used to diagnose proximal DVT or screen high-risk patients after surgery. Serial scans can be used to follow up patients on treatment, untreated patients with calf vein thrombosis and those with a high clinical probability of DVT but an initial normal result. Ultrasound is also used to screen for DVT in patients with suspected PE; however, a negative result does not exclude PE.

Table 1

Differential diagnosis of DVT

Venous disorders

Chronic venous hypertension

– panniculitis

– lipodermatosclerosis

– venous incompetence

Superficial thrombophlebitis

Extrinsic compression of the veins

– tumour

– haematoma

– arterial aneurysm

Orthopaedic disorders

Ruptured Baker's cyst

Muscle strain, injury, tear or haematoma

Compartment syndrome



Lymphatic disorders

Lymphoedema and lymphangitis


Reperfusion oedema or post-bypass surgery

Leg oedema in a paralysed limb

Generalised oedematous state

Fig. 1

Diagnostic approach to patients with clinically suspected DVT


Colour duplex ultrasonography allows direct visualisation of the deep veins and detection of venous flow. It is non-invasive, safe and readily repeatable. Diagnosis of DVT is based on:

– failure of the vein to compress

– presence of a thrombus (Fig. 2)

– absence of venous flow augmentation during compression of the calf

Non-compressibility of the vein is the most sensitive and specific feature of proximal DVT (>95%).

The value of venous scanning may be limited in patients with recurrent DVT if a previous scan is not available. In this situation, progress scanning at 1-2 weeks after starting anticoagulation, and again at cessation of anticoagulation, will provide a useful baseline for future management.

Calf vein thrombosis
Venous ultrasonography was previously reported to be less sensitive (<60%) for calf vein thrombosis. The results have improved with recent studies (75-85% sensitivity) using experienced staff and a standard protocol.2 If calf DVT is found, there are two options. Firstly, postoperative patients or patients with ongoing risk factors can be treated until progress scanning confirms non-propagation of the thrombus. Secondly, in patients who are fully ambulant and whose risk factors are no longer present, anticoagulant may be withheld, but progress scanning after 5-7 days is necessary to detect propagation. If a calf DVT is not demonstrated, serial ultrasound scanning is the preferred diagnostic approach in patients with a high clinical probability of DVT.

Until the mid 1980s, the diagnosis of DVT was largely dependent on venography. Although being replaced by ultrasonography, it is still regarded as the `gold standard' for diagnosing DVT. Radio-opaque contrast medium is injected into a dorsal vein on the foot. A DVT is present if there is an intraluminal filling defect which remains constant in shape and location in at least two different projections.

Unfortunately, the test is invasive, technically difficult and up to 20% of venograms are inadequate. Adverse effects include pain, superficial phlebitis, hypersensitivity reactions, cutaneous necrosis and chemical cellulitis from extravasation, and rarely the precipitation of acute renal failure.

Venography may still have a role in diagnosing acute DVT in patients with recurrent thrombosis, screening for DVT in asymptomatic high-risk patients and diagnosing abdominal (iliac or inferior vena cava (IVC)) venous thrombosis. It can be used to assess the venous anatomy before the insertion of an IVC filter.

Table 2

Risk factors for and clinical probability of DVT

Risk factors for DVT

Age >40 years

Recent major surgery

Immobility: bed rest >3 days, intensive care, long distance travel

Previous history of DVT or PE

Venous disorders such as varicose veins, or foreign bodies such as indwelling catheters

Strong family history of DVT or PE (2 first degree relatives)

History of recent trauma or surgery, especially of pelvis or lower limbs


Paralysis of lower limb(s), plaster immobilisation

Severe obesity

Oral contraceptives, pregnancy, puerperium

Known thrombophilia:

– deficiency of protein C, protein S or antithrombin III

– activated protein C resistance or presence of Factor V Leiden

– positive anticardiolipin antibody or lupus anticoagulant

Associated medical conditions:

– cardiac disorders: heart failure, acute myocardial infarction

– chest infection

– inflammatory bowel disease

– nephrotic syndrome

– Behcet's syndrome

– polycythaemia, paroxysmal nocturnal haemoglobinuria

– homocystinaemia

– paraproteinaemia

Clinical probability of DVT

High probability:

two or more risk factors present and no alternative diagnosis

Low probability:

no risk factor with alternative diagnosis

Intermediate probability:

all other combinations

Impedance plethysmography (IPG)
IPG measures the volume changes of the limb in response to the inflation and deflation of a thigh cuff. It is non-invasive, safe and requires minimal training. The test can be done at the patient's bedside and can be easily repeated for serial measurements.

The test is highly sensitive (92%) and specific (95%) for symptomatic proximal DVT, and has a high positive predictive value (90%). A normal result essentially excludes the diagnosis of proximal DVT.

The sensitivity of IPG is low in calf vein thrombosis (20%) and in screening for DVT in asymptomatic postoperative high-risk patients (22%).

Fig. 2

A subtotal occlusive thrombus in the right popliteal vein is shown in this colour Doppler image.


Iodine-125 fibrinogen scan
This was useful in diagnosing calf vein thrombosis (sensitivity 80% and specificity 75%). The diagnosis was made when there was a 20% increase in isotope counts compared to the corresponding site on the other leg. The product is no longer available. There was a risk of blood-borne virus transmission because the fibrinogen was derived from human plasma.

Magnetic resonance venography
This is a relatively new technique with a sensitivity and specificity >90% for DVT. It is not a routine test as it is expensive and not widely available. Nevertheless, it is potentially useful in diagnosing DVT involving the pelvis, abdomen and upper extremities.

D-dimer, a product of fibrinolysis, is elevated in venous thromboembolism. This blood test has a high sensitivity and high negative predictive value (>95%), but a low specificity and low positive predictive value (<50%). The test is therefore useful if it is normal. A normal result together with a normal non-invasive test (ultrasonography or IPG) reliably excludes DVT. A newer rapid d-dimer test (SimpliRED) is being developed for use in emergency situations.

There have been major advances in the diagnostic tests for DVT. In experienced hands, colour duplex ultrasonography has replaced the conventional contrast venography as the investigation of choice. It can be used in the screening and surveillance of patients at high risk of DVT or in those with suspected PE.

Further reading
Koopman MM, van Beek EJ, ten Cate JW. Diagnosis of deep vein thrombosis. Prog Cardiovasc Dis 1994;37:1-12.

Stacey MC. Diagnosis of deep vein thrombosis [editorial; comment]. Med J Aust 1995;163:509-10.

Allen PL. The role of ultrasound in the diagnosis of deep vein thrombosis [editorial]. Br J Hosp Med 1997;57:546-7.

Conner G, McGrath M. The outpatient diagnosis and management of the patient with deep vein thrombosis. St Vincent's Clinic Proceedings 1997;5:30-6.

Thomas DA, deBoisblanc BP, Summer WR. Venous thromboembolism. A contemporary diagnostic and therapeutic approach. Postgrad Med 1997;102:179-81, 185-7, 191-4.


Tay Jam Chin

Fellow, Vascular Medicine, St Vincent's Hospital, Sydney

Michael McGrath

Senior Staff Specialist and Consultant, Vascular Medicine, St Vincent's Hospital, Sydney