Diagnostic approach to patients with suspected deep venous thrombosis of the lower limbs
(Portuguese PDF version)

Hamilton Almeida Rollo,1 Veronica Barreto Fortes,2 Archângelo Tarciso Fortes Junior,2 Winston Bonetti Yoshida,3 Sidnei Lastória,1 Francisco Humberto de Abreu Maffei,4

1. Physician. Assistant professor, Vascular Surgery, Department of Surgery and Orthopedics, Faculdade de Medicina de Botucatu (FMB), Universidade Estadual Paulista (UNESP), Botucatu, SP, Brazil.
2. Master in Vascular Surgery, Department of Surgery and Orthopedics, FMB/UNESP, Botucatu, SP, Brazil.
3. Adjunct professor, Vascular Surgery, Department of Surgery and Orthopedics, FMB/UNESP, Botucatu, SP, Brazil.
4. Professor, Vascular Surgery, Department of Surgery and Orthopedics, FMB/UNESP, Botucatu, SP, Brazil.

Hamilton Almeida Rollo
Vascular Surgery
Department of Surgery and Orthopedics - Medical School
CEP 18618-970 - Botucatu, SP, Brazil
Phone: +55 14 3811.6269/3811.6092
Fax: +55 14 3815.7428
E-mail: hrollo@fmb.unesp.br


Deep venous thrombosis is a relatively common disease, which can present pulmonary embolism as a complication in its acute phase, and later the post-thrombotic syndrome. Thus, diagnosis should be made as soon as possible, in order to prevent or minimize such complications. Several studies have shown that the symptoms and the clinical signs are inaccurate for the deep venous thrombosis diagnosis and that complementary exams are necessary. As an attempt to simplify the patients' assessment, Well et al., in 1997, developed a clinical prediction index that combines symptoms, signs and risk factors for deep venous thrombosis and managed to make a simpler approach through an association of this index with the complementary exams. Phlebography has been considered the gold standard of complementary exams. However, since it is an invasive exam and thus subject to complications, other diagnostic methods were introduced aiming at making the diagnostic approach simpler and less invasive. Doppler ultrasound, duplex scan, impedance plethysmography, computed tomography, and blood tests such as the D-dimer are some of the available methods for assessing the patient with suspicion of deep venous thrombosis. Among them, duplex scan has shown excellent accuracy and it is currently widely accepted as the first choice test for approaching the patient with deep venous thrombosis. Several authors have suggested an association of diagnostic methods to simplify and make the assessment of such patients more cost-effective, leading to the introduction of a wide range of diagnostic strategies. The different diagnostic methods used for assessing deep venous thrombosis are discussed, as well as a review of the literature on the accuracy, advantages and disadvantages of these methods.

Key-words: venous thrombosis, diagnosis, lower extremity.
Palavras-chave: trombose venosa, diagnóstico, extremidade inferior.

J Vasc Br 2005;4(1):79-92

Deep venous thrombosis (DVT) is the third cardiovascular disease most frequent in the USA. Anderson et al. estimated around 170,000 new cases of DVT or pulmonary embolism (PE) a year, and 90,000 relapses over the same period, resulting in at least 13,000 deaths every year.1

Also in the USA an incidence of 0.8 cases per 1,000 inhabitants/year2 was estimated, and in Sweden Nylander & Olivercrana estimated an incidence of 0.9 cases per 1,000 inhabitants/year.3 In our environment, Maffei estimated 0.6 cases per 1,000 inhabitants/year, based on DVT cases confirmed by phlebography or duplex scanning (DS).4 In a review work using metanalysis in 2003, Fowkes et al.5 estimated a worldwide DVT incidence of 0.5 cases per 1,000 inhabitants/year.

DVT, which is quite common in hospitals, affects 84 people per 100,000 every year,1 being the most common cause of morbidity/mortality in surgical patients and responsible for 300,000 to 600,000 hospitalizations a year.6 DVT is present in 20% to 35% of intra-hospital mortality and is associated to PE in 10% to 20% of cases, according to necropsy-based studies.7-10 These data are similar to those found by Maffei et al.,11 in a work done with necropsies in patients who died at the Hospital das Clínicas da Faculdade de Medicina de Botucatu, in which they found PE in 19.1%. Among these patients, the cause of death was PE in 3.7% of cases.

DVT is a disease that has been widely studied. Its cause may be analyzed based on Virchow's triad described in 1860. Stasis, endothelium injury and hypercoagulability are, independently or combined, the factors involved in the etiopathogeny of DVT. They relate to the main known risk factors. Immobilization, paresis or paralysis, and prolonged rest lead to blood stasis. Orthopaedic, gynecological and abdominal surgeries create vessel wall injuries. Neoplasia, thrombophilia and the use of oral contraceptives or hormone replacement therapy, among others, are causes of hypercoagulability.4

The typical complaint by the patient with DVT is the presence of edema and pain in the lower limb, which may involve the leg or the whole limb, depending on the site and extension of the thrombus. The presence of triggering factors, previous DVT, and family history should be investigated. Initial manifestation can be thoracic pain, dyspnea, hemoptysis, or shock suggesting pulmonary thromboembolism.

The physical examination might reveal edema, erythema, dilatation of the superficial venous system, temperature increase, and muscular tenderness painful in response to palpation.

However, there are affections that can occur with a similar clinical picture, such as cellulitis and Baker's cyst, which may lead to a diagnostic confusion.

Until the development of reliable complementary examinations, the DVT diagnosis was made clinically with low accuracy. In 1940 Bauer introduced phlebography to confirm the DVT diagnosis, but due to technical difficulties it did not initially receive general acceptance.12 This only occurred in the 1960's due to an improvement on radiological and contrast techniques. Haeger, in the late 1960's showed that in 54% of patients receiving treatment for DVT the venous system had normal phlebography.13 Thus, when approaching patients with DVT, it is important to confirm the diagnosis. If DVT is left untreated, the patient is subject to its complications, such as PE and the post-thrombotic syndrome (PTS), which are the most frequent complications. PTS is present in 10% of cases after 2 years and 30% after 5 years, and it is the cause of disability in patients at a productive age.14

Such DVT complications can be minimized by early anticoagulant therapy. On the other hand, despite being highly efficient in preventing progression of thrombosis, embolization, and disease recurrence, the anticoagulant therapy is associated to complications such as hemorrhage in approximately 5% of cases and heparin-induced thrombocytopenia in around 1% of cases.15 Moreover, the anticoagulant therapy implies in blood tests and frequent returns to hospital.

Because of that, whenever possible, the anticoagulant therapy should be administered after the diagnosis is confirmed by objective examinations. Thus, the unnecessary patient exposure to anticoagulation risks is avoided.


The clinical diagnosis of DVT can be problematic, since several venous thrombi do not provoke the typical symptoms of the disease because they do not totally obstruct the vein or because there is a development of the collateral circulation. Besides, as we have seen, 30% to 50% of patients with DVT symptoms and signs do not present the disease, because these symptoms and signs can occur in other affections.4

Haeger, in 1969 studied 72 patients with clinical suspicion of DVT. When phlebography was performed, he found the presence of thrombus in only 46% of these patients. Among the symptoms and signs assessed by him, calf pain was the most sensitive (90%), while dilated superficial veins and Lowenberg's sign had the highest specificity (82% and 85%, respectively).13 However, none of the signs individually presented good sensitivity and specificity and, even among patients with high suspicion, positive and negative predictive values were 0.55 and 0.66, respectively. The author concluded that the clinical signs were not reliable for the DVT diagnosis.

Richards et al.16 found a DVT occurrence of 40% diagnosed by phlebography in patients with clinical suspicion. Calf pain was the most sensitive symptom (62%) and painful palpable cord in the deep venous system tract was the most specific (98%) and the less sensitive (10%). In a similar study, Sandler et al.17 found 58% of DVT. When symptoms and signs were assessed, they found a good sensitivity for pain (86%), edema (97%), and temperature increase (72%). However, these symptoms and signs presented low specificity.

Rollo et al.18 performed a phlebographic study in 424 patients with clinical suspicion of DVT and found 68.6% of patients with confirmed diagnosis. In another study done by Lastória et al.19, calf pain, edema, and muscular tenderness (86.7%) were the most frequently found symptoms and signs in 188 patients with DVT diagnosed by phlebography.

Such facts could suggest that the clinical assessment was of little importance in patients with suspicion of DVT. Nevertheless, the symptoms, signs, and risk factors, as a group, could be used to determine the clinical probability of DVT, which can be useful when approaching these patients.


Wells et al., in 1995,20 developed a clinical model to classify patients concerning the probability of presenting DVT. They classified this probability based on criteria to which they gave a score. Such criteria were divided into three categories: symptoms and signs, risk factors, and probable differential diagnosis. According to the scoring, they were classified into high, moderate, and low probability of presenting DVT (Table 1).

click hereTable 1 - Clinical model to determine the probability of presenting deep venous thrombosis proposed by Wells et al.20

Major criteria
Active neoplasia
Paralysis, paresis or recent plaster immobilization of lower limbs
Recent rest in bed for more than 3 days and/or major surgery in the last 4 weeks
Increase in sensitivity along the deep venous system veins
Edema of all the limb
Calf edema (>3 cm) compared to the normal leg
Family history of deep venous thrombosis (two or more first-degree relatives)
Minor criteria
History of recent trauma in the symptomatic leg
Major pitting edema on the affected leg (unilateral)
Superficial collateral veins
Previous hospitalization over the past 6 months
Clinical Probability
= three major criteria and no differential diagnosis
= two major criteria + = two minor criteria + no differential diagnosis
1 major criterion + = 2 minor criteria + a differential diagnosis
1 major criterion + = 1 minor criteria + no differential diagnosis
0 major criterion + = 3 minor criteria + a differential diagnosis
0 major criterion + = 2 minor criteria + no differential diagnosis
All other combinations


In order to make its use in clinical practice easier, such model was modified after an evaluation using analysis and logistic regression and, in 1997, the outcome of its application in 593 patients with clinical suspicion of DVT was published.21 Risk factors for DVT included active cancer, plaster immobilization, paralysis, and major surgery, as well as clinical characteristics, such as edema of the limb, difference in calf diameter higher than 3 cm, presence of visible superficial venous circulation, and pain in deep vein tracts. For each of the criterion present in the patient, one point would be added to the score. When the patient presented a differential diagnosis more likely to be cellulitis, lymphedema, Baker's cyst or muscle hematoma, for instance, two points would be subtracted from the score (Table 2).

click hereTable 2 - Modified clinical model to determine the probability of presenting deep venous thrombosis proposed by Wells et al.21

Clinical characteristics Score
Active cancer 1
Paresis, paralysis or plaster immobilization of lower limbs 1
Immobilization (> 3 days) or major surgery (up to 4 weeks) 1
Increase in sensitivity along the deep venous system veins 1
Edema of all the limb 1
Calf edema (> 3 cm) compared to the normal leg 1
Major pitting edema on the affected leg (unilateral) 1
Superficial collateral veins 1
Most likely differential diagnosis -2

According to the presented model, patients would be classified concerning the probability of presenting DVT into three groups:

    1 Low probability when the score is equal or lower than 0.
    2 Moderate probability when the score is 1 or 2.
    3 High probability when the score is equal or higher than 3.

Kahn, in a review article, compared three other prediction models published on the literature to Wells' model, concluding that the latter presents a better methodology and a higher potential of becoming useful in the management of the patient with suspicion of DVT.22 This clinical model was reproduced by several authors.23-25 In a recent study performed in Switzerland, Miron et al.26 developed a new model and compared it to Wells' model, applying both models in 270 cases of suspicion of DVT. Both models proved to be efficient for classifying patients as to the risk of presenting DVT. However, the authors26 found some disadvantages in Wells' test, namely, the fact that it excludes patients with previous DVT and it does not take into account a family history of DVT.

Wells et al., in 2003,27 performed a new modification in the prediction model, including previous DVT confirmed by objective examination among risk factors. The classification was simplified and patients with a score lower than two were considered possibly without DVT, and those with score equal or higher than two possibly with DVT. This model was applied in 1082 patients associated to DS and to the D-dimer (DD). Around 18% of included patients presented a history of DVT.


Non-invasive methods

Ultrasound Doppler

The continuous-wave ultrasound Doppler is a non-invasive examination at a low cost, easy to perform, that can be repeated with no restrictions. For this reason, it has been considered in the work-up for DVT. Doppler allows the verification of the presence of venous flow by detecting the spontaneous and phasic sound signal with respiration in the projection of veins to be assessed, and increasing this signal in response to distal compression or sudden proximal decompression. It is based on the alterations in such parameters when DTV is present.

Since it is an examination in which it is not possible to see the vessels, it is subject to false negatives in cases of distal thrombosis, duplication of femoral or popliteal veins, partial thrombosis, and isolated thrombi in inaccessible veins. On the other hand, it is subject to false positive when there is extrinsic compression or previous thrombosis.28 Sumner & Lambeth found sensitivity of 93%, specificity of 86%, positive predictive value of 80%, and negative predictive value of 95% for DVT.29

Despite its limitations, the continuous-wave ultrasound Doppler has shown, in experienced hands, good accuracy for assessing proximal veins and, therefore, can be useful for assessing patients with suspicion of DVT.

Impedance plethysmography (IPG)

IPG is a technique for indirectly measuring the venous drainage index in response to venous flow occlusion and release by inflation and deflation of a thigh cuff. The measurement is made through electrodes placed on the calf, which monitor and record on a thermosensitive paper the modifications in impedance.

The inflation of the cuff with a venous occlusion proximal to the electrodes reduces the impedance as the leg becomes engorged with blood, which is an electrical conductor. In a normal individual, when the cuff is deflated, there is a fast blood drainage, which is reflected by an increase in electrical impedance. When there is a venous obstruction due to thrombi or extrinsic compression, there is a delay in blood drainage that accounts for a slower and progressive increase in impedance in response to the cuff release, since the limb remains engorged with blood for a longer period of time.

According to Stouffer et al.30, the positive predictive value for this exam in occlusive DVT in thigh is 90%, being less sensitive in calf veins, where it detects only 20% of thrombi.

Agnelli et al.31 carried out a prospective study assessing IPG use for DVT diagnosis in symptomatic and asymptomatic patients, comparing it to phlebography. They found sensitivity around 84% for symptomatic patients and 18% for asymptomatic patients, and specificity of 83% and 92%, respectively, suggesting that it is inefficient for detecting little extensive thrombosis, without an important repercussion in venous flow.

Cogo et al.32 found, in a review study where IPG was assessed in comparison to phlebography, sensitivity of 90% and specificity of 96%.

Similar to Doppler, IPG is subject to false negatives in partial DVT, in the presence of double femoral or popliteal veins and in the thrombosis of only one of the leg veins. Venous obstruction in the pelvis due to lymphadenomegaly, pregnancy and obesity, increase in central venous pressure, and peripheral arterial disease can cause false positive.

IPG is being replaced by DS. However, it can be used in the follow-up of symptomatic patients with negative initial result for assessing a possible progression of distal DVT to proximal veins, and in the reassessment of patients with DVT previously diagnosed.

Duplex scanning

The imaging ultrasound or real-time ultrasound, associated to Doppler velocimeter (DS or eco-Doppler), offers a simultaneous assessment of the vessel image and the characteristics of the blood flow under Doppler spectral analysis or color Doppler. Such fact allows a better assessment of the venous flow and makes the examination faster and easier, since it offers a better visualization of the veins, making the identification of non-occlusive thrombi and small-caliber veins easier. The main ultrasonographic criteria evaluated are compressibility, presence of flow, and visualization of echogenic material in the lumen of the vein.

In the last decade it was demonstrated that the imaging ultrasound has a very high accuracy for proximal DVT diagnosis (popliteal, femoral, and iliac veins) when compared to phlebography. Cogo et al.32 found, in a review study where DS was compared to phlebography, sensitivity of 95% and specificity of 93% in outpatients and sensitivity of 97% and specificity of 95% in hospitalized patients.

Birdwell et al.33 found negative predictive value of 99.3% and positive predictive value of 94.4% when assessing femoral and popliteal veins in 601 patients with suspicion of DVT comparing DS to phlebography.

However, this accuracy falls when the DVT diagnosis is made distally (on the leg veins). Mattos et al.34 carried out a prospective study in orthopaedic surgery postoperative patients, in which DS and phlebography were performed in 174 symptomatic and asymptomatic patients. Among symptomatic patients there was 100% of sensitivity and 98% of specificity of the duplex scanning in the proximal segment, and 94% of sensitivity and 75% of specificity in the distal segment. Among asymptomatic patients there was an equal specificity. DS, on the other hand, was less sensitive, with 67% for proximal thrombi and 56% for distal thrombi. In a prospective study comparing DS to phlebography in four venous segments, Reis35 found sensitivity of 52.2 a 53.6; 86.7; 100 and 81.8% in the leg, knee, thigh and pelvis, respectively. He reported specificity of 57.1 a 66.7; 75; 80 e 87.5% for the same segments. Cogo et al.32 reported sensitivity of 75% for distal DVT. Kearon et al.36, in a metanalysis study, found sensitivity of 73% for this segment. Hirsh et al.15 reported sensitivity and specificity around 70% in this segment. Mattos et al.,37 in a study in which 696 limbs with suspicion of DVT were assessed, reported visualization of the three pairs of veins in the leg in 655 (94%), concluding that it is a good examination for assessing distal DVT. On the other hand, Wolf et al.38 performed DS on 699 limbs and reported that calf veins were not seen in 118 examinations (23.4%).

The risk of PE as a consequence of DVT isolated from the leg veins seems to be smaller. Nonetheless, there is a risk of up to 20% of undetected distal thrombosis progression for proximal segments. Such progression occurs more frequently within 7 days.39 Kakkar et al.40 observed 23% of DVT propagation from distal to proximal, detected by the fibrinogen test confirmed by phlebography. In a survey performed by Philbrick et al.,41 they found 10% of distal DVT progression in symptomatic patients and 29% of DVT recurrence in symptomatic patients with distal DVT. Due to these findings, some authors believe that the DS should be repeated within 24 to 48 hours and after 7 days, aiming at detecting such thrombi in progression and occasionally undetected in the first duplex scanning.36,38,42,43

There seems to have been an improvement in the diagnostic precision for DVT in leg veins with the use of better image acquisition equipment and Power Doppler. Forbes & Stevenson44 compared the use of Power Doppler to phlebography in 50 patients with clinical suspicion of DVT. Power Doppler showed a sensitivity and specificity of 100% and 79%, respectively, positive predictive value of 71%, and negative predictive value of 100%.

Stevens et al.45 have recently assessed reliability and security of the DS complete examination (assessment of all deep veins of the lower limb, from the ankle to the inguinal region) in 445 symptomatic patients for DVT (first episode). Among patients with negative DS for DVT, 375 were followed-up for 3 months and the occurrence of thromboembolic phenomena was verified during this period. Most patients (372 or 99.2%) did not present thromboembolism, suggesting that the DS is very reliable for diagnosing symptomatic DVT.

The role of DS for diagnosing DVT is well established. It is the best non-invasive method for diagnosing DVT, replacing phlebography in the assessment of patients with suspicion of the disease. It presents several advantages, such as lower cost, besides dismissing the risk of contrast and radiation. Its accuracy has improved as researchers become more experienced and devices are made with better image and more resources.

Semi-invasive and invasive methods

Computed tomography (CT)

The contrast-enhanced CT can be used for diagnosing DVT. The usual tomographic finding for DVT is the image of a vessel with no contrast of its interior, surrounded by a halo with a parietal reinforcement. It presents sensitivity from 89% to 100% and specificity from 94% to 96% and has some advantages when compared to phlebograhy. Among them, lower volume of contrast, the contrast is injected into a peripheral vein of the upper limb, and it is a faster examination.46Peterson et al.47 performed a study comparing CT to DS in patients with suspicion of DVT and PE and found sensitivity of 71% and specificity of 93%.

CT is considered to be better than phlebography for diagnosing venous thrombosis in abdominal, pelvic, and thoracic veins. It is indicated when there is suspicion of involvement of these sites. It is also better for differentiating recent from old thrombi, as well as detecting adjacent abnormalities (extrinsic compression).30

Despite the advantages, its use is restrict, since it is a high-cost method that is not always available in many hospitals.

Magnetic nuclear resonance

Phlebography using magnetic resonance (MR) is an examination that offers good accuracy for diagnosing DVT and has the advantage of obtaining simultaneous images of both lower limbs, showing the venous system in practically all its extension. It allows thrombi detection in pelvic veins and in the inferior vena cava and, if necessary, images of thoracic veins in only one session. Moreover, it can be performed on pregnant women and in patients using plaster immobilization.

In order to assess lower limbs the contrast can be injected in the foot, as in the conventional phlebography, with the advantage of not being nephrotoxic and not causing pain when injected. The MR can offer subsides for a differential diagnosis, such as extrinsic compression and data about thrombus characteristics. Its sensitivity varies from 87% to 97% and the specificity varies from 93% to 97%.48,49

The MR has limitations. For instance, it cannot be performed on patients with metal implants or paced, patients who suffer from claustrophobia or are in a severe general condition, besides not being available in all hospitals.

A technique for obtaining a direct thrombus image has been recently described, without using the contrast, known as MRDTI (Magnetic Resonance Direct Thrombus Imaging). Visualization is possible due to the transformation of hemoglobin in the red cells present in the thrombus in metahemoglobin, which presents strong paramagnetic properties, with five free electrons. Thus, it works as a contrast allowing DVT detection in any segment and PE with a single examination.50 In a comparative study to phlebography, it showed sensitivity of 96% and specificity of 90%, with no significant difference in the accuracy among distal, proximal, and pelvic segments.51

Fibrinogen uptake test

The 125 I-fibrinogen uptake test was developed in the 1960's and validated by several studies in the 1970's, when it served as a lever for assessing primary prophylaxis in risk patients.52 Its use allowed us to know the DVT incidence in several risk situations. The examination detects the incorporation of labeled fibrinogen on the formation thrombus.52 It was initially described as a good accuracy examination for distal DVT with lower sensitivity for thigh veins. It was further accepted as the standard examination for DVT investigation and as a prophylaxis assessment method in hospitalized patients.53,54

Besides the low sensitivity for proximal thrombi, this examination does not detect pelvic thrombi and its use is contraindicated in pregnant patients.

The fibrinogen uptake test has not been used currently, due to the risk of transmitting diseases, such as hepatitis C and AIDS.4

Blood tests

The use of a peripheral blood sample for DVT diagnosis could be considered the ideal method. Several proteins, as a product of fibrin degradation, thrombin anti-thrombin complexes, and inhibitors of plasminogen activation are in an increased concentration on the blood of patients with DVT. However, blood tests for detecting these substances are difficult to be performed, which avoids its routine use. Besides, they have limited sensitivity and specificity for diagnosing DVT.55

Among these examinations we should highlight the D-dimer (DD), which has proven to be useful for diagnosing DVT.

DD is a specific derivative product from the fibrin and generated by the degradation of the fibrin matrix in fresh thrombi. Since the DD is present in any situation in which there is fibrin formation and degradation, it is not a very specific index for DVT and it can be present in several situations, such as recent surgery, trauma, cancer, and sepsis. However, it is very sensitive and has proven to be useful as an associated test for the exclusion diagnosis.

Several studies have recently reviewed the sensitivity and specificity of the tests in order to determine the DD as a method for DTV exclusion. Bounameaux et al.56 evaluated the ELISA test in a review study, reporting sensitivity of 97% and specificity of 45%. Perone et al.25, in a review study, reported sensitivity between 95% and 97% and specificity between 37% and 45% for the conventional ELISA. Perrier et al.57 reported negative predictive value of 99% with the semiquantitative ELISA method (VIDAS Rapid ELISA), which is easier and faster to perform than the conventional ELISA.

Two other DD determination methods were evaluated. The whole-blood agglutination test (SimpliRED) and the latex agglutination test. These tests are easy to perform, faster and cheaper than the ELISA. They are also less sensitive, although more specific than the ELISA. SimpliRED's sensitivity in clinical studies varied from 84% to 94%, while the latex agglutination test showed sensitivity lower than 80%.43,58,59 Dale et al.60 compared three different tests for determining the DD and found sensitivity of 98% for ELISA, 95% for the immunofiltration assay (NycoCard D-dimer) and 73% for the latex agglutination test.

In a study comparing seven methods for determining DD, three conventional ELISA tests, three quick tests and the latex agglutination test, there was a similar sensitivity (98%-96%) for the ELISA tests and two of the quick tests (VIDAS D-Di and Instant IA D-Di). For the NycoCard D-dimer there was a sensitivity of 82%. The latex agglutination test presented the lowest sensitivity (74%).59

The DD has proven to be a promising associated test for diagnosing DVT. Michiels et al.61 believe that for patients with low probability of DVT, negative DD, and negative DS, there would be no need for the series of ultrasound examinations. Arcelus et al.24 assessed the association of the whole-blood agglutination test, DS, and risk determination for DVT and their results suggest that patients with low probability of DVT and negative DD might dismiss the DS. Wells et al.27 showed that its use associated to a clinical prediction model makes the approach to patients with suspicion of DVT simpler, reducing the need for performing DS.


As mentioned earlier, the DVT diagnosis depends on the use of objective complementary examinations. Phlebography is still considered the gold standard among complementary examinations for diagnosing DVT. The examination, when performed under proper conditions and by a skilled examiner, offers an excellent assessment of the deep venous system, showing the absence of DVT through the uniform distribution of the contrast on the veins and the presence of DVT when the contrast fails to fill the vessel, determining the thrombus location and extension.

Nevertheless, it is an invasive examination, which causes discomfort for the patient and can present some complications such as hypersensitivity to contrast. It may even trigger DVT in some patients. Albrechtsson & Olsson62 described the occurrence of thromboembolic complications after phlebography in 7% of patients, suggesting that thrombosis and PE might be caused by the examination. Rollo et al.18 found 29.8% of side effects and minor complications in 424 patients submitted to phlebography, with only one case of severe complication (anaphylaxis).

Phlebography presents some contraindications. It should be avoided on pregnant women with previous history of contrast allergy. In case of patients with renal insufficiency, it should be performed carefully.

Moreover, around 20% of hospitalized patients do not present conditions of being submitted to the examinations with proper techniques, such as, for instance, patients in a severe condition admitted to the intensive care unit.63

We should add to these limitations the fact that in 1.5%-5% of cases it might be difficult to interpret the phlebography, creating the need for another complementary examination.

Thus, phlebography has been progressively replaced by non-invasive methods, initially plethysmography, and more recently by imaging ultrasonography.


The association of Wells' clinical prediction model to complementary exams, such as IPG, DS, DD, and/or phlebography, has been tested by several authors.

Anand et al.,65 in a review study, assessed the importance of patient stratification concerning the risk of presenting DVT associated to complementary examinations, suggesting that the use of clinical prediction protocols that include risk factor and physical examination, along with complementary examinations, simplifies the diagnosis for DVT.

In the assessment of patients with suspicion of DVT and PE, Perrier et al.57used an algorithm combining the determination of the clinical probability for DVT, DD, DS, and pulmonary scintigraphy. They obtained 94% of definite diagnosis for DVT and PTE without the need for more invasive examinations.

Michiels et al.61 proposed that the use of the clinical prediction model associated to the DD test by the fast ELISA method followed by DS is safe and cost-effective for dismissing or confirming the diagnosis for DVT. For patients with low probability (LP) and negative DD, DS would not be required; in patients with moderate probability (MP) DS should be initially performed and, if negative, DD is then made. The negative DD excludes DVT; if positive, the differential diagnosis should be dismissed and, if not detected, DS should be repeated within 7 days. According to the authors there is no advantage in performing the DD on patients with high probability of DVT. In such cases DS should be performed.

Perone et al.25 compared four strategies for diagnosing DVT that used the determination of the probability of presenting DVT, series of DS and DD in different approaches. They concluded that the most cost-effective approach would be the combination of clinical probability and DD with a single ultrasound examination.

Kearon et al.66 assessed 445 patients with suspicion of DVT, but patients with LP and negative DD were not submitted to DS. In a 3-month follow-up it was found only one case of DVT, suggesting that this therapeutic approach would be justified for patients under discussion.

Other authors agree with the approach in which patients classified as low probability of presenting DVT and negative DD could dismiss the DS, thus reducing the investigation cost.67-69

Dryjski et al.70 prospectively assessed 66 patients with suspicion of DVT in emergency departments. These patients were classified as to DVT probability, according to Wells' model, and those with high probability were submitted to DS. Those with MP and LP were first submitted to the DD and then to DS. They found 9.6% of DVT cases, all among patients with high probability. The sensitivity and the negative predictive value for the DD associated to the probability pre-test were both 100%. However, for the DD separately, they were 80% and 95%, respectively. Based on the mentioned results, they proposed that for patients with LP and MP with negative DD, there would be no need for performing DS. Nonetheless, the number of studied cases was very small.

Kraaijnhagen et al.71 performed a prospective study with 1756 patients with suspicion of DVT. All were classified as to DVT probability, according to Wells' model and then submitted to the DD and DS. Their results suggest that for patients with LP and negative DD, it would be safe to dismiss the DS. For all other patients with negative DD and initial DS, the series of examinations would not be required.

Cornuz et al.72 in a similar approach determined that the negative predictive value for the DD is 96%, but in the group of patients with LP for DVT this value is 100%, showing that the association of the probability pre-test with the DD is cost-effective, improving the accuracy of such examination.

Several authors found similar results with this approach, suggesting that the application of the clinical model has become the easiest, safest, cheapest and less invasive diagnosis for DVT.70,71,73 In a review study, Kelly & Hunt74 defend that it is safe to dismiss the DS and leave patients with LP who present negative DD untreated. For patients with MP, this approach is only safe if more sensitive tests are used, such as the ELISA.

Wells et al.27 have recently used the clinical prediction model with a new modification, which includes, as one of the items in the score, the previous DVT confirmed by a complementary examination, allowing the inclusion of such patients in the study. This new model classified patients into two groups: possibly with DVT when the score is equal or higher than 2, and possibly without DVT when the score is lower than 2. It was applied associated with DS and DD in 566 patients and associated with only DS in 530 patients. The authors concluded that the use of such model along with DD simplifies the diagnosis for DVT, reducing the number of DS performed, without compromising the patient's safety.


The diagnosis for DVT is mainly based on complementary examinations. Among the available examinations, the most widely used is the DS, due to its accuracy, low cost, availability and patient's tolerance.

More recently, it has been demonstrated that the use of clinical prediction models associated with DD and imaging examinations has helped to make the diagnostic process easier and more cost-effective, thus highlighting the importance of clinical assessment when managing these patients. Among the approaches proposed by several authors, the use of the probability test along with DS and DD, as seen in Figure 1, has been more used.71,74

click hereFigure 1 - Algorithm for the diagnostic approach of deep venous thrombosis (DVT) using Dimer-D (DD), imaging ultrasonography (IU) or duplex scanning (DS) and Wells et al.'s model, 1997.

According to what can be seen in this review, based on the application of Wells et al.20 model in our service and the modifications proposed by Wells et al.27 in their clinical prediction model published in 2003, we began to adapt the diagnostic algorithm shown in Figures 2 and 3 for the diagnosis of DVT in the service of Vascular Surgery of the HC da Faculdade de Medicina de Botucatu - UNESP. In Figure 2 the algorithm is based on the clinical model with DS and in Figure 3 the algorithm shows the approach when DD is available.

click hereFigure 2 - Algorithm for the diagnostic approach of deep venous thrombosis (DVT) using imaging ultrasonography (IU) or duplex scanning (DS) and Wells et al.'s model, 2003.

click hereFigure 3 - Algorithm for the diagnostic approach of deep venous thrombosis (DVT) using Dimer-D (DD), imaging ultrasonography (IU) or duplex scanning (DS) and Wells et al.'s model, 2003.


1. Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worchester DVT Study. Arch Intern Med 1991;151:933-8.

2. Gillum RF. Pulmonary embolism and thrombophlebitis in the United Estates. Am Heart J 1987;114:1262-4.

3. Nylander G, Olivercrana H. The phlebography pattern of acute leg thrombosis within a defined urban population. Acta Dis Child 1991;66:713-7.

4. Maffei FHA. Trombose Venosa Profunda dos Membros Inferiores: Incidência, Patologia, Fisiopatologia e Diagnóstico. In: Maffei FHA, Lastoria S, Yoshida WB, Rollo HA. Doenças Vasculares Periféricas. 3ª ed. São Paulo: Medsi; 2002. p. 1363-86.

5. Fowkes FJ, Price JF, Fowkes FG. Incidence of diagnosed deep vein thrombosis in the general ppopulation: systematic review. Eur J Vasc Endovasc Surg 2003;25:1-5.

6. National Institutes of Health Consensus Conference. Prevention of venous thrombosis and pulmonary embolism. JAMA 1986;256:744-9.

7. Diebold J, Lohrs U .Venous thrombosis and pulmonary embolism. A study of 5039 autopsies. Path Res Pract 1991;187:260-6.

8. Saeger W, Genzkow M. Venous thrombosis and pulmonary embolisms in post-mortem series: probable causes by correlations of clinical data and basic diseases. Path Res Pract 1994;190:394-9.

9. Lindblad B, Sternby NH, Berqvist D. Incidence of venous thromboembolism verified by necropsy over 30 years. BMJ 1991;302:709-11.

10. Yoo HH, Paiva SA, Silvaira LV, Queluz TT. Logistic regression analysis of potential factor for pulmonary thromboembolism. Chest 2003;123:813-21.

11. Maffei FHA, Falleiros ATS, Venezian LA, Franco MF. Contribuição ao estudo da incidência e anatomia patológica do tromboembolismo pulmonar em autópsias. Rev Ass Med Bras 1980;26:7-10.

12. Bauer G. A venographic study of thrombo-embolic problems. Acta Chir Scand 1940;84:1-75.

13. Haeger K. Problems of acute deep venous thrombosis: the interpretation of sings and symptoms. Angiology 1969;20:219-23.

14. Prandoni P, Villalta S, Bagatella P, et al. The clinical course of deep vein thrombosis. Prospective long term follow up of 528 symptomatic patients. Haematologica 1997;82:423-8.

15. Hirsh J, Raschke R, Warketin T. Heparin: mechanism of action, pharmacokinetics, dosing considerations, monitoring, efficacy and safety. Chest 1995;108:S258-75.

16. Richards KL, Armstrong JJ, Tikoff G. Noninvasive diagnosis of deep vein thrombosis. Arch Intern Med 1976;136:1091-6.

17. Sandler DA, Martin JF, Duncan JS. Diagnosis of deep-vein thrombosis: comparison of clinical evaluation, ultrasound, plethysmography, and venoscan with venogram. Lancet 1984;2:716-9.

18. Rollo HA, Maffei FHA, Lastória S, Yoshida WB, Castiglia V. Uso rotineiro da flebografia no diagnóstico da trombose venosa profunda dos membros inferiores. Cir Vasc Angiol 1986;2:7-14.

19. Lastória S, Maffei FHA, Rollo HA, Yoshida WB, Massetti MCV, Mattar L. Quadro clínico e tratamento da trombose venosa profunda dos membros inferiores. Estudo prospectivo de 188 pacientes. Resumo publicado nos anais do XXVI Congresso Brasileiro de Angiologia, Fortaleza, 1985.

20. Wells PS, Hirsh J, Anderson DA, et al. Accuracy of clinical assessment of deep-vein thrombosis. Lancet 1995;345:1326-30.

21. Wells PS, Anderson DR, Bormanis J, et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997;350:1795-8.

22. Kahn S. The clinical diagnosis of deep venous thrombosis: integrating incidence, risk factors, and symptoms and signs. Arch Intern Med 1998;158:2315-23.

23. Anderson DR, Wells PS, Stiell I, et al. Thrombosis in the emergency department. Arch Intern Med 1999;159:477-82.

24. Arcelus JI, Caprini JA, Sehgal LR, Lipsitz E, Reyna JJ. The role of risk assessment and rapid whole-blood D-dimer testing for the diagnosis of deep vein thrombosis in symptomatic patients. JP 2002;2:57-63.

25. Perone N, Bounameaux H, Perrier A. Comparison of four strategies for diagnosing deep vein thrombosis: a cost-effectiveness analysis. Am J Med 2001;110:33-40.

26. Miron M-J, Perrier A, Bounameaux H. Clinical assessment of suspected deep vein thrombosis: comparison between a score and empirical assessment. J Intern Med 2000;247:249-54.

27. Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003;349:1227-35.

28. Sumner DS. Diagnosis of deep vein thrombosis. In: Rutherford RD. Vascular Surgery. 4th ed. Philadelphia: W. B. Saunders; 1995. p. 1698-743.

29. Sumner DS, Lambeth A. Reliability of Doppler ultrasound in the diagnosis of acute deep venous thrombosis both above and below the knee. Am J Surg 1979;138:205-10.

30. Stouffer GA, Sheahan RG, Lenihan DJ, Gupta R. Deep venous thrombosis: a review of the pathophysiology, clinical features, and diagnostic modalities. Am J Med Sci 2001;322:358-64.

31. Agnelli G, Cosmi B, Radicchia S, et al. Features of thrombi and diagnostic accuracy of impedance plethysmography in symptomatic and assymptomatic deep vein thrombosis. Thromb Haemostasis 1993;70:266-9.

32. Cogo A, Lensing AW, Wells P, Prandoni P, Büller HR. Noninvasive objective tests for diagnosis of clinically suspected deep-vein thrombosis. Haemostasis 1995;25:27-39.

33. Birdwell BG, Raskob GA, Whitsett TL, et al. Predictive value of compression ultrasonography for deep vein thrombosis in symptomatic outpatients: clinical implications of the site of vein noncompressibility. Arch Intern Med 2000;160:309-13.

34. Mattos MA, Londrey GL, Leutz DW, et al. Color-flow duplex scanning for the surveillance and diagnosis of acute deep venous thrombosis. J Vasc Surg 1992;15:366-76.

35. Reis FB. Diagnóstico da trombose venosa profunda dos membros inferiores: Comparação prospectiva entre a ultra-sonografia dúplex com a flebografia ascendente [dissertação]. Escola Paulista de Medicina, Universidade Federal de São Paulo: São Paulo; 2000.

36. Kearon C, Julian JA, Newman TE, Ginsberg JS. Noninvasive diagnosis of deep vein thrombosis. McMaster Diagnostic Imaging Practice Guidelines Initiative. Ann Intern Med 1998;128:663-77.

37. Mattos MA, Melendres G, Sumner DS, et al. Prevalence and distribution of calf vein thrombosis in patients with symptomatic deep venous thrombosis: a color-flow duplex study. J Vasc Surg 1996;24:738-44.

38. Wolf B, Nichols D, Duncan JL. Safety of a single duplex scan to exclude deep venous thrombosis. Br J Surg 2000;87:1525-8.

39. Cogo A, Lensing AW, Prandoni P, Hirsh J. Distribution of thrombosis in patients with symptomatic deep vein thrombosis. Implications for simplifying the diagnostic process with compression ultrasound. Arch Intern Med 1993;153:2777-80.

40. Kakkar VV, Flanc C, Howe CT, Clarke MB. Natural history of postoperative deep vein thrombosis. Lancet 1969;2:230-2.

41. Philbrick JT, Becker DM. Calf deep venous thrombosis. A wolf in sheep's clothing? Arch Intern Med 1988;148:2131-8.

42. Cogo A, Lensing AW, Koopman MM, et al. Compression ultrasonography for diagnostic management of patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ 1998;316:17-20.

43. Hirsh J, Lee A. How we diagnose and treat deep vein thrombosis. Blood 2002;99:3102-10.

44. Forbes K, Stevenson AJ. The use of power Doppler ultrasound in the diagnosis of isolated deep venous thrombosis of the calf. Clin Radiol 1998;53:752-4.

45. Stevens SM, Elliott CG, Chan KJ, Egger MJ, Ahmed KM. Withholding anticoagulation after a negative result on duplex ultrasonography for suspected symptomatic deep venous thrombosis. Ann Intern Med 2004;140:985-91.

46. Faria RC. Aplicações da angiotomografia no diagnóstico das patologias venosas. In: Maffei FHA, Lastoria S, Yoshida WB, Rollo HA. Doenças Vasculares Periféricas. 3ª ed. São Paulo: Medsi; 2002. p. 519-33.

47. Peterson DA, Kazerooni EA, Wakefield TW, et al. Computed tomographic venography is specific but not sensitive for diagnosis of acute lower-extremity deep venous thrombosis in patients with suspected pulmonary embolus. J Vasc Surg 2001;34:798-804..

48. Polak JF, Fox LA. MR assessment of the extremity veins. Seminars in Ultrasound, CT and MRI 1999;20:36-46.

49. Montgomery KD, Potter HG, Helfet DL. Magnetic resonance venography to evaluate the deep venous system of the pelvis in patients who have an acetabular fracture. J Bone Joint Surg 1995;77:1639-49.

50. Moody AR. Magnetic resonance direct thrombus imaging. J Throm Haem 2003;1:1403-9.

51. Fraser D, Moody A, Martel A, Morgan P, Davidson I. Diagnosis of lower-limb deep venous thrombosis: a prospective blinded study of magnetic resonance direct thrombus imaging. Ann Intern Med 2002;136:89-98.

52. Kakkar VV. The diagnosis of deep vein thrombosis using the 125I-fibrinogen test. Arch Surg 1972;104:152-9.

53. Lensing AW, Hirsh J. I-Fibrinogen leg scanning: reassessment of its role for the diagnosis of venous thrombosis in post-operative patients. Thromb Haem 1993;69:2-7.

54. Maffei FHA, Almeida MRH, Yoshida WB. O teste do fibrinogênio marcado com iodo 125 no diagnóstico das tromboses venosas profundas dos membros inferiores. F Med Br 1981;82:491-6.

55. Kakkar VV. Venous thrombosis today. Haemostasis 1994;24:86-104.

56. Bounameaux H, de Moerloose P, Perrier A, Reber G. Plasma measurement of D-dimer as diagnostic aid in suspected venous thromboembolism: an overview. Thromb Haemost 1994;72:1-6.

57. Perrier A, Desmarais S, Miron M-J, et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet 1999;353:190-5.

58. Lee AY, Ginsberg JS. The role of D-dimer in the diagnosis of venous thromboembolism. Curr Opin Pulm Med 1997;3:275-9.

59. Elias A, Aptel I, Huc B. D-dimer test and diagnosis of deep vein thrombosis: a comparative study of 7 assays. Thromb Haemost 1996;76:518-22.

60. Dale S, Gogstad GO, Brosstad F, et al. Comparison of three D-dimer assays for the diagnosis of DVT: ELISA, latex an imunofiltration assay (NycoCard D-dimer). Thromb Haemost 1994;71:270-4.

61. Michiels JJ, Freyburger G, Graaf FV, Janssen M, Oortwijn W, van Beek EJ. Strategies for the safe effective exclusion and diagnosis of deep vein thrombosis by the sequential use of clinical score, D-dimer testing, and compression ultrasonography. Sem Thromb Hemost 2000;26:657-67.

62. Albrechtsson U, Olsson C-G. Thrombotic side-effects of lower-limb phlebography. Lancet 1976;1:723-4.

63. Wheeler HB, Anderson Jr FA. Diagnostic methods for deep vein thrombosis. Haemostasis 1995;25:6-26.

64. Björgell O, Nilsson PE, Jarenros H. Isolated non-filling of contrast in deep leg vein segments seen on phlebography, and a comparison with color Doppler ultrasound, to assess the incidence of deep leg vein thrombosis. Angiology 2000;51:451.

65. Anand SS, Wells PS, Hunt D, Brill- Edwards P, Cook D, Ginsberg JS. Does this patient have deep vein thrombosis? JAMA 1998;279:1094-9.

66. Kearon C, Ginsberg J, Douketis J, et al. Management of suspected deep venous thrombosis in outpatients by using clinical assessment and D-dimer testing. Ann Intern Med 2001;135:108-11.

67. Janes S, Ashford N. Use of a simplified clinical scoring system and D-dimer testing can reduce de requirement for radiology in the exclusion of deep vein thrombosis by over 20%. Br J Haem 2001;112:1079-82.

68. Corris PA. A practical approach to the diagnosis of venothromboembolism. Clin Med 2001;1:274-81.

69. Bucek RA, Koca N, Reiter M, Haumer M, Zontsich T, Minar E. Algorithms for the diagnosis of deep vein thrombosis in patients with low clinical pretest probability. Thromb Res 2002;105:43-7.

70. Dryjski M, O'Brien-Irr MS, Harris L, Hasset J, Janicke D. Evaluation of a screening protocol to exclude the diagnosis of deep vein thrombosis among emergency department patients. J Vasc Surg 2001;34:1010-5.

71. Kraaijnhagen RA, Piovella F, Bernardi E, et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002;162:907-11.

72. Cornuz J, Ghali WA, Hayoz D, Stoianov R, Depairon M, Yersin B. Clinical prediction of deep venous thrombosis using two risk assessment methods in combination with rapid quantitative D-dimer testing. Am J Med 2002;112:198-203.

73. Wells P, Anderson D, Ginsberg J. Assessment of deep vein thrombosis or pulmonary embolism by the combined use of clinical model and noninvasive diagnostic tests. Sem Thromb Hemost 2000;26:643-56.

74. Kelly J, Hunt BJ. A clinical probability assessment and D-dimer measurement should be the initial step in the investigation of suspected venous thromboembolism. Chest 2003;124:1116-9.

J Vasc Br - Official Publication of the Brazilian Society of Angiology and Vascular Surgery