HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fraser, J. D. Articles by Anderson, D. R. Search for Related Content PubMed PubMed Citation Articles by Fraser, J. D. Articles by Anderson, D. R.

Deep Venous Thrombosis: Recent Advances and Optimal Investigation with US¹

¹ From the Departments of Radiology (J.D.F.) and Medicine (D.R.A.), Queen Elizabeth II Health Sciences Centre, 1278 Tower Rd, Halifax, Nova Scotia, Canada B3H 2Y9. Received February 12, 1998; revision requested March 18; final revision received September 9; accepted October 19. D.R.A. is a Research Scholar of the Canadian Heart and Stroke Foundation. Address reprint requests to J.D.F.

Index terms: Embolism, pulmonary, 60.72, 944.442 • Extremities, thrombosis, 93.442 • Veins, thrombosis, 93.442 • Veins, US, 93.12981, 93.12983, 93.12984

Deep venous thrombosis (DVT) is a common clinical problem. Compression ultrasonography (US) remains the imaging procedure of choice for the investigation of patients with suspected DVT. It is a highly sensitive and specific test for the diagnosis of proximal DVT in symptomatic patients. Management approaches that rely on two negative compression US studies obtained 1 week apart have proved to help safely exclude the diagnosis of DVT. The performance of an assessment of clinical pretest probability and/or a D-dimer assay may reduce the need to perform serial US examinations. The performance of bilateral compression US in patients with suspected pulmonary embolism may be used to decrease the need for pulmonary angiography, especially if combined in management algorithms that include D-dimer testing and consideration of clinical pretest probability. Compression US also appears accurate for the diagnosis of DVT involving the upper extremity. Compression US is much less sensitive for diagnosing DVT following high-risk surgical procedures such as total hip or knee arthroplasty, and its use as a screening test is not recommended in this setting.

Deep venous thrombosis (DVT) is a common clinical problem that complicates many medical and surgical disorders (1,2). It can cause morbidity in itself due to acute pain and swelling of the affected limb, and it may also cause structural damage to the valves of the deep veins that results in the postphlebitic syndrome. If not recognized, deep venous thrombi can extend and embolize to the pulmonary arterial circulation. Pulmonary embolism can cause sudden death or, if nonfatal, result in shortness of breath and chest discomfort (3). In the United States, the annual combined incidence of DVT and pulmonary embolism is at least 70 per 100,000 individuals (4).

Fortunately, effective therapy is available for the treatment of DVT (1,5). Antithrombotic therapy reduces the morbidity of this disorder and the risk of it causing pulmonary embolism. However, since the clinical signs and symptoms of DVT are nonspecific, it is important to promptly perform objective testing to confirm the diagnosis and enable the institution of safe and effective therapy (6).

The objectives of this article are twofold. The first is to update radiologists about new information regarding the clinical presentation, diagnosis, and treatment of DVT since this topic was last reviewed in Radiology in 1993 (7). The second is to focus on current knowledge about the optimal use of ultrasonography (US), the imaging procedure of choice for the diagnosis of DVT. We will review US techniques and discuss the utility of US in the diagnosis of the initial episode of DVT, postoperative DVT, recurrent DVT, DVT associated with pulmonary embolism, and upper extremity DVT.

DEFINITIONS

US has been carefully evaluated in two clinical settings as an imaging technique for DVT of the lower extremity (1,3,6,7). The first setting is as a diagnostic test in patients presenting with symptoms and signs that are suspected to be caused by DVT. In this article, such patients will be referred to as symptomatic. In the second setting, US has been evaluated as a screening test for DVT in high-risk clinical situations, such as following total hip arthroplasty. These patients usually have no specific symptoms or signs of DVT and in this article will be referred to as asymptomatic.

Some studies have evaluated the sensitivity and specificity of US for the diagnosis of DVT compared to those of the reference standard, venography (1,3,6). Most studies report separately the accuracy of US for the diagnosis of proximal DVT, which is defined as thrombosis occurring within the popliteal, superficial femoral, and/or common femoral veins (1,3,6). In this article, these studies will be referred to as accuracy studies.

Other studies have evaluated the safety of relying on US alone for making the diagnosis of DVT of the lower extremity (1,7). Patients in these studies underwent one or more US examinations for suspected DVT. If the US study remained negative, no further testing for DVT was performed, and patients were not treated with antithrombotic therapy. To determine the safety of these approaches, the patients in whom the diagnosis of DVT was considered excluded were followed up for 3–6 months for the development of symptomatic venous thromboembolic complications. In this article, these studies will be referred to as management studies.

CLINICAL PRESENTATION OF SYMPTOMATIC DVT

Patients who come to medical attention because of symptoms of lower extremity DVT will usually present in one of two ways. The first is with symptoms of calf-popliteal vein DVT. Most patients with acute DVT will initially develop symptoms of pain and swelling in the calf of one leg (8). The symptoms tend to increase with ambulation and improve with rest. There may be associated increased warmth, redness, and tenderness of the calf area. On average, a patient's symptoms will persist for about 7 days before the patient comes to medical attention (9,10). During that time, symptoms usually worsen. The pain and swelling may become more severe and progress proximally up the leg to the popliteal fossa and medial thigh area.

Venographic correlates of this clinical syndrome will demonstrate DVT of the calf veins that may extend to some degree into the more proximal leg veins (7). These thrombi are contiguous and, depending on the extent of progression, may involve the popliteal, superficial femoral, common femoral, and external iliac veins. The clinical and venographic findings suggest that most deep venous thrombi start in the veins of the calf and with time migrate proximally. Fewer than 20% of patients with symptomatic DVT have thrombi that involve the calf veins alone (6,8).

The second form of presentation is with symptoms of iliofemoral DVT. Patients with isolated iliofemoral DVT first develop symptoms of pain in the buttock and/or groin region (6). With time, the pain will extend into the medial thigh, and swelling of the proximal leg will develop. If untreated, the entire leg may become swollen, painful, and dusky in color, and prominent collateral superficial veins may develop. This syndrome is referred to as phlegmasia cerulea dolens (11,12).

Venographic correlates of iliofemoral DVT will demonstrate, in early stages, thrombus isolated to the external iliac and common femoral veins (6,8). With more extensive disease, contiguous extension of the clot will move distally to involve the superficial femoral and the popliteal veins.

Fewer than 10% of patients with DVT will have isolated iliofemoral disease (7), and this syndrome tends to occur in certain well-recognized clinical situations. DVT in the peripartum period frequently occurs in the iliofemoral region, and in over 90% of cases, it will involve the left leg (13), likely due to compression of the left common iliac vein by the right iliac artery during pregnancy. DVT associated with a pelvic mass or recent pelvic surgery is typically found in the iliofemoral veins (11). Finally, DVT occurring in association with oral contraceptive use or the antiphospholipid antibody syndrome may first develop in the iliofemoral system (14). The antiphospholipid antibody syndrome is a condition caused by the development of autoantibodies that interact with phospholipid surfaces. Patients with this syndrome may develop arterial or venous thrombosis, thrombocytopenia, a livido reticularis skin rash, or fetal miscarriage (15,16).

DVT usually develops in only one leg at a given time. Bilateral calf popliteal DVT is occasionally seen in patients with metastatic adenocarcinoma. Iliofemoral DVT may result in bilateral findings if the thrombus extends proximally to involve the inferior vena cava.

ASYMPTOMATIC DVT

Surprisingly, most patients who develop DVT are asymptomatic. Screening venography performed after major surgical procedures such as total hip or total knee arthroplasty will demonstrate that as many as 50%–80% of patients have DVT if prophylaxis is not used (17). Even with the use of effective prophylaxis such as low-molecular-weight (LMW) heparin or warfarin, 15%–45% of patients will develop venographic evidence of DVT following joint arthroplasty (17). Although most postoperative deep venous thrombi are isolated to the calf veins, 20%–50% may involve the more proximal venous system (17). Most patients with postoperative DVT have no specific symptoms, likely because the thrombi are very small (in some cases, <1 cm in length), and they often do not cause vein occlusion (18). Postoperative thrombi do not necessarily follow the typical anatomic distribution seen in symptomatic patients. Postoperative DVT may involve an isolated portion of one of the proximal leg veins, likely due to local trauma from the surgical procedure (18), and most resolve spontaneously without causing specific symptoms or complications (17,19).

PATHOGENESIS AND RISK FACTORS FOR DVT

Many patients who develop DVT have well-defined risk factors that are associated with this condition. These risk factors include recent malignancy, major surgical procedures, trauma, prolonged immobilization, pregnancy or use of oral contraceptives, underlying inflammatory states, or a previous history of venous thromboembolism (2,17). During the past decade, there have been a number of congenital and acquired hypercoagulable syndromes identified that are associated with an increased risk of DVT (13,17,20,21). These are listed in Table 1.

TREATMENT OF DVT

LMW Heparin
There has been a major development in the initial treatment of patients who develop acute DVT. Formerly, patients with acute DVT were admitted to a hospital, placed at bed rest, and given an intravenous infusion of unfractionated heparin for 7–14 days (25,26). LMW heparin has recently become available in Canada and the United States for the treatment of DVT. LMW heparin is a derivative of unfractionated heparin but has several advantages over the parent compound (27). In comparison to unfractionated heparin, LMW heparin has a longer half-life, better bioavailability, and a very predictable dose response. In addition, LMW heparin is less likely to cause heparin-induced thrombocytopenia than is unfractionated heparin (28). Clinical trials of patients with DVT or pulmonary embolism have demonstrated that LMW heparin administered subcutaneously in a fixed, weight-based dose is as safe and effective as unfractionated heparin administered intravenously with dose adjustments made by using standardized nomograms (29–31). In fact, two studies have demonstrated the safety and feasibility of using LMW heparin to treat DVT on an outpatient basis (32,33). LMW heparin is an important treatment advance, but it also has implications for management algorithms for patients with suspected DVT or pulmonary embolism.

Radiologists need to be aware of several practical considerations when LMW heparin is being used for the treatment of venous thromboembolism (27). First, LMW heparin is given by means of subcutaneous injection, and its anticoagulant effect is prolonged (lasting 12–24 hours, depending on preparation and dose). Second, the anticoagulant effect of LMW heparin is not completely neutralized by protamine sulphate. Third, LMW heparin does not cause a consistent prolongation of the activated partial thromboplastin time, and this test should not be used to monitor its effect. For these reasons, in some situations, such as before invasive procedures in which direct control of hemostasis is not possible, it may be desirable to convert patients from LMW heparin to unfractionated heparin.

Thrombolytic Therapy
Thrombolytic therapy is not commonly used for the treatment of DVT (1,34). Although thrombolytics cause more rapid clot lysis than parenteral heparin, data from randomized controlled trials have not demonstrated that their use reduces the incidence or the severity of postthrombotic syndrome (17,35,36).

Although thrombolytic therapy would be considered warranted in patients with venous gangrene secondary to an extensive proximal DVT, such cases are very rare. On the basis of data from uncontrolled studies, some clinicians have advocated consideration of the use of thrombolysis for patients with acute iliofemoral or extensive proximal DVT, in whom symptoms have been present for less than 48 hours (1,34). Radiologists need to be cautious in selecting such patients because thrombolytic therapy carries the risk of serious and even fatal bleeding, whereas DVT, once treatment is initiated, rarely results in a patient's death (37).

DIAGNOSTIC METHODS FOR DVT

Clinical Diagnosis
Before the 1970s, the diagnosis of DVT was made entirely on clinical grounds. With the availability of venography, it became recognized that errors were made when the diagnosis of DVT was based on the clinical examination alone (38,39). In fact, individual symptoms and signs were proved to be both insensitive and nonspecific for the diagnosis of DVT, and it became apparent that objective testing must be performed to confirm this diagnosis (38,39). Subsequently, the clinical diagnosis fell into such disregard that its utility was believed only to be a trigger for the consideration of the need for objective testing. It became regarded as completely inaccurate for the diagnosis of DVT.

Studies in recent years have revisited the accuracy of the clinical diagnosis for DVT. It has been shown that explicit clinical criteria can be used to accurately categorize patients with suspected DVT into high, moderate, and low pretest probability categories (40). With such criteria, patients considered at high clinical pretest probability have over a 75% prevalence of DVT confirmed by objective testing. On the other hand, patients considered at low pretest probability but in whom the diagnosis cannot be excluded on clinical grounds alone have a less than 5% prevalence of DVT. These criteria combine signs and risk factors for DVT with consideration of an alternative diagnosis that may account for a patient's symptoms. A simple nine-point clinical criteria scoring system has been developed to determine pretest probability for DVT, and this has proved to be a useful adjunct to noninvasive testing for the treatment of patients with suspected DVT (41).

Venography
Ascending venography was the first imaging procedure available for the diagnosis of DVT and has been regarded as the reference-standard technique (42). Venography remains the only diagnostic test that enables reliable detection of DVT isolated to the calf veins, the iliac veins, or the inferior vena cava. Venography is also the most accurate method for the diagnosis of asymptomatic thrombi (6,43,44). However, it has a number of limitations that make it less practical and attractive than noninvasive methods (6). Venography requires a meticulous technique by skilled radiologists. It should be performed on a tilting fluoroscopic table, and this necessitates the transfer of patients to the radiology department. It requires a cooperative patient who can be examined in a semierect position and who has adequate venous access on the foot of the affected leg. Often large volumes of contrast medium, up to 200 mL, are required for adequate visualization of the entire deep venous system. Results of recent studies have shown that even at centers in which venography is performed in research studies, 10%–20% of venographic studies failed to depict some segment of the venous system, making these images suboptimal for interpretation (40,45).

In addition to the technical challenges associated with obtaining an adequate venogram, there are both minor and serious adverse side effects associated with this invasive procedure. These include local pain, skin reactions, and postinjection superficial phlebitis. Nausea, vomiting, and dizziness may also occur. Even with use of lower osmolar contrast agents, minor side effects occur in about 20% of patients undergoing venography (46,47). Serious side effects include local skin necrosis from extravasation of contrast medium, severe allergic reactions to contrast medium, impaired renal function, and postinjection DVT. Clinically important venous thrombosis after venography has been reported in up to 2% of patients in whom conventional ionic contrast agents were used but is likely to be less common with the use of nonionic contrast medium (46).

Venography is contraindicated in patients with renal failure, those with a history of severe reaction to contrast material, and those unable to partially weight bear on the unaffected extremity during table tilting. Despite the fact that venography remains the most accurate diagnostic method for DVT in asymptomatic patients, it is not considered a suitable screening test because of the invasive nature of the procedure (6,42). In addition, with the development of accurate noninvasive diagnostic tests, conventional venography is no longer the diagnostic test of choice in the initial evaluation of patients with clinically suspected DVT.

NONINVASIVE TESTING

In the 1970s and 1980s, four noninvasive techniques were developed to help diagnose DVT in the lower extremity and potentially avoid the need for venography. The first three used indirect methods to diagnose DVT; these methods were based on either changes in venous hemodynamics (Doppler US, impedance plethysmography) or the presence or absence of fibrin formation (iodine 125 fibrinogen scanning) (48–51). Although these techniques (alone or in combination) appeared sufficiently sensitive to usually avoid the need for venography, the specificity of abnormal test results was relatively poor (52,53). ¹²⁵I fibrinogen leg scanning has largely been abandoned due to concerns about the safety of using a diagnostic technique derived from a blood product.

Real-time B-mode US is the most recently available noninvasive test for the diagnosis of DVT (54–56). This technique provides direct visualization of the deep venous structures and has proved to be the most sensitive and specific noninvasive test for the diagnosis of DVT involving the proximal leg veins (51,57). US can be performed rapidly and without the need for extensive technologist training, provided the studies are restricted to the proximal deep venous system. It is also a portable technique that allows for the assessment of critically ill patients at the bedside.

US FOR THE DIAGNOSIS OF LOWER EXTREMITY DVT

Techniques
US of the lower extremity deep venous system is performed with the patient in the supine position, preferably with the head of the bed raised 20°–30° to promote venous pooling in the leg (6,54). The leg is externally rotated and slightly flexed at the knee. A 5–10-MHz linear transducer is required. Duplex Doppler or color Doppler imaging or both are helpful to identify venous vessels but are not mandatory (6,58,59). The transducer is placed transversely in the groin area to identify the common femoral vein, just medial to the common femoral artery. In the absence of DVT, gentle pressure should cause the vein lumen to collapse and the anterior and posterior walls to become superimposed (Fig 1). In the presence of DVT, it is not possible to collapse the vein lumen even with the application of sufficient pressure to occlude the adjacent artery. During the procedure, the transducer is moved distally along the deep venous system, and light compression is applied at 1-cm intervals. The procedure extends from the most proximal aspect of the common femoral vein, along the superficial femoral vein and popliteal vein until the popliteal vein divides into the posterior tibial and peroneal branches (calf trifurcation). At each level, the vein is assessed for the presence or absence of compressibility.

View larger version (182K): [in this window] [in a new window]   Figure 1a. (a) Transverse US image of the left common femoral vein (arrowhead) and smaller adjacent artery (arrow). (b) With compression, in the absence of thrombosis, the walls of the vein are completely apposed while the artery (arrow) remains patent.

View larger version (170K): [in this window] [in a new window]   Figure 1b. (a) Transverse US image of the left common femoral vein (arrowhead) and smaller adjacent artery (arrow). (b) With compression, in the absence of thrombosis, the walls of the vein are completely apposed while the artery (arrow) remains patent.

Compression of the veins deep within the adductor canal is sometimes difficult due to the deep course of the vein through the muscles. Compression in this segment of the distal superficial femoral vein is aided by placing one hand underneath the medial aspect of the distal thigh and compressing the vein between the fingers and the transducer. This not only aids in compression of the vein but also often brings the vein in closer proximity to the transducer head allowing for better visualization.

If the patient is mobile, the popliteal vein is best assessed in the lateral decubitus or prone position with the knee flexed passively to approximately 10°–15° to avoid collapse of the vein. If the patient is unable to move from the supine position, the popliteal vein can usually be adequately assessed by raising the affected leg sufficiently to place the transducer behind the knee, which again should be slightly flexed. In the popliteal fossa, the popliteal vein lies superficial to the artery and can be easily collapsed with gentle pressure, sometimes making localization difficult.

Compression US Evaluation of the Calf Veins
There is considerable controversy about the need to diagnose DVT isolated to the calf veins (1,6). In most studies, US imaging has been limited to the proximal venous system. However, some investigators believe examination of calf vessels is a useful procedure (60,61). The anatomic locations and US appearances of the calf veins are shown in Figure 2.

View larger version (74K): [in this window] [in a new window]   Figure 2a. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (173K): [in this window] [in a new window]   Figure 2b. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (126K): [in this window] [in a new window]   Figure 2c. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (125K): [in this window] [in a new window]   Figure 2d. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (147K): [in this window] [in a new window]   Figure 2e. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (108K): [in this window] [in a new window]   Figure 2f. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (169K): [in this window] [in a new window]   Figure 2g. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

View larger version (173K): [in this window] [in a new window]   Figure 2h. Location of the calf veins. (a) Location of the paired calf veins just above the ankle. ATV = anterior tibial veins, F = fibula, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (b) Transverse US image just above ankle demonstrates the paired posterior tibial veins (V) and posterior tibial artery (A) imaged from a posteromedial approach. Note there is inadequate venous flow velocity to visualize with color Doppler without flow augmentation. (c) Longitudinal image with Doppler interrogation of one of the posterior tibial veins during flow augmentation. This can be used to help confirm the location of the vein and suggests patency. (d) Color Doppler US demonstration of the paired peroneal veins (V) and artery (A) just above the ankle during augmentation via a posterolateral approach. (e) Paired anterior tibial veins (V) and artery (A) during flow augmentation via an anterior approach. (f) Location of the paired calf veins at the midcalf level. ATV = anterior tibial veins, F = fibula, IM = interosseous membrane, PTV = posterior tibial veins, PV = peroneal veins, T = tibia, TN = tibial nerve. (g) Transverse color Doppler image of the paired larger posterior tibial veins (arrows) and smaller peroneal veins (arrowheads) at the level of the midcalf during flow augmentation imaged from a posterior medial approach. (h) Transverse US image of the paired anterior tibial veins (arrowheads) during augmentation imaged via an anterior approach. The vessels are often difficult to visualize at or proximal to this level because of their deep location just anterior to the interosseous membrane between the tibia and fibula.

The paired peroneal veins lie laterally in the lower leg just posterior to the fibula at the level of the ankle. More proximally, these veins are located deeper in the calf musculature and accompany the peroneal artery adjacent to the fibula. Once identified, these vessels can be followed along with the posterior tibial veins as they both have parallel courses along the posterior tibialis muscle, with the peroneal veins lying slightly anterolaterally. More proximally, the two paired veins form the common tibial trunk, which becomes the popliteal vein. Although adequately evaluated by means of a medial approach, occasionally a lateral approach with the leg flexed at the knee will be helpful to image segments of the peroneal veins.

The anterior tibial veins originate at the dorsum of the foot just anteromedial to the tibia. They continue proximally coursing lateral to the tibia and deep to the tibialis anterior and extensor muscles. At the level of the midcalf, they lie anteriorly to the interosseous membrane. The anterior tibial veins usually cannot be followed above the proximal third of the calf, and their connection with the popliteal vein is usually not readily visible.

In addition to the three paired deep calf veins, there are gastrocnemius muscular vein branches that join directly into the popliteal veins and soleal veins; these latter veins extend from the soleal sinuses within the deep musculature to the posterior tibial or peroneal veins and may harbor thrombus (56).

The three groups of paired calf veins, once localized, are evaluated by using compression US in a similar fashion to that of the proximal veins. Augmentation with manual compression and color flow imaging of the lower calf is useful to localize and evaluate the veins for flow (62). The small caliber of the veins and their varied anatomic locations make the examination for calf DVT a time-consuming process.

Diagnostic US Criteria for DVT
In the absence of DVT, the entire deep venous system between the common femoral vein and the trifurcation of the popliteal vein should collapse with complete apposition of the vein walls during gentle compression. The inability to completely compress the vein lumen is the principal criterion for the diagnosis of DVT (6,7,54,63,64) (Fig 3a–3c). In patients with DVT, several adjunctive US findings may be observed that are less sensitive and less specific than vein compressibility as diagnostic parameters. Acute DVT usually causes distention of the involved vein, and Doppler evaluation may reveal absence of flow (Fig 3d). If there is incomplete obstruction of the vein lumen, there is usually loss of the phasic respiratory venous flow pattern and a continuous flow wave may be observed that is minimally affected by augmentation. Acute DVT is often anechoic and cannot be distinguished from a normal vein. With time, the clot usually will become echoic. In the absence of DVT, internal echos due to artifact of slow-flowing blood may be observed within vein lumina. Thus, lack of vein compression is required to confirm the presence of DVT (6).

View larger version (169K): [in this window] [in a new window]   Figure 3a. Case of a 34-year-old woman who presented early in the postpartum period with buttock, groin, and upper thigh pain. (a) Transverse US image of the common femoral vein (arrow) shows a noncompressible segment with distention of the vein from acute thrombus. This also involved the proximal superficial femoral vein. Distance between cursors (1, dotted line) is 1.07 cm. (b, c) US images demonstrate normal compressibility of popliteal segment. (b) Transverse image of the popliteal artery (red) and vein (blue) shows spontaneous venous flow. (c) With compression, there is complete apposition of the vein walls, while the artery (arrowhead) is patent; this excludes popliteal vein thrombosis. (d) Longitudinal color Doppler image along the left iliac artery (yellow and blue) and thrombosed left iliac vein (arrowheads). Note that a lower frequency transducer was required to visualize the vessels due to their deeper location in the pelvis. (e) Longitudinal color Doppler image demonstrates the patent inferior vena cava (orange) in this patient.

View larger version (97K): [in this window] [in a new window]   Figure 3b. Case of a 34-year-old woman who presented early in the postpartum period with buttock, groin, and upper thigh pain. (a) Transverse US image of the common femoral vein (arrow) shows a noncompressible segment with distention of the vein from acute thrombus. This also involved the proximal superficial femoral vein. Distance between cursors (1, dotted line) is 1.07 cm. (b, c) US images demonstrate normal compressibility of popliteal segment. (b) Transverse image of the popliteal artery (red) and vein (blue) shows spontaneous venous flow. (c) With compression, there is complete apposition of the vein walls, while the artery (arrowhead) is patent; this excludes popliteal vein thrombosis. (d) Longitudinal color Doppler image along the left iliac artery (yellow and blue) and thrombosed left iliac vein (arrowheads). Note that a lower frequency transducer was required to visualize the vessels due to their deeper location in the pelvis. (e) Longitudinal color Doppler image demonstrates the patent inferior vena cava (orange) in this patient.

View larger version (96K): [in this window] [in a new window]   Figure 3c. Case of a 34-year-old woman who presented early in the postpartum period with buttock, groin, and upper thigh pain. (a) Transverse US image of the common femoral vein (arrow) shows a noncompressible segment with distention of the vein from acute thrombus. This also involved the proximal superficial femoral vein. Distance between cursors (1, dotted line) is 1.07 cm. (b, c) US images demonstrate normal compressibility of popliteal segment. (b) Transverse image of the popliteal artery (red) and vein (blue) shows spontaneous venous flow. (c) With compression, there is complete apposition of the vein walls, while the artery (arrowhead) is patent; this excludes popliteal vein thrombosis. (d) Longitudinal color Doppler image along the left iliac artery (yellow and blue) and thrombosed left iliac vein (arrowheads). Note that a lower frequency transducer was required to visualize the vessels due to their deeper location in the pelvis. (e) Longitudinal color Doppler image demonstrates the patent inferior vena cava (orange) in this patient.

View larger version (192K): [in this window] [in a new window]   Figure 3d. Case of a 34-year-old woman who presented early in the postpartum period with buttock, groin, and upper thigh pain. (a) Transverse US image of the common femoral vein (arrow) shows a noncompressible segment with distention of the vein from acute thrombus. This also involved the proximal superficial femoral vein. Distance between cursors (1, dotted line) is 1.07 cm. (b, c) US images demonstrate normal compressibility of popliteal segment. (b) Transverse image of the popliteal artery (red) and vein (blue) shows spontaneous venous flow. (c) With compression, there is complete apposition of the vein walls, while the artery (arrowhead) is patent; this excludes popliteal vein thrombosis. (d) Longitudinal color Doppler image along the left iliac artery (yellow and blue) and thrombosed left iliac vein (arrowheads). Note that a lower frequency transducer was required to visualize the vessels due to their deeper location in the pelvis. (e) Longitudinal color Doppler image demonstrates the patent inferior vena cava (orange) in this patient.

View larger version (204K): [in this window] [in a new window]   Figure 3e. Case of a 34-year-old woman who presented early in the postpartum period with buttock, groin, and upper thigh pain. (a) Transverse US image of the common femoral vein (arrow) shows a noncompressible segment with distention of the vein from acute thrombus. This also involved the proximal superficial femoral vein. Distance between cursors (1, dotted line) is 1.07 cm. (b, c) US images demonstrate normal compressibility of popliteal segment. (b) Transverse image of the popliteal artery (red) and vein (blue) shows spontaneous venous flow. (c) With compression, there is complete apposition of the vein walls, while the artery (arrowhead) is patent; this excludes popliteal vein thrombosis. (d) Longitudinal color Doppler image along the left iliac artery (yellow and blue) and thrombosed left iliac vein (arrowheads). Note that a lower frequency transducer was required to visualize the vessels due to their deeper location in the pelvis. (e) Longitudinal color Doppler image demonstrates the patent inferior vena cava (orange) in this patient.

View larger version (195K): [in this window] [in a new window]   Figure 4a. US images of acute proximal DVT. (a) Transverse image of the superficial femoral artery (A) and vein (V) shows noncompressible vein due to acute thrombus. (b) Longitudinal image at the same level demonstrates lack of spontaneous venous flow. Red represents flow within the adjacent superficial femoral artery. (c) Longitudinal color Doppler image during augmentation demonstrates flow (blue) around the acute thrombus. Red represents arterial flow within the adjacent superficial femoral artery.

View larger version (138K): [in this window] [in a new window]   Figure 4c. US images of acute proximal DVT. (a) Transverse image of the superficial femoral artery (A) and vein (V) shows noncompressible vein due to acute thrombus. (b) Longitudinal image at the same level demonstrates lack of spontaneous venous flow. Red represents flow within the adjacent superficial femoral artery. (c) Longitudinal color Doppler image during augmentation demonstrates flow (blue) around the acute thrombus. Red represents arterial flow within the adjacent superficial femoral artery.

View larger version (68K): [in this window] [in a new window]   Figure 4b. US images of acute proximal DVT. (a) Transverse image of the superficial femoral artery (A) and vein (V) shows noncompressible vein due to acute thrombus. (b) Longitudinal image at the same level demonstrates lack of spontaneous venous flow. Red represents flow within the adjacent superficial femoral artery. (c) Longitudinal color Doppler image during augmentation demonstrates flow (blue) around the acute thrombus. Red represents arterial flow within the adjacent superficial femoral artery.

Report of US Examination
Since US is a real-time test and is not videotaped in most radiology departments, reports are critical for documenting the extent and location of DVT. This documentation is important not just for the initial diagnosis but also for providing a baseline evaluation in anticipation of the potential reinvestigation of a patient at a later date for suspected recurrent DVT. A diagram outlining the extent of the thrombosis that can become a part of the patient's permanent record can be very helpful in this regard. Such a diagram can also indicate the maximal compressed diameters of the popliteal and common femoral veins.

Radiologists must be wary of a clinician's limited knowledge of the deep venous anatomy. For instance, many physicians are unaware that the superficial femoral vein is actually a deep venous structure and that thrombosis involving this vein may be considered as superficial phlebitis by the unsuspecting clinician. The term superficial femoral vein either should be avoided in the radiology report or at least the conclusion section of the report should clearly indicate the presence of a DVT if the clot involves the superficial femoral vein (65).

Appearance of Chronic DVT
Over time, the US appearance of DVT will evolve. In some areas of the vein, the thrombus may become increasingly echogenic and the intima of the vein wall may thicken and become echogenic and resistant to compression (Fig 5) (66). However, other areas of the vein may revert to normal both in appearance and in response to compression. By 12–24 months after an acute DVT, about 50% of patients will have complete US resolution of thrombus and normally compressible proximal leg veins (66–68). This is associated with restoration of antegrade venous flow and frequently with the development of venous reflux in a previously occluded vein. The contrasting appearances of the acute and chronic DVT are outlined in Table 2.

View larger version (178K): [in this window] [in a new window]   Figure 5a. Appearance of chronic DVT. (a) Transverse US image of the proximal superficial femoral vein just below the bifurcation of the common femoral artery demonstrates echogenic thrombus (arrow) within the vein. (b) Transverse image at the same level during compression shows noncompressible superficial femoral vein (V). Note mild flattening (arrow) of the wall of the femoral artery (A) with compression. (c) Color Doppler image at the same level demonstrates spontaneous venous flow around the thrombus (T) within the partially recanalized superficial femoral vein (V). Arterial flow is seen within the superficial and deep femoral arteries (A). (d) Longitudinal view of the superficial femoral vein with color Doppler flow US demonstrates irregular recanalization of the vein (arrows) with spontaneous flow around the thrombus (T), and arterial flow within the adjacent artery (arrowhead). (e) Longitudinal color Doppler flow image of the superficial femoral vein (V) demonstrates spontaneous phasic flow around the thrombus (T) within the partly recanalized vein (V). Flow is also noted in the adjacent artery (A).

View larger version (187K): [in this window] [in a new window]   Figure 5b. Appearance of chronic DVT. (a) Transverse US image of the proximal superficial femoral vein just below the bifurcation of the common femoral artery demonstrates echogenic thrombus (arrow) within the vein. (b) Transverse image at the same level during compression shows noncompressible superficial femoral vein (V). Note mild flattening (arrow) of the wall of the femoral artery (A) with compression. (c) Color Doppler image at the same level demonstrates spontaneous venous flow around the thrombus (T) within the partially recanalized superficial femoral vein (V). Arterial flow is seen within the superficial and deep femoral arteries (A). (d) Longitudinal view of the superficial femoral vein with color Doppler flow US demonstrates irregular recanalization of the vein (arrows) with spontaneous flow around the thrombus (T), and arterial flow within the adjacent artery (arrowhead). (e) Longitudinal color Doppler flow image of the superficial femoral vein (V) demonstrates spontaneous phasic flow around the thrombus (T) within the partly recanalized vein (V). Flow is also noted in the adjacent artery (A).

View larger version (120K): [in this window] [in a new window]   Figure 5c. Appearance of chronic DVT. (a) Transverse US image of the proximal superficial femoral vein just below the bifurcation of the common femoral artery demonstrates echogenic thrombus (arrow) within the vein. (b) Transverse image at the same level during compression shows noncompressible superficial femoral vein (V). Note mild flattening (arrow) of the wall of the femoral artery (A) with compression. (c) Color Doppler image at the same level demonstrates spontaneous venous flow around the thrombus (T) within the partially recanalized superficial femoral vein (V). Arterial flow is seen within the superficial and deep femoral arteries (A). (d) Longitudinal view of the superficial femoral vein with color Doppler flow US demonstrates irregular recanalization of the vein (arrows) with spontaneous flow around the thrombus (T), and arterial flow within the adjacent artery (arrowhead). (e) Longitudinal color Doppler flow image of the superficial femoral vein (V) demonstrates spontaneous phasic flow around the thrombus (T) within the partly recanalized vein (V). Flow is also noted in the adjacent artery (A).

View larger version (135K): [in this window] [in a new window]   Figure 5d. Appearance of chronic DVT. (a) Transverse US image of the proximal superficial femoral vein just below the bifurcation of the common femoral artery demonstrates echogenic thrombus (arrow) within the vein. (b) Transverse image at the same level during compression shows noncompressible superficial femoral vein (V). Note mild flattening (arrow) of the wall of the femoral artery (A) with compression. (c) Color Doppler image at the same level demonstrates spontaneous venous flow around the thrombus (T) within the partially recanalized superficial femoral vein (V). Arterial flow is seen within the superficial and deep femoral arteries (A). (d) Longitudinal view of the superficial femoral vein with color Doppler flow US demonstrates irregular recanalization of the vein (arrows) with spontaneous flow around the thrombus (T), and arterial flow within the adjacent artery (arrowhead). (e) Longitudinal color Doppler flow image of the superficial femoral vein (V) demonstrates spontaneous phasic flow around the thrombus (T) within the partly recanalized vein (V). Flow is also noted in the adjacent artery (A).

View larger version (51K): [in this window] [in a new window]   Figure 5e. Appearance of chronic DVT. (a) Transverse US image of the proximal superficial femoral vein just below the bifurcation of the common femoral artery demonstrates echogenic thrombus (arrow) within the vein. (b) Transverse image at the same level during compression shows noncompressible superficial femoral vein (V). Note mild flattening (arrow) of the wall of the femoral artery (A) with compression. (c) Color Doppler image at the same level demonstrates spontaneous venous flow around the thrombus (T) within the partially recanalized superficial femoral vein (V). Arterial flow is seen within the superficial and deep femoral arteries (A). (d) Longitudinal view of the superficial femoral vein with color Doppler flow US demonstrates irregular recanalization of the vein (arrows) with spontaneous flow around the thrombus (T), and arterial flow within the adjacent artery (arrowhead). (e) Longitudinal color Doppler flow image of the superficial femoral vein (V) demonstrates spontaneous phasic flow around the thrombus (T) within the partly recanalized vein (V). Flow is also noted in the adjacent artery (A).

Compression US of the entire proximal deep venous system may be technically difficult to perform in some patients. These include markedly obese patients; patients with tense, swollen extremities; burn patients; and patients who have had recent lower extremity surgery. Fortunately, even in these patients, the most common sites for DVT (ie, the common femoral and popliteal veins) are readily identifiable due to their superficial location. Duplicated vein segments may pose a problem if thrombus within one limb of the duplicated segment is not visualized and compression of the nonthrombosed duplicated segment is interpreted as a negative examination. On occasion, one can be fooled by the presence of a large collateral vein; however, this can usually be avoided by carefully identifying the normal course of the deep veins in association with the adjacent arteries.

ACCURACY OF US FOR THE DIAGNOSIS OF DVT

Symptomatic Patients
US is a very accurate test for the diagnosis of DVT in symptomatic patients. Authors of independently performed accuracy studies in which venography was used as the reference standard have reported that the sensitivity and specificity of compression US exceed 95% and 98%, respectively, for the diagnosis of DVT involving the proximal leg veins (46–48,70).

Studies evaluating the accuracy of compression US for the evaluation of calf vein DVT have been relatively few, and the results have been highly variable. Sensitivity ranges have been between 11% and 100%, and specificity ranges have varied between 90% and 100% (60,61,70,71). A meta-analysis of methodologically high-quality studies reported the sensitivity of US for the diagnosis of DVT isolated to the calf veins to be 73% (70). The rates of technically inadequate studies were higher than those observed in evaluation of proximal DVT, as great as 40% in one study (71).

Asymptomatic Patients
US has been extensively evaluated as a screening test performed in asymptomatic patients in high-risk settings for DVT. A meta-analysis that was restricted to methodologically high-quality accuracy studies showed that the sensitivity of compression US screening for proximal DVT was only 62% following total hip or knee arthroplasty (72). Lower sensitivity rates were observed for isolated calf vein thrombosis. The sensitivity of postoperative US screening was not increased with use of color or Doppler imaging (58,73). Likely explanations for the low sensitivity of US screening in asymptomatic patients include small size of the thrombus, the unpredictable location of the thrombi, and their nonocclusive nature. The differences in the US appearance of DVT in symptomatic and asymptomatic patients are outlined in Table 3.

Although US is less sensitive for the diagnosis of proximal DVT in the postoperative setting, some have continued to advocate its potential role as a screening test following high-risk surgical procedures; the rationale being that thrombi may be detected while they are small and antithrombotic treatment initiated to prevent the development of clinically important complications. This hypothesis was recently evaluated in a randomized controlled trial that compared the outcome of patients who underwent a screening compression US procedure of the proximal venous system with those who had sham (or placebo) US performed prior to hospital discharge following total hip or knee arthroplasty (19). All patients in that study received warfarin prophylaxis, and screening venography was not performed. The authors reported that the rate of symptomatic DVT and pulmonary embolic complications in a 3-month follow-up period was identical for patients who had a normal US study at hospital discharge as for patients who did not have a US examination performed at all (19). On the basis of these results, routine US screening cannot be recommended following joint arthroplasty, at least in patients receiving DVT prophylaxis.

DIAGNOSTIC APPROACH FOR THE FIRST EPISODE OF SUSPECTED DVT

Patients with clinically suspected DVT should undergo compression US of the proximal venous system. Because of the very high specificity of this test, a positive compression US result is sufficiently predictive that treatment can be initiated in patients who have no history of DVT (1,6,34). The dilemma is how to treat patients who have a negative initial US study since about 10%–20% of patients with symptomatic DVT will have thrombosis isolated to the veins of the calf (8) and at least 20%–30% of calf vein DVT will eventually extend to the proximal venous system, at which point the risk of clinically important pulmonary embolic complications occurring is much greater (74,75).

To detect the proximal extension of calf vein thrombosis, the approach of serial US testing has evolved. Patients with an initial negative study have antithrombotic therapy withheld and then undergo at least one additional follow-up compression US study over a 1-week period. Patients who have proximal extension of DVT detected may have antithrombotic treatment initiated, whereas those with negative serial US studies are presumed not to have DVT accounting for their leg symptoms. Management studies have demonstrated that serial US testing is a relatively safe strategy for investigating patients with suspected DVT since those with negative serial US studies have a less than 1% risk of developing symptomatic proximal DVT or pulmonary embolism in a 3-month follow-up period (76,77).

The drawback of the serial US testing approach is that very few patients (1%–2% in two recent studies) with suspected DVT who have a negative initial US study will be confirmed to have proximal DVT at serial testing (76,77). In one study, the authors estimated that it costs at least $5,000 per proximal DVT detected by performing an additional US test (77). In addition, in a decision analysis study, the authors estimated that it would cost over $390,000 per life year saved from fatal pulmonary embolism to perform a second and $3.5 million per life year saved to perform a third US serial examination in patients whose initial US study was negative (78). Performing calf imaging in all patients with negative proximal examination results is likewise not desirable since the rate of inadequate scans would mitigate the need for additional testing in many patients. In addition, the positive predictive value of US for the diagnosis of calf vein DVT is quite low given the low prevalence of DVT and the lower specificity of US in this setting.

It would be desirable to identify a low-risk group of patients in whom the diagnosis of DVT could be safely excluded on the basis of a single negative US study of the proximal venous system. Approaches that potentially could be used to avoid performing unnecessary US tests would be incorporation of clinical pretest probability and/or D-dimer results into the diagnostic algorithm for patients with suspected DVT.

Clinical Diagnosis
The validation of the accuracy of the clinical diagnosis has resulted in an alternative approach to serial US for the investigation of patients with suspected DVT. Rather than consider all patients as having an equal likelihood of DVT, determination of pretest likelihood can be used to identify high-risk individuals for whom a more aggressive diagnostic approach may be indicated and low-risk individuals in whom minimal diagnostic testing needs to be performed.

A recent management study has confirmed that consideration of the clinical pretest probability can be used to obviate serial US testing in a subset of patients with suspected DVT. Patients with a low clinical pretest probability as determined by explicit clinical criteria had DVT safely excluded on the basis of a single negative US test (41). Fewer than 0.5% of patients in this subgroup developed venous thromboembolic complications in a 3-month follow-up period. This study also showed that consideration of the clinical diagnosis could be used to identify a subgroup of patients who were at increased risk of having a false-negative US result. About 18% of patients who clinicians thought were highly likely to have DVT and whose US studies were negative were found to have a proximal DVT at venography. However, in each case, the thrombosis involved only the popliteal vein and arguably could have be been safely detected with a serial US test a week later (41).

D-Dimer Test
Several different serologic markers of thrombosis have been evaluated for their predictive value for the diagnosis of DVT. The test that has been studied most extensively and appears to be the most useful is the D-dimer test. D-Dimer is a breakdown product of the cross-linked fibrin blood clot. Several D-dimer assays have been validated to be sensitive but nonspecific markers of DVT and pulmonary embolism (79–84). Therefore, a positive test has a very low predictive value and is not helpful at "ruling in" the diagnosis of DVT. However, studies have reported negative predictive values of over 97% by using D-dimer testing for the diagnosis of DVT (84,85), and clinical trials have focused on the potential value of using the negative predictive value of the D-dimer test to exclude the diagnosis of DVT.

It is important to recognize that the negative predictive value of the D-dimer test varies depending on a patient's clinical probability for DVT. In one study, the overall negative predictive value of the D-dimer test was 98.3% (95% CI: 96.2%, 100%) (85). In patients with a low pretest probability for DVT, the predictive value appeared to rise to 99.5% (95% CI: 97.3%, 100%). However, in patients with a high clinical probability for DVT, the negative predictive value of the D-dimer test was only 85.7% (95% CI: 42.0%, 99.6%) (Table 4).

Ongoing clinical trials are also evaluating the safety of obviating any imaging procedure in patients who are at low pretest probability for DVT and who have a negative D-dimer result. A negative D-dimer result should not prevent the performance of diagnostic testing in patients with moderate or high probability for DVT because of the lower predictive value of a negative result in these patient populations.

There are many different commercially available D-dimer assays. In laboratories in which hemostasis testing is performed, the D-dimer test has been used for many years as a diagnostic test for disseminated intravascular coagulation. Some of these D-dimer assays are of insufficient sensitivity to be used as diagnostic tests for DVT. It is important for physicians relying on the negative predictive value of the D-dimer assay to be sure that the test being used in their local hospitals is one that has been validated for the diagnosis of DVT or pulmonary embolism. Management algorithms for patients with suspected DVT are presented in Figure 6.

View larger version (18K): [in this window] [in a new window]   Figure 6. A–C, Management algorithms for patients with suspected DVT. A, Serial US; B, US and clinical probability; C, US, clinical probability, D-dimer test. * = alternativ