Noninvasive Venous Studies

Each year in the United States, approximately 200,000 patients die from a pulmonary embolus. DVT can be found in about 80% of patients with a pulmonary embolus. See Figure 88-1 for the most common sites for DVT; the incidence increases with age, and DVT is more common in women. One third to one half of patients older than 40 years of age who experience an acute myocardial infarction, a hip fracture, major surgery (especially orthopedic, pelvic, or urologic), or a stroke develop venous thrombi. Box 88-1 lists traditional risk factors in hospitalized patients. Lower limb DVT affects 1% to 2% of hospitalized patients. In addition, as a result of previous DVT, the prevalence of postphlebitic sequelae in the adult population is estimated to be 5%.

Figure 88-1 Six most common sites of deep venous thrombosis (DVT) in the lower body. 1, Left iliac vein; 2, common femoral vein; 3, termination of deep femoral vein (profunda femoris); 4, popliteal vein at adductor canal; 5, posterior tibial vein; 6, intramuscular veins of calf.

Box 88-1 Traditional Risk Factors for Deep Venous Thrombosis (Especially in Hospitalized Patients)

Acute myocardial infarction

Acute respiratory failure

Acute stroke with paresis

Blood type (persons with type A may be at higher risk than those with type O)

Fractures, especially spine, pelvis, long bone fractures, or multiple fractures

Heart disease, especially congestive heart failure

Hypercoagulable states

Local injury to veins

Malignancy

Mechanical ventilation

Myeloproliferative disorders

Nephrotic syndrome

Oral contraceptive use

Paroxysmal nocturnal hemoglobinuria

Persons older than 40 years of age

Persons with paralysis or otherwise immobilized

Postoperative state from major surgery

Pregnancy or postpartum, especially postcesarean section

Previous DVT or pulmonary embolism

Trauma

Ulcerative colitis or Behçet’s syndrome

Venous stasis

Early diagnosis of DVT is important because approximately 50% of untreated proximal DVT cases will result in a pulmonary embolism. Diagnosis of DVT is also important to minimize long-term complications such as venous stasis or ulceration from chronic venous insufficiency. Accurate diagnosis is crucial to limit anticoagulation therapy to those who really need it. Venous thrombi usually arise at bifurcations and in valve cusps. An aging thrombus can adhere to the vein wall and damage or destroy nearby valves. The two most important valves for controlling venous hydrostatic pressure are those of the proximal superficial femoral vein and the distal popliteal vein. Destruction of these valves is more likely to lead to sequelae. The goal is to diagnose DVT before a thrombus either embolizes or becomes extensive enough to permanently damage these or any other valves.

Clinical diagnosis of acute DVT, without the benefit of radiographic or noninvasive techniques, has been reported to be notoriously inaccurate for years, with only about a 50% accuracy rate. However, this accuracy was probably underestimated because it was based on older studies performed on seriously ill, hospitalized patients. Since then, various scoring systems have been developed in an attempt to predict pretest likelihood of DVT in ambulatory patients. The best known and studied is the Wells scoring system, which was first proposed in 1995 and updated in 2003. Although patients in this study were ambulatory, they were seen in either the emergency department or hospital, so these data may not be as applicable to patients in a primary care clinic. However, the Wells system has since been studied in 1082 ambulatory patients presenting to 5 major academic medical centers. Of these, 495 patients were thought likely to have DVT, whereas 587 were categorized as unlikely. Diagnostic evaluation followed by 3 months of observation confirmed DVT or pulmonary embolism in 28% of those thought likely to have DVT and in only 5.5% of those deemed unlikely to have DVT (Table 88-1). Combining the Wells system with further diagnostic studies can be used to virtually exclude DVT. For instance, a negative D-dimer test result in those thought unlikely to have DVT effectively excluded DVT (<1%) during the 3-month follow-up.

TABLE 88-1 Wells Scoring System for Predicting Deep Venous Thrombosis

Clinical Variable	Score*
Active cancer (ongoing treatment or active within the last 6 mo or palliative care for cancer)	1
Paralysis, paresis, or recent plaster immobilization of the lower extremities	1
Recently bedridden for 3 or more days, or major surgery within the last 12 wk requiring regional or general anesthesia	1
Localized tenderness along the distribution of the deep venous system	1
Entire leg swelling	1
Calf swelling at least 3 cm larger in circumference than that of the asymptomatic leg, measured 10 cm below the tibial tuberosity	1
Pitting edema confined to the affected leg	1
Distended collateral superficial veins (not varicosities)	1
Previously documented DVT	1
Alternative diagnosis at least as likely as DVT	−2

DVT, deep venous thrombosis.

* Scoring method: if 1 or less, DVT unlikely; if 2 or greater, DVT likely.

From Wells PS, Owen C, Doucette S, et al: Does this patient have deep vein thrombosis? JAMA 295:199–207, 2006.

Alternatives for Diagnosis of Deep Venous Thrombosis, Their Limitations, and Evidence Supporting Their Use

1 Contrast venography is regarded as the gold standard for diagnosis of DVT; however, it is not without its own risks—including allergic reactions, congestive heart failure, acute renal insufficiency, and postvenography syndrome. (After venography, postvenography syndrome affects 10% to 20% of patients. Although it usually causes only transient discomfort in the calf for 24 to 48 hours, and in most cases it resolves without treatment, it may actually progress to DVT.) In addition, venography is not easily repeated, and it is usually performed only in one limb. Venography may be impossible to perform in patients with poor venous access, especially obese patients and those with severe edema or cellulitis. It may also be difficult to perform in urgent situations without proper support staff. Furthermore, contrast venography cannot be performed in 20% to 25% of patients because of previous DVT. Although a study of postmortem contrast venograms reported a sensitivity rate of 95% and a specificity rate of 97% for diagnosis of DVT, a more recent multicenter study evaluating the degree of interobserver variation questions these high sensitivities and specificities. For these reasons, as well as the fact that compression and duplex ultrasonography have become more readily available, they have basically replaced contrast venography for making the diagnosis of DVT.

2 Compression ultrasonographic imaging (high-frequency, B-mode, real-time) has limitations in obese and asymptomatic postoperative patients. Ultrasonography’s ability to visualize the venous system above the inguinal ligament (i.e., pelvic, iliac veins) or distal to the popliteal vein is also limited. That said, for symptomatic proximal DVT, sensitivities ranging from 93% to 100% and specificities ranging from 97% to 100% have been reported since the 1980s (Appelman and colleagues, 1987; Vogel and colleagues, 1987; Cronan and colleagues, 1987; Lensing and colleagues, 1989). Compression ultrasonography is less operator dependent than duplex scanning, and this may explain why it is the most common technique used in large urgent-emergent care centers and after hours in hospitals. Because it is a less expensive technique, many centers have developed algorithms for its use; they either use compression ultrasonography alone in low-risk patients or use compression ultrasonography as a screen to determine whether a duplex scan is needed. Although approximately 10% of isolated calf DVTs will be missed with compression ultrasonography, calf DVT by itself is not life-threatening and may only need to be followed to exclude proximal progression. Benefits of compression ultrasonography include the ability actually to visualize the veins, valves, and thrombus. Compression ultrasonography may be necessary in special cases in which (other than obese or asymptomatic postoperative patients) impedence plethysmography (IPG) has unclear results or limitations. Compression ultrasonography is also usually performed during duplex scanning.

3 Duplex ultrasonographic scanning combines velocity measurements using Doppler technology with ultrasonographic imaging. Duplex scanning may be used to confirm findings from compression ultrasonography. Duplex scanning is accurate and reproducible, and compared with venography, its sensitivity and specificity for DVT are more than 90%. Both the positive and negative predictive values for DVT are in the 90% to 95% range. Studies have shown that color enhancement of the Doppler velocities also improves accuracy in areas where compression may be difficult, such as at the inguinal ligament, the adductor canal, or in the calf veins. It may also be used to determine the patency of the pelvic and iliac veins and the inferior vena cava.

4 Electrical impedance plethysmography was previously the most extensively studied and commonly used noninvasive technique. It remains the least expensive and least operator-dependent alternative to venography. IPG provides a functional evaluation of the venous system for outflow obstruction. Compared with contrast venography, a sensitivity of 92% and a specificity of 95% were reported for IPG from studies in the 1990s. However, the range for sensitivity dropped to 66% in one study. High false-positive rates are found in the presence of certain conditions (e.g., obesity, congestive heart failure, external venous compression from gravid uterus in pregnancy) and chronic DVT. IPG is inaccurate in the diagnosis of thrombi in calf veins, profunda femoris, or internal iliac veins. It is also limited for diagnosing nonoccluding DVT, asymptomatic or moderately symptomatic DVT, and, therefore, postoperative DVT. IPG is limited for diagnosing clots in paired veins or determining the progression of disease. For these situations, duplex or color Doppler imaging has mostly replaced IPG. That said, a number of prospective studies (Hull and colleagues, 1985; Huisman and colleagues, 1986; Heijboer and colleagues, 1993) have demonstrated that serial negative IPGs are sufficiently sensitive to justify withholding anticoagulation, even in symptomatic patients.

5 Hand-held (pocket) Doppler (with or without recorded velocities) can be used to assess venous function. Pooled data from several studies found an overall sensitivity of 84% and specificity of 88% for detection of lower extremity DVT in symptomatic outpatients. However, individual studies report sensitivities ranging from 31% to 100% and specificities ranging from 59% to 100%, thereby indicating the shortcoming of this technique and the fact that it is very operator dependent (Turnbull and Dymowski, 1989). Consequently, hand-held Doppler is probably most useful when combined with another study such as compression ultrasonography or IPG, and was therefore included in the protocols of many of the original noninvasive venous studies. When combined with compression ultrasonography or IPG, hand-held Doppler can add information about the calf veins.

6 D-dimer, a degradation product of cross-linked fibrin, is usually elevated in patients with DVT. After determining pretest likelihood, the results of a high-sensitivity D-dimer assay can be used in place of IPG, ultrasonography, or duplex scanning in patients unlikely to have DVT and is especially useful in symptomatic patients with a suspected first episode of DVT. It can also be checked in addition to IPG, ultrasonography, or duplex scanning. A D-dimer test is most effective in excluding DVT in outpatients because hospitalized patients frequently have other conditions that cause an elevated result (i.e., false positives). Increasing age also increases the likelihood of false-positive results. Although there are at least seven commercial assays available, the enzyme-linked immunosorbent (ELISA) technique is considered a high-sensitivity test. From a recent meta-analysis (Wells and colleagues, 2006), a negative high-sensitivity D-dimer test combined with a low-probability score (pretest likelihood) resulted in a 0.1% likelihood of DVT during 3-month follow-up without using ultrasonography. Although semiquantitative slide agglutination assays are probably not accurate enough to use for exclusion of DVT (low sensitivity), newer quantitative agglutination methods (moderate sensitivity) have increased the accuracy to acceptable levels (<0.5% during 3-month follow-up) in patients unlikely to have DVT by the Wells scoring system. Likewise, bedside assays that use whole blood are also now available and are considered to be of moderate sensitivity and useful when combined with the Wells scoring system (Fig. 88-2).

7 Radionuclide scintigraphy or magnetic resonance venography (MRV) may be useful in selected patients. Scintigraphy uses radiolabeled albumin or tagged red blood cells as the contrast for venography. However, it may be unreliable in the calf and not helpful in those with previous DVT. Radiolabeled autologous platelets or peptides can also be used to detect active thrombus formation. These techniques may be helpful in symptomatic patients with a previous history of DVT. MRV is particularly useful in the diagnosis of portal, inferior vena caval, pelvic vein, or calf vein DVT. However, MRV is highly operator dependent, expensive, and not always available in urgent situations. It should probably be reserved for cases in which scintigraphy is contraindicated. Both radionuclide scintigraphy and MRV are beyond the scope of this textbook.

Figure 88-2 Algorithm for incorporating Wells scoring system into testing for deep venous thrombosis (DVT). (1) If DVT is unlikely (see Table 88-1) and the moderate or high-sensitivity D-dimer test (or impedence plethysmography [IPG], or both) result is negative, no further testing is required. (2) Symptomatic patients with an abnormal compression ultrasonography (CUS) or duplex study result can be treated without any further testing, or (3) if the patient has a negative CUS or duplex study result, yet DVT is likely, options include venography or serial CUS or duplex studies (i.e., repeat in 1 week). This algorithm provides a safe and cost-effective manner of excluding DVT. ^*Because this algorithm has not been studied prospectively and there is some subjectivity in using the Wells scoring system, some clinicians will always perform another noninvasive test before excluding DVT.