In managing patients with lower limb venous thrombosis it is of utmost importance to have a good understanding of the venous system of the lower leg (see Fig. 27.1). A source of ongoing confusion and a “potentially lethal misnomer” is the term “superficial femoral vein” which describes a vein that is not superficial [16]. The femoral vein is contained within the deep muscle compartments bound by the muscle fascia and is a “deep” vein. DVTs above the knee (but including involvement of the popliteal vein) are considered proximal deep venous thromboses of the lower extremity.

Critically ill ICU patients have many factors which increase the risk of DVT. The risk of DVT can be quantified according to the modified Caprini score (see Table 27.2) [17–19]. Using this risk assessment tool most ICU patients fall into the high risk or very high risk group. Consequently ALL ICU patients require DVT prophylaxis (mechanical, pharmacologic, or both) on admission to the ICU. Despite universal guidelines recommending DVT prophylaxis in all ICU patients, in the XPRESS trial only 50 % of the patients were receiving DVT prophylaxis [20]. Ho et al. demonstrated that omission of thromboprophylaxis within the first 24 h of ICU admission was associated with an increased risk of mortality (OR, 1.22; CI 1.15–1.30, p = 0.001) [21].

While few head-to-head randomized studies have been performed, sequential compression devices (SCDs), unfractionated heparin (UFH) and low molecular weight heparin (LMWH) appear to be equally effective in reducing the risk of DVT in low to moderate risk patients [22–25]. In high risk patients (orthopedic patients) SCDs appear to be as effective as LMWH in preventing DVTs [26–28]. In high risk patients LMWH and fondaparinux appear to be more effective in preventing asymptomatic DVT than UFH [29–31]. The combination of SCDs and LMWH appear to act additively in reducing the risk of DVT [31–33]. Graduated compression stockings, however, appear to have no role in preventing DVT [32]. These data suggest that in high to very high risk patients a combination of pharmacologic prophylaxis (if not contraindicated) together with SCD’s are indicated [19]. As is evidenced from Table 27.2 trauma patients are at a very high risk of DVT. LMWH has been demonstrated to be safe and effective in trauma patients and is the approach recommended by the East Surgical Group and the current ACCP guidelines [34, 35]. In trauma patients at high risk of bleeding the placement of a removal IVC filter may be a suitable strategy [36].

In a meta-analysis comparing any heparin formulation to placebo the risk of DVT and PE was 50 % lower with heparin [37]. Trials testing UFH used only bid dosing. Although there are no direct comparisons of bid versus tid UFH in any population, indirect comparisons suggest that their effects are similar on thrombosis and bleeding. The Canadian Critical Care Trial group performed a multi-center trial, in which they compared dalteparin (a LMWH) to unfractionated heparin (at a dose of 5,000 IU twice daily) in 3,764 patients [22]. There was no significant difference between the rates of proximal DVT, which occurred 5.1 % of patients receiving dalteparin versus 5.8 % receiving UH. The proportion of patients with PE was significantly lower with dalteparin (1.3 %) than with UH (2.3 %); the explanation for this finding and its implications are unclear. There was no significant difference in the rates of major bleeding or death in the hospital.

The use of sequential compression devices (SCDs) and graduated compression stockings (GCS) in the prevention of DVT has been controversial. However, as discussed above SCDs appear to be as effective as UF and LMWH in reducing the risk of DVT. In an observational propensity adjusted study Arabi et al. demonstrated that the use of SCDs was associated with a significantly lower VTE incidence compared with no mechanical thromboprophylaxis (RR 0.45; 0.22–0.95; p = 0.04) [38]. In this study GCS were not associated with a lower risk of VTE. Vignon randomized 407 patients with a high risk of bleeding to receive SCDs and GCS or GCS alone for 6 days during their ICU stay [39]. By day 6, the incidence of TED was 5.6 % in the SCD + GCS group and 9.2 % in the GCS group (RR 0.60; CI 0.28–1.28; p = 0.19). Lacut et al. randomized 151 patients who had suffered an intracerebral bleed to SCD + GCS or GCS alone [40]. Asymptomatic DVTs were detected in 4.7 % in the SCD + GCS group and 15.9 % in the GCS group (RR 0.29; CI 0.08–1.00). The CLOTS 3 trial randomized 2,876 immobile patients who had suffered a stroke to SCDs or no SCDs [41]. The primary outcome (DVT in the proximal veins detected on a screening CDU or symptomatic DVT in the proximal veins within 30 days of randomization) occurred in 8.5 % of patients allocated SCD and 12.1 % of patients allocated no SCD; an absolute reduction in risk of 3.6 % (95 % CI 1.4–5.8). The CLOTS 1 study randomized 2,518 patients who had suffered a stroke to thigh-length GCS or no GCS [42]. The primary outcome occurred in 10.0 % patients allocated to thigh-length GCS and in 10.5 % who did not receive GCS. These studies provide robust data that SCD reduce the risk of DVT while GCS do not.

The current American College of Chest Physicians (ACCP) guidelines suggest “For acutely ill hospitalized medical patients at increased risk of thrombosis, we recommend anticoagulant thromboprophylaxis with LMWH, UH bid, UH tid, or fondaparinux (Grade 1B)” [43]. Furthermore, they recommend “For acutely ill hospitalized medical patients at increased risk of thrombosis who are bleeding or at high risk for major bleeding, we suggest the optimal use of mechanical thromboprophylaxis with SCD. When bleeding risk decreases, and if VTE risk persists, we suggest that pharmacologic thromboprophylaxis be substituted for mechanical thromboprophylaxis” (Grade2B). The reader is referred to the “American College of Chest Physicians evidence-based clinical practice guidelines on the Prevention of venous thromboembolism (9th Edition)” for a comprehensive review on this topic [43]. Decousus et al. evaluated the independent risk factors for bleeding in a large cohort of medical patients [44]. These authors then developed a bleeding score risk to quantitate the risk of bleeding (see Table 27.3). Patients with a bleeding risk score ≥7 were at an increased risk of bleeding. While this scoring system has not been independently validated it provides a useful tool to weigh the risks and benefits of pharmacological DVT prophylaxis in medical patients [45].

It is important to note that LMWHs and fondaparinux are renally excreted (and therefore should not be used when GFR < 35 mL/min), will accumulate in renal failure, and that the anticoagulant activity of these drugs are not easily reversed. Prophylactic vena-caval filters have very limited utility for DVT prophylaxis; their use is associated with significant long term sequela (recurrent DVT and post-phlebitic syndrome). Vena-caval filters are frequently placed in trauma patients, because of the perceived risk of pharmacologic prophylaxis. However, a large randomized clinical trial supports the fact that pharmacologic prophylaxis is both safe and effective in these patients [34]. Furthermore, there are no randomized prospective evaluations of the use of vena-caval filters in this setting.

Despite adequate prophylaxis, DVTs develop in between 5 and 10 % of ICU patients [20, 22, 46]. Furthermore, these DVTs’ are frequently “clinically silent”. These observations have led to the idea of routinely screening ICU patients for DVT with compression ultrasound. However, an economic and decision analysis by Sud and colleagues concluded that “appropriate prophylaxis provides better value in terms of costs and health gains than routine screening for deep vein thromboses” [47]. In this study weekly screening Doppler compression ultrasound cost more than $200,000 per QALY gained.

Superficial phlebitis refers to the clinical findings of pain, tenderness, induration and/or erythema in one of the superficial veins due to inflammation, infection and/or thrombosis. The term “superficial venous thrombosis” is preferred when the presence of clot is confirmed (by ultrasonography) and the term “superficial phlebitis” in the absence of venous thrombosis. Patients with inherited thrombophilic states have an increased risk of superficial venous thrombosis [59, 60]. In addition, obesity and immobilization increase the risk of superficial venous thrombosis. The differential diagnosis includes DVT, cellulitis, lymphangitis, insect bite and erythema nodosum. Superficial and deep venous thrombosis can occur together because of direct extension of the superficial venous clot into the deep venous system; therefore, all patients with suspected superficial venous thrombosis should undergo CUS examination [61]. Superficial venous thrombosis of the great saphenous vein, particularly when the clot extends to the sapheno-femoral junction, in is associated with an increased risk of DVT [61, 62]. In a large prospective, epidemiologic study, Decousus and coauthors reported that among 844 patients with superficial venous thrombosis 25 % had concomitant DVT or PE [63]. Furthermore, among the 600 patients without DVT or PE at enrollment, 10 % developed thromboembolic complications at 3 months despite the fact that 90 % received anticoagulants.

A systemic review which included 24 studies involving almost 2,500 cases concluded that treatment with LMWH should be considered to prevent thromboembolic events and the extension and/or recurrence of superficial venous thrombosis [64]. This review suggested treatment with an “intermediate dose of LMWH for at least a month.” Furthermore, the review concluded that “the optimal dose and duration of anticoagulation requires further investigation and the role of NSAIDS alone or in combination with LMWH remains uncertain” [64, 65]. In patients with thrombosis of the great saphenous vein it may be prudent to repeat CUS before stopping treatment with LMWH.

The diagnosis of PE is one of the more challenging dilemmas in clinical medicine. Scoring systems have been used to stratify patients as having a high or low risk venous thromboembolism. For suspected pulmonary embolism, two scores are widely used: the Wells score [69] and the revised Geneva score [70]. The Wells score can be used to diagnose suspected deep vein thrombosis [71]. These scoring system are presented in Tables 27.6, 27.7, and 27.8. These rules have similar predictive accuracy [72]. While an arterial blood gas analysis, electrocardiogram (ECG) and chest radiograph (CXR) should be performed in all patients suspected of having a PE, these tests have a low specificity for the diagnosis of PE. Fibrin D-dimer is a degradation product of cross-linked fibrin, and its concentration increases in patients with acute venous thromboembolism. When assayed by a quantitative ELISA or by automated turbidimetric assays, D-dimer is highly sensitive (more than 95 %) in excluding acute DVT or PE, usually below a threshold of 500 μg/L; i.e. D-dimer has a very high negative predictive value [73]. In elderly patients (50 years and older) a threshold of age x 10 allows for a larger number of patients to be ruled out for PE with a low likelihood of false negative tests [74]. Compression ultrasonography (CUS) has largely replaced venography as the main imaging procedure to diagnose DVT. Ultrasound has a lower sensitivity and specificity for the detection of calf vein thrombosis than it does for proximal DVT.

Pulmonary angiography has traditionally been considered the gold standard with which to compare other methods, however, angiography is invasive, costly and not readily available in most hospitals. Ventilation/perfusion scanning (V/Q) currently has a limited role in the diagnosis of PE. While a completely normal V/Q scan rules out a PE and a high probability scan effectively rules in PE, most V/Q scans are of indeterminate probability and neither rule in or rule out a PE. CT pulmonary angiography (CTPA) has largely replaced V/Q lung scintigraphy as the main imaging modality in suspected PE. CTPA has the additional advantage of allowing visualization of the lung parenchyma allowing the formulation of alternative diagnoses. Single detector CTPA has a sensitivity of only about 70 %. Multidetector CTPA is more sensitive than single-detector CT angiography. In the PIOPED II study (conducted between 2001 and 2003) multidetector CTPA had a sensitivity 83 % and a specificity 96 %; this increased to 90 % and 95 % respectively with CT venography (of the lower limbs) [75]. Consequently, CUS of the lower extremities was recommended in patients with a negative CTPA to improve the diagnostic sensitivity. However, more recently the added benefit of CUS with a negative CTPA has been questioned [73]. Righini et al. compared two diagnostic strategies: clinical probability assessment and either D-dimer measurement and CTPA (DD-CT strategy) or D-dimer measurement, CUS and CTPA [76]. In the DD-US-CT strategy, D-dimer was measured only in patients with a low or intermediate clinical probability on the revised Geneva score. In these patients, pulmonary embolism was ruled out by a negative D-dimer test without further testing. When the D-dimer concentration was greater than 500 ng/mL, CUS was performed in both legs, and patients with a proximal DVT were given anticoagulant drugs without further testing. Patients without proximal DVT underwent CTPA and were treated if positive for PE. The DD-CT strategy was similar except that a CUS was not performed. The primary outcome was the 3-month thromboembolic risk in patients who were left untreated on the basis of the exclusion of PE by diagnostic strategy; the primary outcome occurred in 0.3 % in the DD-US-CT group and 0.3 % in the DD-CT group. These data suggest that CUS may not be necessary in all patients with a suspected PE and a negative CTPA. However, it would appear prudent to perform a CUS in those patients who are admitted to hospital and require further diagnostic workup. Isolated subsegmental abnormalities, which are reported in 10–20 % of CTPA’s may be due to PE causing symptoms or incidental PE that is not responsible for the patient’s symptoms or may be a false positive finding [77]. Consequently, it is uncertain if patients with these findings should be treated. Treatment is generally recommended in those patients with clinical evidence of PE (high D-dimer, etc) and a low risk of bleeding [77].