Management and Prevention of Venous Thromboembolism Including Surgery and the Pregnant State

Surgical Patients

In general, screening for VTE with duplex ultrasound or other modalities is not advocated for surgical populations with the exception of the high-risk trauma patient who has received suboptimal prophylaxis (6). The incidence of both DVT and PE is sharply reduced when anticoagulant prophylaxis is used regardless of the risk group (Table 1). Several concerns have slowed the wider application of appropriate prophylaxis including bleeding risk and cost. However, numerous reports have demonstrated cost-effectiveness (7, 8) and small or no increases in major bleeding (9, 10) with pharmacological prophylaxis (7, 8). Prevention is obviously more effective than screening, clinical diagnosis, and treatment.

The type and intensity of prophylaxis is administered according to risk level from low to highest (Table 1). Accordingly, the intensity of the preventive measures escalates as risk factors increase. Patients at a low risk level are not given any specific measures, whereas patients at the highest risk level are given extensive preventive measures. Several antithrombotic regimens are summarized in Table 2, which includes worldwide experience and regimens. In the United States, only enoxaparin (Lovenox; Aventis; Strasbourg, France) and dalteparin (Fragmin; Pharmacia & Upjohn; Peapack, NJ) have been approved by the FDA for prophylaxis. In the general surgery population, the meta-analysis by Claggett and Reisch of 49 pooled studies revealed that unfracationated heparin q8h is superior to q12h of dosing (9).

In the gynecologic, urologic, orthopedic, and general surgery population, low-dose unfractionated heparin (LDUH) and LMWH have been extensively studied. LMWH has the advantage of being administered once daily and is less likely to cause heparin-induced thrombocytopenia and thrombosis than standard heparin (11). Antithrombotic therapy or prophylaxis can also be used (with caution) in patients having spinal puncture or epidural catheters placed for regional anesthesia or analgesia (12).

For all patient groups, aspirin was not recommended because of more effective alternative agents. Aspirin has the appeal of being inexpensive and easy to administer with few side effects but has generally been found to be ineffective in preventing VTE, especially in general surgery and orthopedic patients. Pooling more than 30 antiplatelet trials of various scientific designs flawed the Antiplatelet Trialists' Collaboration meta-analysis, which showed a significant reduction of DVT and PE (by 37% and 71%, respectively); e.g., none of the studies utilized contrast venography for outcomes. The Pulmonary Embolism Prevention (PEP) trial assessed 4088 hip and knee arthroplasty patients and found no benefit with aspirin use for either arterial or venous events (13, 14). The 6th American College of Chest Physicians (ACCP) consensus also reviewed the PEP collaborative group's findings as they pertained to hip fracture patients and did not recommend the routine use of aspirin as thromboprophylaxis in this population (12). So clearly, the use of aspirin plays no role in DVT prevention.

In two large trials of 4483 patients evaluating LMWH versus adjusted-dose warfarin in patients undergoing hip arthroplasty, the incidence of DVT was statistically significantly reduced in the LMWH group. There was an increase in excessive minor bleeding with the LMWH dalteparin (Fragmin®; Pharmacia & Upjohn, Bridgewater, NJ) administered preoperatively. This adverse event did not occur with enoxaparin (15, 16).

Both LMWH and warfarin sodium carry American College of Chest Physicians (ACCP) grade 1A recommendations (the strongest possible) for elective hip and knee replacements. LMWH can be initiated 12 hours before surgery, 12–24 hours after surgery or 4–6 hours after surgery at half the usual high-risk dose and then continued with the usual high-risk dose the following day. Warfarin sodium is an effective and safe oral anticoagulant that requires an international normalized ratio (INR) of 2.0–3.0 for both treatment and prophylaxis and so should be initiated preoperatively in most instances (12).

The optimal duration of anticoagulant prophylaxis remains unknown. In the orthopedic trials, 7–14 days was the usual thromboprophylactic period, with subsequent trials demonstrating a 50% reduction of total and proximal DVT with extended prophylaxis (17). With the length of stays often less than 5 days, out-of-hospital prophylaxis for both clinical benefit and cost effectiveness needs to be assessed. The ACCP recommends at least 7–10 days of prophylaxis.

Medical Patients

Similar to major surgery, acute hospitalization for a medical indication poses a substantial risk of thromboembolic complications. Prophylaxis is felt to be underutilized in this population due to the lack of clear evidence gained from studies using experimental and flawed methodologies including the recruitment small numbers of nonrandomized, heterogeneous patients.

One of the primary reasons hospitalized patients are felt to be vulnerable to thromboembolic complications is the restricted mobility experienced during acute illness, as well as additional risk factors that may be accentuated by the diseases themselves. The common acute illnesses, which significantly increase the risk of DVT and PE, include congestive heart failure of New York Heart Association (NYHA) Functional Class III or IV, acute respiratory or complicated chronic respiratory insufficiency, acute infectious diseases excluding septic shock, acute arthritic episodes of the lower extremities, or spinal cord injury (18, 19).

Currently, enoxaparin is the only agent approved by the FDA for the prophylaxis of VTE in acutely ill medical patients. For patients with ischemic stroke and impaired mobility, the routine use of unfractionated heparin (UFH), LMWH, or the heparinoid danaparoid carries ACCP grade 1A evidence (12,20–22). If anticoagulant prophylaxis is contraindicated, consider mechanical measures with external stocking or intermittent pneumatic compression.

Samama et al demonstrated that the risk of venous thromboembolism in hospitalized patients, including those with cancer, could be reduced by nearly one third with a once-daily regimen of a LMWH (23). The Phase III multicenter study enrolled more than 1100 patients and supported the threshold concept for anticoagulation. Thromboemboli were reduced by approximately 60% in the 40 mg once-daily enoxaparin group (23). Contrary to the surgical population at moderate risk, for whom 20 mg of enoxaparin is effective, there seems to be a threshold in acutely ill medical patients.

For pharmacological prophylaxis, both UFH and LMWH are available. The standard of care in the United Stated is the administration of 5000 U of UFH twice daily. Although low-dose UFH is used as prophylaxis against thrombosis, it cannot be considered a validated control treatment for medical patients. The few studies supporting its use include small numbers of patients (24–27), and the results of two studies that have evaluated mortality among medical patients given 5000 U of UFH subcutaneously twice daily are conflicting (20, 27). In addition, the recommendations of consensus conferences are not definitive (20). In sharp contrast, the results for once daily LMWH are clearly and convincingly proven. LMWH has significantly reduced the incidence of DVT in hospitalized medical patients.

Pregnancy

Pregnancy is in itself a hypercoagulable condition and venous thromboembolism remains an important cause of maternal mortality. Various risk factors and physiological changes favor the formation of venous thrombosis (Table 3). Inherited and acquired thrombophilia are also associated with recurrent pregnancy loss (28). Certainly, an anticoagulant may be required during pregnancy either for the prevention of thromboembolic disease in patients already on long-term antithrombotic treatment (i.e., valvular prostheses) or for the prevention of complications of risk factors such as hereditary or acquired thrombophilia.

Most pulmonary embolisms occur postpartum, most frequently in association with Cesarean section procedures. Given the limited data in the literature, no strong evidence exists about individual risk, selection of agents and dosing regimens, or how long thromoprophylaxis should continue. All guidelines are empiric grade C recommendations (Table 4).

Because warfarin crosses the placenta and can cause embryopathy in any trimester, this agent should be avoided during pregnancy. The LMWHs appear to be at least as effective as UFH for thromboprophylaxis and have a lower risk of bleeding, heparin-induced thrombopenia, and possibly osteoporosis (29).

Like UFH, LMWH is cleared through the kidney and could be subject to changes in pharmacokinetics in pregnancy. However, several studies have shown that LMWH does not cross the placenta in any trimester (30–32), and, despite the absence of licensing, LMWH is widely used during pregnancy. In most cases, a monitoring of anti-Xa activity is unnecessary.

Summary

Appropriate prophylaxis utilizing both pharmacological and nonpharmacological modalities with an evidence or knowledgebased approach should be applied to many hospitalized patient groups. A risk stratification scheme for the likelihood of VTE development will identify those patients who will benefit from prophylaxis in this setting and should also protect provider and hospitals from legal liability once implemented throughout the institution. Guidelines to prevent VTE have been widely distributed and generally have been assumed to be effective.