The pilot SIGN guideline,²² considering published evidence up to early 1994, identified major elective (e.g. total hip or knee replacement) or traumatic (e.g. hip fracture) orthopaedic surgery as "high risk", for both asymptomatic DVT (incidence 40-80%) and fatal PE (incidence 1-10%) in the absence of prophylaxis. As discussed in section 1.3, the recent literature has established that fatal PE is uncommon after total hip or knee replacement in the absence of routine anticoagulant prophylaxis.¹¹³

5.1 Total hip or knee replacement

Table 6 combines data from control groups (placebo injections or no specific prophylaxis) in meta-analyses of randomised trials of antiplatelet drugs (Table 3) or low-dose subcutaneous heparins (Table 5) with recent data from UK cohort studies in major elective orthopaedic surgery (total hip or knee replacement). Without specific anticoagulant prophylaxis, there is a high incidence of asymptomatic DVT (50% detected by routine venography 7-14 days after surgery) at the site of surgery (femoral vein at hip replacement, popliteal vein at knee replacement) and a significant incidence of nonfatal, clinical DVT and PE (up to 5% in hip replacement; up to 14% in knee replacement). The overall risk of fatal pulmonary embolism is about 0.4%(95% CI 0.1-1.0%).¹¹³ These risks of clinical and fatal VTE are similar to those in major general or gynaecological surgery (Tables 3,5 and 6), despite the higher incidence of asymptomatic DVT in major orthopaedic surgery (total DVT 50%; proximal DVT 16%).¹¹⁴

Recent data suggests that total mortality after total hip or knee replacement in the absence of pharmacological prophylaxis may be lower than the estimate from the control groups in antiplatelet or heparin trials (1% in Table 6), at 0.3%.^113,114 Obviously the incidence of fatal PE must be lower than the incidence of total mortality.¹¹⁴

More randomised trials have been performed in total hip replacement compared to total knee replacement. The effects of prophylaxis appear similar in both groups of patients and have been combined as "major elective orthopaedic surgery" in meta-analyses (Tables 4,5 and 6), and in this guideline. Detailed references are given in recent reviews.^19,20,113 The Scottish national arthroplasty register may provide additional information on VTE risk and prophylaxis in the future.

All patients undergoing hip or knee arthroplasty should be enrolled in the Scottish national arthroplasty register.

Table 6: Summary of incidence of VTE events, mortality and major bleeding in control groups in meta-analyses of antiplatelet drugs (Table 3) or low-dose subcutaneous heparins (Table 5); and in recent UK cohort studies without routine heparin prophylaxis; in major elective orthopaedic surgery (total hip or knee replacement)

5.1.1 MECHANICAL PROPHYLAXIS

As noted in section 3.2, there is evidence from pooled analyses of small randomised trials that GECS, IPC or foot pumps reduce the risk of asymptomatic DVT in elective orthopaedic surgery, including both hip and knee replacement. IPC may be less effective in reducing the risk of proximal DVT,¹¹³ while foot pumps may be more effective in reducing the risk of proximal DVT. ^{86,87,88,89,90,91,92,93} Evidence level 1++,2++

Patients undergoing total hip or knee replacement (or other elective major orthopaedic surgery) can be considered for mechanical prophylaxis (GECS±IPC or foot pumps).

5.1.2 ANTIPLATELET DRUGS (ASPIRIN)

In a meta-analysis of randomised trials,³¹ antiplatelet drugs (usually aspirin) reduced the risk of asymptomatic DVT and of PE by about one-third in patients undergoing elective orthopaedic surgery, as in all surgical trials (see Table 3). While not statistically significant, there was a similar reduction in total mortality. Aspirin may be more effective in reduction of proximal DVT than warfarin.¹¹⁷ The excess risk of bleeding after total hip replacement appears low and there appears to be no significant increased risk of wound haematoma or infection.³¹ Evidence level 1++

A recent cohort study observed low risks of fatal PE (0.13%), nonfatal PE (0.94%) and symptomatic DVT (1%) in patients with THR receiving aspirin and mechanical prophylaxis (GECS or IPC).¹¹⁸ Evidence level 2

Patients undergoing total hip or knee replacement (or other elective major orthopaedic surgery) can be considered for aspirin (150 mg orally, started before surgery and continued for 35 days).

5.1.3 HEPARINS

In a meta-analysis of randomised trials, both UFH and LMWH reduced the risk of asymptomatic DVT by about 50% in patients undergoing elective orthopaedic surgery, as in all surgical trials. They had no effect on symptomatic PE or mortality and there was a non-significant trend to increased risk of major bleeding (see Table 5). There was a statistically significant reduction in fatal PE (as in all surgical trials); however the incidence of fatal PE in the control groups of these studies (1.5%) was appreciably higher than that in recent studies without routine pharmacological prophylaxis (0.1-1.0%).¹¹⁴ Evidence level 1++

In recent UK cohort studies of total hip replacement and total knee replacement in which there was a high incidence of spinal and epidural anaesthesia, and routine mechanical prophylaxis, use of heparin prophylaxis was not associated with significant reductions in clinical VTE or mortality.^113,114 The additional benefit of UFH or LMWH compared to routine early mobilisation, mechanical prophylaxis and aspirin is therefore unclear.¹¹⁴ Evidence level 2+

Patients undergoing total hip or knee replacement (or other elective major orthopaedic surgery) can be considered for UFH or LMWH.

See section 7 for timing of initial dose for patients undergoing spinal or epidural blocks.

The routine duration of UFH or LMWH prophylaxis is until discharge from hospital (usually 7-15 days). However, in contrast to non-orthopaedic surgery, there is a high risk of recurrent asymptomatic DVT when venography is repeated at 4-5 weeks after surgery.¹¹⁹ This may explain why heparin prophylaxis is no more effective than mechanical prophylaxis or aspirin, which can be readily continued for 35 days after surgery.³¹

LMWH prophylaxis can also be continued for 4-5 weeks after surgery, and was more effective than conventional LMWH (or warfarin) prophylaxis for 7-15 days in reducing risks of asymptomatic DVT and symptomatic VTE in a recent meta-analysis.¹¹⁹ However, the risks of symptomatic VTE (1.5% with 4-5 weeks LMWH and 3.3% with 7-15 days LMWH) in this meta-analysis were higher than those observed in cohort studies using routine mechanical prophylaxis ± aspirin.^114,118 The role of extended LMWH prophylaxis is therefore unclear. Because of its logistic problems and costs, it should be reserved for very high-risk patients (e.g. previous VTE and/or multiple risk factors).¹⁹ Evidence level 2+

The duration of UFH or LMWH prophylaxis should be 7-15 days after lower limb arthroplasty, extended to 4-5 weeks in very high-risk patients.

5.1.4 ORAL ANTICOAGULANTS

Warfarin (commenced before or immediately after surgery and adjusted to target INR 2.5 [range 2.0-3.0]) is similarly effective to UFH or LMWH prophylaxis in prevention of asymptomatic DVT and symptomatic VTE after elective total hip or knee replacement, and is widely used in North America.^19,113 Problems include the need for regular monitoring of the INR, and increased risk of postoperative bleeding. It is not widely used in the UK,⁸⁰ but could be considered in high-risk patients as an alternative to UFH or LMWH (e.g. in patients on long term warfarin: see section 3.8). Evidence level 1+

Patients undergoing major orthopaedic surgery (e.g. total hip or knee replacement) can be considered for warfarin (target INR 2.0-3.0), e.g. those already receiving warfarin.

5.1.5 SUMMARY

Patients undergoing total knee or hip replacement (or other elective major orthopaedic surgery) should receive thromboprophylaxis: mechanical (GECS±IPC, foot pumps), pharmacological (aspirin or heparin or warfarin), or both.

5.2 Hip fracture surgery

Hip fracture surgery (see SIGN guideline 56: Prevention and management of hip fracture in older people)¹²⁰ carries high risks of asymptomatic DVT (45%), symptomatic DVT (1-11%), symptomatic PE (3-13%) and fatal PE (1-7%) in the absence of VTE prophylaxis (see Tables 3 and 5).¹¹⁴

The risks of fatal PE and total mortality (7%)³¹ are higher than in elective arthroplasty, probably due to the higher mean age of hip fracture patients. Early surgery (within 24 hours) reduces the risk of DVT¹²¹ and of fatal PE.²⁵ Evidence level 2+

Early surgery (within 24 hours) is recommended where possible to reduce the risk of DVT and fatal PE after hip fracture.

5.2.1 MECHANICAL PROPHYLAXIS

A meta-analysis of four randomised controlled trials of mechanical methods (two trials of IPC and two of foot pumps; no trials of GECS were identified) involving 422 patients observed that the incidence of asymptomatic DVT was reduced from 19% to 6%.⁸⁴ There was insufficient data to establish the effects of these devices on symptomatic VTE or mortality. In the PEP trial,³¹ use of GECS (by 30% of patients, non-randomised) was not associated with reduction in symptomatic VTE. Evidence level 1+

Mechanical prophylaxis (IPC or foot pumps) should be considered to reduce the risk of asymptomatic DVT after hip fracture. There is no evidence for the efficacy of GECS in hip fracture patients.

5.2.2 ANTIPLATELET DRUGS (ASPIRIN)

A meta-analysis of randomised controlled trials (mainly the PEP study of aspirin³¹) in patients undergoing surgery for hip fracture observed that aspirin reduced the risk of asymptomatic DVT (42% to 36%), symptomatic DVT (1.5% to 1.0%), all PE (1.6% to 0.8%), and fatal PE (0.8% to 0.4%), with no effect on total mortality (see Table 3). Evidence level 1++

The excess risk of bleeding was small (one additional transfused bleed per 1,000 patients who were not receiving concomitant heparin prophylaxis).³¹ Evidence level 1++

All patients with hip fracture should receive aspirin (150mg orally, started on admission and continued for 35 days) unless contraindicated.

5.2.3 HEPARINS

A meta-analysis of UFH and LMWH in hip fracture surgery showed that heparins reduced the risk of asymptomatic DVT from 39% to 24% (NNT=6.5). Unlike elective arthroplasty (see section 5.1), no studies of recurrent asymptomatic DVT (venography at 4-5 weeks) or prolonged prophylaxis were identified. There was insufficient data to establish the effects of heparins on symptomatic VTE, mortality, or bleeding⁸⁴ (see Table 5). Evidence level 1++

In a multivariate analysis of predictors of death in a multicentre regional audit, mortality was lower among patients receiving pharmacological prophylaxis for VTE.¹²² However, use of heparin prophylaxis (18% UFH, 26% LMWH, non-randomised) was not associated with reduction in symptomatic VTE in the PEP trial.³¹ As with elective arthroplasty, the additional benefit of UFH or LMWH compared to routine early mobilisation, mechanical prophylaxis and aspirin is therefore unclear.¹¹⁴ Evidence level 2+

Heparin should be reserved for selected patients at high risk of VTE after hip fracture due to: multiple risk factors (see section 2) contraindications to routine mechanical prophylaxis and/or aspirin.

5.2.4 ORAL ANTICOAGULANTS AND DEXTRANS

These methods also reduce the risk of VTE after hip fracture surgery,¹⁹ but are not widely used in the UK due to logistic problems, and to risks of bleeding (oral anticoagulants) and anaphylaxis (dextrans).

5.3 Knee arthroscopy

The incidence of asymptomatic DVT after knee arthroscopy is not well defined.^{123,124,125,126} At present there is insufficient evidence to justify routine prophylaxis in patients undergoing knee arthroscopy.

5.4 Trauma

The risk of DVT and fatal PE varies greatly among trauma patients; those with spinal cord injury, multiple trauma or lower limb fractures have high risks of asymptomatic DVT on screening (about 50%) and fatal PE (0.5-2.0%).¹⁹

A recent meta-analysis¹²⁷ observed no significant efficacy for UFH in prophylaxis of VTE in trauma patients. There are limited randomised trials of LMW heparins in trauma patients.¹⁹ LMWH (enoxaparin) reduced the risk of asymptomatic DVT compared to control or UFH^128,129 in patients with major trauma, without increased overt intracranial bleeding. Mechanical prophylaxis (IPC or foot pump) may be almost as effective in prophylaxis of asymptomatic DVT.^130,131 Evidence level 1+

Two studies have shown that outpatient LMWH reduced the incidence of asymptomatic DVT in patients with plaster cast immobilisation.^132,133 Evidence level 1+

At present, there is no evidence that any prophylactic method reduces the risk of clinical VTE or mortality in trauma patients. Following the PEP trial,³¹ future trials of aspirin in trauma patients should also be considered. Aspirin should be considered in patients in whom LMWH and mechanical prophylaxis are not possible.

In patients with spinal cord injury, major lower limb fractures or multiple trauma, LMWH prophylaxis can be considered, unless contraindicated (e.g. by risk of intracranial bleeding).

In patients with contraindications to LMWHs, mechanical prophylaxis can be considered (e.g. IPC or foot pump).

In patients in whom LMWH is contraindicated and mechanical prophylaxis is not feasible (e.g. patients in plaster casts) , aspirin (150mg/day), started on admission and continued for 35 days can be considered.