1.1 Epidemiology of venous thromboembolism

Deep vein thrombosis (DVT) of the lower limbs is a common disease, often asymptomatic, but presenting with clinical symptoms (leg pain or swelling) in about 1 per 1,000 people per year in the general population. Complications include pulmonary thromboembolism (PE) and post-thrombotic leg syndrome (PLS). DVT has multiple causes (see Table 1).

Asymptomatic DVT is defined as DVT detected by screening with ¹²⁵I fibrinogen scanning, ultrasound, or ascending venography.^1,2,3,4,5

Symptomatic DVT (leg pain or swelling) results from occlusion of a major leg vein. It requires specific investigation and treatment⁶ which in hospitalised patients may delay discharge, or require readmission to hospital.

Pulmonary embolism, which in 90% of cases results from an asymptomatic DVT,^6,7 may present as sudden death, breathlessness, faintness, collapse, or chest pain.⁶ Nonfatal PE in hospitalised patients may delay discharge, or require readmission to hospital. Fatal PE is under-diagnosed, because of the non-specificity of symptoms and signs prior to death, which may be attributed to myocardial infarction, pneumonia, or other pathology. About 10% of hospital deaths (1% of all admissions) were attributable to PE in the UK in one study from the 1980s.⁷ More recent studies have continued to highlight the significant contribution of PE to in-hospital mortality,^{8,9,10,11,12,13} especially after emergency surgery when prophylaxis is often omitted.⁹

Post-thrombotic leg syndrome (chronic leg pain, swelling, dermatitis, ulcers) is a consequence of destruction of leg vein valves by DVT. Leg ulcers are observed in 2-10% of patients approximately 10 years after their first symptomatic DVT.^{14,15,16,17,18} About 0.2% of the general population have venous leg ulcers.

Venous thromboembolism (VTE) is defined as DVT±PE.

1.2 Rationale for prophylaxis

The risk of VTE is increased tenfold in patients who are hospitalised after trauma, surgery or immobilising medical illness, and also in pregnant and puerperal women. Screening studies for asymptomatic DVT, using contrast venography or radiolabelled fibrinogen scanning, as well as autopsy studies have shown that DVT is common in such patients. In many of these patients, DVT remains asymptomatic but in others it can cause morbidity and mortality.^1,7,19,20 The rationale for prophylaxis is based on its efficacy, the clinically silent nature of VTE, its high prevalence in hospitalised, pregnant or puerperal patients, and its potentially disabling or fatal consequences.^{18,19,20,21,22,23,24} Clinical diagnosis is difficult,⁶ and treatment with full-dose anticoagulant therapy has risks, especially of bleeding.⁶

There is evidence that routine prophylaxis reduces morbidity, mortality and costs in hospitalised patients at risk of DVT and PE, as highlighted in national and international guidelines.^19,20,21,22 In contrast, screening for asymptomatic DVT, and its treatment, are expensive, insensitive and not cost-effective compared to routine prophylaxis in at-risk patients.^19,20,21,22

1.3 Development and review of the SIGN guideline

The need for a national guideline on prophylaxis was highlighted by a study of fatal PE in surgical patients in Scotland up to 1995, which showed that 56% of patients who died of PE did not receive antithrombotic prophylaxis, despite having major risk factors and no contraindications to standard antithrombotic regimens.⁸ VTE is probably an escalating public health problem, due to the increasing age of the population.^19,25

Following publication of the SIGN guideline in 1995,²² studies have reported high interest from hospital trusts in implementation of the guideline²⁶ and high compliance with the trust local protocol/guideline.^27,28,29

The original SIGN guideline²² recommended prophylaxis primarily according to efficacy in reducing risk of asymptomatic DVT (detected by routine screening), a widely accepted "surrogate" endpoint frequently employed in randomised clinical trials (RCTs). Further reasons to consider this endpoint in developing the guideline were that asymptomatic DVT is the usual precursor of both fatal PE^7,12 and PLS;¹⁵ and that prophylactic methods which reduce the incidence of asymptomatic DVT also reduce the incidence of symptomatic DVT^30,31 and of symptomatic PE, including fatal PE, to a similar extent.^{2,3,4,5,30,31}

Since the publication of this guideline,²² concern has been raised, especially by UK orthopaedic surgeons, as to the appropriateness of routine pharmacological antithrombotic prophylaxis. The validity of asymptomatic DVT as a surrogate endpoint in RCTs has been questioned because of the discrepancy between high incidence of venographic DVT and low risk of clinical outcomes such as death, e.g. in elective hip and knee replacement surgery.^{31,32,33,34,35,36,37,38,39} The risks and benefits of anticoagulant prophylaxis in orthopaedic patients need to be balanced. For instance, low molecular weight heparins (LMWHs) have been associated with an increased risk of vertebral canal haematoma after spinal and epidural block in recent reports from North America.^40,41,42 Recent analyses of the efficacy of heparin prophylaxis in medical patients with acute myocardial infarction⁴³ or acute ischaemic stroke,⁴⁴ have also raised questions as to the role of routine heparin in addition to routine aspirin therapy in such patients.⁶

1.4 Major changes to the guideline

The revised guideline includes new evidence, systematic reviews, and consensus statements published from 1994-2001. In addition:

Patient groups are no longer categorised as high, moderate and low risk.²² Instead:

• where possible the incidences of asymptomatic and symptomatic DVT and PE and mortality in patients given no antithrombotic prophylaxis are defined. These baseline incidences of thromboembolic risk are established from observational studies, as well as control groups in randomised trials (which constitute a selected population)⁴⁵
• The efficacy of prophylaxis on a range of outcome measures is considered, including
  1. 1. asymptomatic DVT
  2. 2. symptomatic VTE
  3. 3. total mortality (multifactorial and hence insensitive to change in only one component)
• The balance of average bleeding risk versus average reduction in thromboembolic morbidity and mortality is considered.

1.5 How to use the guideline

The guideline begins by identifying patient groups at risk of VTE (section 2) and describing the available methods of prophylaxis (section 3), with general recommendations about efficacy, safety, and how they should be used. Appropriate methods of prophylaxis for specific patient groups are considered in subsequent sections.

This guideline is not intended to be construed or to serve as a standard of medical care. Standards of care are determined on the basis of all clinical data available for an individual case and are subject to change as scientific knowledge and technology advance and patterns of care evolve. The ultimate judgement regarding a particular clinical procedure or treatment plan must be made in light of the clinical data presented by the patient and the diagnostic and treatment options available. However, it is advised that significant departures from the national guideline or any local guidelines derived from it should be fully documented in the patient's case notes at the time the relevant decision is taken.

1.6 Differences from other national guidelines

This guideline differs significantly from the most recent North American guideline¹⁹ in recommending aspirin as an effective prophylaxis in surgical patients (section 3.3). The North American guideline¹⁹ does not recommend aspirin because other prophylactic methods (e.g. heparins) are more effective in reducing asymptomatic DVT. The SIGN guideline development group notes that aspirin is as effective as heparins in reducing the risk of fatal pulmonary embolism, which is the most clinically relevant endpoint in prophylaxis of VTE. Both aspirin (Table 3) and heparins (Table 5) reduce the overall absolute risk of fatal PE in surgical patients from 0.6% to 0.2%. Furthermore, aspirin carries a lower relative risk of major bleeding (1.24, 95% CI 1.12 - 1.37) compared to heparin (1.75, 95% CI 1.43-2.01: Tables 3 and 5).