Intended for healthcare professionals


Thromboprophylaxis for adults in hospital

BMJ 2007; 334 doi: (Published 17 May 2007) Cite this as: BMJ 2007;334:1017
  1. David A Fitzmaurice, professor of primary care,
  2. Ellen Murray, research fellow
  1. Department of Primary Care and General Practice, University of Birmingham, Birmingham B15 2TT
  1. d.a.fitzmaurice{at}

    An intervention that would save many lives is still not being implemented

    The evidence that pharmacological thromboprophylaxis can reduce the rate of venous thromboembolism by 60-65% is compelling.1 2 3 Last month the United Kingdom's National Institute for Health and Clinical Excellence (NICE) published guidelines on venous thromboembolism in patients having surgical procedures,4 which are summarised in this week's BMJ.5 The risks to surgical patients, particularly those undergoing orthopaedic procedures, are well known, but most people who develop venous thromboembolism in hospital are medical patients.

    The prevention of venous thromboembolism in adult patients in hospital was the main challenge to patient safety in 2001, according to a technical assessment by the Agency for Healthcare Research and Quality in the United States.6 In 2005, the UK government's Health Select Committee reported that venous thromboembolism caused more than 25 000 potentially preventable deaths a year, and probably half of these deaths resulted from admission to hospital.7

    Despite all this evidence, mortality due to venous thromboembolism after hospital admission is still at least 10 times greater than the more widely publicised mortality due to methicillin resistant Staphylococcus aureus (MRSA). Overall, the number of deaths from venous thromboembolism in the UK each year is five times greater than the combined total number of deaths from breast cancer, AIDS, and road traffic incidents. Indeed a revised estimate, based on an epidemiological model using extrapolation from European data, suggests that about 60 000 deaths from venous thromboembolism occur annually in the UK.8 Autopsy data indicate that about 10% of deaths in hospital are due to pulmonary embolism.9

    Despite the considerable evidence base for thromboprophylaxis, it is poorly implemented in the UK. A combination of factors may be responsible—as a result of poor education, health professionals' lack awareness of this condition; venous thromboembolism is often a silent disease (80% of deep vein thromboses are subclinical); and venous thromboembolism often occurs after discharge from hospital. Prescribing costs may also be a barrier to the use of thromboprophylactic drugs, but this is not clear.

    The Health Select Committee reported two years ago that thromboprophylaxis was not effectively implemented in the UK—as few as 20% of eligible patients were receiving appropriate prevention. The committee recommended that NICE should produce its planned guidelines on venous thromboembolism for surgical procedures more quickly. It also recommended that an independent expert working group be set up to investigate how current best practice and guidance on venous thromboembolism could be promoted and implemented and what resources might be needed to support delivery of any strategy through existing structures. This committee was to report to the chief medical officer in July 2006.

    The expert working group's report and the chief medical officer's response were published last month.10 11 The expert group recommended that, on admission to hospital, all adults should have a risk assessment for venous thromboembolism that is formally documented and incorporated into the hospital's system for the Clinical Negligence Scheme for Trusts.12 The group also recommended that the Department of Health should set core standards aimed at ensuring 100% compliance with risk assessment for thromboprophylaxis. Moving on to prevention, the report stated that aspirin should not be used for thromboprophylaxis as it is less effective than other agents, such as low molecular weight heparin. The chief medical officer has brought the report to UK doctors' attention and has set up another committee to implement the recommendations of the report.

    The consultation phase for the NICE guidelines was highly contentious because the draft guidelines emphasised mechanical prophylaxis—using compression stockings and, during surgery, inflatable boots—rather than drugs. Indeed, concerns about the way NICE reached its recommendations partly led to the Health Select Committee's decision some months ago to review NICE.13 The published NICE guidelines review the same evidence as that in the expert working group's report and, while both agree that aspirin should not be used, NICE has retained the emphasis on mechanical rather than chemical means of thromboprophylaxis. Furthermore, NICE classes patients aged over 60 as being at high risk rather than those aged over 40.

    The Health Select Committee's report two years ago provided an opportunity to change practice. Meanwhile, more than 25 000 people may have died needlessly each year because of the failure to implement simple thromboprophylaxis in UK hospitals.

    Summary of expert working group's recommendations on thromboprophylaxis for adults in hospital

    Medical patients
    • Particularly those admitted for longer than four days, who have reduced mobility with either severe heart failure, respiratory failure, inflammatory illness, or cancer: heparin, preferably low molecular weight heparin

    High risk surgical or orthopaedic patients
    • Mechanical prophylaxis and low molecular weight heparin or fondaparinux

    Intermediate risk surgical patients
    • Mechanical prophylaxis and low molecular weight heparin or fondaparinux

    Low risk surgical patients
    • Mechanical prophylaxis and early mobilisation


    • Competing interests: DF was a member of the expert working group on the prevention of venous thrombosis in hospitalised patients. DF and EM have received research income from Leo Laboratories.

    • Peer review and provenance: non-commissioned; externally peer reviewed.


    View Abstract