Article Text

Download PDFPDF

Pulmonary embolism guidelines: will they work?
  1. A C Miller1,
  2. D A R Boldy2
  1. 1Chairman, BTS Pulmonary Embolism Working Party
  2. 2Chairman, BTS Standards of Care Committee
  1. Correspondence to:
    Dr A C Miller, Mayday Hospital, Croydon CR7 7YE, UK;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Updated guidelines on the management of suspected acute pulmonary embolism are published in this issue of Thorax. CT pulmonary angiography has become the central imaging investigation in the diagnosis of pulmonary embolism, but it is important that its use is carefully controlled to avoid unnecessary investigations in overworked radiology departments.

Two years ago several members of the British Thoracic Society independently suggested that, in the light of recent publications, the Standards of Care Committee should update their 1997 advice on suspected acute pulmonary embolism (PE). This issue of Thorax contains the results of this endeavour, with evidence sufficiently robust now to allow these to be called guidelines (see pp 470–83). Those familiar with the previous publication, to which the new document frequently refers, will recognise that the same basic structure has been used.

Previously, the main recommended option in the face of a non-diagnostic ventilation-perfusion (V/Q) scan was pulmonary angiography, an invasive test that was little used before and probably little more after. This advice has been made redundant by developments in CT pulmonary angiography (CTPA), with strong evidence that, even though a negative result may not entirely exclude PE, it does make anticoagulation unnecessary. There is no doubt that CTPA should now be considered the central imaging investigation in suspected PE, and many acute hospitals are developing experience of the techniques.

However, this adds a considerable workload to radiology departments already struggling to cope with the increased imaging requirements for cancer staging, not helped by the fact that PE is confirmed in only 20–35% of those where it has been suspected. This has led to different strategies for reducing the number of unnecessary CTPAs, but these too have their problems.

Some hospitals continue to use V/Q scanning as a way of obtaining a definite answer, but 30–50% of such patients will require CTPA anyway, which delays definitive investigations and lengthens hospital stay, and others do not have nuclear medicine on site. Indeed, were CTPA easily and rapidly available, V/Q scanning would become largely obsolete.

“There is no doubt that CTPA should now be considered the central imaging investigation in suspected PE”

Another approach is to restrict CTPA to patients who do not have the combination of low clinical probability and a negative d-dimer test. Since this halves the number requiring lung imaging it is an attractive option, but practical experience shows that there are pitfalls. Firstly, although the 1997 clinical probability assessment system has been retained (now on good evidence), the fact that it is very simple does not mean that junior doctors use it properly, as many senior clinicians and radiologists have discovered. Secondly, the seductive simplicity and low cost of d-dimer assays mean that, in many emergency departments, the test is done on a high proportion of patients who arrive with chest pain and/or breathlessness, even when another diagnosis is obvious, similar to the way that cardiac enzymes are often misused. The end result is that, where the d-dimer test happens to be positive, it is assumed that this makes PE more likely after all.

For these guidelines to work in practice without widespread abuse of the recommendations, the following policies should be introduced:

  1. Clinical probability must be recorded in the admission notes and be included on every request form for CTPA, a policy that would be fruitful to audit regularly.

  2. d-dimer tests may only be requested by junior doctors with at least 6 months post-registration experience of acute medicine. They should not be performed when CTPA is required anyway.

  3. Request forms for CTPA which do not state clinical probability and (where appropriate) d-dimer results are automatically rejected by the radiology department, accompanied by a strong telephone message.

  4. Where V/Q scanning remains part of the hospital’s management algorithm, there should be a system in place whereby an inconclusive result leads to an automatic CTPA on the same day, without a further request having to be generated by the clinician.

  5. There should be at least one interested physician and radiologist who together review and refine both the hospital’s policy and its application in practice.

These principles might appear draconian and unenforceable, yet several hospitals have been adopting them successfully for some time, finding that they stimulate good clinical thinking, reduce the number of unnecessary imaging tests, speed up the process of reaching a correct diagnosis, and are educationally valuable. Indeed, with the widespread changeover from unfractionated to low molecular weight heparin, an increasing number of patients are being quickly investigated and, where necessary, managed without being hospitalised.

Although the document is aimed primarily at junior physicians in UK Accident & Emergency departments, almost all the underpinning evidence comes from continental Europe and North America. Allowing for variations in resources, organisation and clinical practice in different countries and debates around interpretation of current data, it broadly reflects current international thinking and should prove useful elsewhere.

View Abstract

Linked Articles