Managing pulmonary embolism

Pulmonary embolism is a great masquerader. It presents with numerous non-specific signs and symptoms that may mimic a variety of other clinical conditions. Once pulmonary embolism is suspected diagnostic and therapeutic procedures are highly dependent on the clinical presentation of the patient, the local resources, and the expertise of the doctor treating the patient. Many algorithms have been established to prevent underdiagnosing and over-diagnosing the disease as both carry a substantial risk of fatality. Unfortunately a universally accepted approach to the management of pulmonary embolism is still missing.

In 1997 the British Thoracic Society (www.britthoracic.org.uk) published a practical strategy for managing suspected pulmonary embolism, to bridge the gap between clinical research and routine management.¹ Over the past six years more evidence has been generated, and a major update now means that these recommendations can be converted into practical guidelines for daily use.² Four major issues in pulmonary embolism management deserve particular attention: d-dimer testing, computed tomography pulmonary angiography, thrombolytic treatment, and low molecular weight heparins.

The greatest challenge for the attending doctor is to guess how likely it is that a particular patient has pulmonary embolism. Defining the pretest probability sets the stage for the use of d-dimer testing: it determines which tests are ordered and how they are to be interpreted. This, in combination with a properly validated d-dimer test, helps to triage patients who do or do not need further imaging.

Plasma d-dimers have an excellent negative predictive value. Combined with low clinical pretest probability of pulmonary embolism a negative d-dimer test enables the doctor to omit further imaging safely.³ This is unfortunately not the case if the pretest probability of pulmonary embolism is high. Several assays are now available but cannot be applied interchangeably. Clinicians should be aware of the characteristic performance of the test used in their institution before integrating it into diagnostic algorithms.⁴

Combining clinical assessment, d-dimer, ultrasonography, and lung scan leads to a diagnosis in most outpatients with suspected venous thromboembolism; it is safe and may even be cost effective.⁵

In daily practice, however, the high number of non-diagnostic or false positive scan results, especially in patients with prior cardiopulmonary diseases, makes further investigations often inevitable. Spiral computed tomography has advanced in many institutions to the first line imaging technique to rule in pulmonary embolism. It also can establish probable alternative diagnoses. A negative result from spiral computed tomography rules out clinically significant pulmonary embolism with similar accuracy as normal lung scanning or a negative pulmonary angiogram.⁶ The accuracy of spiral computed tomography in detecting isolated subsegmental emboli is still under debate, and the clinical relevance of such emboli is unclear. The use of multislice scanners may improve diagnostic accuracy.⁷ The low rate of recurrence supports the accuracy of a negative result from spiral computed tomography to exclude pulmonary embolism.⁸

Intuitively it seems to make sense to apply thrombolytic treatment in patients with massive pulmonary embolism. Only one trial has, however, shown a survival benefit; this trial enrolled only eight patients with pulmonary embolism and shock. Another multicentre trial found that patients treated with heparin alone were significantly more likely to receive rescue thrombolysis than patients treated with alteplase plus heparin (24.6% v 10.2%, P=0.004). Mortality, however, was not reduced (2.2% v 3.4%, P=0.71).⁹ Since the rate of fatal haemorrhage with thrombolytic treatment is about 2.1%,¹⁰ its use in haemodynamically stable patients is still being debated.

Low molecular weight heparins have revolutionised the early treatment of haemodynamically stable patients with pulmonary embolism. Both dosing and administration are simple, and laboratory monitoring is not required except in obese patients, pregnant women, and patients with renal insufficiency. Weight adjusted treatment with low molecular weight heparins is also safe and effective in the outpatient setting, in patients with submassive pulmonary embolism.¹¹ Efficacy and safety in the initial treatment of massive pulmonary embolism has not been studied yet.

Guidelines developers must make a careful judgment about the generalisability, applicability, consistency, and clinical impact of the evidence to create a clear link between the evidence and recommendation. Most of the recommendations are grade B (high quality systematic review of observational studies directly applicable to the target population with overall consistency of results or results extrapolated from systematic reviews of randomised controlled trials). This grade will come to be regarded as the best achievable in many areas.¹² The current guidelines are very reader and user friendly. Concise summaries and distinct diagnostic algorithms can be completed within a few hours. These give clear advice for more junior doctors and are applicable to all healthcare systems in the developed world. Unfortunately we know that there are many guidelines out there, and more often than not they are being ignored. Let us hope that these guidelines make the difference.