• Marc Blondon, Marc Righini, Drahomir Aujesky, Grégoire Le Gal, and Arnaud Perrier.
• Division of General Internal Medicine, Department of Internal Medicine, University Hospitals of Geneva, Geneva, Switzerland. marc.blondon@hcuge.ch
• Chest. 2012 Sep 1;142(3):697-703.

BackgroundThe diagnostic workup of pulmonary embolism (PE) may take several hours. The usefulness of anticoagulant treatment while awaiting the results of diagnostic tests has not been assessed. The objective of this study was to compare the risks and benefits of bid low-molecular-weight heparin vs no treatment in patients with suspected PE.MethodsWe developed a decision tree with the following outcomes: mortality related to untreated and treated PE, mortality due to major hemorrhage, and intracranial bleeding. The timeframe extended from the suspicion of PE to its confirmation or exclusion. Most probabilities were derived from data from the Computerized Registry of Patients with VTE (RIETE). We estimated the incidence of bleeding by categories of clinical prediction rules of PE from a recent diagnostic management study of PE. Uncertainty was assessed through one-way and probabilistic sensitivity analyses.ResultsThe model favored preemptive anticoagulation if the diagnostic delay was > 6.3 h, > 2.3 h, and > 0.3 h (Revised Geneva low, intermediate, and high probability) and > 8.1 h and > 1.7 h (Wells unlikely and likely). With a diagnostic delay of 6 h, the absolute mortality reduction with anticoagulation was 0%, 0.02%, and 0.1% for low, intermediate, and high clinical probability, respectively. In one-way sensitivity analyses, the mortality of untreated PE was the most critical variable. Probabilistic analyses reinforced the superiority of anticoagulation in intermediate- and high-probability patients and suggested that low-probability patients might not benefit from treatment after diagnostic delays of < 6 to 8 h.ConclusionsOur model suggests that patients with intermediate and high/likely probabilities of PE benefit from preemptive anticoagulation. With a low probability, the decision to treat may rely on the expected diagnostic delay.

25 keywords...

hide…