Appropriateness of diagnostic management and outcomes of suspected pulmonary embolism.

Background: International guidelines include several strategies for diagnosing pulmonary embolism with confidence, but little is known about how these guidelines are implemented in routine practice.

Objective: To evaluate the appropriateness of diagnostic management of suspected pulmonary embolism and the relationship between diagnostic criteria and outcome.

Design: Prospective cohort study with a 3-month follow-up.

Setting: 116 emergency departments in France and 1 in Belgium.

Patients: 1529 consecutive outpatients with suspected pulmonary embolism.

Measurements: Appropriateness of diagnostic criteria according to international guidelines; incidence of thromboembolic events during follow-up.

Results: Diagnostic management was inappropriate in 662 (43%) patients: 36 of 429 (8%) patients with confirmed pulmonary embolism and 626 of 1100 (57%) patients in whom pulmonary embolism was ruled out. Independent risk factors for inappropriate management were age older than 75 years (adjusted odds ratio, 2.27 [95% CI, 1.48 to 3.47]), known heart failure (odds ratio, 1.53 [CI, 1.11 to 2.12]), chronic lung disease (odds ratio, 1.39 [CI, 1.00 to 1.94]), current or recent pregnancy (odds ratio, 5.92 [CI, 1.81 to 19.30]), currently receiving anticoagulant treatment (odds ratio, 4.57 [CI, 2.51 to 8.31]), and the lack of a written diagnostic algorithm and clinical probability scoring in the emergency department (odds ratio, 2.54 [CI, 1.51 to 4.28]). Among patients who did not receive anticoagulant treatment, 44 had a thromboembolic event during follow-up: 5 of 418 (1.2%) patients who received appropriate management and 39 of 506 (7.7%) patients who received inappropriate management (absolute risk difference, 6.5 percentage points [CI, 4.0 to 9.1 percentage points]; P < 0.001). Inappropriateness was independently associated with thromboembolism occurrence (adjusted odds ratio, 4.29 [CI, 1.45 to 12.70]).

Limitations: This was an observational study without evaluation of the risk for overdiagnosis.

Conclusions: Diagnostic management that does not adhere to guidelines is frequent and harmful in patients with suspected pulmonary embolism. Several risk factors for inappropriateness constitute useful findings for subsequent interventions.