Clinical Question

Can a simple clinical decision tool safely decrease the use of computed tomographic angiography (CTA) in patients with suspected pulmonary embolism?

Bottom Line

In this study, using a simplified algorithm in patients with suspected pulmonary embolism can safely decrease the number of CTAs. (Level of Evidence = 1b)

Synopsis

Over the past several years, the addition of d-dimers to emergency department laboratory panels and the widespread access of rapid diagnostic tools (i.e., CTA) have resulted in a decrease in the prevalence of pulmonary embolism to approximately 10%. These researchers from the Netherlands prospectively evaluated a simple algorithm prediction model and also compared it with the Wells prediction model in 3,465 consecutively recruited patients with suspected pulmonary embolism. The simple model (called YEARS) assesses three factors: clinical signs of deep venous thrombosis, hemoptysis, and whether pulmonary embolism is the most likely diagnosis.

In addition to a d-dimer test result, each patient was scored using the YEARS and Wells models. If a patient had no YEARS items and a d-dimer result of less than 1,000 ng per mL, the authors ruled out pulmonary embolism (1,320 patients fit this category). If a patient had no YEARS items and a d-dimer result of 1,000 ng per mL or greater, the clinician ordered a CTA (423 patients). If a patient had one or more YEARS items and a d-dimer result of less than 500 ng per mL, pulmonary embolism was ruled out (331 patients), but if the d-dimer result was 500 ng per mL or greater, the clinician ordered a CTA (1,391 patients). The researchers followed patients in whom a pulmonary embolism was ruled out for three months. Approximately 13% of the patients were diagnosed with pulmonary embolism at the outset, a handful of patients were anticoagulated for reasons other than thromboembolism (e.g., atrial fibrillation), and 18 (0.6%) of the 2,946 remaining patients were subsequently diagnosed with pulmonary embolism. Six of those were fatal. Based on the YEARS algorithm, 48% of the patients did not need a CTA compared with 34% using the Wells model.

Study design: Decision rule (validation)

Funding source: Self-funded or unfunded

Setting: Uncertain

Reference:van der HulleTCheungWYKooijSet alYEARS study groupSimplified diagnostic management of suspected pulmonary embolism (the YEARS study): a prospective, multicentre, cohort study [published correction appears in Lancet. 2017;390(10091):230]. Lancet2017;390(10091):289–297.