External validation of the YEARS diagnostic algorithm for suspected pulmonary embolism

Abstract

Background: Validated diagnostic algorithms are used to manage patients with suspected pulmonary embolism (PE). The recently published YEARS study proposed a simplified diagnostic strategy to reduce the use of computed tomography pulmonary angiography. Objectives: To externally validate this strategy in an independent cohort. Methods: We analyzed data from three previous prospective cohort studies of outpatients with suspected PE. We retrospectively applied the YEARS algorithm. The three YEARS clinical criteria are: clinical signs of deep vein thrombosis, hemoptysis, and PE as the most likely diagnosis. If zero YEARS criteria are met, a D-dimer < 1000 ng/mL will rule out PE. If ≥1 YEARS criteria are met, a D-dimer < 500 ng/mL will rule out PE. Results: Of the 3314 patients, 731 (22.1%) had PE. Applying the YEARS diagnostic algorithm, 1423 (42.9%) patients could have had PE ruled out without imaging. Of these patients, 17 (1.2%; 95% confidence interval 0.8-1.9) were diagnosed with PE at initial testing. All 17 had no YEARS item and a D-dimer < 1000 ng/mL. All 17 had a D-dimer level above their age-adjusted cutoff. Among the 272 patients with no YEARS criteria and a D-dimer < 1000 ng/mL but above their age-adjusted D-dimer cutoff, PE was diagnosed in 6.3% (17/272; 95% confidence interval 3.9-9.8). Conclusion: We provide external validation of the YEARS diagnostic algorithm in an independent cohort. The rule appears to safely exclude PE. However, caution is required in patients with no YEARS item and a D-dimer < 1000 ng/mL but above their age-adjusted D-dimer cutoff.