A retrospective evaluation of the age-adjusted D-dimer versus the conventional D-dimer for pulmonary embolism

Abstract

 

The conventional D-dimer cut-off value of at least 500 μg FEU/l has good sensitivity but poor specificity for identifying pulmonary embolism. An elevated age-adjusted D-dimer value (age in years × 10 μg FEU/l) for patients at least 50 years old has been recommended as a better cut-off with adequate sensitivity and improved specificity for identifying pulmonary embolism compared with the conventional value. We retrospectively reviewed 3117 patient encounters in which a D-dimer was ordered. The D-dimer value, age of the patient, and the computed tomography radiology report was evaluated. The sensitivity and specificity of the age-adjusted D-dimer was calculated using bootstrapping. With an assumed 99% sensitivity for the conventional D-dimer cut-off the specificity was 39.2% [95% confidence interval (CI): 37.5–41.0%]. The sensitivity of the age-adjusted D-dimer was 91.8% (95% CI: 83.8–97.2%) with a specificity of 51.0% (95% CI: 49.1–53.1%). The sensitivity of the age-adjusted D-dimer was unacceptably low compared with the conventional D-dimer cut-off.