Multicenter Evaluation of the YEARS Criteria in Emergency Department Patients Evaluated for Pulmonary Embolism. (11th September 2018)

• Record Type: Journal Article
• Authors: Kabrhel, Christopher; Van Hylckama Vlieg, Astrid; Muzikanski, Alona; Singer, Adam; Fermann, Gregory J.; Francis, Samuel; Limkakeng, Alex; Chang, Ann Marie; Giordano, Nicholas; Parry, Blair
• Editors: Runyon, Michael S.
• Is Part Of: Academic emergency medicine. Volume 25:Number 9(2018)
• Page Start: 987
• Page End: 994
• Publication Date: 2018-09-11
• Subjects: Emergency medicine -- Periodicals; 616.02505
• DOI: 10.1111/acem.13417
• Languages: English
• ISSNs: 1069-6563

Abstract

Abstract: Background: It may be possible to safely rule out pulmonary embolism (PE) in patients with low pretest probability (PTP) using a higher than standard D‐dimer threshold. The YEARS criteria, which include three questions from the Wells PE score to identify low‐PTP patients and a variable D‐dimer threshold, was recently shown to decrease the need for imaging to rule out PE by 14% in a multicenter study in the Netherlands. However, the YEARS approach has not been studied in the United States. Methods: This study was a prospective, observational study of consecutive adult patients evaluated for PE in 17 U.S. emergency departments. Prior to diagnostic testing, we collected the YEARS criteria: "Does the patient have clinical signs or symptoms of DVT?" "Does the patient have hemoptysis?" "Are alternative diagnoses less likely than PE?" with YEARS (+) being any "yes" response. A negative D‐dimer was <1000 mg/dL for YEARS (–) patients and <500 mg/dL for YEARS (+) patients. We calculated test characteristics and used Fisher's exact test to compare proportions of patients who would have been referred for imaging and patients who would have had PE "missed." Results: Of 1, 789 patients, 84 (4%) had PE, 1, 134 (63%) were female, 1, 038 (58%) were white, and mean (±SD) age was 48 (±16) years. Using the standard D‐dimer threshold, 940 (53%) would not have had imaging, with two (0.2%, 95% confidence interval [CI] = 0.02%–0.60%) missed PE. Using YEARS adjustment, 1, 204 (67%, 95% CI = 65%–69%) would not have been referred for imaging, with six (0.5%, 95% CI = 0.18%–1.1%) missed PE, and using "alternative diagnoses less likely than PE" adjustment, 1, 237 (69%, 95% CI = 67%–71%) would not have had imaging with six (0.49%, 95% CI = 0.18%–1.05%) missed PE. Sensitivity was 97.6% (95% CI = 91.7%–99.7%) for the standard threshold and 92.9% (95% CI = 85%–97%) for both adjusted thresholds. Negative predictive value (NPV) was nearly 100% for all approaches. Conclusions: D‐dimer adjustment based on PTP may result in a reduced need for imaging to evaluate possible PE, with some additional missed PE but no decrease in NPV.