ABSTRACT
BACKGROUND: Assessment of pulmonary embolism (PE) remains a diagnostic and investigative burden to emergency departments. The decision of which D-dimer cut-off to use in low-risk patients remains controversial. AIMS: To compare the sensitivity and specificity of varying D-dimer cut-offs in the diagnosis of PE for Wells low-risk patients. METHODS: Retrospective review of patients presenting to a tertiary emergency department over 42 months who had a D-dimer performed for PE risk stratification. Wells scores were calculated for each patient, those with Wells score of ≤4 ('PE unlikely') were analysed. Four D-dimer thresholds were compared, including traditional threshold (≥0.5 µg/mL), age-adjusted (≥age in years × 0.01 µg/mL), doubled-traditional threshold and YEARS criteria. RESULTS: During the study period, 2291 D-dimers were ordered for suspected PE, of which 2125 were low risk for PE. Of these low-risk patients 46 (2.2%) were found to have a PE. The sensitivity and specificity for each D-dimer threshold were traditional threshold (95.6% and 65.6%), age-adjusted (93.5% and 71.7%), doubled traditional (69.6% and 85.5%) and YEARS criteria (80.4% and 84.0%). Utilising an age-adjusted threshold, YEARS criteria or doubled-traditional threshold would have resulted in 70, 217 and 245 fewer imaging investigations. CONCLUSIONS: The prevalence of PE in this low-risk cohort was very low. Utilising an age-adjusted D-dimer would have reduced imaging tests performed while maintaining good sensitivity. Although The YEARS criteria and doubled-traditional threshold would have reduced scanning considerably both had sensitivities of less than 90%.
