113: Evaluation of Three D-Dimer Methods Reveals Age-Stratified Thresholds Maintain High Negative Predictive Value for Excluding Pulmonary Embolism While Improving Specificity

Abstract

Pulmonary embolism (PE) is a serious condition that can lead to death or significant morbidity. Thus PE is actively excluded in many patients presenting to the ED with vague complaints including pain and dyspnea. The high negative predictive value (NPV) of D-dimer can aid the clinician by excluding pulmonary embolism in patients at low risk of the disease. We conducted a clinical validation of two d-dimer assays: the Siemens Innovance, and the Biosite Triage and compared them with our current assay, the Siemens Advance. We also sought to assess whether more stringent criteria can safely be applied to younger patients suspected of PE to limit diagnostic imaging. We prospectively studied 1375 patients suspected of PE presenting to our emergency department (ED). ED physicians assessed clinical pretest probability for PE using the Modified Wells model. Patients with Wells score < 4.5 had a d-dimer blood test. Surplus specimens from a subset of our patients were used to generate d-dimer results with the Innovance and Triage assays. Patients with positive d-dimer test (Advance≥1.2μg/mL), meeting Kline Criteria, or at high likelihood for PE, received imaging, either CT pulmonary angiography (CTPA) or ventilation/perfusion (V/Q) scintigraphy. Radiologists were blinded to clinical score and d-dimer test results. PE positive patients had a positive CTPA or intermediate or high probability V/Q scan. Follow-up to exclude false negative results included phone calls and review of hospital and death records. Sensitivity, specificity, PPV and NPV were calculated at different d-dimer thresholds and age ranges. The agreement between the three d-dimer tests was determined using the Spearman Correlation statistic. The Triage and Innovance methods compared favorably with correlation statistics of r=0.88. The Advance method did not compare as well; r=0.48 versus Triage and 0.52 versus Innovance). The table (1st 3 lines) displays the very high NPV for all three assays, but shows the significantly (p< 0.001) improved specificity of both the Innovance and Triage assays compared with the Advance assay.Tabled 1 The 2nd half of the table shows that the cutoff values for excluding PE can be significantly increased for patients under age 50 while maintaining 100% NPV. This results in significantly higher specificity and thus decreased imaging (radiation exposure) and cost. Using cutoffs for the Innovance (0.5 μg/mL) and Triage (400 ng/mL) assays, 41.4% and 28.2% of patients would be imaged. When increasing the thresholds to 1.6 μg/mL (Innovance) and 800 ng/ml (Triage), 10.6% and 16.0% of patients would be imaged, respectively. For those under age 50, only 3.6% or 6.9% of patients (Innovance or Triage, respectively) would have been imaged. Innovance and Triage d-dimer assays result in excellent exclusion of PE while increasing specificity compared to the Advance assay. A higher d-dimer threshold can be safely applied to patients under age 50, resulting in fewer imaging studies, and significantly reducing hospital costs and radiation exposure.