OBJECTIVE: To determine the role of four ELISA D-dimer assays in the exclusion of pulmonary embolism. DESIGN: Blinded comparison using pulmonary angiography and/or lung scintigraphy as a reference method. SETTING: A secondary and tertiary referral centre. PATIENTS AND METHODS: Consecutive patients with suspected pulmonary embolism underwent lung scintigraphy, followed by angiography if a non-diagnostic result was obtained. Comorbid conditions resulting in elevated plasma D-dimer levels were defined a priori. Cut-off levels for 100% sensitivity was determined. A decision-analytic model was used to determine effectiveness and costs in the management pulmonary embolism. MAIN OUTCOME MEASURES: The exclusion efficacy of the various assays at a sensitivity of 100% and cost-effectiveness. RESULTS: A total of 179 patients were included (78 inpatients and 101 outpatients; 74 patients had comorbid conditions). Pulmonary embolism could be adequately excluded in between 8% and 18% of all patients, an in between 3% and 7% and 11% and 27% of inpatients and outpatients, respectively, depending on the assay used. D-dimer assays could exclude pulmonary embolism in <5% of patients with comorbid conditions, whereas this increased to 14-32% in outpatients without comorbid conditions. A cost-effectiveness analysis showed a cost reduction of 10% at a specificity of 30%, largely due to a 28% decrease in angiography requirements. Furthermore, for every 2% decrease to sensitivity, one per 1000 evaluated patients would die as a result of inadequately treated pulmonary embolism. CONCLUSION: D-dimer ELISA assays may have a role in the exclusion of pulmonary embolism in symptomatic outpatients, where the application may reduce angiography by 30% and costs by 10%.
OBJECTIVE: To determine the role of four ELISA D-dimer assays in the exclusion of pulmonary embolism. DESIGN: Blinded comparison using pulmonary angiography and/or lung scintigraphy as a reference method. SETTING: A secondary and tertiary referral centre. PATIENTS AND METHODS: Consecutive patients with suspected pulmonary embolism underwent lung scintigraphy, followed by angiography if a non-diagnostic result was obtained. Comorbid conditions resulting in elevated plasma D-dimer levels were defined a priori. Cut-off levels for 100% sensitivity was determined. A decision-analytic model was used to determine effectiveness and costs in the management pulmonary embolism. MAIN OUTCOME MEASURES: The exclusion efficacy of the various assays at a sensitivity of 100% and cost-effectiveness. RESULTS: A total of 179 patients were included (78 inpatients and 101 outpatients; 74 patients had comorbid conditions). Pulmonary embolism could be adequately excluded in between 8% and 18% of all patients, an in between 3% and 7% and 11% and 27% of inpatients and outpatients, respectively, depending on the assay used. D-dimer assays could exclude pulmonary embolism in <5% of patients with comorbid conditions, whereas this increased to 14-32% in outpatients without comorbid conditions. A cost-effectiveness analysis showed a cost reduction of 10% at a specificity of 30%, largely due to a 28% decrease in angiography requirements. Furthermore, for every 2% decrease to sensitivity, one per 1000 evaluated patients would die as a result of inadequately treated pulmonary embolism. CONCLUSION: D-dimer ELISA assays may have a role in the exclusion of pulmonary embolism in symptomatic outpatients, where the application may reduce angiography by 30% and costs by 10%.