[Usefulness of fast ELISA determination of D-dimer levels for diagnosing pulmonary embolism in an emergency room].

OBJECTIVE To determine the sensitivity and negative predictive value of D-dimer levels measured by fast enzyme-linked immunoabsorbent assay (ELISA) in pulmonary embolism. PATIENTS AND METHODS Prospective study of consecutive patients with suspicion of pulmonary embolism attended in the Emergency Room of the Hospital de La Princesa in Madrid, Spain. Thromboembolism was diagnosed with an algorithm established in the hospital, and D-dimer levels were determined by fast ELISA (VIDAS D-dimer Assay) in each patient suspected of pulmonary embolism. Patients with negative findings from a test not considered a reference method for thromboembolism were followed for 3 months. RESULTS Of 132 patients with clinical suspicion, 28 (21.2%) were positive and 104 (78.7%) were negative for embolism. D-dimer levels were below 0.5 microg/mL in 31 patients, 30 of whom did not have pulmonary thromboembolism whereas 1 did. D-dimer levels were abowe 0..5 microg/mL in 101 patients; thromboembolism did not occur in 74 of these but was reported in the remaining 27. For a value of 1 microg/mL, 66 patients had values below the cut off, 3 of whom presented pulmonary embolism. The remaining 66 patients had D-dimer levels above or equal to 1 microg/mL; 25 of them had a positive diagnosis for embolism and 41 had a negative diagnosis. Sensitivity and negative predictive values were 96.4% (95% confidence interval [CI], 79.8%-99.9%) and 96.8% (95% CI, 81.5%-98.8%), respectively, at a cut off of 0.5 microg/mL; and 89.2% (95% CI, 70.6%-97.2%) and 95.45% (95% CI, 86.4%-98.8%), respectively, at a cut off of 1 microg/mL. CONCLUSIONS In an emergency room, thromboembolism can be excluded if plasma levels of D-dimer measured by fast ELISA are below 0.5 microg/mL because of the high negative predictive value at this cut off.