[Value of the association of D-dimer measurement and the evaluation of clinical probability in a non-invasive diagnostic strategy of pulmonary embolism]. / Intérêt de l'association du dosage des D-dimères et de l'évaluation de la probabilité clinique dans une stratégie diagnostique non invasive de l'embolie pulmonaire.

New diagnostic tools in suspected pulmonary embolism complete the classical diagnostic strategy of pulmonary scintigraphy and pulmonary angiography to limit the indications of these two invasive investigations. In a prospective series of 204 consecutive patients with suspected pulmonary embolism the association of D-dimer measurement and clinical probability was assessed for the exclusion of the diagnosis of pulmonary embolism. The D-DI Liatest is a new generation, unitary, rapid and quantitative latex test with a comparative diagnostic performance to that of the reference ELISA test, and well adapted to emergency situations.The clinical probability was assessed by a quantitative score based on past history, clinical symptoms and signs. The positive diagnosis of pulmonary embolism was made by spiral CT scanner and/or pulmonary angiography, associated with Duplex ultrasonography of the leg veins in nondiagnostic results. The prevalence of pulmonary embolism was 42.6% and the absence of anticoagulation in patients considered not to have pulmonary embolism was associated with a thrombo-embolic incidence of 0.9% at 3 months. Fifty-six patients had D-dimer concentrations equal or inferior to the threshold of 500 microg/L; the sensitivity was 99% and the specificity 47% with a negative predictive value of 98% to 100% in cases with a low clinical probability. D-dimer measurement is reliable and has a high cost-benefit value in ambulatory patients with suspected of pulmonary embolism and is even more valuable when the clinical probability of this diagnosis is low.