Contribution of a new, rapid, quantitative and automated method for D-dimer measurement to exclude deep vein thrombosis in symptomatic outpatients

Ninety-nine consecutive outpatients with symptoms suggestive of deep vein t hrombosis (DVT) were tested for the presence of D-dimer using a new rapid m ethod (BC D-Dimer) based on the agglutination principle and performed on a BCT analyzer. Dimertest Gold EIA and VIDAS D-Dimer devices were used as com parative methods. Venography was performed in all patients, with DVT diagno sed in 39 of them (34 proximal). The BC D-Dimer test proved to be rapid, au tomated and well suited for individual tests with a good reproducibility (c oefficient of variation % 1.5-5.3). Its performance was comparable with tha t of the other methods, as indicated by the areas under the receiver operat ing characteristic curves (Gold EIA 0.95; VIDAS 0.95; BC D-Dimer 0.91) and the coefficients of correlation (Pearson's coefficients from 0.832 to 0.876 ). On the basis of the Kappa coefficients, there was quite a good concordan ce between the three tests. At the cut-off levels that provided the highest sensitivity (97.4%, for all the tests), the negative predictive values wer e similar for all methods, at over 95%. The corresponding specificity range d from 62.7 to 81.7%. In conclusion, this study shows that the new method c an be included in prospective clinical trials to test the utility of D-dime r measurement in combination with other non-invasive diagnostic procedures in the management of the diagnosis of DVT. (C) 1999 Lippincott Williams & W ilkins.