Diagnostic value of the D-dimer test in deep vein thrombosis: Improved results by a new assay method and by using discriminate levels

Previous studies have suggested that D-dimer testing reliably selects patients for whom duplex sonography should be performed for diagnosis of deep vein thrombosis (DVT). However, the interassay correlation is poor. Therefore, we tested four D-dimer methods for their ability to rule out DVT, including the Miniquant test, a new D-dimer assay method. Sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) were calculated vs. duplex sonography. Twenty-nine of 108 (27%) patients with suspected DVT were diagnosed as having DVT by sonography. The Vidas enzyme-linked immunoabsorbent assay (ELISA) test, the Miniquant turbidimetric test and the latex agglutination test for D-dimer all provided discriminate values for achieving 100% sensitivity and 100% NPV. D-dimer results demonstrated higher specificity and PPV in the outpatient vs. the inpatient group. This probably reflects the higher proportion of inpatients with comorbid conditions, such as malignancy and postsurgery status, in whom D-dimer results show very poor specificity and PPV. The new Miniquant turbidimetric assay performed as well as the more established ELISA method. We conclude that the D-dimer tests were shown to possess the necessary sensitivity and NPV to be useful in screening patients with suspected DVT. A negative D-dimer test in selected patients could be helpful in reducing the number of sonograms performed for diagnosis of DVT. (C) 2001 Elsevier Science Ltd. All rights reserved.