UCL Discovery

Abstract

We evaluated the performance of eight d-assays for the exclusion of deep vein thrombosis (DVT); Biopool AutoDimer, Biopool MiniQuant, bioMèrieux MDA d-Dimer, VIDAS, Dade Behring d-Dimer Plus, Trinity Biotech AMAX, NycoCard d-dimer and IL Test d-Dimer. The assays were evaluated both as stand-alone tests, and in combination with pretest probability (PTP). d-dimer assays and PTP assessment were performed on 410 patients presenting to the emergency department with suspected acute DVT. DVT was diagnosed in 76 of 410 patients (18·5%) by compression ultrasound or other imaging techniques, as required. Receiver operator characteristics analysis established optimum cut-off values and these were compared with manufacturer's cut-off values where provided. As stand-alone tests, the assays varied immensely regarding cut-off value, negative predictive value (NPV 93–100%) and specificity (0–67%). At least one patient with confirmed DVT had a low d-dimer level by each method: to achieve 100% sensitivity it would be necessary to reduce cut-off values to levels below clinical usefulness. When low d-dimer was used in combination with PTP, six of eight methods achieved ≥98% NPV, with a diagnosis of DVT excluded in 16–44% of patients without the requirement for diagnostic imaging. The highly variable diagnostic performance of these d-dimer assays means that some assays are unsuitable for certain diagnostic strategies. However, our data suggest that the combination of sensitive d-dimer assays with an assessment of PTP may be used to exclude a diagnosis of DVT.