The use of a fixed high sensitivity to evaluate five D-dimer assays' ability to rule out deep venous thrombosis: A novel approach

Abstract

Suspected deep venous thrombosis (DVT) is difficult to refute without complex diagnostic algorithms and expensive testing. We analysed five d-dimer assays' utility for exclusion of suspected DVT during a prospective clinical cohort trial, choosing a highly sensitive cut-off value at which to compare the assays. Assays were performed on 436 consecutive patients who were referred with symptoms that suggested a first episode of DVT. Venous thromboembolism (VTE) was defined as positive findings on comprehensive duplex ultrasonography or any episode, or complication of VTE detected during 3 months of clinical follow-up. All five assays were performed in 377 patients. At a highly sensitive cut-off value, all five assays reliably excluded DVT in the study population. While the choice of a highly sensitive cut-off value reduced the specificity of all the assays, the change in specificity differed between tests. Our findings suggest that a second-generation d-dimer assay could be used as a stand-alone test to rule out suspected DVT when a highly sensitive cut-off value is chosen. These findings should be subjected to a prospective management study, as a small reduction in sensitivity from our findings could result in a clinically relevant decrease in negative predictive value. © 2005 Blackwell Publishing Ltd.