Diagnostic Accuracy of D-Dimer Testing for Deep Vein Thrombosis: A Meta-Analysis

Abstract

D-dimer testing is widely used to rule out deep vein thrombosis (DVT) in patients with suspected thromboembolism of the veins when used in conjunction with clinical pre-test probability tests. This meta-analysis summarises the results of 15 instances of diagnostic accuracy research published in 2020-2025 that tested the D-dimer assays (laboratory-based and point-of-care) using lower extremity DVT, either as a single test or as a clinical decision rule (Wells, YEARS, age-adjusted thresholds). A total of 18400 patients were analysed across the reported studies, and they include both single-centre diagnostic assessments to large multicentre cohort studies. D‑dimer has shown a very high sensitivity in ruling out proximal DVT but with a low specificity, and both are heterogeneous, so its positive predictive power is limited, which leads to a high prevalence of false positive results in older and comorbid patients. The sensitivity and specificity were estimated to be 0.96 and 0.52, pooled. Since the techniques were appropriately utilized, age-adjusted and clinical-probability-adjusted strategies produced meaningful specificity increases, with sensitivity held constant. A high heterogeneity (I 2 large) was observed, which was contributed to by the assay type, cut-off levels, patient setting (emergency department or inpatient/ post-surgery), and DVT prevalence. In turn, D -D-dimer testing should be combined with a proven pre-test probability instrument in clinical use, and a low- or middle-probability patient with negative D -D-dimer safely rules out DVT in most modern cohorts. At least wider use of age-adjusted cut-offs, and careful local validation of assays.