Inappropriate use of D-dimer and impact on the test characteristics for deep vein thrombosis exclusion

This study aimed to investigate the prevalence of the inappropriate use of D-dimer testing for deep vein thrombosis (DVT) exclusion, and to evaluate the diagnostic value of D-dimer testing compared between testing in appropriate and inappropriate conditions. A retrospective chart review of patients with D-dimer requests at Siriraj Hospital, Bangkok, Thailand during January-September 2017 was performed. Data relating to D-dimer requests, Wells score, imaging investigations, and coexisting conditions that cause elevated D-dimer without correlation to DVT were analyzed. From 2196 cases, there were 300 D-dimer requests for DVT diagnosis. Of those, 5 (1.7%) and 158 (53.6%) requests were considered to be absolutely inappropriate D-dimer test requests due to the absence of pre-test Wells score evaluation and testing in DVT-likely patients, respectively. One-hundred and forty-eight (49.3%) requests were considered to be the relatively inappropriate D-dimer test requests due to testing in patients with coexisting conditions that included malignancies, sepsis, surgery, trauma, arterial thrombosis, and pregnancy. Three patients had more than one of these conditions. Comparatively higher negative predictive values (NPVs) of D-dimer assay for DVT exclusion were observed when used in DVT-unlikely (100% versus 80%), and in patients without coexisting conditions (100% versus 66.7%). The present study revealed the rate of inappropriate D-dimer testing for DVT exclusion to be high and this may lead to missed DVT diagnosis, overuse of imaging study, and unnecessary cost of D-dimer testing. Appropriate patient selection before test request can improve the NPV of D-dimer for DVT exclusion.