The present study sought to evaluate the discriminative and predictive ability of D-dimer for pediatric limb DVT. Children aged 28 days-18 years requiring imaging to rule out limb DVT, as per the treating clinical team, were enrolled in the study. The outcome was ultrasound proven DVT. D-dimer levels were obtained around the time of imaging. Receiver operating characteristic (ROC) curves and logistic regression models were used for data analyses. In total, 296 patients were enrolled between 2017-2020; 204 patients were diagnosed with DVT [DVT(+)]. Median D-dimer levels were 2.3 μg/mL FEU (25th -75th percentile 0.9-3.9) among DVT(+) and 1.9 μg/mL FEU (25th -75th percentile 0.8-4.0) among DVT(-) patients (p=0.60). The area under the ROC curve (AUC) was 0.52 [95% confidence interval (CI) 0.45-0.59]. The odds ratio for D-dimer levels was 1.00 (95 % CI 0.99-1.01), holding confounders constant. In a sub-group exploratory analysis including 23 patients with no underlying conditions or co-morbidities, the AUC curve was 0.90 (95% CI 0.76-1.00). In this prospective cohort study of consecutive children with suspected limb DVT, D-dimer levels had poor discriminative and predictive ability for DVT. However, D-dimer levels showed better discriminative and predictive ability for DVT in an exploratory sample of patients with no underlying conditions or co-morbidities at the time of diagnosis. This article is protected by copyright. All rights reserved.
Maria Laura Avila, Nour Amiri, Eleanor Pullenayegum, Victoria Abigail Sealey, Riddhita De, Suzan Williams, Jennifer Vincelli, Leonardo R Brandão