Pulmonary embolism in patients with COVID-19 and D-dimer diagnostic value : a retrospective study

Abstract

Background D-dimer levels are significantly higher in COVID-19 patients with Pulmonary Thromboembolism (PTE) as compared to those without PTE, but its clinical utility is still uncertain. Purpose To determine the D-dimer performance for ruling out PTE in patients with COVID-19. We also assessed clinical and laboratory factors associated with the presence of PTE on CT Pulmonary Angiogram (CTPA). Methods Retrospective study involving all patients who presented at a tertiary care hospital from March 2020 to May 2021 with severe acute respiratory syndrome from COVID-19, who underwent CTPA and had D-dimer collected within 48 hours from CTPA. The D-dimer ability to classify patients with or without PTE according to CTPA was evaluated. Results A total of 697 patients [382 (54.8%) men; mean (SD) age, 59 (20.5) years] were included, of which 71.5% required intensive care admission, 32.4% had PTE, and 35.6% died during hospitalization. PTE was independently associated with mortality [42.5% vs. 32.3%; p = 0.038]. D-dimer levels were higher in patients with PTE [9.1 (3.9; 20) vs. 2.3 (1.2; 5.1); p < 0.001]. Using the D-dimer cutoff of 0.5 μg/mL or above, sensitivity was 98.2% and specificity 5.7%. The 0.3 μg/mL threshold was associated with 100% of sensitivity for the presence of PTE, with which 99.1% of patients had increased values. ROC curve AUC was 0.77, demonstrating moderate discriminative power of D-dimers to detect PTE. Conclusions D-dimer levels are higher among COVID-19 hospitalized patients with PTE as compared to those without PTE and have moderate discriminative power to detect PTE, but its use to exclude PTE in this population may have limited clinical utility.