Marcelle Pignanelli, MD
Resident Physician
Flushing Hospital Medical Center
Flushing, NY
Disclosure information not submitted.
Joseph Monye, M.D.
Resident Physician
Flushing Hospital Medical Center, United States
Disclosure information not submitted.
Logan Danielson, M.D.
Resident Physician
Jamaica Hospital Medical Center, United States
Disclosure information not submitted.
Jenny Chung, D.O.
Resident Physician
Jamaica Hospital Medical Center, United States
Disclosure information not submitted.
Kelly Cervellione, M.A., MPhil
Clinical Research Director
MediSys Health Network
Richmond Hill, New York, United States
Disclosure information not submitted.
Mikayla Fuchs, B.S.
Clinical Research Associate
MediSys Health Network, United States
Disclosure information not submitted.
Craig Thurm, M.D.
Director of Pulmonary Medicine
Jamaica Hospital Medical Center, United States
Disclosure information not submitted.
Karen Beekman, M.D.
Program Director Internal Medicine
Flushing Hospital Medical Center, United States
Disclosure information not submitted.
R. Jonathan Robitsek, Ph.D.
Research Director Department of Surgery
Jamaica Hospital Medical Center, United States
Disclosure information not submitted.
Title: Decreasing Utilization of CT Pulmonary Angiograms in COVID-19 Positive Patients using D-Dimer Levels
Introduction: There is equipoise about when to obtain CTPA in patients with COVID-19. D-Dimer values are used in patients without COVID-19 to help assess pre-test probability of PE. D-dimer is often elevated in COVID-19 regardless of presence of PE. We aimed to identify and validate a D-dimer value that is sensitive for assessing pre-test probability in COVID-19 patients. This could help reduce utilization, costs, and chance of spread of infection. During influx of COVID-19 cases, cutoffs may be clinically useful in prioritizing use of critical care resources.
Methods: The study utilized an initial sample (S1) to identify a cutpoint and an independent sample (S2) for
validation. Patients admitted (Jan-May 2021) to 2 community hospitals with COVID-19 who underwent
CTPA to rule out PE and had D-dimer within 48 hours prior to scan were included. Cutpoint analyses
were conducted to determine optimal sensitivity and negative predictive value (NPV).
Results: S1 included 68 patients. Median (min, 25ile, 75ile, max) D-dimer was 2026 (220; 1193; 4730; 106004) and 11 (16%) were CTPA+ for PE. Cutpoint analysis showed a D-dimer of 1710 had sensitivity=1.0, specificity=0.49, AUC=0.75, and NPV= 1.0. Using this cutpoint, 27 patients would not have undergone CTPA, with no PEs missed. S2 included 383 patients; 75 (19%) were CTPA+ for PE. Median D-dimer was 1789 (215; 1020; 5312; 128000). Cutpoint analysis showed a D-dimer of 1791 provided sensitivity=.79, specificity=.57, AUC=.73, and NPV=.92. Using this cutpoint, 59 patients would not have undergone CTPA, yet 16 PEs would have been missed.
In S2, those with PE were more likely to have signs/symptoms of DVT (X2=6.46, p=.01), have had recent surgery (X2=3.92, p=.05), and be on therapeutic dose anticoagulant at the time of D-dimer (39% vs. 53%, p=.04). There were no differences between groups in age, gender, history of DVT/PE, hemoptysis, immobilization, malignancy, tachycardia, or ICU admission.
Conclusions: Overutilization of CTPA has been observed in hypoxic COVID-19 patients. Using a D-Dimer cut-off between 1710 and 1790 provided high sensitivity and NPV for PE. Using D-Dimer cut-offs may improve diagnostic efficacy of CTPAs, reduce hospital costs, and decrease spread of infection. Further research to increase specificity and PPV using additional clinical information is ongoing.