Chest CT Characteristics of Patients Admitted with COVID-19 Pneumonia

Abstract

Rationale: The novel coronavirus disease-19 (COVID-19) has presented major challenges for global health systems. Given limited availability of diagnostic testing and delays in test results during the first wave of the pandemic, our hospital used computed tomography (CT) to risk stratify patients with suspected COVID-19. The aim of this study was to describe the various patterns of disease on chest CT and relate them to chest x-ray (CXR) findings. Methods: This is a retrospective review of 559 symptomatic patients infected with COVID-19 (diagnosed by real-time reverse transcription polymerase chain reaction) admitted from March 2020 to May 2020 at Temple University Hospital (Philadelphia, PA) who received admission CXR and chest CT scans that were performed within 24 hours of admission. Scans were independently reviewed by a group of radiologists. CXRs was interpreted as “consistent with COVID-19” if there were lower lobe peripheral opacities. Chest CTs were evaluated for the presence of ground glass opacities, consolidations, interlobular septal thickening, centrilobular nodules, and crazy paving pattern. Chest CT was also assessed for background lung disease (emphysema, interstitial lung disease). Results: Of the 559 patients, median age was 58 years old, 55.5% were female, and 56.7% were African American. Median BMI was 31.61. Median duration of symptoms at time of chest imaging was 5 days. 153 (27.4%) of patient's admission CXR was not consistent with COVID-19. Of those, 124 (81%) had abnormalities on chest CT. Median number of lobes involved with disease on CT was 3.8 and 317 patients (56.7%) had all 5 lobes with disease. The most common abnormalities found were ground glass opacities (n=507, 90.7%), consolidations (n=224, 40%) and centrilobular nodules (n=127, 22.7%). Less common findings included pleural effusion (n=62, 11.8%), lymphadenopathy (n=55, 9.8%), pericardial effusion (n=24, 4.2%), and pneumothorax (n=3, 0.53%). Of note, 82 (14.7%) patients were found to have emphysema, and 2 (0.35%) were found to have interstitial lung disease. Conclusion: We present one of the largest reviews of CT scans in patients admitted for COVID-19. The majority of our population had significant burden of disease on CT at time of presentation. Ground glass opacities and consolidations were the predominant findings. Most patients did not have background emphysema or interstitial lung disease. The fact that many patients with normal CXR had abnormalities on chest CT highlights the utility of chest CT in evaluating patients with COVID-19.