COVID-19 pneumonia and its lookalikes: How radiologists perform in differentiating atypical pneumonias
- PMID: 34700092
- PMCID: PMC8524806
- DOI: 10.1016/j.ejrad.2021.110002
COVID-19 pneumonia and its lookalikes: How radiologists perform in differentiating atypical pneumonias
Abstract
Purpose: To examine the performance of radiologists in differentiating COVID-19 from non-COVID-19 atypical pneumonia and to perform an analysis of CT patterns in a study cohort including viral, fungal and atypical bacterial pathogens.
Methods: Patients with positive RT-PCR tests for COVID-19 pneumonia (n = 90) and non-COVID-19 atypical pneumonia (n = 294) were retrospectively included. Five radiologists, blinded to the pathogen test results, assessed the CT scans and classified them as COVID-19 or non-COVID-19 pneumonia. For both groups specific CT features were recorded and a multivariate logistic regression model was used to calculate their ability to predict COVID-19 pneumonia.
Results: The radiologists differentiated between COVID-19 and non-COVID-19 pneumonia with an overall accuracy, sensitivity, and specificity of 88% ± 4 (SD), 79% ± 6 (SD), and 90% ± 6 (SD), respectively. The percentage of correct ratings was lower in the early and late stage of COVID-19 pneumonia compared to the progressive and peak stage (68 and 71% vs 85 and 89%). The variables associated with the most increased risk of COVID-19 pneumonia were band like subpleural opacities (OR 5.55, p < 0.001), vascular enlargement (OR 2.63, p = 0.071), and subpleural curvilinear lines (OR 2.52, p = 0.021). Bronchial wall thickening and centrilobular nodules were associated with decreased risk of COVID-19 pneumonia with OR of 0.30 (p = 0.013) and 0.10 (p < 0.001), respectively.
Conclusions: Radiologists can differentiate between COVID-19 and non-COVID-19 atypical pneumonias at chest CT with high overall accuracy, although a lower performance was observed in the early and late stage of COVID 19 pneumonia. Specific CT features might help to make the correct diagnosis.
Keywords: Atypical; Bacteria; COVID-19; CT; Fungal; Viral.
Copyright © 2021 Elsevier B.V. All rights reserved.
Conflict of interest statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Figures





References
-
- Waller J.V., Allen I.E., Lin K.K., Diaz M.J., Henry T.S., Hope M.D. The limited sensitivity of chest computed tomography relative to reverse transcription polymerase chain reaction for severe acute respiratory syndrome coronavirus-2 infection: a systematic review on COVID-19 diagnostics. Invest. Radiol. 2020;55(12):754–761. doi: 10.1097/RLI.0000000000000700. - DOI - PMC - PubMed
MeSH terms
LinkOut - more resources
Full Text Sources
Medical