Evaluation of an association of radiological findings and severity of the disease in patients with the new coronavirus infection (COVID-19)

Abstract

Rationale: Interpretation of lung abnormalities identified by computed tomography (CT) in patients with COVID-19 could be controversial in some cases. At present, there is no highly reliable algorithm for assessment and prediction of the disease coursed based on CT findings.Aim: To identify an association of the radiological findings in COVID-19 and its clinical manifestations.Materials and methods: This observational retrospective cohort study included 92 patients, categorized into three groups according to their clinical severity (mild COVID-19 29 patients, moderate COVID-19 33 patients, and severe COVID-19 30 patients). Chest CT was performed in all patients at admittance to the hospital and at day 10 of their hospital stay.Results: Almost all patients with severe COVID-19 (28 patients, 96.6%) demonstrated an increase in the damaged lung parenchyma volume at the second CT. The risk of clinical deterioration in these patients was 15.037-fold higher, compared to that in the patients with a stable volume of lung lesions. The area of pulmonary lesions at the first CT demonstrated its good prognostic ability (ROC area under the curve 0.831, sensitivity 87.5%, specificity 70.0%, p<0.001) to predict clinical deterioration. The presence of bronchial dilation in the total patient group significantly (p<0.01) correlated with an increase of the pulmonary lesion area. Clinical deterioration was found in 5 patients (62.5%) with bronchial dilatation.Conclusion: CT patterns in COVID-19 patients do not always correlate with clinical severity of the disease. Therefore, lung CT cannot be used for prediction of the COVID-19 course as a single method without clinical and laboratory assessments.

Keywords

• covid-19
• novel coronavirus infection
• computed tomography
• viral pneumonia