Evaluation of the RSNA consensus: is it sufficient for the diagnosis of COVID-19 with CT?

Abstract

Purpose: Radiological Society of North America (RSNA) Consensus for coronavirus disease 19 (COVID-19) is developed to evaluate the lung involvement on chest computed tomography (CT) and create a common reporting lexicon. Aim of this study is to determine the frequency of CT features in sex and age groups in patients with COVID-19, compare the findings according to the RSNA consensus classifications, and evaluate the compatibility of the classifications and findings.
 Materials and Methods: Chest CT images of 281 patients with COVID-19 were evaluated. Patients were noted in the appropriate RSNA consensus class. The patients’ data were analyzed by group according to age and sex.
 Results: The main findings included ground-glass opacity, consolidation, and air bronchogram. The common involvement patterns were as follows: bilateral, peripheral, and multifocal. The rates for the typical, atypical, and indeterminate classifications, according to the RSNA consensus, were 63.6%, 9.6%, and 27.0%, respectively. Subpleural fibrous streaking was more frequent in males. Air bronchogram, lymphadenopathy, pleural effusion, subpleural fibrous streaking, bilateral involvement, and a typical classification on CT features were more frequent in the ≥ 65-year age group.
 Conclusion: While the typical appearance classification has results consistent with the findings, we think that the classifications specified as indeterminate and atypical appearance do not show sufficient agreement with the findings and revision is needed for correct diagnostic guidance.