Can chest CT severity scoring in covid-19 act as proxy to clinical classification of disease

Abstract

Purpose: To evaluate chest CT pattern in COVID-19 patients and highlight prognostic utility of CT severity score (CTSS) in grading SARS-CoV-2 disease. Materials and Methods: From May 15th to August 30th2020, retrospective analysis of RT-PCR confirmed 130 COVID-19 patients was done at a tertiary care center in Central India. From the cohort comprising of 81 males and 49 females, 76.92%, 15.38% and 7.69% had clinically mild, moderate and severe disease respectively. CT SS was calculated based on the extent of lobar involvement (0:0%; 1:< 5%; 2:5–25%; 3:26–50%; 4:51–75%; 5:>75%; range 0–5; total score 0–25) and was compared with clinical disease severity using bivariate and agreement analyses. ROC curve was used to evaluate the role of CTSS as a predictor of patients’ disease severity. Results: 59 (45.38%) patients were clinically asymptomatic and 38 (29.23%) patients had no parenchymal abnormality. In the radiologically positive group, 77 (83.69%) patients showed typical CT characteristics. Ground-glass opacification was the main CT pattern seen in 71 (77.17%) patients. Parenchymal changes were commonly observed bilaterally (76.08%) involving posterior more than anterior segments (46.73%) with apicobasilar gradient (lower lobes more than upper lobes). CTSS16 corresponded to mild, moderate and severe disease respectively. The optimal CTSS threshold for identifying moderate-to-severe patients was 8 (area under the ROC curve 0.8718, p