J Korean Med Sci.  2021 Mar;36(8):e51. 10.3346/jkms.2021.36.e51.

Prognostic Implications of CT Feature Analysis in Patients with COVID-19: a Nationwide Cohort Study

  • 1Department of Radiology and Biomedical Research Institute, Pusan National University Hospital, Busan, Korea
  • 2Department of radiology, Soonchunhyang University College of Medicine, Soonchunhyang University Seoul Hospital, Seoul, Korea
  • 3Department of Radiology, Chungbuk National University Hospital, Cheongju, Korea
  • 4Department of Radiology, Pusan National University Yangsan Hospital, Yangsan, Korea
  • 5Department of Radiology, Kosin University Gospel Hospital, Busan, Korea
  • 6Department of Radiology, Chonnam National University Hospital, Gwangju, Korea
  • 7Department of Radiology and Center for Imaging Science, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea


Few studies have classified chest computed tomography (CT) findings of coronavirus disease 2019 (COVID-19) and analyzed their correlations with prognosis. The present study aimed to evaluate retrospectively the clinical and chest CT findings of COVID-19 and to analyze CT findings and determine their relationships with clinical severity.
Chest CT and clinical features of 271 COVID-19 patients were assessed. The presence of CT findings and distribution of parenchymal abnormalities were evaluated, and CT patterns were classified as bronchopneumonia, organizing pneumonia (OP), or diffuse alveolar damage (DAD). Total extents were assessed using a visual scoring system and artificial intelligence software. Patients were allocated to two groups based on clinical outcomes, that is, to a severe group (requiring O2 therapy or mechanical ventilation, n = 55) or a mild group (not requiring O2 therapy or mechanical ventilation, n = 216). Clinical and CT features of these two groups were compared and univariate and multivariate logistic regression analyses were performed to identify independent prognostic factors.
Age, lymphocyte count, levels of C-reactive protein, and procalcitonin were significantly different in the two groups. Forty-five of the 271 patients had normal chest CT findings. The most common CT findings among the remaining 226 patients were groundglass opacity (98%), followed by consolidation (53%). CT findings were classified as OP (93%), DAD (4%), or bronchopneumonia (3%) and all nine patients with DAD pattern were included in the severe group. Uivariate and multivariate analyses showed an elevated procalcitonin (odds ratio [OR], 2.521; 95% confidence interval [CI], 1.001–6.303, P = 0.048), and higher visual CT scores (OR, 1.137; 95% CI, 1.042–1.236; P = 0.003) or higher total extent by AI measurement (OR, 1.048; 95% CI, 1.020–1.076; P < 0.001) were significantly associated with a severe clinical course.
CT findings of COVID-19 pneumonia can be classified into OP, DAD, or bronchopneumonia patterns and all patients with DAD pattern were included in severe group. Elevated inflammatory markers and higher CT scores were found to be significant predictors of poor prognosis in patients with COVID-19 pneumonia.


Chest; Coronavirus; COVID-19; Tomography; X-Ray Computed; Pneumonia; Prognosis
Full Text Links
  • JKMS
export Copy
  • Twitter
  • Facebook
Similar articles
    DB Error: unknown error