Chest Computed Tomography (CT) Severity Scales in COVID-19 Disease: A Validation Study

Med Sci Monit. 2021 May 5:27:e931283. doi: 10.12659/MSM.931283.

Abstract

BACKGROUND Imaging-based quantitative assessment of lung lesions plays a key role in patient triage and therapeutic decision-making processes. The aim of our study was to validate the Total Severity Score (TSS), Chest Computed Tomography Score (CT-S), and Chest CT Severity Score (CT-SS) scales, which were used to assess the extent of lung inflammation in patients with SARS-CoV-2 infection in terms of interobserver agreement and the correlation of scores with patient clinical condition on the day of the study. MATERIAL AND METHODS A total of 77 chest CT scans collected from 77 consecutive patients hospitalized because of SARS-CoV-2 were included. The scans were assessed independently by 2 radiologists aware of the patients' positive results of RT-PCR tests. Each chest CT was assessed according to the 3 scales. To assess the interobserver agreement of CT scan assessments, Cohen's k and intraclass correlation coefficient (ICC) were calculated. RESULTS For the overall assessment, the k was 0.944 and the ICC was 0.948 for the TSS; the kappa was 0.909 and the ICC was 0.919 for the CT-S; and the k was 0.888 and the ICC was 0.899 for the CT-SS. The CT-SS (r=0.627 for Radiologist 1 and r=0.653 for Radiologist 2) revealed the strongest positive correlation with the patient clinical condition as expressed using the Modified Early Warning Score. CONCLUSIONS The interobserver agreement for the 3 evaluated scales was very good. The CT-SS was found to have the strongest positive relationship with the Modified Early Warning Score.

MeSH terms

  • Biomarkers
  • COVID-19 / diagnostic imaging*
  • COVID-19 / virology*
  • Humans
  • Image Processing, Computer-Assisted
  • Observer Variation
  • Radiography, Thoracic*
  • Reproducibility of Results
  • SARS-CoV-2*
  • Severity of Illness Index
  • Tomography, X-Ray Computed* / methods
  • Tomography, X-Ray Computed* / standards

Substances

  • Biomarkers