Abstract
Background: Effective early risk stratification in COVID-19 remains a critical challenge in emergency care, particularly due to the limitations of RT-PCR testing, including delayed processing and false negatives. There is an unmet need for imaging tools that are fast, reliable, and safe for repeated use in acute clinical settings. Methods: In this prospective, multicenter study, over 1000 patients hospitalized with suspected or confirmed COVID-19 were initially screened. A total of 555 patients with PCR-confirmed infection were ultimately included for analysis. All participants underwent low-dose chest CT (LDCT) at admission. Pulmonary involvement was assessed using the chest CT severity score (CTSS) based on a unified protocol. CTSS values were analyzed in relation to ICU admission, in-hospital mortality, demographic data, oxygen saturation, dyspnea scores, and laboratory markers (CRP, LDH, lymphocyte, and neutrophil counts). Imaging was interpreted by board-certified radiologists under harmonized reporting standards. Results: CTSS values ≥13 and ≥15 were significantly associated with ICU admission and in-hospital mortality, respectively (p < 0.01). Strong correlations were observed between the CTSS and CRP, LDH, and dyspnea scores, with negative correlations to oxygen saturation and lymphocyte count. The standardized LDCT protocol ensured consistent image quality and minimized radiation exposure. Conclusions: LDCT combined with the CTSS provides a robust, reproducible, and radiation-sparing method for emergency risk stratification in COVID-19. Its high clinical utility supports deployment in frontline triage systems and future AI-enhanced diagnostic workflows.