Association between oxygen saturation/fraction of inhaled oxygen and mortality in patients with COVID-19 associated pneumonia requiring oxygen therapy.

Background: Coronavirus disease 2019 (COVID-19) can manifest from asymptomatic to acute respiratory distress syndrome (ARDS). COVID-19 associated pneumonia develops into ARDS due to rapid progression of hypoxia. Although arterial blood gas analysis (ABGA) should be implemented to confirm this deterioration, it is not easy to obtain such tests in the COVID-19 environment. Therefore, this study was conducted to determine whether oxygen saturation (SpO2) and SpO2/fraction of inhaled oxygen (FiO2) (SF ratio) predicts ARDS and mortality. Methods: This was a retrospective cohort study that enrolled COVID-19 pneumonia patients requiring oxygen therapy from Feb 2020 to May 2020. Of 100 COVID-19 pneumonia cases, we compared 59 cases of pneumonia requiring oxygen, divided into ARDS and non-ARDS pneumonia requiring oxygen. The factors affecting mortality were investigated. Results: At the time of admission, the SpO2, FiO2, and SF ratios of the ARDS group were significantly different from those of the non-ARDS pneumonia requiring oxygen support group (P <0.001, respectively). With respect to predicting occurrence of ARDS, the SF ratio on admission and the SF ratio at exacerbation showed an overall area under the curve of 85.7% and 88.8% (P < 0.001, respectively). Multivariate Cox regression analysis identified the SF ratio at exacerbation (HR, 0.916; 95% CI, 0.846-0.991; P = 0.029) and National Early Warning Score (NEWS) (HR, 1.277; 95% CI, 1.010-1.615; P = 0.041) as significant predictors of mortality. Conclusions: The SF ratio on admission and the SF ratio at exacerbation can predict occurrence of ARDS. The SF ratio at exacerbation and NEWS has a significant effect on mortality.