Abstract
COVID-19 patients often present with rapidly progressing acute hypoxemic respiratory failure, requiring orotracheal intubation with different prognostic issues. However, ICU specialists lack predictive tools to stratify these patients. We conducted a single-center cross-sectional retrospective study to evaluate if the ROX index, measured under non-invasive oxygenation support, can predict ICU mortality in a COVID-19 intubated patient cohort. This study took place in the division of intensive care at the Geneva University Hospitals (Geneva, Switzerland). We included all consecutive adult patients treated by non-invasive oxygenation support and requiring intubation for acute respiratory failure due to COVID-19 between 9 September 2020 and 30 March 2021, corresponding to the second local surge of COVID-19 cases. Baseline demographic data, comorbidities, median ROX between H0 and H8, and clinical outcomes were collected. Overall, 82 patients were intubated after failing a non-invasive oxygenation procedure. Women represented 25.6% of the whole cohort. Median age and median BMI were 70 (60-75) years and 28 (25-33), respectively. Before intubation, the median ROX between H0 and H8 was 6.3 (5.0-8.2). In a multivariate analysis, the median ROX H0-H8 was associated with ICU mortality as a protective factor with an odds ratio (95% CI) = 0.77 (0.60-0.99); p < 0.05. In intubated COVID-19 patients treated initially by non-invasive oxygenation support for acute respiratory failure, the median ROX H0-H8 could be an interesting predictive factor associated with ICU mortality.