ABSTRACT
High-flow nasal oxygen (HFNO) is recommended as a first-line treatment in patients with acute hypoxemic respiratory failure due to COVID-19. We assessed the effectiveness of HFNO and predictors of failure and death. The medical records of 200 consecutive adult patients treated with HFNO were analysed. Ninety-two patients (46%) were successfully cured, 52 (26%) required noninvasive ventilation, and 61 (30.5%) received intubation. Overall mortality was 40.5%. Risk factors of HFNO ineffectiveness were: SpO2 ≤ 90% with conventional oxygen therapy (HR 0.32, 95% CI 0.19-0.53, p < 0.001), SpO2 ≤ 74% without oxygen therapy (HR 0.44, 95% CI 0.27-0.71, p < 0.001), an age ≥ 60, comorbidities, biomarkers (C-reactive protein, procalcitonin, creatinine, lactate dehydrogenase), duration of symptoms before admission to hospital ≤ 9 days, start of treatment with HFNO ≤ 4 days. The multivariate logistic regression models (age ≥ 60, comorbidities, C-reactive protein concentration and SpO2 with oxygen therapy) revealed a high predictive value of death and HFNO failure (AUC 0.851, sensitivity 0.780, specificity 0.802; AUC 0.800, sensitivity 0.776, specificity 0.739, respectively). HFNO is a safe method for treating acute hypoxemic respiratory failure, with effectiveness reaching nearly 50%. Low values of SpO2 without and during oxygen therapy seem to be good diagnostic tools for predicting death and HFNO failure.