LAUSR

Background and Aims: High-flow nasal oxygen (HFNO) in COVID-19 related acute hypoxemic respiratory failure can reduce the need for intubation, but it is important to identify which patient will fail HFNO to prevent delay for patient intubation. Methods: This is a retrospective cohort study of which 42 confirmed COVID-19 patients were recruited. All subjects was using HFNO and treated at Persahabatan Hospital from March to June 2020. Demographic, clinical and laboratory data before HFNO, vital sign and respiratory index after 24 hours of HFNO was recorded. Bivariate analysis was used to find difference between success or failed outcome. Multivariate analysis was performed to predict HFNO failure. Results: Majority of patients were male (73%), mean age 58.76 years with hypertension and diabetes as the most common comorbidity. HFNO Success outcome is 38.1%. Statistically significant difference (p <0.05) between success and failure group was observed on respiratory rate (23.12 vs 29.85), pulse rate (90.44 vs 100.27), oxygen saturation (98 vs 93), PaO2/Fio2 (139.27 vs 69.68), SpO2/FiO2 (117.58 v 100) and ROX index (4.97 vs 3.32). Vital sign and respiratory index measured after 24 hours of HFNO significantly differed between success and failed outcome on day 7. COX regression analysis showed that ROX index can predict HFNO failure (hazard ratio 0.265, p=0.005, CI 0.105-0.664). Optimal cut off value cannot be determined because of small sample size. Conclusion: Vital sign and respiratory index are significantly worse in HFNO failure group. ROX index can predict HFNO failure.