RESUMO
Introduction: NIV (Non-invasive ventilation) and HFNC (High Flow nasal cannula) are being used in patients with acute respiratory failure. HACOR score has been exclusively calculated for patients on NIV, on other hand ROX index is being used for patients on HFNC. This is first study where ROX index has been used in patients on NIV to predict failure. Aim of the study: This study investigates the comparative diagnostic performance of HACOR score and ROX index to predict NIV failure. Methods: We performed a retrospective cohort study of non-invasively ventilated COVID-19 patients admitted between 1st April 2020 to 15th June 2021 to ICU of a tertiary care teaching hospital located in Central India. We assessed factors responsible for NIV failure, and whether these scores HACOR/ROX index have discriminative capacity to predict risk of invasive mechanical ventilation. Results: Of the 441 patients included in the current study, 179 (40.5%) recovered, while remaining 262 (59.4%) had NIV failure. On multivariable analysis, ROX index > 4.47 was found protective for NIV-failure (OR 0.15 (95% CI 0.03-0.23; p<0.001). Age > 60 years and SOFA score were other significant independent predictors of NIV-failure. The AUC for prediction of failure rises from 0.84 to 0.94 from day 1 to day 3 for ROX index and from 0.79 to 0.92 for HACOR score in the same period, hence ROX score was non-inferior to HACOR score in current study. DeLong's test for two correlated ROC curves had insignificant difference expect day-1 (D1: 0.03 to 0.08; p=3.191e-05, D2: -0.002 to 0.02; p = 0.2671, D3: -0.003 to 0.04; p= 0.1065). Conclusion: ROX score of 4.47 at day-3 consists of good discriminatory capacity to predict NIV failure. Considering its non-inferiority to HACOR score, the ROX score can be used in patients with acute respiratory failure who are on NIV.