Publication:

Predictors of failure of non-invasive ventilation in severe COVID-19 patients

Background Respiratory oxygenation index (ROX index) is a simple parameter that uses bedside parameters like respiratory rate and peripheral oxygen saturation. It has been extensively evaluated to predict success of HFNC in patients with pneumonia and found beneficial in predicting HFNC failure. However, there are very few literatures are available to establish the efficacy of ROX index in predicting NIV success. The aim of this study was to determine whether the ROX Index could predict NIV therapy success in patients with AHRF due to SARS-CoV-2 pneumonia.

Methods After obtaining approval and clearance from the institutional ethics committee of BMCRI, patients who were admitted in Victoria Hospital intensive care unit with acute respiratory failure due to COVID-19 who were initiated on NIV or CPAP based on clinician decision during the study period of April 2020 – July 2021 were included.

Results The optimal cut-off value of ROX index for predicting ventilatory requirement according to the maximum Youden index was ≤3.93. This cut-off value showed a sensitivity of 89.83% and a specificity of 92.86%, and the AUC was 0.965 (95% CI 0.909 to 0.992, P<0.0001).

Conclusions When compared between ROX index, SPO2 at Room air and CT Score, it was observed that ROX index had the highest, AUROC, sensitivity, and specificity. This indicates that ROX index can be used as a predictive variable for NIV failure or success.