Study dataset.

Patients with mild-to-moderate coronavirus disease 2019 (COVID-19)-associated acute respiratory distress syndrome (ARDS) can be treated with a high-flow nasal cannula (HFNC). The use of the respiratory rate-oxygenation (ROX) index, calculated as the ratio of oxygen saturation (SpO2)/fractional oxygen (FiO2) to respiratory rate, in the first few hours after HFNC initiation can help identify patients who fail HFNC therapy later. However, few studies have documented the use of the ROX index during the period of HFNC therapy. Therefore, we aimed to demonstrate the diagnostic performance of the ROX index when calculated throughout the HFNC therapy period and to determine the best cut-off point for predicting HFNC failure. We conducted a retrospective study of patients with COVID-19-associated ARDS who commenced HFNC at the Faculty of Medicine Vajira Hospital, Navamindradhiraj University, Thailand, between April 1 and August 30, 2021. We calculated the ROX index every 4 h throughout the HFNC therapy period and defined HFNC failure as a subsequent endotracheal tube intubation. The performance of the ROX index was analyzed using the area under the receiver operating characteristic curve (AUC). We applied the ROX index ≤ 4.88 to predict HFNC failure and obtained a new ROX cut-off point using Youden’s method. In total, 212 patients with COVID-19 treated with HFNC were included in the study. Of these, 81 patients (38.2%) experienced HFNC failure. The ROX index ≤ 4.88 demonstrated a reasonable performance in predicting HFNC failure (AUC, 0.77; 95% confidence interval [CI], 0.72–0.83; p<0.001). However, compared with the original cut-off point of ≤ 4.88, the new ROX index cut-off point of ≤ 5.84 delivered optimal performance (AUC, 0.84; 95% CI, 0.79–0.88; p<0.001), with a significantly better discriminative ability (p = 0.007). In conclusion, a ROX index ≤ 5.84 was found to be optimal for predicting HFNC failure in patients with COVID-19-associated ARDS.