RESUMO
High-flow nasal cannula (HFNC) is a relatively recent therapy that has been used to treat respiratory failure. Until now, the criterion for failure requiring escalation to other forms of ventilatory support has remained unclear. This study evaluated how the ROX index predicts the success or failure of HFNC in infants with bronchiolitis. A prospective, observational, multicenter study was conducted in 2 pediatric ICUs. The data were collected at 7 moments. Patients were categorized into failure and success groups according to HFNC. A total of 102 infants were included, 18(17.6%) of whom failed HFNC therapy. For the ROX index, significant differences were observed between the failure 5.8(95%CI 4.7-7.1) and success 7.7(95%CI 7.2-8.2) groups (p = 0.005) at the 12 h evaluation. According to the analysis of the performance of the ROX index, the AUC at 12 h was 0.716(95%CI 0.591-0.842; p = 0.016). The best cutoff range for the ROX index at 12 h was 6.50-7.18, with a sensitivity of 42% and a specificity of 66% at the cutoff of 6.50, and a sensitivity of 92% and a specificity of 54% at the cutoff of 7.18. We concluded that the ROX index could be effective at predicting the failure of HFNC therapy in infants with bronchiolitis beginning at 12 h after installation.