Predictors of high-flow nasal cannula failure in pediatric patients with acute respiratory distress

BACKGROUND: Heated humidified high-flow nasal cannula (HFNC) has gained popularity recently and is considered a standard respiratory support tool for pediatric patients with acute respiratory distress. However, data are limited on the bedside parameters that can predict HFNC failure in pediatric patients. PURPOSE: To evaluate the performance of SpO(2)/FiO(2) (SF) ratio, pediatric respiratory rate-oxygenation (pROX) index, and clinical respiratory score (CRS), for predicting the HFNC outcomes. METHODS: This prospective observational study included 1- month to 15-year-old patients with acute respiratory distress who required HFNC support. The HFNC setting, vital signs, CRS, and treatment outcomes were recorded. Data were analyzed to determine the predictors of HFNC failure. RESULTS: Eighty-two children participated in the study, 16 of whom (19.5%) did not respond to HFNC treatment (failure group). Pneumonia was the main reason for intubation (62.5%). Predictors of HFNC failure at 12 hours were: SF index ≤166 (sensitivity, 62.5%; specificity, 87.8%; area under the curve [AUC], 0.75), pROX index <132 (sensitivity, 68.7%; specificity, 84.8%; AUC, 0.77), and CRS ≥6 (sensitivity, 87.5%; specificity, 96.9%; AUC, 0.92). CONCLUSION: The CRS was the most accurate predictor of HFNC failure in pediatric patients. A CRS ≥ 6 at 12 hours after HFNC initiation and pROX, a newly modified parameter, are helpful indicators of HFNC failure.