Background High‐flow nasal cannula (HFNC) therapy is widely used for children with bronchiolitis, but its optimal role remains uncertain. Our institution created and later revised a clinical pathway guiding HFNC initiation and weaning. Methods A retrospective review of 1690 bronchiolitis encounters was conducted. Trends in the duration of HFNC and hours spent weaning HFNC as proportions of the monthly hospital length of stay (LOS) for bronchiolitis, hospital LOS, and escalation of care were compared using interrupted time series (ITS) models across three study periods: Baseline (HFNC managed at provider discretion), Intervention 1 (pathway with initiation at 0.5 L/kg/min and escalation up to 2 L/kg/min), and Intervention 2 (revised pathway, initiation at the maximum rate of 2 L/kg/min). Both pathway iterations provided titration and weaning guidance. Maximum respiratory scores were used to adjust for case severity. Results After adjustment for severity and time, both HFNC duration and HFNC weaning time (as a proportion of monthly LOS) decreased at the start of Intervention 1, but subsequently increased. During Intervention 2, both these measures trended downward, returning to baseline. Total LOS did not change in the baseline or intervention periods. Escalation of care did not differ from baseline to the end of Intervention 2. Conclusion Initiating HFNC at higher flow rates with weaning guidance for children hospitalized with bronchiolitis was associated with a reduction in HFNC duration without differences in LOS or escalation of care. These findings suggest that standardization through clinical pathways can limit HFNC duration in bronchiolitis.

BACKGROUND: High-flow nasal cannula therapy (HFNC) is increasingly used for children with bronchiolitis, but its optimal role remains relatively uncertain. Our institution created and later revised a ward-based clinical pathway guiding initiation and weaning of HFNC. METHODS: A retrospective review of 664 encounters was conducted. Total duration of HFNC, duration of weaning from maximum flow rate, hospital length of stay (LOS), and escalation of care were compared across 3 study periods using generalized linear models: Baseline (HFNC managed at provider discretion, 5/1/2015-4/30/2017, n = 215), Intervention 1 (first pathway, initiation at 0.5 L/kg/min and escalation up to 2 L/kg/min, 5/1/2017-2/28/2018, n = 155), and Intervention 2 (revised pathway, initiation at the maximum rate of 2 L/kg/min, 3/1/2018-1/31/2020, n = 294). Both pathway iterations provided specific titration and weaning guidance. Models controlled for initial SpO2 and maximum respiratory rate to adjust for the possibility of differing case severity. RESULTS: After adjustment for severity, total duration of HFNC was significantly reduced in the second intervention, but not the first. Adjusted LOS was shorter in Intervention 1 and Intervention 2 compared to the baseline period. Adjusted weaning time and escalation of care did not differ significantly between periods. CONCLUSION: Pathway implementation was associated with modestly decreased LOS and – when initial flow rates were increased to 2 L/kg/min – reduced duration of HFNC, without differences in escalation of care or weaning time. These findings highlight the potential role of standardization of practice with clinical pathways in safely reducing unnecessary care.