O0441 Oct.5: 4:31pm - Standardized HFNC Protocol for Bronchiolitis with Lower Targeted SpO2 Goal (90-95%) Decreases Hospital and ICU Length of Stay

Background
The AAP recommends targeting oxygen saturation (SpO2) ≥90% for bronchiolitis. High flow nasal cannula (HFNC) is an effective therapy for bronchiolitis, but the effect of applying a specific SpO2 target to infants and children receiving HFNC for bronchiolitis is not known. We conducted a retrospective cohort study of a quality improvement initiative that instituted a standardized SpO2 goal (90-95%) to a non-standardized SpO2 goal to determine the impact on outcomes among critically ill children with bronchiolitis.

Methodology
A standardized protocol of HFNC with SpO2 targets for management of bronchiolitis was implemented at UC Davis Children’s Hospital in March 2019. We retrospectively studied patients hospitalized 1/1/18-12/31/18 as the “non-standardized SpO2” group and 3/17/19-2/17/20 as the “standardized SpO2” group targeting SpO2 90-95%. The months between the two periods were excluded due to a brief pilot protocol in December 2018. Children ≤36 months with bronchiolitis were included if they required HFNC at any point in their hospitalization. Exclusion criteria were prematurity, and cardiac, non-bronchiolitis respiratory, or neurologic diseases. The primary outcome was hospital length of stay (LOS). Secondary outcomes were ICU LOS, duration of oxygen and respiratory support, and adverse events (escalation of respiratory support to CPAP/BiPAP/intubation, death). Data were extracted from the Electronic Medical Record by UC Davis Health Data Provisioning Core using automated data extraction algorithm validated via chart review. Descriptive statistics and group differences were computed using linear regression, density plots and kernel density estimates using the adaptive kernel density estimator and Silverman optimal bandwidth. We estimated between-group differences conditional on calendar month of admission given the seasonality of bronchiolitis.

Results
Demographics (Table) did not differ between the standardized (N=128) and non-standardized SpO2 groups (N=100). Hospital and PICU LOS were shorter in the standardized SpO2 group, 2.2 vs 3.2 days (p<0.001) and 1.3. vs 2.3 days (p<0.001), respectively (Table). Respiratory support on HFNC with FiO2>0.21 was >15 hours shorter in the standardized SpO2 group compared to non-standardized group (p<0.001). The non-standardized SpO2 group spent >22 hours longer (p<0.001) on oxygen when SpO2 was >95% (labeled as over-treated in table). These differences remained significant even after controlling for seasonality of bronchiolitis based on months. Distribution graph of SpO2 relative to fraction of inspired oxygen (SpO2/FiO2 ratio, Figure) showed the standardized SpO2 group was managed with FiO2=0.21 for longer duration, reflecting ability to wean to room air sooner on the standardized protocol. There was no difference in risk of escalation to higher respiratory support between the groups and no deaths.

Conclusion
A standardized HFNC protocol for treatment of bronchiolitis targeting an SpO2 range of 90-95% was associated with decreased hospital and ICU LOS, and lower FiO2 use without an increase risk in adverse events.