2.2-Interventions
Participants who met inclusion criteria were invited to participate in the study, and written informed consent was obtained from the parent/guardian. Then, enrolled participants were assigned to either LFO or HFNCO by using simple randomization. On even-numbered days, patients received LFO, while on odd-numbered days received HFNCO. Masking of the treatment was not possible because of the visual difference between the two interventions.
Participants on LFO received supplemental oxygen via a simple face mask at a range of 6-10 L/min to maintain a SpO2 level between 92%-98%. Oxygen therapy weaned off when the patient sustained the SpO2 above 92% in the ambient oxygen concentration (21%). Participants on HFNCO received supplemental oxygen via Airvo 2 high-flow system (Fisher and Paykel Healthcare), using an age-appropriate Optiflow Junior cannula, at a 2 L*kg/min (between 8 L/min and 25 L/min). The fraction of inspired oxygen (FiO2) was set at 40%. Then, the FiO2was adjusted to obtain the SpO2 levels between 92%-98%. After the starting flow rate continued for a minimum of 4 h, it was decreased by 0.5 L*kg/min per hour according to the patient’s clinical response. When the FiO2 was equal to the ambient oxygen concentration (21%), and the flow rate was decreased to 0.5 L*kg/min, HFNCO was stopped. Our hospital did not have a standardized protocol for weaning off HFNCO therapy at the study time. Therefore, the lead investigator and other researchers created a new protocol to reduce any bias on wean off oxygen (Supplemental p3-7). Treatment failure was considered at 4 h when at least three of the following five criteria were present: i) An increase or no changes in HR relative to baseline; ii) An increase or no changes in RR relative to baseline; iii) Persistence of low SpO2 (<92%) in the three near and independent measurements despite receiving FiO2 >40% (HFNCO) and at a maximum 10 L/min (LFO); iv) An increase or no changes in CRS relative to baseline; v) Arterial blood partial pressure of carbon dioxide ≥60 mm/Hg. Children who failed on HFNCO were transferred to the ICU. Those who deteriorated on LFO were trialed HFNCO (at 2 L*kg/min) as a rescue treatment in the general ward. The patients who did not respond to HFNCO were transferred to the ICU. Fifteen days after discharge, we carried out follow-up phone interviews with parents or caregivers (Supplement Table 2; p8,9).
Participants were followed with a cardiorespiratory monitor and pulse oximetry. The baseline assessments including HR, RR (measured by counting one minute; and corrected according to age and body temperature20,21), SpO2 (determined by the value of pulse oximetry with an appropriate waveform sustaining for 3 s), CRS, blood pressure, and body temperature were performed in room air. Those measurements were recorded at the 1, 2, 4, 12, and 24 h; then were evaluated once a day until discharge. In addition, venous blood gas was obtained at baseline and 4 h. The multiplex real-time polymerase chain reaction was performed on nasopharyngeal swab specimens at admission to show the viral pathogens (respiratory syncytial virus-A and -B, rhinovirus, parainfluenza virus [1-4], influenza A, B, and A(H1N1)pdm09, human metapneumovirus, adenovirus, coronavirus [OC43, 229E, NL63, and HKU1], human bocavirus, and enterovirus).
The primary outcome was the time that HR and RR returned to the normal range for age and the time that CRS took to regress from severe to moderate bronchiolitis or from moderate to mild bronchiolitis. Secondary outcomes included changes in HR, RR, and CRS from randomization to an average of 96 h. We also analyzed the duration of oxygen therapy, LOS, treatment failure, ICU admission, and adverse events (AEs).