• General Paediatrics
• Respiratory

We read with interest the recent article by Martin and Kneyber: Is there a role for high-flow nasal cannula oxygen therapy to prevent endotracheal intubation in children with viral bronchiolitis?1 Respiratory support using high flow nasal cannula oxygen (HFNCO) in severe bronchiolitis has been used in various settings including a general paediatric ward.2

We would like to share our experience of using HFNCO for the last 3 years (2010–2011, 2011–2012 and 2012–2013) in a paediatric ward, working to a strict protocol. Infants with bronchiolitis who met the criteria (SaO₂<90% in FiO₂ of 0.4, RR≥60 breaths/min) were offered a trial of HFNCO. Patients were carefully monitored and observations were documented on a Paediatric Early Warning Score chart. Escalation to other interventions such as nCPAP or invasive ventilation was not delayed.

A total of 45 patients between the ages of 5 days and 15 months satisfied the criteria for HFNCO. Eighteen children had potentially significant comorbidities. There were six well ex-preterm babies, two with chronic lung disease and on home oxygen, four with Trisomy 21 (two with repaired congenital heart disease), three with global developmental delay, two with underlying congenital heart disease and one with congenital hypotonia.

Eleven children required escalation of respiratory support (5 out of 27 infants with no comorbidities and 6 out of 18 with significant comorbidities). Five (three previously well and two with comorbidities) required intubation within the first 2 h of commencing HFNCO. Six required nCPAP, but two of these were subsequently ventilated. Four of the six needing nCPAP were weaned on to HFNCO within 12–24 h. Thirty-four patients continued on HFNCO.

There was a decrease in heart rate (median 171 to 136) and respiratory rate (median 79 to 53) and improvement in PH (median of 7.32 to 7.38) and PCO₂ (median of 7.7 to 6.6 kPa), within 4 h of initiating HFNCO.

Infants spent a median length of 48 h on HFNCO (18 h to 7 days). Decrease in oxygen requirements was also noted in the first 4 h. Most infants tolerated nasogastric feeds and four required intravenous fluids. No adverse effects were identified.

We conclude that in our experience this simple technique has decreased the number of unstable infants that were previously transferred to a PHDU for nCPAP3 (criteria for initiating HFNCO were same as those for nCPAP). Patients were managed locally on the paediatric ward with minimal extra nursing resources.

HFNCO can be safely delivered on a paediatric ward.