Over recent years, the use of high flow nasal cannula in the treatment of bronchiolitis in infants has increased. Whilst it used to be mainly used in PICU, it is now widely used in EDs and on the wards. The recent PARIS trial examined whether delaying starting high flow in infants with bronchiolitis led to a worse outcome (it didn’t). See Alasdair Munro’s excellent analysis here.
But is high flow actually useful in these patients, and if so when? Should we be using it in our Emergency Departments at all?
The PREDICT research group published an updated systematic review this month in the Journal of Paediatrics and Child Health.
O’Brien S, Craig S, Babl FE, Borland ML, Oakley E, Dalziel SR; Paediatric Research in Emergency Departments International Collaborative (PREDICT) Network, Rational use of high-flow therapy in infants with bronchiolitis. What do the latest trials tell us?’ A Paediatric Research in Emergency Departments International Collaborative perspective, J Paediatr Child Health. 2019 Jul;55(7):746-752.
The basics
High flow nasal cannula (HFNC) is a method of delivering humidified, heated gas with flow rates through the nasal passages at higher rates than is achieved through standard nasal oxygen. It can be delivered at 2-3L/min to a max of 60L/min, compared to standard oxygen therapy which is at 2-3L/min.
The idea is that HFNC reduces upper airway resistance and provides some positive pressure.
What papers were examined?
The authors included 34 papers since 2015, which included three RCTs.
What RCTs have been done on HFNC in bronchiolitis?
There are three since 2015 and one prior to this in 2014 (1891 patients in total). Three were conducted on inpatients/ED patients, and one is on PICU patients.
There are some barriers to drawing conclusions as they varied so much in which patients they included and how they assessed outcomes.
Patients:
- Franklin et al – included if sats <92-94%
- Kepreotes et al – excluded if sats <90% and included if sats>94%
- Campana et al – used a non-validated clinical score (which includes sats) but no data about what the sats were
Flow rates and oxygen:
- Franklin et al – 2L/kg
- Kepreotes et al – 1L/kg
- Campana et al – 6-8L/min (so approximately 1L/kg but not specifically defined)
Outcomes:
- Franklin et al – treatment failure (i.e. escalation of care)
- Kepreotes et al – time needing oxygen, and escalation of care (slightly different definition to Franklin et al)
- Campana et al – non-validated clinical score as primary outcome
Milesi et al included infants with bronchiolitis requiring PICU admission and randomised them to HFNC or nCPAP
What did the RCTs conclude about HFNC?
Franklin et al and Kepreotes et al reported fewer treatment failures (although read this post as to why that isn’t necessarily of clinical significance).
Importantly the RCTs found no difference in PICU admissions, intubation rates, length of time on oxygen, or length of stay between those on HFNC and those on standard oxygen therapy.
However, a number of these patients had failure of treatment on standard oxygen therapy and were rescued to high flow (61% out of the 200 who needed rescue HFNC were successfully rescued).
Two thirds of infants in these studies did not require escalation of care to HFNC.
Is HFNC safe?
Milesi et al and Campana et al reported no adverse events.
Kepreotes et al reported four (two in each arm): one HFNC and two standard therapy patients had a brief period without oxygen when the tubing disconnected; and one patient on HFNC inhaled some condensation from the circuit.
Franklin et al reported two pneumothoraces (one in each arm).
What are the author’s recommendations on when to use HFNC?
There are times where babies with bronchiolitis can have physiological deterioration due to discomfort, mucus plugging, hunger, or hypoxia. These types of deterioration may result in a medical review, but often resolve themselves and are transient. Occasionally the deterioration is a true clinical deterioration.
The definition of treatment failure in these papers relied on various measures, but these differed between RCTs. It is clear there is a lack of consensus on when we need to initiate rescue therapy.
The authors of this systematic review recognise that early warning scores are helpful, but also importantly remind us that a response to standard oxygen therapy won’t be immediate, and in fact an improvement in physiological parameters in response to the initiation of oxygen therapy should be seen in four hours.
Bottom line
- Infants with bronchiolitis and hypoxia should be started on 2L oxygen via nasal cannula.
- HFNC should only be started if there is deterioration after this has been administered.
- There is no evidence for using HFNC for work of breathing in infants with no hypoxia.
- There is no evidence for using HFNC as an early treatment for bronchiolitis in ED.
