Woodhead 2006.

Methods	Randomised crossover trial
Participants	30 infants admitted to neonatal intensive care unit, intubated, planned to extubate to HFNC
Interventions	Randomised to one modality for 24 hours after extubation then switched to other modality Humidified HFNC ‐ VapothermTM (mean 3.1 L/min); n = 15 Non‐humidified HFNC (mean 1.8 L/min); n = 13
Outcomes	Need for intubation (no pre‐specified criteria); nasal mucosa examination; pneumothorax or pneumomediastinum
Notes	Data from the first crossover period only were included in the review Flow rates differed significantly between interventions Funding source unclear
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number table, stratified by weight
Blinding (performance bias and detection bias) Need for intubation	High risk	No set criteria for intubation
Blinding (performance bias and detection bias) Nasal damage	Low risk	'Masking' of intervention; unclear how effective
Other bias	Low risk