Failure of Humidified High-Flow Nasal Cannulae: A Retrospective Analysis

Background: The use of Humidified High-Flow Nasal Cannulae (HHFNC) as an alternative mode of non-invasive ventilation in neonates has become widespread in neonatal practice. Identification of factors contributing to the failure of HHFNC is therefore important.

Objective: To report clinical and demographic characteristics of the infants that fail treatment with HHFNC.

Methods: The records of all infants admitted to the neonatal unit of King’s College Hospital (London, UK) over 2 years from 01/01/2015 to 31/12/2016 who went onto HHFNC were reviewed. Information was taken from hospital notes and the electronic patient data system BadgerNet. Comparison was made between the babies who failed on HHFNC and those who successfully weaned off HHFNC. Failure was defined as the need to switch to continuous positive airway pressure (CPAP) or intubation and mechanical ventilation. Success was defined as weaning off all respiratory support or to supplemental oxygen delivered by low flow nasal cannulae. In cases where the babies had multiple episodes of treatment with HHFNC, only the first episode was reviewed.

Results: Two hundred and twenty-six infants were identified; data were obtained for 154 . Of these, 134 were suitable for use. Infants that failed HHFNC had a lower mean birthweight compared to those who succeeded (1376g versus 1603g respectively, p=0.035), higher mean FiO₂ when starting HHFNC (0.34 vs 0.28 respectively, p=0.004), and more frequently had a positive blood culture at the time of starting HHFNC (9% vs 2% respectively, p=0.001). Multivariate regression analysis of birth weight, blood culture and FiO₂, using failure of HHFNC as the outcome, revealed that the mean FiO₂ at the time of starting HHFNC was statistically significant (p<0.05), whilst the birthweight and the incidence of positive blood culture were not.

Conclusion: Failure of HHFNC is more common among infants with a higher oxygen requirement at the point when HHFNC is started.