Objectives: To 1) describe patterns of use of high-flow nasal cannula therapy, 2) examine differences between patients started on high-flow nasal cannula and those started on noninvasive ventilation, and 3) explore whether patients who failed high-flow nasal cannula therapy were different from those who did not.
Design: Retrospective analysis of data collected prospectively by the Paediatric Intensive Care Audit Network.
Setting: All PICUs in the United Kingdom and Republic of Ireland (n = 34).
Patients: Admissions to study PICUs (2015-2016) receiving any form of respiratory support at any time during PICU stay.
Interventions: None.
Measurements and main results: Eligible admissions were classified into nine groups based on the combination of the first-line and second-line respiratory support modes. Uni- and multivariate analyses were performed to test the association between PICU and patient characteristics and two outcomes: 1) use of high-flow nasal cannula versus noninvasive ventilation as first-line mode and 2) high-flow nasal cannula failure, requiring escalation to noninvasive ventilation and/or invasive ventilation. We analyzed data from 26,423 admissions; high-flow nasal cannula was used in 5,951 (22.5%) at some point during the PICU stay. High-flow nasal cannula was used for first-line support in 2,080 (7.9%) and postextubation support in 978 admissions (4.5% of patients extubated after first-line invasive ventilation). High-flow nasal cannula failure occurred in 559 of 2,080 admissions (26.9%) when used for first-line support. Uni- and multivariate analyses showed that PICU characteristics as well as patient age, primary diagnostic group, and admission type had a significant influence on the choice of first-line mode (high-flow nasal cannula or noninvasive ventilation). Younger age, unplanned admission, and higher admission severity of illness were independent predictors of high-flow nasal cannula failure.
Conclusions: The use of high-flow nasal cannula is common in PICUs in the United Kingdom and Republic of Ireland. Variation in the choice of first-line respiratory support mode (high-flow nasal cannula or noninvasive ventilation) between PICUs reflects the need for clinical trial evidence to guide future practice.