Abstract
OBJECTIVE: To characterize real-world high-flow nasal cannula use and outcomes. METHODS: A retrospective observational study using an electronic health record registry of all adult patients in a five-hospital system between 2017 - 2025. We identified all high-flow nasal cannula episodes, defined as periods of high-flow nasal cannula use containing breaks no longer than 6 hours. We describe key measurements, including high-flow nasal cannula episode duration, intubation rates, and death or hospice discharge rates; secondary outcomes included intensive care unit admission, lengths of stay, and discharge location. We used adjusted hierarchical logistic regression to evaluate variation across hospitals. RESULTS: 28,269 high-flow nasal cannula episodes from 17,519 individual patients over 19,313 hospitalizations were identified. Average high-flow nasal cannula use increased from 176 episodes/month in 2017 to 269 episodes/month by 2024 (p < 0.001). Median episode duration was 13.9 [interquartile range 4.3 - 36] hours, 24.5% of episodes were followed by intubation, and 29.3% by death or hospice discharge; 83.6% of high-flow nasal cannula use was escalation respiratory therapy, with the remainder used within 24 hours of extubation. Illness severity, admitting service, hospital, and unit type were associated with the odds of intubation and the duration of high-flow nasal cannula episodes; 16% of the variation in mortality and 20% of the variation in intubation were attributable to hospital-level variation. CONCLUSION: High-flow nasal cannula is used in multiple hospital settings and contexts. The use of high-flow nasal cannula has increased between 2017 and 2025. Outcomes of high-flow nasal cannula use are influenced by both patient characteristics and the contexts in which it is used.