Predictors of failure in infants with viral bronchiolitis treated with high-flow, high-humidity nasal cannula therapy*

Pediatr Crit Care Med. 2012 Nov;13(6):e343-9. doi: 10.1097/PCC.0b013e31825b546f.


Objectives: Viral bronchiolitis is an acute infection and inflammatory disease of the respiratory tract, with infants typically presenting with the most severe symptoms. Medical management of bronchiolitis is mostly supportive. Several preliminary studies suggest potential benefit from the use of high-flow nasal cannula systems. Although high-flow nasal cannula is a well-established modality in the newborn intensive care unit, its use in the pediatric intensive care unit for acute respiratory failure is far less established. The objective of this study was to identify any laboratory and clinical variables that may predict high-flow nasal cannula failure in management of bronchiolitis in the pediatric intensive care unit.

Design: The study design was a retrospective chart review of all patients admitted to the pediatric intensive care unit from 2006 to 2010 with a diagnosis of viral bronchiolitis. Inclusion criteria included the initiation of high flow nasal cannula therapy at the time of admission and age ≤ 12 months. Exclusion criteria were intubation prior to admission, age >12 months, and the presence of a tracheostomy.

Patients: A total of 113 patients with viral bronchiolitis met the inclusion criteria.

Setting: Academic free standing Children's Hospital in the Midwest.

Interventions: Retrospective chart review.

Measurements and main results: The data were analyzed by comparing those patients who responded to high-flow nasal cannula (n = 92) with those who were nonresponders to high-flow nasal cannula and required intubation (n = 21). No differences were noted between the groups for age, sex, or ethnicity. Mean weight and weight-for-corrected-age percentiles were significantly lower for patients who failed high-flow nasal cannula (p = .016 and .031, respectively), but weight-for-corrected-age percentile was not significant in logistic regression controlling for other variables. Respiratory rate prior to the initiation of high-flow nasal cannula also correlated strongly with respiratory deterioration (p < .001). The PCO2 was significantly higher for both before (p < .001) and after (p < .001) initiation of therapy in the nonresponder group. Pediatric Risk of Mortality III scores for the patients who failed high-flow nasal cannula were significantly higher (p < .001) than those of patients who tolerated this therapy.

Conclusions: History of prematurity and the patient's age did not increase a patient's risk of failure. Nonresponders to high-flow nasal cannula therapy were on the onset, more hypercarbic, were less tachypnic prior to the start of high-flow nasal cannula, and had no change in their respiratory rate after the initiation of high-flow nasal cannula therapy. Nonresponders had higher pediatric risk of mortality scores in the first 24 hrs.

MeSH terms

  • Body Weight
  • Bronchiolitis, Viral / therapy*
  • Carbon Dioxide / blood
  • Catheterization
  • Critical Care*
  • Female
  • Humans
  • Humidity
  • Infant
  • Infant, Newborn
  • Infant, Premature
  • Logistic Models
  • Male
  • Nose
  • Oxygen / blood
  • Oxygen Inhalation Therapy*
  • Predictive Value of Tests
  • Respiratory Rate
  • Retrospective Studies
  • Severity of Illness Index
  • Treatment Failure


  • Carbon Dioxide
  • Oxygen