Nasal continuous positive airway pressure (nCPAP), widely used in neonatal intensive care is also more and more used in infants with severe acute bronchiolitis. There is no evidence that nCPAP improves outcome, but several studies showed that it reduces work of breathing (WOB), improves gas exchange and decreases intubation rate. High flow nasal prong (HFNP) therapy, also used in neonatal units, has recently been suggested for bronchiolitis management. This technique allows warmed and moist gas administration protecting nasal mucosa, and preventing variations of inspired gas FiO2. HFNP also creates a continuous positive pressure, reducing WOB, and seems much more comfortable and better tolerated by babies than nCPAP. There are no studies comparing both techniques for bronchiolitis management. The main objective of this study is to compare WOB with nCPAP and HFNP therapy in infants with severe bronchiolitis in pediatric intensive care unit. Secondary purpose is to compare efficacy and tolerance of both techniques. We propose to lead an open, single center, crossover randomized study. Each patient will receive the two treatments. Three consecutive periods will be studied: first one (minimum 2 hours), with the first technique CPAP or HFNP according to randomization (treatment 1); second period (2 hours) with the second technique (treatment 2); Third period (6 hours) with the second technique. The primary endpoint will be the comparison of WOB (estimated by the calculation of the transdiaphragmatic pressure-time product) at the end of the two firsts periods. Efficacy and tolerance of each technique will be evaluated and compared by compared measuring respiratory rate, heart rate, FiO2, SpO2, transcutaneous PCO2, modified Wood score, and EDIN score during the third period. Nine subjects by groups are required.

Inclusion Criteria:

• Infants less than 3 months
• Infants hospitalized in the pediatric intensive care unit of the hospital in northern Marseille for an episode of acute viral bronchiolitis, defined by the French consensus conference of 2000 (1) clinical signs of infection of the upper airways with tachypnea, brake expiratory and / or wheezing auscultation, signs of struggle
• Infants with acute respiratory distress score with Wood changed> 3 and requiring ventilatory Support
• Infants whose parents or legal guardians have accepted their participation in the research and signed informed consent.
• With Social Security

Exclusion Criteria:

• - Clinical condition requiring immediate intubation before the start of treatment to the:

  • severity of respiratory distress: FiO2> 60% to maintain SpO2> 92%
  • hemodynamic instability
  • the occurrence of apneas désaturantes (respiratory pauses greater than 20 seconds duration associated with desaturation <80% and / or bradycardia <80/min)
  • the presence of neurological disorders with altered consciousness

• Presence of co-morbidities:

  • chronic respiratory failure
  • heart