125 IN VITRO AND IN VIVO EVALUATION OF A NEONATAL HIGH-FLOW NASAL CANNULA SYSTEM

• Published in: Journal of investigative medicine 2007-01, Vol.55 (1), p.S97
• Main Authors: Lampland, A. L., Meyers, P. A., Worwa, C. T., Plumm, B. J., Mammel, M. C.
• Format: Article
• Language: English
• Subjects: Continuous positive airway pressure; Respiratory therapy
• ISSN: 1081-5589; 1708-8267

BackgroundHumidified high flow nasal cannulae (HHNC) are used in NICUs to deliver oxygen and continuous positive airway pressure. Few data exist regarding these systems.ObjectiveAnalyze pressures delivered by the Fisher & Paykel RT329 HHNC system at varying gas flows both in vitro with varying cannula sizes and air leaks and in vivo in a neonatal piglet model.MethodsIn vitro: Pressure and flow derived at different points in the RT329 system were measured with a flow analyzer (Fluke Biomedical, Everett, WA) at 0%, 30%, and 50% leaks. Fixed variables included initial flow (0.5-6 lpm) and cannula size. Continuous variables included pressure and flow at various points in the system. In vivo: We measured esophageal pressures (Pes) generated at flow levels ranging from 1 to 6 lpm in seven newborn piglets.ResultsIn vitro: Pressure delivery within the system was limited by the pop-off valve (40 cmH2O); flow delivery was preserved to the point of nasal cannula insertion. Pressure and flow delivery at the cannula end was limited at > 2 lpm. With all cannula sizes, introduction of a 30% and 50% leak resulted in > 96% and > 98% decrease in delivered pressure, respectively. With all cannulae, removal of the pop-off valve resulted in a direct increase in pressure delivery as the flow was increased: 2 lpm > 80 cmH2O, 3 lpm > 140 cmH2O, 4 lpm > 210 cmH2O, 5 lpm > 220 cmH2O, and 6 lpm > 280 cmH2O. In vivo: As HHNC flow was increased from 1 lpm to 6 lpm, mean Pes increased, but extreme variation suggests a variable leak (figure values are mean ± SD).ConclusionIn vitro using the RT329 system with no leak, we saw high pressures, only limited by the pop-off valve, at flows > 2 lpm. Addition of a 30 to 50% leak at the cannula end dramatically reduced delivered pressures to < 2 cmH2O, regardless of initial flow. In vivo, we saw variable but generally increasing Pes as flow increased. Because of variable leaks, cannula size, and patient variation, it may be impossible to predict pressure delivery during HHNC use without direct measurement.