Mechanisms of nasal high flow therapy in newborns

Abstract

In newborns, it is unclear how nasal high flow (NHF) generates positive airway pressure. In addition, the reported benefits of NHF such as reduction in work of breathing may be independent of airway pressure. The authors hypothesized that during NHF the area of leak and the flow determine airway pressure and that NHF can reduce the required minute ventilation to maintain gas exchange. In response to NHF, pressure was measured in the upper airways of 9 newborns and ventilation was measured in another group of 17 newborns. In a bench model, airway pressures were measured during NHF with different prong sizes, nare sizes, and flows. The airway pressures during 8 L/min NHF were greater when a larger cannula versus a smaller cannula was used (P < 0.05). NHF reduced minute ventilation in 16 of 17 neonates, with a mean decrease of 24% from a baseline of 0.66 L/min (SD 0.21) (P < 0.001), and was unrelated to changes in airway pressure; arterial oxygen saturation by pulse oximetry ($\text{S}{\text{p}}_{{\text{O}}_2}$) and tissue CO₂ were unchanged. In the bench model, the airway pressure remained <2 cmH₂O when <50% of the "nare" was occluded by the prongs. As the leak area decreased, because of a smaller nare or a larger cannula, the airway pressure increased exponentially and was dependent on flow. In newborns NHF using room air substantially reduced minute ventilation without affecting gas exchange irrespective of a decrease or an increase of respiratory rate. NHF generates low positive airway pressure that exponentially increases with flow and occlusion of the nares.NEW & NOTEWORTHY In healthy newborns, nasal high flow (NHF) with room air reduced minute ventilation by one-fourth without affecting gas exchange but, in contrast to adults, produced variable response in respiratory rate during sleep. During NHF, pressure in the upper airways did not exceed 2 cmH₂O at 8 L/min (3.4 L·min^-1·kg^-1) and was unaffected by opening of the mouth. NHF can generate higher pressure with larger prongs that decrease the leak around the cannula or by increasing the flow rate.

Keywords: CPAP; nasal high flow; neonate; ventilation; work of breathing.

Fig. 1.

Raw recordings of the pressure (cmH₂O) in the nasal cavity of a spontaneously breathing neonate during no therapy, nasal high flow (NHF) at 8 L/min via smaller and larger cannulas, and bubble continuous positive airway pressure (bCPAP) set to 5 cmH₂O. Recordings were obtained with mouth closed and mouth open during each therapy.

Fig. 2.

Bar graphs showing the mean (SD) airway pressure (A) and dynamic range of pressure (peak expiratory pressure − peak inspiratory pressure) (B) during no therapy, bubble continuous positive airway pressure (bCPAP) set to 5 cmH₂O, and nasal high flow (NHF) set to 8 L/min via a larger cannula (NHF L) and a smaller cannula (NHF S) with mouth open and mouth closed. *Significant difference vs. bCPAP; P < 0.05; ^#significant difference vs. NHF L, P < 0.05. NS, not significant.

Fig. 3.

Ventilation parameters during the application of nasal high flow (NHF) at 8 L/min in sleeping neonates. NHF reduced minute ventilation (A) in all neonates, but the respiratory rate (B) and tidal volume (C) responses were variable. The group means are shown as bar graphs, and individual data points are shown by lines. NS, not significant.

Fig. 4.

Relationship between change in respiratory rate (RR, breaths/min) and change in minute ventilation (V̇e, L/min) in response to nasal high flow of 8 L/min in 17 neonates. A greater decrease in respiratory rate was associated with a greater reduction in V̇e.

Fig. 5.

Results from the bench experiment. Top: raw recordings of the chamber pressure (cmH₂O) and breathing flow (L/min) in the bench model; the gray line indicates recordings during no nasal high flow (NHF), and the black line indicates recordings during NHF at 8 L/min via the smaller cannula (A) or the larger cannula (B). Bottom: positive end-expiratory pressure (PEEP, cmH₂O) and dynamic range of pressure (maximum − minimum, cmH₂O) at different flow rates and nare diameters. The symbols indicate the raw data, and the lines indicate the exponential growth curve created from the data. The figure demonstrates that the pressure and pressure swings increase as the area of “nare” occlusion by the NHF cannula increases.

Fig. 6.

The graph shows the change in airway pressure, %, that may occur with a change in the leak area (%), as predicted by Poiseuille’s law, Q = πr⁴P/8ηL, rearranged for pressure (P) as the solution and assuming that length (L) and fluid viscosity (η) remain the same across scenarios and are therefore removed so that the solution is dependent on the radius (r) and flow (Q). The law predicts that a high leak will result in low airway pressure regardless of the flow rate. As the area of the leak is reduced, the airway pressure will increase exponentially. For the same prong-to-nare ratio the airway pressure will be greater in the smaller nare because of a smaller leak area. Circles represent diameters of prongs and “nares,” and the gray area between the circles represents leak.