PEEP Generated by High-Flow Nasal Cannula in a Pediatric Model

Abstract

Background:

High-flow nasal cannulas (HFNC) have been increasingly used in pediatric critical care settings. Different mechanisms have been theorized as to how HFNC reduces work of breathing. The likely primary mechanism by which HFNC reduces work of breathing is by generating positive end-expiratory pressure (PEEP). However, there are limited data assessing the PEEP delivered by moderate gas flows (8 to 50 L/min), which are used most commonly in pediatric patients.

Methods:

Pediatric upper airway models were created with a 3D printer and were connected to an ASL 5000 lung simulator (version 3.5, IngMar Medical, Pittsburgh, Pennsylvania). Respiratory system mechanics for patient body weights of 4 kg to 75 kg were simulated by the ASL 5000. Age/weight-specific flows were delivered via five Teleflex Comfort Flo HFNC devices (Research Triangle Park, NC) using a high flow generating system with a Teleflex Neptune humidifier. Pressure throughout the simulated airway were measured at HFNC flows of 6 L/min to 60 L/min with 25%, 50%, and 75% air leak to simulate open-mouth breathing.

Results:

PEEPs of 1.5 to 36 cm H₂O were delivered by HFNC flows of 8 to 50 L/min. In general, for each specific cannula, increasing the flow and decreasing the air leak resulted in higher levels of PEEP (p <0.001 and > 10% difference). Using appropriate flows for each patient model's weight, increasing model size from 4 kg to 15 kg was positively associated with increasing levels of PEEP (p <0.001 and > 10% difference). However, the 20 kg, 35 kg, and 75 kg models generated less PEEP than the 15 kg model, but the 75 kg model generated more PEEP than either the 20 or 30 kg models (p <0.001 and > 10% difference). Airway and alveolar pressure (PEEP) trended similarly.

Conclusions:

High flow nasal cannulas deliver a varying amount of PEEP at the alveolar level with flows of 8 to 50 L/min. With each specific sized cannula and model weight, increasing flow and decreasing leak resulted in the generation of greater PEEP. PEEP levels differed across cannulas and model weights at the same leak level, likely partially related to differences in the nasal interface between the HFNC device and the model nares.