Abstract Rationale Nasal high flow (NHF) is an established form of respiratory support that delivers heated, humidified air with or without supplemental oxygen via a nasal cannula at flow rates that typically exceed the patient’s peak inspiratory flow. NHF is a flow-controlled therapy in which the generated pressure cannot be directly measured or monitored; it varies depending on flow rate, upper airway anatomy, mouth position, and breathing pattern. This study was designed to predict the generated pressure based on leak area and flow, to support individualized treatment strategies. Method Twelve healthy participants were enrolled. Each had their nares photographed and underwent magnetic resonance imaging (MRI) of the nasopharynx. Anatomically accurate nasal models were 3D-printed from the MRI scans and attached to a standard nasopharynx in a closed-mouth configuration. Positive end-expiratory pressure (PEEP) was measured in bench-top experiments using symmetrical and asymmetrical NHF cannula interfaces at flow rates ranging from 10 to 60 L/min. Nares images were analyzed to calculate the leak area around the prongs using digital image analysis (ImageJ, National Institutes of Health, Bethesda, USA). Based on bench-top data, a simplified equation was derived to estimate PEEP: PEEP = Q2/Aleak1.5 where Q is the flow rate in L/min, and Aleak is the total cross section leak area of the nares in mm2. This equation was tested using both symmetrical and asymmetrical cannulas (Optiflow and Optiflow Duet, Fisher & Paykel Healthcare, New Zealand). Results Only three models exceeded a mean root mean square error (MRMSE) of 1.5 cmH2O during testing with the asymmetrical interface, and two with the symmetrical interface. The predicted PEEP showed good agreement with measured pressures across both symmetrical and asymmetrical cannula types. The mean bias was −0.53 cmH2O, with 95% limits of agreement ranging from −3.38 to 2.33 cmH2O. The overall MRMSE was 0.85 cmH2O (Figure 1), indicating a low average deviation between predicted and measured values. These results support the reliability of the prediction model for estimating airway pressure under different flow conditions and cannula designs. Conclusion A simple equation combining nares leak area, measured by simple digital image analyses, and flow rate may enable estimation of PEEP during NHF. This abstract is funded by: Fisher & Paykel Healthcare
Vieira et al. (Fri,) studied this question.