695 - Comparison of Pressure and Bubbling Sound Properties Across Five Simulated Bubble CPAP Devices

Publication Number: 4679.695

Jasmine Chapdelaine, McGill University Faculty of Medicine and Health Sciences, Montreal, PQ, Canada; Sarah T. Wu, McGill University Faculty of Medicine and Health Sciences, Pointe-Claire, PQ, Canada; Amanda Gross, McGill University Faculty of Medicine and Health Sciences, Montreal, PQ, Canada; Emily R. Campbell, Research Institute- McGill University Health Centre, Monteal, PQ, Canada; Robert Kearney, McGill University Faculty of Medicine and Health Sciences, Montreal, PQ, Canada; Guilherme Santanna, McGill University Health Center, MONTREAL, PQ, Canada; Wissam Shalish, McGill University Health Center, Montreal, PQ, Canada

Background: Bubble continuous positive airway pressure (bCPAP) is a mode of non-invasive respiratory support characterized by the delivery of pressure oscillations through a water-submerged expiratory limb. As air escapes, bubbles are created and the resulting sounds have been shown to correlate with delivered pressure in a single-device simulation study. However, considerable variability exists among bCPAP device designs, which may alter the properties of both sounds produced and pressure delivered.

Objective: (A) Describe pressure and sound properties across 5 commercially available bCPAP devices; (B) Evaluate and model the relationship between bubbling sounds and pressure across these devices.

Design/Methods: Each bCPAP system was connected to a neonatal manikin via a standardized prong interface. Pressure was measured by a catheter in the expiratory tubing and bubbling sounds were recorded by a microphone near the water canister (Fig.1). Pressure (5-8 cm H2O) and flow (6, 8, 10 L/min) settings were varied for a total of 12 combinations, each recorded for 2 minutes in random order and repeated 5 times per device to ensure replicability. Pressure (mean), pressure oscillation (standard deviation), and sound amplitude (root mean square [RMS]) metrics were computed. Metrics were compared across devices using the Kruskal-Wallis test and visualized with heatmaps. The relationship between pressure and sound metrics was plotted and evaluated using Pearson correlation coefficients (r). Individual linear regression models modeled this relationship (with adjusted coefficient of determination [R2]).

Results: Pressure and sound metrics were found to differ significantly across all 5 devices and all 12 setting combinations (p < 0.05). Across pressure and flow setting combinations, the calculated metrics revealed highly variable pressure, pressure oscillation, and sound amplitude patterns between and within each device (Fig. 2). Across all devices, a strong relationship was found between sound amplitude and pressure metrics (Fig. 3A and 3B). Individual linear regression models predicted both mean pressure and pressure oscillations from system settings and sound recordings (Fig. 3C and 3D).

Conclusion(s): Across tested bCPAP devices, pressure and sound characteristics varied markedly. Nonetheless, sound and pressure metrics were strongly correlated, enabling accurate prediction of measured pressure from bubbling acoustics for each device. This study confirms the potential for acoustic-based monitoring across commonly used bCPAP devices and highlights the need for clinical validation.

Figure 1: Experimental Setup
A 36-week neonatal manikin with anatomically correct upper airways was connected via a standardized prong interface (Babi.Plus nCPAP kit, size 2 prongs) to each of the 5 bCPAP devices: Fisher & Paykel Bubble CPAP, Seattle PAP Plus (Dräger), Vayu Bubble CPAP, the original Bubble Babi.Plus (Respirologics) and its new iteration with an additional 7.3 mm hole in the expiratory limb (New Babi.Plus). Bubbling sounds were recorded with a lavalier microphone at the level of the water canister. A Mikro-Tip (Millar) pressure catheter placed into the proximal expiratory tubing measured the pressure into the system, as depicted by the raw pressure signals for each device.

Figure 2: Heatmaps Comparing Sound and Pressure Metrics
(A) For each CPAP setting at a set flow of 8 L/min, sound amplitude and mean pressure measured at the expiratory limb (y-axis) are compared across the five tested devices (x-axis). (B) For each flow setting at a set CPAP of 6 cmH2O, sound amplitude and pressure oscillations (y-axis) are presented across devices. The horizontal color gradient, with yellow representing lower values and dark blue representing higher values, allows for comparison across devices. The vertical gradient, from white to opaque, depicts the variable patterns across system settings.

Figure 3: Relationship between Pressure and Sound Metrics
For the trials at a flow of 8 L/min, mean pressure measured at the expiratory limb and sound amplitude were plotted and exhibited strong correlations across all devices (r = 0.832-0.995). Most devices demonstrated positive correlations, whereby an increase in sound relates to greater pressure delivery, except for the negative correlation of the Vayu. Similarly, Figure 3B illustrates the strong correlations (r = 0.973-1.000) between pressure oscillations and sound amplitude, for the trials at a pressure of 6 cmH2O. All devices displayed positive correlations, with the exception of Seattle PAP Plus, suggesting that greater bubbling sounds relate to less oscillations. Individual linear regression models were created for each device to predict both mean pressure (C) and pressure oscillations (D) from set CPAP level, set flow level, and bubbling sound amplitude. These metrics were accurately estimated (R2 = 0.647-0.996) across devices.