Traditionally, pulse oximetry had enabled limited and indirect respiratory monitoring. Because such devices measure only peripheral oxygen saturation, their use created the potential for delaying complication detection, with possible subsequent health risks for the patient. For example, pulse oximetry is unable to directly detect hypoventilation or apnea, especially in patients undergoing procedural sedation while receiving supplemental oxygen [4] [5].

A superior monitoring approach involves the breath-to-breath measurement of the concentration of carbon dioxide (CO₂) in exhaled respiratory gas, which has gained ready acceptance, particularly with endorsement from the American Society of Anesthesiologists (ASA) for use of end-tidal capnography (EtCO₂) as a standard of care for moderate and deep procedural sedation [6] [7].

Although capnography has greater efficiency than pulse oximetry for effective detection of hypoventilation and apnea, accurate and consistent measurements of the EtCO₂ during minimally invasive procedures under deep sedation have historically been challenging [8]. This difficulty results from the capnography port of the nasal cannula being open to air, causing atmospheric gases to be entrained and sampled [9]. Additionally, delivery of supplemental oxygen to patients, particularly at flows >5 liters per minute (L/min), causes a “wash-out” or dilution of the sample of exhaled CO₂ and results in either a falsely low reading or no reading at all [10].

Recent prospective randomized controlled trials (RCTs) report up to 54% of all patients experience severe hypoxemia secondary to sedation-related UAO and respiratory depression [11]. Although passive oxygenating devices can provide higher concentrations of oxygen, they are incapable of generating positive pressure to maintain airway patency. Continuous Positive Airway Pressure (CPAP) equipment has been shown to relieve UAO by creating a pneumatic stent [12]. However, their utility is limited by the machine’s very large size and relatively greater expense, and the high oxygen flows required to maintain pressure also dilute EtCO₂ sampling [13] [14].

A recent RCT comparing the SuperNO₂VA^TM nasal PAP ventilation device (Vyaire Medical, Inc., United States) vs. nasal cannula with capnography during deep sedation documented a significantly higher minute ventilation and reduction in the incidence of severe hypoxemia in the SuperNO₂VA^TM nasal PAP ventilation device cohort compared to the nasal cannula with capnography cohort [15]. However, the design of the SuperNO₂VA nasal PAP ventilation device had the disadvantage of being unable to capture EtCO₂, especially in patients who exhale from their mouths, which also results in false apnea alarms.

A solution that offers the ability to monitor EtCO₂ and deliver supplemental oxygen is the novel SuperNO₂VA^TM Et Nasal Mask (Vyaire Medical, Inc., United States). This completely sealed nasal PAP device provides positive pressure to maintain upper airway patency without the use of capital equipment. The SuperNO₂VA^TM Et Nasal Mask (Figure 1) also is designed to capture EtCO₂ exhaled from both the patient’s mouth and nose. Combining capnography with positive pressure in a single device may prove to be a methodology to further improve patient outcomes in deep sedation as opposed to passive oxygenation techniques with capnography.

The objectives of this study were to validate the capability of the SuperNO₂VA^TM Et to capture EtCO₂ exhaled from the nose and the mouth, provide 20 cm H₂O positive pressure, quantify leak rates, and summarize the performance testing compared to a predicate device.

A simulated patient setup was used to compare the accuracy of CO₂ measurements within the SuperNO₂VA Et Nasal Mask and a predicate device, the Smart CapnoLine® Plus, Adult/Intermediate CO₂ Oral-Nasal Set (Medtronic, United States).

The Device Under Test (DUT), either the SuperNO₂VA Et or Oral-Nasal Set, was placed on a face surrogate and breathing simulation was provided by a Large Animal Volume Controlled Ventilator Model 613 (Harvard Apparatus, United States). This device is suitable for humans up to 50 kg (110 lb) and enables an adjustable VT from 30 to 700 milliliters (ml) per stroke and an adjustable respiratory rate from 7 to 50 breaths per minute (BPM). The concentration of CO₂ flowing through the surrogate nose and mouth was set using a digitally controlled flow meter and CO₂ source, and verified using a CO₂ monitor (Dräger Narkomed 6400). A Datex-Ohmeda 5250 Respiratory Gas Anesthesia Monitor (General Electric Healthcare, United States) connected to the EtCO₂ sampling port was used to monitor CO₂. Testing assessed eight combinations of Input CO₂ (1% ± 0.25%; 5% ± 0.5%); breath rate and VT (12 BPM/500 ml; 20 BPM/300 ml); and O₂ flow rates (1 L/min; 5 L/min). Table 1 lists the combinations of Input CO₂, Breath Rate/VT, and O₂ Flows that were tested. After a 3-min stabilization period to reach steady-state, the CO₂ waveform of the sensor connected to the EtCO₂ sampling port was recorded for 16 seconds via an analog port of an oscilloscope (Tektronix TBS2000, United States).

To evaluate the performance of the Oral-Nasal Set and SuperNO₂VA Et when a patient is breathing exclusively nasally or orally, the same set of eight tests were repeated while simulating nasal breathing and oral breathing. Three trials were performed for each of the eight test conditions and breathing type (i.e., nasal or oral).

In addition, leak rate and ability to hold a positive pressure for five minutes were tested for three SuperNO₂VA Et Nasal Masks and, as a comparator, a full-face anesthesia mask (Ventlab^TM inflatable anesthesia mask VR5100; SunMed, United States). The DUT was placed on a surrogate face and sealed with 10 pounds of force. To determine the leak flow rate, the O₂ flow rate was slowly reduced until a minimum flow was achieved while still maintaining a positive pressure of 20 cm H₂O.

Absolute and relative errors between the CO₂Max, defined as maximum CO₂ during the 16-second trial, and the Input CO₂ were quantified for each DUT.

A b s o l u t e E r r o r = C O 2 M a x D U T − I n p u t C O 2

R e l a t i v e E r r o r = ( C O 2 M a x D U T − I n p u t C O 2 ) I n p u t C O 2 ∗ 1 0 0 % .

Negative errors correspond to an underestimation of CO₂. Unpaired t-tests compared CO₂Max errors between the two devices for tests with Input CO₂ of 1% and 5%. Unpaired t-tests were also performed to compare CO₂Max errors between the two devices at O₂ Flows of 1 L/min and 5 L/min. Paired t-tests were performed to compare CO₂Max errors during Nasal Breathing and Oral Breathing trials for each of the two devices. As a comparator, DUT accuracy was measured against the specifications of the International Organization for Standardization (ISO 80601-2-55:2018) requirements for the basic safety and essential performance of a respiratory gas monitor intended for continuous operation with a patient, defined as ± (0.43%vol + 8% of gas level) [16].

The SuperNO₂VA Et Nasal Mask had lower CO₂Max errors than the Oral-Nasal Set for all eight conditions (Figure 2).

For 1% Input CO₂, CO₂Max errors were significantly larger for the Oral-Nasal Set, −0.12%Vol ± 0.03%Vol (−12.2%Vol ± 3.3%Vol, mean ± SD), compared to the SuperNO₂VA Et Nasal Mask, −0.01%Vol ± 0.02%Vol (−1.3%Vol ± 2.2%Vol) (p = 0.0005). All 12 trials for the Oral-Nasal Set and the SuperNO₂VA Et Nasal Mask met the ISO accuracy specification.

For 5% Input CO₂, the Oral-Nasal Set significantly underestimated CO₂Max error, −0.93%Vol ± 0.16%Vol (−18.6%Vol ± 3.2%Vol), compared to the SuperNO₂VA Et Nasal Mask, −0.08%Vol ± 0.06%Vol (−1.5%Vol ± 1.2%Vol) (p < 0.0001). At 5% Input CO₂, eight of the 12 trials for the Oral-Nasal Set failed to meet the ISO accuracy specification, while all SuperNO₂VA Et Nasal Mask met the specification.

To examine the effect of O₂ Flow on performance of the two devices, results from trials with O₂ Flow of 1 L/min were compared to trials with O₂ Flow of 5 L/min (Figure 3). Trials with the SuperNO₂VA Et had significantly lower errors than the Oral-Nasal Set with O₂ Flows of 1 L/min (0.01%vol vs. 0.11%vol, p = 0.0032) and 5 L/min (−0.03%vol vs. −0.14%vol, p = 0.0032). The difference in performance was even larger with an Input CO₂ of 5%. Specifically, the SuperNO₂VA Et errors were significantly less at both 1 L/min (−0.04%vol vs. 0.91%vol, p < 0.0001) and 5 L/min (−0.11%vol vs. −0.95%vol, p = 0.0002).

The same set of eight tests were repeated while simulating Nasal Breathing and Oral Breathing for each of the two devices (See Figure 4). For the Oral-Nasal Set, CO₂Max measurements were significantly lower for the Oral Breathing compared to Nasal Breathing trials for Input CO₂ concentrations of 1% (paired t-test, p = 0.0005) and 5% (p = 0.0091). For the SuperNO₂VA Et, there was no

significant difference in CO₂Max measurements for Nasal Breathing and Oral Breathing trials for both Input CO₂ concentrations (1%: p = 0.33, 5%: p = 0.064). At an Input CO₂ of 5%, the Oral-Nasal Set had 10 out of the 12 Nasal Breathing trials and 9 out of 12 Oral Breathing trials outside of the ISO error bound (shaded green region).

Both the SuperNO₂VA Et Nasal Mask and the full-face anesthesia mask successfully held a pressure of 20 cm H₂O for three, 5-minute trials. The SuperNO₂VA Et Nasal Mask had a leak rate of 2.0 L/min for all three samples compared to the mean leak rate of 2.7 (range: 2.5 - 3.0 L/min) for the anesthesia mask (Table 2).