Reduced Effective Oxygen Delivery and Ventilation with a Surgical Facemask Placed under Compared to over an Oxygen Mask: A Comparative Study

Objectives Consensus guidelines for perioperative anesthesia management during the COVID-19 pandemic recommend that patients wear a facemask in addition to their oxygen mask or nasal cannulae following tracheal extubation, where this is practical. The effects on effective oxygen delivery and ventilation of a surgical facemask under compared to over an oxygen (O2) mask are unclear. Design Single-center, comparative pilot study. Setting. Endoscopy procedure room at a major academic hospital. Subjects Five healthy anesthesiologists. Interventions. Using a carbon dioxide (CO2) sampling line positioned at the lips, the fraction of inspired O2 (FiO2), fraction of expiratory O2 (FeO2), expiratory end-tidal CO2 (EtCO2), and respiratory rate (RR) were measured under the following conditions: (1) a surgical facemask only, (2) a surgical facemask under an O2 mask, (3) an O2 mask only, and (4) a surgical facemask over an O2 mask. Measurements and Main Results. The sampled fractional expired oxygen (FeO2) at the lips was significantly lower when the surgical facemask was under compared to when over the O2 mask (27.9± 1.68 vs. 49.9 ± 6.27, p = 0.001), while there was no significant difference in inspired oxygen (FiO2). The sampled expiratory EtCO2 was significantly higher when the surgical facemask was under the O2 mask compared to when over the O2 mask (28.3 ± 8.5 vs. 23.5 ± 7.6, p = 0.026). The RR was not significantly different when the surgical facemask was under compared to over the O2 mask. Conclusions Effective oxygen delivery and ventilation was reduced (lower FeO2 and increased EtCO2) when a surgical facemask was placed under compared to over an O2 mask.


Introduction
Surgical patients routinely require supplementary oxygen (O 2 ) by facemask after undergoing anesthesia. Administration begins in the operating room, continues during transport to the recovery room, and extends for a period during recovery. Patients with COVID-19 undergoing surgery also commonly need supplementary O 2 for a period of time following anesthesia yet pose a risk of viral spread. A simple O 2 mask which is routinely used at the time of extubation and during the postoperative period has been shown to increase aerosolization of infectious viral particles that can be detected up to a distance of 0.4 meters at normal flow rates [1]. Given the infectious risk, consensus guidelines for anesthesia management during the COVID-19 pandemic recommend that patients wear a surgical facemask in addition to their O 2 mask or nasal cannulae following tracheal extubation, where practical [2]. An unanswered question is whether the facemask should be placed over an O 2 mask or if it should be worn underneath the O 2 mask and if there is any difference in O 2 delivery and carbon dioxide (CO 2 ) elimination between the two configurations. A published letter assessing a single individual indicates that inspired O 2 concentration may be equivalent with the two mask configurations; however, expired O 2 concentration and carbon dioxide levels were not assessed [3]. We hypothesized that a surgical facemask worn underneath the O 2 mask would both decrease the expired fraction of expiratory O 2 concentration and carbon dioxide (CO 2 ) levels. We sought to examine the various effects on oxygenation and ventilation of patients with the surgical facemask under or over the O 2 mask and showed for the first time the consequences of reduced fraction of expiratory oxygen (FeO 2 ) and increased expiratory end-tidal CO 2 (EtCO 2 ) with the surgical facemask under compared to over an O 2 mask in health volunteers. Answering these questions could be important for surgical patients following extubation and during recovery from anesthesia, especially those with underlying respiratory impairment while we continue to minimize risk for aerosolization.

Methods
e study cohort consisted of five healthy anesthesiologists from our institution that volunteered to participate in the study. We performed a power calculation based on the Blinks et al.'s [3] study and used an expected mean and standard deviation of the differences in the EtCO 2 of 15 and 6, respectively, and an alpha of 0.05 and a desired power of 0.8. e study received IRB exemption as a quality improvement project. Using a Drager Apollo Anesthesia machine in the endoscopy procedure room, we used a gas sampling line positioned at the lips to record the fraction of inspired oxygen (FiO 2 ), fraction of expiratory oxygen (FeO 2 ), expiratory end-tidal CO 2 (EtCO 2 ), and respiratory rate (RR) under the following conditions: (1) a surgical facemask only, (2) a surgical facemask under an O 2 mask, (3) an O 2 mask only, and (4) a surgical facemask over an O 2 mask. O 2 was delivered at 6 L/min via the oxygen mask. e subjects were under each condition for at least 2 minutes prior to making any measurements. For each condition, five individual measurements were recorded for FiO 2 , FeO 2 , EtCO 2 , and RR.
Descriptive statistics were performed and reported as a mean and standard deviation. Paired t-testswere used to perform comparisons between groups. A two-sided p value of <0.05 was used to denote statistical significance. All statistical analyses were performed using STATA ® release 14.2 (StataCorp LLC, College Station, TX, USA).  Figure 1 shows there was no significant difference in the sampled FiO 2 at the lips when the surgical facemask was under the O 2 mask compared to when the surgical facemask was over the O 2 mask (34.8 ± 6.5 vs. 35.7 ± 10.7, p � 0.885), as well as with the O 2 mask alone (34.8 ± 6.5 vs. 36.3 ± 10.2, p � 0.795).

Results
As expected, the sampled FiO 2 at the lips was significantly lower with the surgical facemask on room air compared to when the surgical facemask was over (19.7 ± 2.6 vs. 35.7 ± 10.7, p � 0.035) or under the O 2 mask (19.7 ± 2.6 vs. 34.8 ± 6.5, p < 0.01), as well as the O 2 mask alone (19.7 ± 2.6 vs. 36.3 ± 19.2, p � 0.040). Figure 2 shows that the sampled FeO 2 at the lips was significantly lower when the surgical facemask was under the O 2 mask compared to when the surgical facemask was over the O 2 mask (27.9 ± 1.68 vs. 49.9 ± 6.27, p � 0.001). e FeO 2 when the subject had only a surgical facemask was significantly higher compared to when the surgical facemask was under the O 2 mask (18.6 ± 1.1 vs 27.9 ± 1.68, p < 0.001). ere was no significant difference in sampled FeO 2 when the surgical facemask was over the O 2 mask compared to the O 2 mask alone (49.9 ± 6.3 vs. 46.1 ± 5.2, p � 0.864). Figure 3 shows that the sampled expiratory EtCO 2 was significantly higher when the surgical facemask was under the O 2 mask compared to when the surgical facemask was over the O 2 mask (28.3 ± 8.5 vs. 23.5 ± 7.6, p � 0.026) and the O 2 mask alone (28.3 ± 8.5 vs. 23.3 ± 7.8, p � 0.010). ere was no significant different in expiratory EtCO 2 with the surgical facemask under the O 2 mask compared to the surgical facemask alone (28.3 ± 8.5 vs. 24.9 ± 8.6, p � 0.36).

RR Is Unchanged Regardless of Surgical Facemask and O 2 Mask Configuration.
e RR was not significantly changed when the surgical facemask was under compared to over the O 2 mask (14.2 ± 2.8 vs. 13.9 ± 4.9, p � 0.88), as well as the O 2 mask only (14.2 ± 2.8 vs. 14.8 ± 3.1, p � 0.70).

Discussion
e ability to provide supplemental oxygen and ensure adequate ventilation is essential for patients recovering from anesthesia and surgery. e COVID-19 pandemic has challenged providers to meet patients' needs for adequate gas exchange while minimizing the risks of infectious aerosolization. While guidelines recommend that patients wear a facemask in addition to their oxygen mask or nasal cannula, there is limited literature on the most effective way of achieving oxygenation and ventilation goals. Our study examines the various effects on oxygenation and ventilation of patients with the surgical facemask under or over the O 2 mask and shows for the first time the consequences of reduced FeO 2 and increased expired EtCO 2 with the surgical facemask under compared to over an O 2 mask in healthy volunteers. ese effects may have significant consequences, particularly in patients with marginal respiratory reserve and the ability to tolerate apneic episodes without significant consequences during the postoperative period such as during transport.
A small number of prior investigations [3][4][5][6][7] have examined the effects of the face mask or nasal cannulae in combination with a surgical mask on oxygen delivery. ese prior studies have been limited by small numbers of subjects, lack of repeated measurements, and focus on specific measures of gas exchange. An initial investigation by Binks et al. measured the FiO 2 at the lips in a healthy volunteer breathing 6 L/ min oxygen via an (Hudson) O 2 mask placed over the top of a surgical mask and breathing 6 L/min oxygen via an O 2 mask placed underneath a surgical mask. e FiO 2 measured for the 2 mask configurations were 0.50 and 0.54, respectively, and the authors concluded that the O 2 mask can be worn over a surgical facemask without compromising the FiO 2 . It is important to note that this study did not evaluate differences in the effects on FeO 2 between the 2 mask configurations which we found were reduced with the mask under rather than over the O 2 mask despite no significant change in measured FiO 2 . e reduced FeO 2 with the surgical facemask under compared to over the O 2 mask is likely the result of the relatively reduced effective flow rate of oxygen and increased entrainment of air resulting in reduced effective O 2 delivery to the lungs. Of note, it is the FeO 2 rather than the FiO 2 that reflects the oxygen stores within the lung and increases the safety factor if apnea occurs [4].
Matsui et al. compared the impact on oxygenation and ventilation in a single subject wearing an oxygen mask above a surgical mask and wearing a nasal cannula below a surgical mask [5]. e authors found that the partial pressure of oxygen in arterial blood (PaO 2 ) decreased when changing from the nasal cannula at 4 L/min to the oxygen mask 4 L/min (from 154 mmHg to 108 mm Hg), recovering once again when returning to the nasal cannula. In contrast, the partial pressure of carbon dioxide in arterial blood (PaCO 2 ) and respiratory rate remained almost unchanged. e authors concluded that oxygen delivered via a nasal cannula worn under a surgical mask might prevent the spread of infection while simultaneously allowing maintenance of a high PaO 2 in patients. Notably, their investigation was limited by the single subject design and lack of repeated measurements.
Montiel et al. examined the impact of placing surgical facemasks on patients that were receiving a high-flow nasal cannula (HFNC) for hypoxemic respiratory failure due to COVID-19 [6]. Among the 21 patients studied, the investigators found the PaO 2 increased from 59 (±6) to 79 mmHg (±16) (p < 0.001) when a surgical facemask was placed over HFNC. e investigators postulated that the improvement in oxygenation could be explained not only by an increased oxygen concentration under the mask but also by a decrease of room air entrainment that is known to dilute the gas     [7]. e investigators found that regardless of the oxygen flow rate, the FiO 2 was higher when wearing a surgical facemask over an oxygen mask compared to wearing a surgical mask under a surgical mask. In our study, we found that FiO 2 was not significantly changed when the surgical facemask was placed under the O 2 mask compared to over the O 2 mask; however, we used a simple O 2 mask, whereas it appears Hamada et al. used a nonrebreather O 2 mask as they were able to achieve nearly an FiO 2 of 80% when the surgical facemask was over their O 2 mask in their study. To our knowledge, our study is the first to examine the impact on oxygenation and ventilation of asurgical facemask placed over compared with under the O 2 mask configurations among multiple subjects with repeated measurements. Importantly, our study demonstrated that ventilation was less effective with the surgical mask under the oxygen mask as indicated by the increase in CO 2 . It might be postulated that the increase in CO 2 would be more pronounced in patients with less effective ventilation (e.g., lung disease or residual anesthetic effects) than what was seen in the healthy volunteers in our study. Our study could also have implications for nonsurgical patients in other areas of the hospital.
Our study was limited by the small number of subjects included, use of healthy subjects only, and fixed oxygen flow rate. Our results might be more pronounced in patients with underlying lung disease or those with respiratory impairment due to the residual effects of anesthetics, analgesics, and paralytics. A more prolonged duration of time with a given mask configuration might also have differing effects.

Conclusions
Effective oxygen delivery and ventilation were impaired (reduced FeO 2 and increased EtCO 2 ) when a surgical facemask was placed under compared to over an O 2 mask. ese findings may have important implications for ensuring adequate gas exchange for patients wearing both oxygen and surgical masks to minimize viral aerosolization. Further studies which include a larger number of participants and those with limited respiratory drive or reserve are needed to more fully evaluate the clinical significance of surgical and oxygen mask configurations on oxygen delivery and ventilation.

Data Availability
All data points are plotted in the figures and are available on request.

Additional Points
Take Home Message.
e COVID-19 pandemic has challenged providers to meet patients' needs for adequate gas exchange while minimizing the risks of infectious   aerosolization. While guidelines recommend that patients wear a facemask in addition to their oxygen mask or nasal cannula, there is limited literature on the most effective way of achieving oxygenation and ventilation goals. Our study examines the various effects on oxygenation and ventilation of patients with the surgical facemask under or over the O 2 mask and shows for the first time the consequences of reduced FeO 2 and increased expired EtCO 2 with the surgical facemask under compared to over an O 2 mask in health volunteers. ese effects may have significant consequences particularly in patients with reducedrespiratory drive or reserve and the ability to tolerate apneic episodes without significant consequences during the postoperative period such as during transport.
Ethical Approval e study received IRB exemption as a quality-improvement project.

Consent
All participants consented to publication.

Conflicts of Interest
e authors declare no conflicts of interest.