Effect of Face Masks on Gas Exchange in Healthy Persons and Patients with Chronic Obstructive Pulmonary Disease.

To the Editor:

Current evidence, from observational studies to systematic reviews and epidemiologic modeling, supports the use of masks by the public, especially surgical masks, for mitigating coronavirus disease (COVID-19) transmission and deaths (1–5). However, public mask use has been heavily politicized with inconsistent recommendations by authorities leading to divided public opinion. Despite evidence to the contrary, an online UK/U.S. survey found that only 29.7–37.8% of participants thought that wearing a surgical mask was “highly effective” in protecting them from acquiring COVID-19 (6). Another reason commonly argued against mask use involves safety concerns, as mask discomfort has been attributed to rebreathing CO2 and hypoxemia, with some even considering that masks are lethal (7).

Objective

To evaluate whether gas exchange abnormalities occur with the use of surgical masks in subjects with and without lung function impairment.

Methods and Findings

To demonstrate the changes in end-tidal CO2 and oxygen saturation as measured by pulse oximetry before and after wearing a surgical mask, we used a convenience sample of 15 house staff physicians without lung conditions (aged 31.1 ± 1.9 yr, 60% male) and 15 veterans with severe chronic obstructive pulmonary disease (COPD) (aged 71.6 ± 8.7 yr, forced expiratory volume in 1 second [FEV1] 44.0 ± 22.2%, 100% male). The patients needed to have a postbronchodilator FEV1 <50% and FEV1/forced vital capacity <0.7 and were enrolled from the pulmonary function laboratory during a scheduled 6-minute walk test ordered to assess the need for supplemental oxygen. In our institution, the 6-minute walk tests are done with arterial blood analysis before and immediately after the walk to assess the need for long-term oxygen. Because of the COVID-19 pandemic, the 6-minute walk tests are done with subjects using a surgical mask. As this was a clinical observation study, exemption from the local institutional review board was obtained. Baseline measures on room air without a mask were performed noninvasively using a Life Sense monitor (model Ls1-9R; Nonin Medical), followed by continuous monitoring using a surgical mask.

At 5 and 30 minutes, no major changes in end-tidal CO2 or oxygen saturation as measured by pulse oximetry of clinical significance were noted at any time point in either group at rest (Table 1). With the 6-minute walk, subjects with severe COPD decreased oxygenation as expected (with two qualifying for supplemental oxygen). However, as a group, subjects with COPD did not exhibit major physiologic changes in gas exchange measurements after the 6-minute walk test using a surgical mask, particularly in CO2 retention.

Table 1.

Physiologic parameters measured in study volunteers before and after surgical mask use

	HR	RR	SpO2	ETCO2	Po2 (mm Hg)	Pco2 (mm Hg)
	(× min−1)	(× min−1)	(%)	(mm Hg)
Healthy house staff*
Baseline values without mask on room air	72.5 ± 10.0 (61 to 88)	17.2 ± 2.5 (13 to 21)	97.5 ± 1.2 (95 to 100)	36.2 ± 2.3 (31 to 40)	NP	NP
Change from baseline
Surgical mask at 5 min at rest	0.50 ± 3.8 (−5.0 to 6.0)	0.52 ± 0.5 (0 to 2.0)	−0.28 ± 0.7 (−1 to 1)	1.06 ± 1.1 (0.8 to 2.5)	NP	NP
Surgical mask at 30 min at rest	−0.64 ± 5.4 (−9.0 to 10.0)	1.13 ± 1.4 (−1.0 to 3.0)	0.10 ± 0.6 (−1 to 1)	0.75 ± 1.0 (0.6 to 2.1)	NP	NP
Subjects with severe COPD†
Baseline values without mask on room air	86.0 ± 9.8 (72 to 103)	20.5 ± 5.4 (12 to 35)	91.3 ± 5.8 (89 to 97)	36.1 ± 3.3 (29.7 to 43.6)	77.2 ± 15.1 (54.4 to 99.6)	38.9 ± 5.6 (30.6 to 50.2)
Change from baseline
Surgical mask at 5 min at rest	−1.83 ± 4.1 (−8.0 to 5.0)	−0.12 ± 3.8 (−6.0 to 4.0)	0.35 ± 1.4 (−1.0 to 3.0)	−1.67 ± 3.9 (−12.0 to 6.0)	NP	NP
Surgical mask at 30 min at rest and before 6MWT	−2.13 ± 6.4 (−14.0 to 8.0)	1.03 ± 5.0 (−7.0 to 9.0)	0.87 ± 2.1 (−2.0 to 5.0)	−1.63 ± 4.3 (−10.0 to 4.0)	NP	NP
Surgical mask after 30 min rest and after 6MWT	11.00 ± 12.0 (−3.0 to 39.0)	3.30 ± 6.2 (−3.0 to18.0)	−2.28 ± 7.3 (−19.0 to 5.0)	−0.14 ± 5.6 (−9.0 to 7.0)	−4.6 ± 13.9 (−15.9 to 27.8)	0.97 ± 2.4 (−2.8 to 5.3)

Definition of abbreviations: 6MWT = 6-minute walk test; COPD = chronic obstructive pulmonary disease; ETCO = end-tidal carbon dioxide; HR = heart rate; NP = not performed; Pco2 = carbon dioxide tension/partial pressure; Po2 = oxygen pressure/tension; RR = respiratory rate; SpO = oxygen saturation as measured by pulse oximetry.

Values shown as mean ± standard deviation (range).

N = 15.

N = 15.

Discussion

Although we did not measure changes in tidal volume or minute ventilation, these data find that gas exchange is not significantly affected by the use of surgical mask, even in subjects with severe lung impairment. Our results agree with a prior observation on 20 healthy volunteers using a surgical mask for 1 hour during moderate work rates, in which mild increases in physiological responses also deemed to be of no clinical significance were observed (8). The discomfort felt with surgical mask use has been ascribed to neurological reactions (increased afferent impulses from the highly thermosensitive area of the face covered by the mask or from the increased temperature of the inspired air) or associated psychological phenomena such as anxiety, claustrophobia, or affective responses to perceived difficulty in breathing (8). These findings are in contrast to the use of N-95 masks, in which carbon dioxide tension/partial pressure (Pco2) may increase in lung-healthy users but without major physiologic burden (9).

We did not intend to compare healthy versus diseased individuals but rather aimed to assess the effect of using a surgical mask in two distinct populations. We acknowledge that our observations may be limited by sample size; however, our population offers a clear signal on the nil effect of surgical masks on relevant physiological changes in gas exchange under routine circumstances (prolonged rest, brief walking). Other studies studying surgical or N-95 masks have used samples of 6–10 individuals per group (8–11). We focused on subjects with severe COPD because they are at a higher risk of CO2 retention compared with subjects with COPD of milder severity or other pulmonary conditions. As shown, we observed a small drop in oxygen pressure/tension in this group, expected based on their disease severity, but not a rise in Pco2 after walking. An ideal setting would have been to allow these individuals to walk without a mask; however, because of the current epidemic, this was not allowed in our institution at the time of the evaluation. The nature of our veteran population precluded us from enrolling women with COPD; however, we do not expect major sex-related physiologic responses when using a surgical mask.

It is important to inform the public that the discomfort associated with mask use should not lead to unsubstantiated safety concerns as this may attenuate the application of a practice proved to improve public health. As growing evidence indicates that asymptomatic individuals can fuel the spread of COVID-19 (12), universal mask use needs to be vigorously enforced in community settings, particularly now that we are facing a pandemic with minimal proven therapeutic interventions. We believe our data will help mitigate fears about the health risks of surgical mask use and improve public confidence for more widespread acceptance and use.