Patient and Health Care Worker Perceptions of Communication and Ability to Identify Emotion When Wearing Standard and Transparent Masks

Key Points

Question  Could transparent masks help to overcome communication barriers associated with widespread mask use among the general population, general health care workers, and health care workers who are deaf or hard of hearing in the United States?

Findings  In this survey study of 1000 members of the general public, 123 general health care workers, and 45 health care workers who are deaf or hard of hearing, participants perceived mask wearing as potentially impairing communication. Respondents reported an improved ability to read emotion with transparent mask use, and transparent masks were generally accepted across all 3 populations surveyed.

Meaning  These findings suggest that transparent masks have the potential to overcome barriers in communication brought on by universal mask wearing during the COVID-19 pandemic.

Importance  Adoption of mask wearing in response to the COVID-19 pandemic alters daily communication.

Objective  To assess communication barriers associated with mask wearing in patient-clinician interactions and individuals who are deaf and hard of hearing.

Design, Setting, and Participants  This pilot cross-sectional survey study included the general population, health care workers, and health care workers who are deaf or hard of hearing in the United States. Volunteers were sampled via an opt-in survey panel and nonrandomized convenience sampling. The general population survey was conducted between January 5 and January 8, 2021. The health care worker surveys were conducted between December 3, 2020, and January 3, 2021. Respondents viewed 2 short videos of a study author wearing both a standard and transparent N95 mask and answered questions regarding mask use, communication, preference, and fit. Surveys took 15 to 20 minutes to complete.

Main Outcomes and Measures  Participants’ perceptions were assessed surrounding the use of both mask types related to communication and the ability to express emotions.

Results  The national survey consisted of 1000 participants (mean [SD] age, 48.7 [18.5] years; 496 [49.6%] women) with a response rate of 92.25%. The survey of general health care workers consisted of 123 participants (mean [SD] age, 49.5 [9.0] years; 84 [68.3%] women), with a response rate of 11.14%. The survey of health care workers who are deaf or hard of hearing consisted of 45 participants (mean [SD] age, 54.5 [9.0] years; 30 [66.7%] women) with a response rate of 23.95%. After viewing a video demonstrating a study author wearing a transparent N95 mask, 781 (78.1%) in the general population, 109 general health care workers (88.6%), and 38 health care workers who are deaf or hard of hearing (84.4%) were able to identify the emotion being expressed, in contrast with 201 (20.1%), 25 (20.5%), and 11 (24.4%) for the standard opaque N95 mask. In the general population, 450 (45.0%) felt positively about interacting with a health care worker wearing a transparent mask; 76 general health care workers (61.8%) and 37 health care workers who are deaf or hard of hearing (82.2%) felt positively about wearing a transparent mask to communicate with patients.

Conclusions and Relevance  The findings of this study suggest that transparent masks could help improve communication during the COVID-19 pandemic, particularly for individuals who are deaf and hard of hearing.

Universal mask wearing serves as a crucial public health measure in preventing the spread of COVID-19.¹ However, studies have shown that masks may negatively affect communication and one’s ability to convey emotions such as empathy, which may impair patient-clinician relationships.^2,3 Additionally, the use of masks presents unique challenges to individuals who are deaf or hard of hearing (DHH). Approximately 17% of all adults living in the United States experience some degree of hearing loss.^1,4,5 Even before COVID-19, those in DHH communities already faced communication barriers in health care settings due to high noise levels and the environmental engineering of health care facilities.⁶ Medical grade masks have been found to affect sound frequency and muffle speech, by as much as 3 to 4 dB for surgical masks and 12 dB for respirators, and also prevent visualization of the lips and facial expressions.⁷ DHH health care workers (HCWs) face increasing barriers to participate in routine aspects of their work.⁸ The potential for prolonged use of face masks after the COVID-19 pandemic, especially in health care settings, heightens the importance of addressing these challenges.^9-12

Transparent masks could help to mitigate these issues.¹³ A few transparent masks have been introduced to the market but are not widely available in most health care settings and may not meet medical grade standards.^14-16 Although there has been significant focus on improving the supply of standard face masks, the widespread adoption of transparent masks remains largely unaddressed. We conducted surveys among a sample of the general population, general HCWs, and HCWs who are DHH to assess the role of widespread mask use on communication. We hypothesized that transparent mask use could improve nonverbal communication and the ability to perceive emotion through facial expressions.

This observational study consisted of 3 pilot cross-sectional online surveys of 3 populations of interest: (1) a survey of opt-in panelist-members from the international survey provider, YouGov, composed of 1.8 million US residents representative of the general population; (2) a survey of general HCWs from 2 urban academic quaternary care centers in Boston, Massachusetts; and (3) a survey of DHH HCWs affiliated with the nonprofit organization Association of Medical Professionals with Hearing Losses (AMPHL). Survey questions were developed based on previously published surveys in the literature and a survey provided by the not-for-profit organization Ideas for Ears.¹⁷ The survey was programmed in REDCap and piloted among members of the study team to test for understanding prior to dissemination. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cross-sectional cohort studies.

Study review and approval was obtained from the institution review board (IRB) of Mass General Brigham (MGB). YouGov conducted the national survey of the general population between January 5 and January 8, 2021. The surveys of HCWs were conducted from December 3, 2020, to January 3, 2021. Informed consent was obtained using an IRB-approved fact sheet, after which participants were presented the survey on the YouGov platform or via REDCap. The YouGov survey was defined as a nonprobability internet panel, and the general HCW and DHH HCW surveys are defined as convenience-based nonprobability samples in accordance with the American Association for Public Opinion Research (AAPOR).¹⁸

National survey respondents from the general population were recruited to the opt-in panel using active volunteer sampling methods via online advertising campaigns (public surveys), permission-based email campaigns, partner-sponsored solicitations, telephone-to-web recruitment, and mail-to-web recruitment. After initial panel recruitment, participants were selected based on representativeness of the general US population. Inclusion criteria included adults older than 18 years living in the United States. Members of YouGov’s opt-in survey panels received incentive in the form of points that can be redeemed for prizes or gift cards. General and DHH HCW respondents were recruited using nonrandomized, nonprobabilistic convenience sampling via departmental email lists, with 1 follow-up reminder email sent to nonresponders approximately 2 weeks after the initial email. Inclusion criteria were HCWs older than 18 years employed through an MGB hospital and/or members of AMPHL. Participants from the HCW populations did not receive incentives or compensation. Participation in all surveys was anonymous, and all data collected were confidential. The participation rate was calculated, and raw results were tabulated. For the national sample through YouGov, weights were applied to ensure that the sample was nationally representative of the general population.

Each survey lasted approximately 15 to 20 minutes (eTables 1-9 in the Supplement). Each survey included a demographic section that gathered information regarding age, sex, race, education, employment status, and household income. Race was classified by YouGov and study authors, and it was assessed to understand how views on mask wearing and mask use may vary demographically. Both surveys of HCWs consisted of additional demographic questions specific to HCWs. The DHH survey included an additional section with questions specific to hearing loss. All surveys included brief video clips showing a study author smiling while wearing a custom-designed transparent mask created by members of the study team and a standard opaque N95 mask (Figure, A).^19,20 Respondents were then asked questions regarding whether they were able to detect the author’s emotion and whether they felt positively toward the use of transparent masks to communicate in a health care setting. They were also asked questions specific to mask wearing and communication as well as mask preference and fit in the context of the COVID-19 pandemic. All survey questions may be found in eTables 1-9 in the Supplement.

Demographic data were described and presented using descriptive statistics (Stata IC version 16 [StataCorp]). To evaluate participant responses in each population to the video clips demonstrating the use of both a transparent and standard mask, a Wilcoxon matched-pairs signed-rank test was used to compare the medians of each population’s responses. Proportions of participant responses across populations for shared questions across all 3 surveys were compared using a Fisher Exact test (2-tailed test, P < .05 for statistical significance). To assess feedback gathered for open-ended response prompts, all responses were collected, and a framework matrix analysis was used to generate key themes that emerged from written qualitative responses (eMethods in the Supplement).²¹

A total of 1279 members of YouGov’s opt-in panel were contacted to participate in the survey. Of those, 1265 participated, and 1180 completed the survey (participation rate, 92.25%). Prior to data collection, sample matching was conducted to generate a nationally representative sample of 1000 participants. The mean (SD) age of participants was 48.7 (18.5) years; 504 participants (50.4%) were men and 496 (49.6%) were women; 117 (11.7%) were Black or African American, 140 (14.0%) Hispanic, and 671 (67.1%) White; a total of 657 (65.7%) had received education beyond high school (Table 1).

A total of 1104 individuals were contacted to participate via MGB’s departmental email lists. Of those who received a recruitment email, 177 participated, and 123 completed the online survey (participation rate, 11.14%). The mean (SD) age of participants was 49.5 (9.0) years; 39 participants (31.7%) were men, and 84 (68.3%) were women; 3 (2.4%) were Black, 3 (2.4%) Hispanic, and 90 (73.2%) White; and a total of 81 (65.9%) had received a postgraduate degree (Table 1). Respondents’ answers to questions regarding health care occupation and mask use are described in eTable 4 in the Supplement.

A total of 196 individuals were contacted to participate via AMPHL’s email lists. Of those, 66 participated in the online survey, and 45 completed the survey (participation rate, 23.95%). The mean (SD) age of participants was 54.5 (9.0) years; 15 participants (33.3%) were men, and 30 (66.7%) were women; 1 (2.2%) was Black, 2 (4.4%) were Hispanic, and 41 (91.1%) were White; a total of 35 (77.8%) had received a postgraduate degree. Respondents’ answers to questions regarding health care occupation and mask use are described in eTable 4 in the Supplement.

Overall, 31 DHH HCW respondents (68.9%) reported a moderately severe to profound hearing loss; 35 (77.8%) and 31 (68.9%) reported preferred methods of communication as auditory input or listening and speechreading, respectively. A total of 38 DHH HCW respondents (84.4%) used an assistive listening device (ALD) for hearing assistance, and 27 (60.0%) reported experiencing mask-fit interference with their ALD. Overall, 32 DHH HCW respondents (71.1%) reported concern that face masks would make living with hearing loss much more difficult. These responses are described in further detail in eTable 8 in the Supplement.

After viewing the video clip of a study author smiling while wearing a standard opaque N95 mask, 201 respondents (20.1%), 25 respondents (20.5%), and 11 respondents (24.4%) in the general, HCW, and DHH HCW populations, respectively, were able to identify the emotion being expressed (237 [20.3%] across all populations). In contrast, after viewing the video of the study author wearing a transparent mask, 781 respondents (78.1%), 109 respondents (88.6%), 38 respondents (84.4%) in the general, HCW, and DHH HCW populations, respectively, were able to identify the emotion being expressed (928 [79.5%] across all populations; P < .001) (Figure, B and eTable 9 in the Supplement). Overall, 450 respondents in the general population (45.0%) felt positively and 366 (36.6%) felt neutrally about interacting with an HCW who was wearing a transparent mask; 76 general HCWs (61.8%) and 37 DHH HCWs (82.2%) felt positively about wearing a transparent mask to communicate with patients (P < .001) (Figure, C and eTable 9 in the Supplement).

In the general population, 518 respondents (51.8%) reported having trouble communicating while wearing a face mask since the COVID-19 pandemic. In the HCW populations, 92 general HCWs (74.8%) and 44 DHH HCWs (97.8%) had trouble communicating while wearing a face mask (P < .001), with 36 DHH HCWs (80.0%) reporting difficulty at a moderate to high level (Table 2). In addition, 416 respondents from the general population (41.6%), 84 general HCWs (69.0%), and 39 DHH HCWs (86.7%) felt communication with others would be more difficult while wearing a standard mask in contrast to a transparent mask (P < .001). Overall, 379 respondents from the general population (37.9%), 75 general HCWs (61.0%), and 41 DHH HCWs (91.1%) felt it would be easier to understand or hear people who wore a transparent mask (P < .001) (Table 2). Additionally, 339 respondents from the general population (33.9%) felt they would be more at ease seeing a HCW who wore a transparent mask, while 75 general HCWs (61.0%) and 34 DHH HCWs (75.6%) felt that patients would be more at ease with a HCW who wore a transparent mask (Table 2). Overall, 302 respondents from the general population (30.2%), 65 general HCWs (52.9%), and 41 DHH HCWs (91.1%) preferred the use of transparent masks over standard masks (P < .001) (Table 2).

Respondents were asked about specific communication challenges the use of face masks had created for them as well as population-specific questions surrounding clinician and patient communication. Respondent rates of specific challenges and thoughts about mask wearing and clinician and patient communication are found in Table 2 and Table 3.

Survey respondents provided feedback regarding challenges and concerns surrounding mask wearing in their daily life in a prompt that asked participants to provide open-ended feedback (eMethods in the Supplement). We discovered 3 important themes in responses: communication, physical discomfort, and effect on work. There was concern in all populations surrounding the loss of facial and other nonverbal cues. General HCWs commented on being unable to interpret a patient’s mood or state of mind and were also concerned about not being heard by patients. DHH HCWs commented on the difficulty of having to communicate that they were DHH to others. Regarding physical discomfort, several in the general population reported concern regarding the fogging up of glasses. General HCWs also noted discomfort from prolonged mask-wearing. Regarding mask wearing and difficulties presented at work, DHH HCWs were most notably affected. Several DHH HCWs reported that they had to switch to telehealth. One respondent reported being laid off/furloughed, while another said they had to retire prematurely. In addition, others said they were limited to working in certain environments where transparent masks were available. Those who remained working in person reported difficulty performing their job and stress and fear of missed communication with their colleagues and patients (eTable 10 in the Supplement).

Respondents were asked a series of questions concerning mask type, preference, and fit. Most respondents in all 3 populations had not used a transparent mask in the last month. Preferences in physical features as well as acceptance of incorporation of advanced technologies into masks varied significantly between populations and are further described in eTable 11 in the Supplement.

Our study found that there is a need to address communication barriers related to mask use, especially among people who are DHH. Furthermore, we found that the use of transparent masks is generally accepted and could help to improve communication in both public and health care settings.

Prior studies performed before COVID-19 have shown that transparent masks improve speech understanding in DHH and normal-hearing populations and that patients perceive greater empathy and trust from surgeons communicating with transparent masks.^13,22 Our study expands on these findings by surveying HCWs in addition to the general public and characterizing the challenges faced by HCWs, especially DHH HCWs, with masks in patient care. Unlike studies performed before the pandemic, the populations surveyed in our study have had considerable personal experience with wearing face masks, and the challenges identified by respondents will help to inform future mask designs.^2,3,7,22 In all 3 survey populations, respondents were more likely to perceive the study author’s smile with a transparent mask compared with a standard mask. In addition, the use of transparent masks was viewed as acceptable, viewed positively, and accepted by a plurality of all survey populations. These findings suggest that transparent masks are acceptable among the broader population and may be an alternative that can be integrated into available mask forms for both the general population setting and in health care settings. However, factors such as discomfort, fogging of the transparent window, and increased reduction in sound quality may be barriers to implementation of transparent masks and should be considered in transparent mask designs.^22,23

Of all populations surveyed, we found that DHH HCWs seemed to experience more difficulties associated with the widespread use of standard masks in everyday life and health care settings at work. DHH HCWs have had to drastically change the way that they work during the pandemic.^8,24 Although this study did not measure other dimensions of communication such as speech interpretation and other grammatical features key in American Sign Language, we believe that our survey findings suggest that these may also be affected by mask use. Further research is necessary to measure this. In addition, some DHH HCWs expressed fear of isolation and lack of independence because of widespread face mask use. Our study did not look at the association between mental health and the use of face masks; however, our findings indicate that further research in this area is needed as well as policies to ensure that communities of people who have a disability are having their communication needs met.

We found that general HCWs agreed that transparent masks would be helpful to them as well, though to a lesser degree. However, general HCWs also have contact with DHH patients. Approximately 72% of people older than 65 years experience hearing loss, and individuals in this age group have higher hospitalization rates.²⁵ This is an important additional population that would benefit from increased availability of transparent masks in health care.

This study has limitations. This was a pilot study using a self-designed survey that has not previously been validated. Further repetitions of this study are needed to test the reliability of our findings. Follow-up studies should also include demonstration of a wider range of emotions beyond a smile. The sample sizes of both general HCW and DHH HCW populations were small and limited to 1 health care system and 1 DHH organization, respectively, which may not be representative of HCWs in the United States as a whole. Further sampling of HCWs nationally, including those who are DHH, is needed. Additionally, a nonrandomized sampling technique was used for both the general HCW and DHH HCW populations; therefore, volunteer, selection, and nonresponse bias were not accounted for. For the general population survey through YouGov, previously validated techniques, such as sample matching and weight adjustment, were used; however, substantial bias can occur with internet-based nonprobabilistic opt-in panels, such as the need for access to the internet and opt-in panel membership. In addition, the national survey was conducted through a national sampling platform of US residents, and respondents’ beliefs and attitudes surrounding mask use may vary depending on their personal experiences. Additionally, we did not assess the representation of DHH people in our survey of the general population. A survey that specifically looks at deafness and hearing loss in the general population would provide further insight into the needs of DHH communities outside of health care settings, including the inherent variety in their communication modalities. Additionally, this study did not evaluate the costs of transparent masks, which could limit their availability.¹⁹ However, we have conducted a prior study comparing costs of multiple respirator-use strategies, including distribution of 1 reusable, transparent respirator to all US health care workers, and found this to be less costly than the current practice of using disposable standard respirators.^14-16,19 Further cost-benefit studies of transparent masks should be performed as they become more commercially available.

The findings of this study suggest that widespread mask use impairs nonverbal communication and the ability to convey emotions. Transparent masks have the potential to alleviate stressors surrounding communication introduced by widespread standard mask wearing. This pilot study provides further support for transparent masks’ utility in supporting nonverbal communication, especially for those who are DHH. Our study suggests that transparent masks are needed by DHH HCWs and are considered acceptable in the general population and general health care settings, suggesting feasibility of implementation.

Accepted for Publication: September 26, 2021.

Published: November 22, 2021. doi:10.1001/jamanetworkopen.2021.35386

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Chu JN et al. JAMA Network Open.

Corresponding Author: Giovanni Traverso MB, BChir, PhD, Department of Mechanical Engineering, Massachusetts Institute of Technology and Division of Gastroenterology, Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA 02139 (cgt20@mit.edu).

Author Contributions: Dr Traverso had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Dr Chu and Ms Collins share co–first authorship.

Concept and design: Chu, Chen, Chai, Dadabhoy, Byrne, Wentworth, DeAndrea-Lazarus, Moreland, Wilson, Ghenand, Hur, Traverso.

Acquisition, analysis, or interpretation of data: Chu, Collins, Chen, Chai, Dadabhoy, Wilson, Booth, Hur.

Drafting of the manuscript: Chu, Collins, Chen, Chai, Dadabhoy, Moreland, Wilson, Hur, Traverso.

Critical revision of the manuscript for important intellectual content: Chu, Chai, Dadabhoy, Byrne, Wentworth, DeAndrea-Lazarus, Moreland, Wilson, Booth, Ghenand, Hur, Traverso.

Statistical analysis: Collins.

Obtained funding: Chai, Traverso.

Administrative, technical, or material support: Chen, Chai, Dadabhoy, Byrne, Wentworth, DeAndrea-Lazarus, Moreland, Wilson, Booth, Ghenand, Traverso.

Supervision: Chu, Chai, Wilson, Hur, Traverso.

Conflict of Interest Disclosures: Dr Chai reported receiving grants from the National Institutes of Health, the Hans and Mavis Psychosocial Foundation, the Defense Advanced Research Projects Agency, the Bill and Melinda Gates Foundation, and e Ink Corporation and receiving personal fees from Biobot Analytics outside the submitted work. Dr Byrne reported being founder of and holding equity in Teal Bio during the conduct of the study; holding equity in Advanced Chemotherapy Technologies outside the submitted work; and holding a patent for Elastomeric masks pending. Dr Wentworth reported being a cofounder of Teal Bio outside the submitted work and having a patent for mask design pending. Dr Moreland, Mr DeAndrea-Lazarus, Dr Wilson, and Ms Booth reported being members of the Association of Medical Professionals with Hearing Losses, a 501c3 nonprofit organization that supported survey dissemination to deaf and hard of hearing health care professionals. Dr Traverso reported having a financial interest in Teal Bio, a biotechnology company focused on developing the next generation of personal protective equipment including clear respirators, and having provisional patent applications surrounding clear masks pending. No other disclosures were reported.

Funding/Support: Dr Chu was supported by grant 5T32DK007191 from the National Institutes of Health. Ms Chen and Mr Ghenand were supported by the MIT Undergraduate Research Opportunities Program. Dr Chai was supported by grants K23DA044874 and R44DA051106 from the National Institutes of Health. Dr Traverso was supported by the Karl van Tassel (1925) Career Development Professorship and the Department of Mechanical Engineering at MIT and the Division of Gastroenterology, Brigham and Women’s Hospital.

Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We would like to thank Ideas for Ears for sharing their survey with us.