Cohort 1 involved 2 hospital systems in the Eastern, Mid-Atlantic part of the United States that purchased and distributed 2 models of EHMRs with an EV to a subset of employees. Because both hospital systems used EHMRs previously (one since 2009 and 2020, respectively), they already added EHMR fit testing and training into their respiratory protection program (RPP). Employees were routinely using EHMRs for 10 to 12 months at the time of data collection. No adaptations were made to the respirator body. However, hospital policy dictated that EVs needed to be covered to prevent possible illness spreading from employees to patients. This protection was normally achieved with a surgical mask being placed over the EV. See previous publications^18,19 for more information about cohort 1.

Cohort 2 involved 32 health care and emergency medical settings across the United States that received EHMR models without an EV or with an EVF from the Strategic National Stockpile (SNS).²⁰ At the time of this study, only 3 EHMR models without an EV or with an EFV were approved by NIOSH. To ensure adequate supplies would be available, the SNS solicited proposals and purchased all 3 EHMR models for public distribution—1 had an EHMR with an EVF and the other 2 did not have an EV. Organizations that requested EHMRs from the SNS could request 1 or all 3 models for use at their organization (pending availability). Four health care and 1 emergency medical setting requested and distributed more than 1 model to employees. Organizations volunteered to provide employees who received an EHMR with NIOSH’s online survey link. In many cases, organizations had no previous experience with EHMRs and had not yet incorporated them into their RPP. However, even if organizations had previous experience with EHMRs and incorporated them into their RPP, the EHMR models received (without an EV or with an EVF) were new to organizations and users. See previous publications for more information about cohort 2.^20,21

An anonymous online survey was available on the respective Research Electronic Data Capture (REDCap) portals to complete during both study periods. The survey for cohort 1 was open between April and June 2021, while the survey for cohort 2 was open between October 2021 and October 2022. Surveys took respondents from each cohort 15 to 20 minutes to complete. Some sections of the survey varied between cohorts, but one was the same for both. This section probed adverse experiences while wearing the EHMR (if applicable) using an abbreviated version of the Respirator Comfort, Wearing Experience, and Function Instrument (R-COMFI) scale.²² Respondents were asked to check (Yes or No) adverse experiences from a list of 10 options: (1) difficulty speaking or being understood (ie, communication challenges); (2) difficult to breathe; (3) bothered by moisture buildup; (4) interferes with other personal protective equipment (PPE); (5) interferes with normal eyeglasses; (6) uncomfortable to wear less than 1 hour; (7) uncomfortable to wear more than 1 hour; (8) feels uncomfortably warm; (9) feels claustrophobic; and (10) causes skin irritation or acne.

Data cleaning and checking within and between the cohorts took place. Within cohort 2, respondents were from health care (53.7%) and emergency first-responder settings (46.3%). Fisher exact test results indicated that physical and physiological experiences that users encountered were not significantly different between these 2 occupational groups in cohort 2 (ie, health care and emergency responder settings), except for feeling claustrophobic while wearing an EHMR (P < .001). Consequently, researchers did not take occupation into account when considering this cohort and retained the group as one for the purposes of this analysis.

For the primary analysis, Fisher exact test was used as a more rigorous analysis for 2 × 2 nonparametric analyses. First, researchers assessed differences in user experiences between the cohorts and second, considered if organizational familiarity with EHMRs may have impacted experiences that were reported. The independent variables included EHMR model distributed (ie, with an EV or without an EV/with an EVF) and organizational experience with EHMRs. The dependent variables included the 10 adverse experiences. Phi coefficient informed the strength of each association tested. Analysis was performed using IBM SPSS 26.²³ A P value less than .05 with the Bonferroni adjustment where necessary represented statistical significance.

The sample consisted of 1,962 respondents (cohort 1, n = 1,080 (55%); cohort 2, n = 882 (45%)). Respondents provided a variety of job titles, including physician, nurse, nursing assistant, environmental services, and facilities management working in intensive care units; emergency departments; operating rooms; pediatrics; and medical/surgical units, firefighter, safety engineer, dentist, dental hygienist, fire chief, industrial hygienist, and emergency paramedic. Table 1 shows job length reported (note minor differences in response options).

In cohort 1, both hospital systems had prior experience overseeing EHMR implementation. In cohort 2, 50% (n = 441) of respondents indicated their organization had prior experience overseeing EHMR use. In other words, 22.5% (n = 441) of the total sample represented organizations that had no prior experience with EHMRs, while the remaining 77.5% (n = 1,521) represented organizations that had experience. Table 2 shows the total time users reported wearing respiratory protection during their shifts.

Table 3 lists the frequency distributions for each statistically significant barrier experienced based on EHMR model and organizational experience. As Table 3 shows, respondents in cohort 1 (EHMR with an EV) were more likely to have communication challenges (82.2%) than were those in cohort 2 (52.8%). Communication challenges and model indicated a strong association (phi = 0.316). A weak but a statistically significant association was found between model and breathing difficulties (phi = − 0.066).

Users in cohort 2 experienced statistically significantly more instances of difficulty breathing, moisture buildup, other PPE interference, being uncomfortable to wear for less than 1 hour, and being uncomfortably warm when compared with cohort 1. Even though the strength of association between model and being uncomfortable to wear for more than 1 hour was statistically significant, the effect was small (phi = 0.121). Considering excess moisture buildup, respondents in cohort 2 reported more discomfort (44.6%) than those in cohort 1 (20.4%), showing a strong relationship between model and excess moisture buildup (phi = − 0.26). Similarly, users in cohort 2 were more likely to express discomfort due to being warm (38.9%) compared with users in cohort 1 (18.9%), rendering a medium effect size (phi = − 0.222).

Respondents who represented organizations with previous EHMR experience were more likely to have communication challenges (75%) than users with organizations that had no experience (48.5%). The association between communication challenges and organizational EHMR experience rendered a medium effect (phi = 0.239). Alternatively, users with organizations that had previous EHMR experience were less likely to express discomfort due to being warm (24.4%) compared with users with organizations that did not have experience (39.9%), indicating a small effect size (phi = − 0.144). Even though the other variables were statistically significant, the strength of association was small. Table 4 shows the significance levels and strength of the association.²⁴

First, as research has thoroughly documented and been reconfirmed in the current study, communication challenges were the most prevalent, regardless of the EHMR model. In the current study, it is important to reinforce that employees in cohort 1 were instructed to wear surgical masks over their EV for source control, which may have introduced not only difficulties communicating but also other adverse effects. Although research⁶ has shown that wearing a surgical mask over an EHMR does not produce clinically significant changes in heart rate and other physiological outcomes, it is possible that this temporary practice caused additional trouble communicating—particularly in being understood by others—as well as general discomfort over prolonged periods of use.

Despite this consideration within cohort 1, communication challenges were still the most prevalent adverse experience regardless of model and organizational experience with EHMRs. Mitigation strategies have been documented such as raising one’s voice and speaking slower.^12,26 Likely due to these ongoing challenges, manufacturers have been exploring the development of EHMR models to improve speech intelligibility.²⁷ Recently, MSA manufactured the first NIOSH Approved EHMR that has a speaking diaphragm and no EV with the hopes of minimizing communication challenges while providing source control.²⁸

The results showed that cohort 2 respondents (EHMRs without an EV or with an EVF) experienced more discomfort at a statistically significantly higher level. These barriers should be considered by both manufacturers and health delivery settings that are considering procurement and use. Specifically, being uncomfortably warm and experiencing excess moisture buildup was encountered by users in cohort 2 approximately 2 times more than those in cohort 1. Although this result may not be surprising and even unavoidable due to some of the work-related scenarios that may require the use of respiratory protection, it demonstrates the need for additional education and expectations during fit testing and training pertaining to excess moisture buildup and potential difficulty breathing. Current OSHA RPP requirements indicate that basic information about EHMR maintenance and storage is required.¹⁷ However, with the increase in models available, it is possible that organizations should include more specific user expectations training depending on the model(s) stockpiled and used by employees.

Also, cohort 1 predominately reported wearing their EHMR for more than 1 hour and reported statistically significantly higher levels of discomfort, whereas cohort 2 predominately reported wearing their EHMR for less than 1 hour and reported statistically significantly higher levels of discomfort for this category. These results suggest that users reported adverse experiences for the scenario in which they were more often experiencing—further supporting the need to improve aspects that contribute to physical discomfort.

This study was novel in that a variety of job roles were included in cohort 2 that had not been studied previously (eg, dental hygienists, dentists, fire-based EMS responders, and paramedics). This inclusion may have impacted some results. First, regarding the inclusion of different job roles in cohort 2, this may explain the results noted above—that users in cohort 2 were approximately 2 times more likely to experience excess moisture buildup and uncomfortable warmth. Specifically, users in cohort 2 also comprised fire-based EMS responders and paramedics, whose job tasks are different and likely demanded more cardiovascular exertion in comparison to HCP, who comprised all of cohort 1. Consequently, selecting certain EHMR models to cater to different job roles and demands may be useful to consider and further study.

Although user experiences varied between the 2 groups, it is important to note that the EHMR models used between the cohorts were physically the same in terms of fit, material, and manufacturer models distributed—the only difference was the configuration of the EV. So, even though users in cohort 2 (EHMR without an EV/with an EVF) reported interference with other PPE statistically significantly more often (36%) in comparison to 23% of users in cohort 1, experiences should not have practically or significantly differed. It is possible that the higher number of participants with less than 2 years of job experience in cohort 2 as well as the incorporation of new job tasks in cohort 2 elicited more adverse experiences among respondents or that there was something specific to their job task that required additional PPE, which elicited potential instances of interference. Consequently, future studies should explore job tenure and roles that may experience more cases of PPE interference to help determine tasks for which EHMRs may not be as practical when considering both worker and patient safety.

Similarly, organizations may determine that in some cases when source control is not a concern, certain job tasks can use EHMRs that have an EV to support user comfort, whereas when EHMRs are going to be used in a sterile environment, EHMRs without an EV or with an EVF be worn. In other words, advances in research and development are affording organizations additional flexibility to accommodate user preferences while ensuring patient safety. To this end, future work is needed with organizations that have varying levels of EHMR experience. In the current study, HCP in more experienced settings may have been more acclimated to wearing EHMRs and thus, might express different opinions. Possible confounders to individual and organizational experience should be proactively identified and if applicable, incorporated into RPPs as an element of emergency preparedness and employee efficacy in relation to the respirator(s) they are provided.

Results must be considered within the current study’s limitations and cannot be generalized to EHMR perceptions in general nor to other types of respiratory protection. First, all participating settings implemented EHMRs at their own pace and with numerous job roles. Because several SARS-CoV-2 variants emerged throughout the study period, distribution, adoption, and length of use likely varied based on when EHMRs were received and incorporated into RPPs. Similarly, cohort 1 had more experience with EHMRs than cohort 2, which could have affected the results.

Specific to cohort 2, those who wore EHMRs without an EV and with an EVF were combined into 1 group. Because it was more important to ensure participating organizations had access to respiratory protection when it was needed, 1 sample was retained for convenience throughout the demonstration project. Although there could have been differences in experiences between these 2 subgroups (eg, users wearing an EHMR without an EV may have experienced significantly warmer or more moisture buildup than those wearing an EHMR with an EVF), some organizations requested both models (without an EV and with an EVF) to compare, making it difficult to discern who may have worn one, or both types. However, a qualitative component of this study was also included²⁹ where managers of participating organizations completed a series of interviews throughout the study. These participants discussed employees’ (perceived) experiences. No notable differences in user experiences regarding communication, comfort, and breathing emerged as a theme when discussing EHMRs without an EVF and with an EVF. However, future research may seek to intentionally explore these differences quantitatively, especially with additional models both without an EV and with an EVF entering the market. Also, given that participating health delivery settings represented different states and geographic regions, protocols might have varied and cannot be directly compared.

Further, employees who volunteered to complete the survey are subject to social desirability bias and self-selection bias, likely indicating these respondents had a priority toward health. It is possible that nonrespondents had varying views of EHMRs that may have swayed the results in some capacity. Also, some of the possible EHMR barriers are subject to inherent bias to the outcome being based on their job tasks (eg, communicating with residents vs patients vs coworkers) or physical characteristics (eg, wearing glasses or baldness). Although this study aimed to recruit and include a variety of locations and respondents with diverse job experiences to address previous limitations,¹⁸ future research should be designed in a way to control for some of these physical characteristics.