The COVID-19 pandemic necessitates the widespread use of personal protective equipment (PPE), not only by medical workers who are in direct contact with patients with COVID-19, but also by hospital personnel in the more peripheral areas of care and maintenance such as receptionists and cleaners and those providing essential services. The number of professionals and service providers who now need to use PPE that is in continuous contact with skin has markedly increased, and this PPE plays an important role in protecting them from infection.1–4
Unexpected occupational skin injuries in medical staff wearing PPE have thus become a topic of concern.5 Several publications document that PPE, especially N95 masks and goggles, cause skin damage or injuries, seriously affecting the health of medical staff.5,6 Tissue damage from PPE manifests as a skin wound under the contours of the gear and has been described as skin tears (STs), device-related pressure injuries (DRPIs), irritant contact dermatitis (including friction injuries), and moisture-associated skin damage (MASD).1,5,7
The main underlying mechanism is tissue deformation by PPE, followed by cell damage, cell death, inflammatory reactions, tissue edema, ischemia, and eventually tissue death.1 In clinical settings, these PPE-related injuries increase the risk of bacteria, fungi, and viruses penetrating the skin and reaching the bloodstream, with potentially fatal results for the affected health professionals.1,7 Measures to protect medical staff from occupational injuries during the pandemic are urgently needed.3,8
The aim of this study was to describe the prevalence of skin injuries caused by medical masks combined with goggles, compare skin disruptions in those wearing N95 masks or surgical masks with goggles, identify the anatomic locations of facial skin injuries, explore the characteristics related to skin disruptions, and describe the prevention strategies used. The authors hope to provide a basis for the protection of medical workers’ facial skin and prevent occupational injuries in China and worldwide.
The authors designed a cross-sectional multicenter survey on skin injuries in medical staff during the COVID-19 pandemic in China. Sample size was determined based on the fact that 85% of medical staff members reported facial injuries in Wuhan in the early stage of the COVID-19 outbreak while wearing protective masks combined with goggles.9 Considering a tolerated absolute error of 2% and a confidence interval (CI) of 95%, the calculated sample size was 1,225 participants. Considering the possibility of 10% invalid questionnaires, a total of 1,362 participants were needed.
Inclusion criteria were (1) medical staff members who wore N95 or surgical masks combined with goggles to work in environments with a moderate risk of exposure to patients with COVID-19 (eg, the ED and screening clinic where patients with suspected cases are treated and triaged) or a high risk (ICUs and wards in designated medical institutions in the epidemic area) according to guidelines published by the Chinese healthcare administration;10,11 (2) age 20 years or older, regardless of sex; and (3) voluntary participation.
Exclusion criteria were (1) medical staff members who only wore N95 or surgical masks or who did not wear PPE or used other alternatives and (2) those who provided incomplete or invalid questionnaires.
The authors designed a questionnaire with 13 items after consulting the relevant literature and guidelines.12 The questionnaire had four general demographic items (sex, age, level of seniority, and occupation), three items about PPE (whether an N95 mask or a surgical mask was worn in combination with goggles, the duration of PPE use, and the occurrence of sweating), four items on skin injuries (types of skin injuries and when they happened, anatomic location, classification), and two items on protective or preventive measures (whether daily protective measures were taken and the type of dressing used).
Before the formal investigation, 30 medical staff members were selected for a pilot survey. It took 4 to 6 minutes for each person to complete the questionnaire. Twenty-seven people (90%) thought the questionnaire was clear and easy to answer. Three persons (10%) thought the type of dressing used was difficult to describe or accurately report because of the lack of relevant knowledge and shortage of dressing material in the early stage of the outbreak. To solve this problem, the authors modified the question regarding dressing type to be multiple choice with the following possible answers: foam dressing (single layer, multilayer, border, no border), hydrocolloid dressing, and other (fill in). After the modification was made, another 30 medical staff members were selected to retest the survey. One hundred percent of them thought that the questionnaire was clear and easy to answer, and each person completed the questionnaire in 3 to 5 minutes.
The authors uploaded the modified questionnaire to the “Questionnaire Star” website and disseminated it to online chat groups of medical teams supporting infectious disease departments, isolation wards, and ICUs in Wuhan via the WeChat social platform on February 8, 2020. Participants voluntarily used their smartphone to answer and submit the questionnaire online through February 22, 2020.
To ensure accurate answers to the questions and enable the identification of different skin injuries, the classification of DRPIs was based on the 2019 International Guide for the Prevention and Treatment of Pressure Ulcer/Injury as stages 1, 2, 3, or 4; deep-tissue injury; or unstageable.12 The 2017 CMS definition was used to identify MASD: “It is caused by sustained exposure to moisture, which can be caused, for example, by incontinence, wound exudate, and perspiration. It is characterized by inflammation of the skin and occurs with or without skin erosion and/or infection.”5 In addition, the 2017 CMS definition was adopted to identify ST: “a result of shearing, friction, or trauma to the skin that causes a separation of the skin layers. They can be partial or full thickness.”5 The classification of ST was based on the International Skin Tear Advisory Panel classification system, which classifies traumatic wounds caused by mechanical forces, including the removal of adhesives. The severity may vary by depth (not extending through the subcutaneous layer).6
Ethics Approval and Consent to Participate
After receiving an explanation of the study, participants provided their informed consent. The questionnaire was completed anonymously. The privacy and confidentiality of the study participants were also strictly maintained.
All the data were downloaded from the “Questionnaire Star” website. Two researchers checked each questionnaire to determine if it was completed and, if the answers were logical, then established a database for the statistical analysis. The Shapiro-Wilk test was used to test the normality of the distributions of the continuous variables. Continuous variables are represented as the means and SDs if normally distributed or as medians and interquartile ranges if nonnormally distributed. Categorical or ordinal variables are represented as frequencies and percentages. Student t tests or Mann-Whitney U tests were used to compare the differences between two groups for continuous or ordinal variables. Pearson χ2 tests or Fisher exact tests were used to compare categorical variables as appropriate.
Variables with P < .10 in the univariate analysis were entered into a multivariate ordinal logistic regression model to establish whether they were independently associated with the number of skin injuries. If the test of parallel lines showed that ordinal logistic regression analysis was not appropriate (P < .05), multinomial logistic regression analysis was used. Both the omnibus test of model coefficients and the test of goodness of fit for the model were performed to determine the validity of the model. Odds ratios and their corresponding 95% CIs for risk factors were calculated on the basis of the model variable coefficients and standard errors, respectively. A P ≤ .05 or less indicated statistical significance. All statistical analyses were performed using SPSS 22.0 for Windows (SPSS, Chicago, Illinois).
In total, 1,620 medical staff members from 145 hospitals in China received the questionnaire, and 1,611 responses were valid (99.4%). In total, 16.8% (271 respondents) were men and 83.2% (1,340 respondents) were women; 14.2% (229 respondents) were doctors and 85.8% (1,382 respondents) were nurses. Most respondents worked in hospitals and departments designated for the treatment of patients with COVID-19. The average age was 32.8 ± 6.9 years, and 65.9% (1,062 respondents) of the respondents were younger than 35 years and 34.1% (549 respondents) were 35 years or older. The average daily wear time (DWT) for PPE was 6.9 ± 2.2 hours, with 88.6% (1,428 respondents) of the respondents reporting a DWT longer than 4 hours and 11.4% (183 respondents) reporting a DWT 4 hours or less. In total, 80.0% (1,289 respondents) reported sweating. The data regarding wearing of protective masks combined with goggles by medical staff are shown in Table 1.
|Wearing Characteristics||N95 and Goggles (n = 1,301)||Surgical Mask and Goggles (n = 310)||Total (n = 1,611)||P|
|Daily wear time||.451|
|> 4 h||1,157 (88.9)||271 (87.4)||1,428 (88.6)|
|≤ 4 h||144 (11.1)||39 (12.6)||183 (11.4)|
|Yes||1,220 (93.8)||69 (22.3)||1,289 (80.0)|
|No||81 (6.2)||241 (77.7)||322 (20.0)|
Among the 1,611 respondents, 1,281 cases of skin injuries were reported, and the overall prevalence of skin injuries in medical staff wearing masks with goggles was 79.5% (95% CI, 77.55–81.49%), including 911 cases of DRPI (56.5%; 711 cases of stage 1 [44.1%], 193 cases of stage 2 [12.0%], 3 cases of stage 3 [0.2%], and 4 cases of deep-tissue injury [0.3%]), 313 cases of MASD (19.4%), and 57 cases of ST (3.5%; all type 1). The N95 masks combined with goggles caused significantly more injuries of all three types than did surgical masks combined with goggles (Table 2). There were 1,103 respondents (68.5%) who reported multiple types of skin injuries, and 386 respondents had skin injuries in four anatomic locations (24.0%), followed by 367 respondents with injuries in three locations (22.8%), 350 with injuries in two locations (21.7%), and 178 with an injury in a single location (11.0%).
|Injury Type||N95 mask and Goggles (n = 1,301)||Surgical Mask and Goggles (n = 310)||Total (N = 1,611)||P|
|DRPI, n (%)||838 (64.4)||73 (23.5)||911 (56.5)||<.001|
|Non-DRPI, n (%)||463 (35.6)||237 (76.5)||700 (43.5)|
|MASD, n (%)||271 (20.8)||42 (13.5)||313 (19.4)||.004|
|Non-MASD, n (%)||1,030 (79.2)||268 (86.5)||1,298 (80.6)|
|ST, n (%)||56 (4.3)||1 (0.3)||57 (3.5)||<.001|
|Non-ST, n (%)||1,245 (95.7)||309 (99.7)||1554 (96.5)|
Among the 1,281 respondents with skin injuries, skin injuries co-occurred on the bridge of the nose, cheek, ear, and forehead (24.0%; 95% CI, 21.76%–25.92%). The prevalence of a single injury on the forehead, ear, cheek, and bridge of the nose ranged from 0.7% to 5.3% (χ2 = 180.500; P < .001). The locations of skin injuries among medical staff wearing protective masks with goggles are shown in Table 3.
|Locations||N95 Mask and Goggles||Surgical Mask and Goggles||Total||Prevalence and 95% Confidence Interval a|
|Nasal bridge (a), n (%)||77 (93.0)||6 (7.0)||83||5.2 (4.07–6.23)|
|Cheek (b), n (%)||46 (81.8)||8 (18.2)||54||3.4 (2.47–4.23)|
|Ear (c), n (%)||25 (85.3)||6 (14.7)||31||1.9 (1.25–2.60)|
|Forehead (d), n (%)||9 (91.7)||1 (8.3)||10||0.6 (0.25–1.00)|
|a + b, n (%)||138 (93.6)||11 (6.4)||149||9.2 (7.83–10.66)|
|a + c, n (%)||65 (87.1)||10 (12.9)||75||4.7 (3.63–5.68)|
|a + d, n (%)||12 (85.7)||2 (14.3)||14||0.9 (0.42–1.32)|
|b + c, n (%)||62 (80.6)||18 (19.4)||80||5.0 (3.91–6.03)|
|b + d, n (%)||25 (92.3)||2 (7.7)||27||1.7 (1.05–2.30)|
|c + d, n (%)||5 (100)||0 (0.0)||5||0.3 (0.04–0.58)|
|a + b + c,n (%)||215 (91.3)||20 (8.7)||235||14.6 (12.86–16.31)|
|a + b + d, n (%)||88 (90.1)||9 (9.9)||97||6.0 (4.86–7.18)|
|a + c + d, n (%)||19 (90.9)||2 (9.1)||21||1.3 (0.75–1.86)|
|b + c + d, n (%)||12 (85.7)||2 (14.3)||14||0.9 (0.42–1.32)|
|a + b + c + d, n (%)||367 (95.1)||19 (4.9)||386||24.0 (21.88–26.05)|
|Total, n (%)||1,165 (91.0)||116 (9.0)||1,281||79.5 (77.55–81.49)|
The differences in the prevalence of injuries between men and women (84.9% vs. 78.4%), nurses and doctors (79.9% vs 78.6%), and those younger than 35 years and those 35 years or older (78.5% vs 81.4%) were not significant (Pall > .05). The prevalence of skin injuries among medical staff wearing N95 masks and goggles (89.5%; 95% CI, 87.88%–91.21%) was higher than among those wearing surgical masks and goggles (37.4%; 95% CI, 32.03%–42.81%). The differences in the prevalence of skin injuries between medical staff who did and did not sweat (94.6% vs 18.9%) and those who reported a DWT over 4 hours versus 4 hours or less (83.8% vs 49.7%) were significant (Pall < .001). Moreover, wearing an N95 mask and goggles, having a DWT over 4 hours, and sweating were associated with more injury sites and a higher prevalence of skin injuries. The characteristics of skin injuries in medical staff caused by wearing protective masks combined with goggles are shown in Table 4.
|Characteristics||n||No. of Skin Injuries||Prevalence
of Skin Injuries
|95% CI for Prevalence||P|
|Masks with goggles||<.001|
|N95 mask and goggles||1,301||155||303||340||367||1165 (89.5)||87.88–91.21|
|Surgical mask and goggles||310||23||47||27||19||116 (37.4)||32.00–42.84|
|>4 h||1,428||165||320||346||359||1,190 (83.3)||81.40–85.27|
|≤4 h||183||13||30||21||27||91 (49.7)||42.41–57.04|
Results of Logistic Regression Analysis
The number of skin injuries was regressed as a dependent variable (none = 0, single location = 1, two locations = 2, three locations = 3, four locations = 4). First, those variables with P < .1 in the univariate analysis (Table 3) were entered into the ordinal logistic regression model. The omnibus test of model coefficients yielded χ2 = 937.894 (degrees of freedom [df] = 4, P < .001), and the results of the tests of the goodness of fit for the model were as follows: Pearson χ2 = 60.507, P = .196; and deviance χ2 = 58.797, P = .241. The results of the parallel lines test were as follows: χ2 = 23.646, df = 9, P = .005.
These results indicated that the ordinal logistic regression model was not adequate. Therefore, the authors conducted a multinomial logistic regression analysis. The model fitting information showed that the likelihood ratio was χ2 = 954.929 (df = 12, P < .001), which meant that the partial regression coefficient of at least one independent variable was not zero at the .001 significance level. In brief, when the reference category of the dependent variable was “no skin injury,” sweating was associated with skin injuries in fewer than two locations, and a DWT over 4 hours and sweating were associated with skin injuries in three locations, whereas the use of an N95 mask and goggles, a DWT over 4 hours, and sweating were associated with skin injuries in four locations. The details are shown in Table 5.
|No. of Skin Injuries a||β||SE||Wald||df||P||OR||95% CI for OR|
|1 location intercept||−4.264||0.465||83.333||1||<.001|
|N95 and goggles||0.082||0.323||0.064||1||.800||1.09||0.58–2.05|
|DWT (>4 h)||0.633||0.343||3.398||1||.065||1.88||0.96–3.69|
|2 locations intercept||−3.889||0.419||90.113||1||<.001|
|N95 and goggles||−0.277||0.276||1.012||1||.314||0.76||0.44–1.30|
|DWT (>4 h)||0.438||0.285||2.360||1||.125||1.55||0.89-2.71|
|3 locations intercept||−5.142||0.518||98.666||1||<.001|
|N95 and goggles||0.055||0.299||0.033||1||.855||1.06||0.59–1.90|
|DWT (>4 h)||1.188||0.355||11.232||1||.001||3.28||1.64–6.58|
|4 locations intercept||−5.775||0.607||90.377||1||<.001|
|N95 and goggles||0.917||0.384||5.699||1||.017||2.50||1.18–5.31|
|DWT (>4 h)||0.749||0.315||5.641||1||.018||2.12||1.14–3.93|
Among the 1,301 respondents who wore N95 masks and goggles, 548 respondents (42.1%) took preventive measures. In contrast, of the 310 respondents who wore surgical masks and goggles, only 49 respondents (15.8%) took preventive measures; this difference was significant (P < .001). The preventive measures taken were mainly intended to protect facial skin by applying various dressings and oils. The details are shown in Table 6.
|Prevention Measures||N95 Mask and Goggles (n = 1,301)||Surgical Mask and Goggles (n = 310)||Total, n (%)|
|None, n (%)||753 (57.9)||261 (84.2)||1,014 (62.9)|
|Foam dressing, n (%)||54 (4.2)||6 (1.9)||60 (3.7)|
|Hydrocolloid dressing, n (%)||240 (18.4)||8 (2.6)||248 (15.4)|
|Applying oil, n (%)||94 (7.2)||20 (6.5)||114 (7.1)|
|Dressing and oil, n (%)||106 (8.1)||6 (1.9)||112 (7.0)|
|Homemade agent, n (%)||54 (4.2)||9 (2.9)||63 (3.9)|
|Total, n (%)||1,301 (100.0)||310 (100.0)||1,611 (100.0)|
|χ 2/P||χ 2 = 86.886||P < .001|
The study showed that the overall prevalence of skin injuries among 1,611 medical staff members wearing protective masks with goggles was as high as 79.5% (95% CI, 77.52–81.49; Table 3), which mirrors the previously reported prevalence of skin lesions on nasal bridges and cheeks in frontline healthcare workers from PPE (83.1% and 78.7%, respectively).5 However, the prevalence of DRPIs (56.5%), MASDs (19.4%), and ST (3.5%) differed. Three types of skin injuries were reported in these results, although desquamation, erythema, maceration, fissures, papules, and pressure injuries caused by PPE were reported in other studies.5,6
The prevalence of skin injuries among medical staff reported in this study was much higher than the prevalence of DRPI reported for critical care patients from mechanical ventilation masks, tubes, splints, and so on (12%-14%);13 was similar to the prevalence of MASD of 20% reported in community-dwelling patients; and was higher than those reported in patients in acute (6%) and long-term care (17%).14,15 However, these rates of ST were much lower than those reported in hospitals and long-term care settings (up to 22%).16 This is possibly because masks and goggles directly compress the nose, cheeks, and forehead, which are areas generally lacking in subcutaneous fat, and the irritation increases with increasing duration of PPE use.
According to medical staff, the local pressure exerted by the PPE was uncomfortable after 2 hours and became increasingly uncomfortable with additional time spent wearing the PPE. Further, 80.0% of the medical staff members sweated when they wore masks and goggles. This sweat was mainly produced by the skin of the scalp and face and collected under the masks and goggles. Exhaled water vapor also collected there, which increased the local humidity and changed the microclimate of the skin;10,17 the longer the DWT, the higher the local humidity. Given the combined effects of pressure and humidity and the external forces generated by the rapid removal of masks and goggles, DRPI, MASD, and ST often occurred on the face, and multiple types of damage occurred in various locations.
The prevalence of skin injuries among medical staff with a DWT longer than 4 hours was higher than that among medical staff members with a DWT 4 hours or less. N95 masks were associated with more local pressure and sweating than surgical masks, leading to a higher prevalence of skin injuries (89.5% vs 37.4%, P < .001). This explanation reinforces a recent conceptual framework for pressure injuries updated in 2014.17 It is very important to clarify the role of protective masks combined with goggles in the development of skin injuries before devising measures to protect medical staff members’ skin and prevent injuries.
Stage 1 (Figures 1A, B) and stage 2 DRPI (Figures 1C, D) as defined in the 2019 guidelines10 accounted for 78.1% (711 cases) and 21.2% (193 cases) of the injuries, respectively. Stage 1 is the most common category of DRPI experienced by medical staff members, which differs from the results reported in critical care patients, in whom DRPI stage 2 is the most common (76.9%).18
The analysis of MASD in this study showed that it was primarily related to sweating and dampness caused by wearing PPE. This is classified as intertriginous dermatitis (ITD), a type of MASD that occurs in skin folds and is commonly found in the inframammary, pannus, groin, perianal, and interdigital areas.14,15 The combination of excess moisture from perspiration, limited air flow, and friction between opposing epidermal surfaces can lead to ITD.14 However, the authors found that the characteristics of MASD in this study were mainly maceration, erythema, erosion, or ulcers in areas in which sweat gathered, including the cheeks, nose bridge, and neck. The ITD on the facial skin mainly manifested as a result of the accumulation of sweat under masks and goggles. Because PPE creates a closed environment and sweat does not evaporate, reported characteristics in this study are different from those reported in the literature.14,15
Univariate analysis showed that the use of protective masks with goggles, prolonged DWT, and sweating were possibly related to the development of skin injuries (Pall < .001). The multinomial logistic regression indicated that sweating increased the risk of injuries in one to four locations (95% CI, 16.23–60.02 for one location; 95% CI, 38.22–239.04 for four locations), and wearing an N95 mask combined with goggles and a DWT over 4 hours increased the risk of skin injuries in four locations (95% CI, 1.18–5.31 and 1.14–3.93, respectively; Table 5). Accordingly, effective measures are needed to prevent and manage sweating and shorten DWT to prevent skin injuries caused by protective masks combined with goggles.
Significantly more medical staff wearing N95 masks combined with goggles took protective measures compared with medical staff wearing surgical masks combined with goggles (42.1% vs 15.8%, P < .001). This may be related to the significantly increased sweating associated with the use of N95 masks combined with goggles (Table 1, P < .001), because more sweat would increase pain. In addition, the N95 masks exert more pressure than surgical masks. Half of the medical staff reported that they used preventive measures to reduce pain, such as the application of hydrocolloid dressings (Figure 2) and multilayer, bordered foam dressings under masks and goggles (Figure 3). Almost all preventive measures were taken after the initial skin pain or injury, which may also explain why the prevalence of skin injuries in the study was 79.5% and even higher in the group wearing N95 masks (89.5%). Pursuing preventive measures before skin pain or injury occurs could possibly improve outcomes.
Strengths and Limitations
Because of the emergency nature of and the quick response to the COVID-19 outbreak in China, and because medical staff were mostly evacuated from the epidemic area in Wuhan within 2 months, it was difficult to observe and collect data on skin protective measures taken by medical staff wearing protective masks with goggles. However, this project received a high rate of valid responses, likely because the research was supported by reputable organizations; further, the research benefited from broad support of frontline medical staff, many of whom had experienced skin injuries.
This study was limited by the typical faults of a self-reported survey. Investigators did not obtain additional information on the use of oils as protectants, such as the location and frequency of application. The best products for facial skin are lotions or creams that are noncomedogenic. These limitations provide opportunities for further research.
The prevalence of skin injuries among medical staff wearing protective masks combined with goggles was very high, especially among those wearing N95 masks combined with goggles. Multinomial logistic regression indicated that sweating was associated with skin injuries in fewer than two locations; a DWT over 4 hours and sweating were associated with skin injuries in three locations; and use of an N95 mask, a DWT over 4 hours, and sweating increased the risk of skin injuries in four locations. Preventive strategies should focus on sweating and the use of protective masks combined with goggles for longer than 4 hours.
[Tratto da: www.journals.lww.com ]