Chinese bivariate fit test panel shows wider bizygomatic breadth and shorter morphological facial length compared to the panel of NIOSH[11-12]. The face shape of subjects in this study appeared to be shorter and narrower than that of both NIOSH and Chinese panels according to Figure 3 and Figure 4. Chen et al. found that the Chinese bivariate fit test panel was able to accommodate at least 95% of the surveyed Chinese workers in 2006[13]. However, the percentage of subjects in this study fell into the Chinese fit test panel is not as much as that in previous research. This may due to regional differences in facial dimensions. In this study, morphological facial length, transversal arc, head length and minimum forehead size were different significantly between six native place groups. It was showed that facial dimension difference existed between Australian-born, United Kingdom-born, and other foreign-born groups by an anthropometric survey of Australian Male[14]. Lim et al[15] found that there were significant differences in eight dimensions including bizygomatic breadth, minimum frontal breadth, bigonial breadth, menton-sellion length, interpupillary distance, head breadth, nose protrusion, nose breadth, nasal root breadth and subnasal-sellion length between birthplace and ethnicity in Malaysia. Statistically significant differences in facial anthropometric dimensions were noted between four racial/ethnic groups including Caucasian, African-Americans, Hispanic, and other (mainly Asian) in the USA[8]. Therefore, in view of the differences in facial dimensions caused by region or ethnicity, it is suggested that the sample of subjects from various ethnicity or native places should be included as much as possible to ensure the representativeness of fit test panel. Moreover, a large sample national population database of facial dimensions is necessary to be established for the design and selection of personal protective equipment in Chinese.
Facepiece fit has been described as the major contributor to particle leakage into FFRs[16]. Fit test is an effective method to select a proper FFR with the tight facial seal for individuals. A fit test is mandatory before initial use of a respirator and then conducted annually, for example in the USA[10]. According to Chinese standards, qualitative or quantitative fit tests have been recommended for screening suitable respirators for users[17-18]. In this survey, except for Wenjian medical N95 respirator with an insufficient sample size, the PRs of the remaining FFRs ranged from 17.6% to 90.8%, which has shown a wide variation for different subjects. Lee et al.[19] suggested that specifying a fit test PR of at least 90% of randomly selected wearers is a criterion for a certified respirator. It is difficult to ensure that the ideal fit can be achieved when different individuals using the same FFR model. WHO recommended that health workers should be fit tested before using a respirator to ensure that they are wearing the correct size[20]. The necessity of this recommendation by WHO was verified by this study again. Furthermore, supervision and training of correct wearing of a FFR should be strengthen in hospitals due to the high exposure risk of respiratory infectious disease. A significant proportion of community population would like to wear FFRs when taking public transportation such as the subway, bus and plane or visit hospitals during the epidemic of respiratory infectious diseases, such as COVID-19. Whether to wear correctly to get seal is one of the key information of science popularization in community population. It was indicated that widely used N95 FFRs in China didn't fit well and can't provide desired protection for respirator users[21]. This phenomenon also can be seen in other Asian countries, including Korea and Iran[22-23]. Community groups are not as easy as health care workers and other occupational personnel to conduct fit tests at the present stage. As a result, to formulate quick methods to guide the selection of FFRs for community population is helpful to promote the effect of respiratory protection. And some researchers argued that after more accurate studies were performed, facial size categories, on the basis of facial dimensions, could be a useful tool to assist in the selection of adequately fitting respirators for people in the situation having no fit testing requirements[24]. An important factor in the optimal fit of FFRs is consideration of the anthropometric dimensions of individuals in the target population. Face size categories could match respirator sizing reasonably well for some models of half-mask respirators[3,25]. Han et al.[26] showed that face width, bitragion-menton arc (bitragion-gnathion arc) and nose protrusion(nasal depth) had to be the primary consideration for half-face mask respirator design. OESTENSTAD et al.[27] put forward that nasal dimensions should be considered when selecting a respirator for an individual wearer. To identify the key facial dimensions that affect fit of respirator in the target population is a prerequisite for applying these dimensions to the design and selection. Zhang et al.[1] found the significant effects of morphological facial length and bitragion-submandibular arc on fit factors. Brazile et al.[28] said the only ficial dimension correlated with respirator fit was nose protrusion. The bitragion-ophryon arc, bitragion-submandibular arc, head breadth and nasal height influenced fit test result in this study. It is suggested that the effects of facial dimensions on fit of FFRs may vary in different populations, for what requiring sufficient investigation to clarify the correlation between facial dimension and fit in the target population. In addition, age was another influential factor to fit in this study, which was consistent with Maxim's findings[29] that individuals aged 18-29 were significantly more likely to pass compared to colleagues aged 30-59. Older people are likely to fail the fit test and that may be related to the decreased elasticity of their facial tissue and the in-tight fitting of the mask. Providing multiple alternative models of FFRs is one way to achieve a satisfactory level of fit. Of 4,198 female and 2,089 male health care workers tested in Australian, 55% passed the first FFR, 21% required 2 and 23% required testing on 3 or more models, 93.3% were successfully fitted totally[29]. Four models of medical N95 respirators were available which ensured 100% of 678 health care workers passes the fit tests, while not all the 107 community people could find a fit one when two FFRs were offered in this study.