Masks and respirators for prevention of respiratory infections: a state of the science review

Abstract

SUMMARYThis narrative review and meta-analysis summarizes a broad evidence base on the benefits-and also the practicalities, disbenefits, harms and personal, sociocultural and environmental impacts-of masks and masking. Our synthesis of evidence from over 100 published reviews and selected primary studies, including re-analyzing contested meta-analyses of key clinical trials, produced seven key findings. First, there is strong and consistent evidence for airborne transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and other respiratory pathogens. Second, masks are, if correctly and consistently worn, effective in reducing transmission of respiratory diseases and show a dose-response effect. Third, respirators are significantly more effective than medical or cloth masks. Fourth, mask mandates are, overall, effective in reducing community transmission of respiratory pathogens. Fifth, masks are important sociocultural symbols; non-adherence to masking is sometimes linked to political and ideological beliefs and to widely circulated mis- or disinformation. Sixth, while there is much evidence that masks are not generally harmful to the general population, masking may be relatively contraindicated in individuals with certain medical conditions, who may require exemption. Furthermore, certain groups (notably D/deaf people) are disadvantaged when others are masked. Finally, there are risks to the environment from single-use masks and respirators. We propose an agenda for future research, including improved characterization of the situations in which masking should be recommended or mandated; attention to comfort and acceptability; generalized and disability-focused communication support in settings where masks are worn; and development and testing of novel materials and designs for improved filtration, breathability, and environmental impact.

Keywords: SARS-CoV-2; infection prevention and control; masks; meta-analysis; methodology; narrative review; non-pharmaceutical interventions; respirators; respiratory infections.

Fig 1

Synthesis of the measured fit factors and protection factors for various types of face coverings, based on the 12 primary studies shown in the key. Reproduced under Creative Commons license from Schmitt and Wang (118). “Protection factor” and “fit factor” are similar constructs. Both provide quantitative estimates of protection based on objective testing (fit factor takes more account of real conditions of use). Studies shown used one or the other. References: O’Kelly et al. (a) (119), Coffey et al. (120), Lee et al. (121), Oberg and Brosseau (77), Duncan et al. (122), Pauli et al. (123), Lindsley et al. (124), De-Yñigo-Mojado et al. (114), O’Kelly et al. (b) (125), Fakherpour et al. (126), Lawrence et al. (127), and van der Sande (128).

Fig 2

Median of the total inward leakage over all participants for different cases. Adapted under Creative Commons license from Bagheri et al. (132). Further details are given in the original source. FFP2 w/o adj., participant wears FFP2 respirator without adjustment; FFP2 with adj., FFP2 with adjustment to fit; FFP2 + surgical, FFP2 with additional surgical mask over it; FFP2 adh. tape, adhesive tape applied to increase fit of FFP2; surg., surgical mask.

Fig 3

Forest plot of community trials: medical masks vs control (no masks). For references, see Table 3. There is some heterogeneity between studies. In some trials, infection or ILI symptoms in an index case was a pre-requisite for recruitment of family members (e.g., Cowling 2008, Suess 2012 and MacIntyre 2009) while in other trials, the intervention was triggered when a first case was confirmed in a non-household setting (Aiello 1 2010, Aiello 2 2012). In others (Abaluck 2022, Bundgaard 2021) there was no pre-requisite for exposure and the intervention was applied to general community (Abaluck 2022). The rate of infection is expected to be higher in settings such as households with an index case. Two primary prevention trials (Cowling 2008 and Suess 2012) also included "source control" – i.e., the index case wearing a mask in addition to their contacts). This was a mixed intervention, but we included them in the analysis because they also included primary prevention in contacts. a. Cowling 2008, PCR positive case numbers are calculated from rates provided in the paper in Table 2 and approximated to nearest whole number (e.g., Medical/surgical mask arm: 0.07*61 = 4 cases, Control arm: 0.06*205 = 12 cases). b. Cowling 2008, Clinical definition 1 in the paper included fever ≥ 38°C, hence placed under Influenza-like illness; case numbers are calculated from rates provided in the paper in Table 2 and approximated to nearest whole number (e.g., Medical/surgical mask arm: 0.18*61 = 11 cases; Control arm: 0.18*205 = 37 cases).

Fig 4

Forest plot of community trials: hand hygiene + medical masks vs control. For references, see Table 3. Among the community trials there is some heterogeneity between studies. In some trials, infection or ILI symptoms in an index case was a pre-requisite for recruitment of family members (e.g., Cowling 2009, Suess 2012), while in other trials, the intervention was triggered when a first case was confirmed in a non-household setting (Aiello 1 2010, Aiello 2 2012). a. Cowling 2009, Clinical definition 1 in the paper included fever ≥ 38°C, hence placed under Influenza-like illness.

Fig 5

Forest plot of trials in health-care workers: any use of N95 vs medical masks. For references, see Table 4. a. MacIntyre 2011 combined values for fit-tested and not-fit tested arms = All N95 arm. b. MacIntyre 2013 (targeted N95 arm) vs control arm was continuous use of medical masks.

Fig 6

Forest plot of trials in health-care workers: continuous use of N95 vs medical masks. For references, see Table 4. a. MacIntyre 2011 combined values for fit-tested and not-fit tested arms = All N95 arm.

Fig 7

Model of mask effectiveness in different levels of population assortativity. Reproduced under Creative Commons license from Fisman et al. (23). Assortativity is the tendency of individuals to interact preferentially with those who are most like themselves. The basic reproduction number (number of secondary cases produced by a primary case in the absence of immunity or control interventions) is on the left-side vertical axis; mask uptake in the population is on the right-side vertical axis with lower uptake at the top of the figure, high uptake at the bottom of the figure, and intermediate uptake in the middle of the figure. Mask effectiveness in reducing transmission from masked, infectious individuals is on the bottom horizontal axis with the highest effectiveness to the right and lower effectiveness to the left. Assortativity (the tendency of like to mix with like) is on the top horizontal axis and ranges from random (non-assortative) mixing on the left to highly assortative mixing (with individuals strongly preferring to interact with people like them) at the right. Pink-shaded areas indicate expected effective reproduction numbers above 1, where epidemic growth will continue. Blue-shaded areas represent combinations of parameter values where the reproduction number is reduced below 1 (the threshold where an epidemic will continue to grow). This is easier to achieve with higher mask effectiveness, a lower baseline reproduction number, higher mask uptake, and lower assortativity. The more unmasked people preferentially associate with other unmasked people, the less likely epidemic control is to occur.