As the COVID-19 pandemic progresses, major health agencies around the globe, such as the World Health Organization (WHO), the United States Center for Disease Control and Prevention (US CDC), and the Infectious Disease Society of America (IDSA), have recommended the wearing of face masks or face coverings as a critical mitigation strategy to limit the spread of COVID-19 (Source: WHO, CDC, IDSA).
The rationale for using masks to limit the spread of SARS-CoV-2 is that monitoring symptomatic patients alone is not adequate; it is also critical to stop virus transmission from asymptomatic patients. To this end, mask use can serve as an important and effective mitigation strategy. A recent systematic review and meta-analysis, which included 13 studies from 7 countries that tested 21708 at-risk individuals, estimated that the proportion of asymptomatic COVID-19 cases was about 17% [95% confidence interval (CI): 14% to 20%], meaning almost 1 in every 5 individuals infected with SARS-CoV-2 showed no symptoms (Byambasuren et al., 2020). The risk of transmission from those who are asymptomatic is relatively low; however, asymptomatic patients may still present a significant risk to the community because they are more likely to be out in public than remain isolated at home (Nogrady, 2020).
Since the outbreak, many studies have been conducted to investigate the protective effects of mask use in limiting the transmission of COVID-19, almost all of which are observational studies. On November 18th, 2020, a long awaited randomized controlled trial (RCT), assessing whether recommending surgical mask use in the community could reduce wearers' risk for SARS-CoV-2 infection, has released its results (Bundgaard et al., 2020).
To our knowledge, this RCT, called DANMASK-19 (Danish Study to Assess Face Masks for the Protection Against COVID-19 Infection), is the only RCT published with regard to mask use for the purpose of preventing COVID-19 transmission. However, DANMASK-19 did not find a statistically significant difference in the occurrence of SARS-CoV-2 infection between participants who were recommended to wear masks outside of home and those who were not.
These DANMASK-19 findings have received great attention, and those who remain skeptical about universal mask policies are championing this RCT as the decisive evidence to prove the uncertain efficacy of mask use in the prevention of the COVID-19 transmission. However, can we really conclude that mask wearing does not limit COVID-19 transmission in community settings based on the DANMASK-19 findings? In this OE Original, we dive deep into this RCT, along with evidence from non-interventional studies, to examine current evidence on the efficacy of mask use in preventing the transmission of COVID-19 in community settings.
The DANMASK-19 RCT
The general information for DANMASK-19 is presented in Table 1.
The risk of bias (RoB) of DANMASK-19, assessed based on the Cochrane criteria, is shown in Figure 1. There was a high RoB in items regarding blinding participants, research personnel, and outcome assessors. It is understandable that due to the nature of the intervention, it was not possible to implement blinding. Usually, we rate high RoB in the incomplete outcome data item if data from more than 20% of participants is lost. Noticeably, in DANMASK-19, about 19.3% of participants did not complete the study, and as a result, their outcome data were unavailable. This percentage (19.3%) is very close to our threshold (20%) for high RoB.
The hypothesis of DANMASK-19 was that the recommendation to wear face masks in public space, compared to no recommendation, reduces the wearers’ risk of infection of SARS-CoV-2 by a relative risk reduction (RRR) of 50% in a setting where community use of masks was not common and mask recommendations were not among the public health mandates already in effect, such as hand-washing and social distancing.
Results from DANMASK-19 were inconclusive and could neither accept nor reject the hypothesis due to lack of statistical significance. Specifically, infection with SARS-CoV-2 occurred in 1.8% of the participants in the mask group [42/2392] and 2.1% in the control group [53/2470]. The risk of contracting SARS-CoV-2 infection among the mask wearers was lower than that of the control participants; however, there was no statistical significance [odds ratio (OR): 0.82; 95% confidence interval (CI): 0.54 to 1.23]. The absolute risk reduction (ARR) between mask wearers and the control group was -0.3 [95% CI: -1.2 to 0.4 percentage point], and the RRR was about 14% reduction in infection [46% reduction to 23% increase], which was lower than the threshold (50%) stated in their hypothesis.
We have identified several study limitations that might have led to the underestimation of the protection of mask wearing. First, the core research question of DANMASK-19 was to determine whether mask wearing could protect the wearers from community infection of COVID-19. DANMASK-19 considered eligible participants as adults without current or prior confirmed COVID-19 or symptoms suggestive of COVID-19; however, the RCT did not provide any explicit information on whether they had excluded participants who were already infected with COVID-19 but asymptomatic at the time of inclusion. Indeed, randomization could have balanced this characteristic between mask and control groups. Yet without exclusion of these asymptomatic participants from inclusion, researchers may have underestimated the protective effects of mask wearers.
Second, DANMASK-19 showed that in the mask wearing group, around 46% of participants followed the recommendation to wear masks completely, 47% followed as recommended predominantly, and 7% not as recommended. The RCT authors concluded that the impacts of non-compliance were not significant by conducting a sensitivity analysis in which the 7% of participants who did not wear masks as recommended were excluded. The sensitivity analysis did not show any significant change in results [mask vs. control: 1.8% vs. 2.1%; OR: 0.84; 95% CI: 0.55 to 1.26]. However, there is a possibility of overestimating adherence to the use of masks. DANMASK-19 used self-reporting of adherence, which is the most common method but also a problematic one (Stirratt et al., 2015). By using the self-reporting of adherence, researchers may overestimate the degree to which the participants executed their assigned intervention due to reasons such as memory biases and vulnerability to social desirability (Stirratt et al., 2015).
Third, DANMASK-19 investigated individual participants who wear masks while outside in the community. However, this design might have ignored the important role and potentially huge impacts of household transmission of COVID-19. Examining 24 studies, a systematic review found that the risk of infection with SARS-CoV-2 due to household contact was over 10 times higher than the risk due to other contacts [OR: 10.72; 95% CI: 5.70 to 20.17; P < 0.001] (Lei et al., 2020). As for DANMASK-19, not considering the household transmission in the mask group might have resulted in an underestimate of the protection from mask wearing. Although it is impossible to ask participants to wear masks at home in a trial, it may be helpful to provide infection information about the members who were not participating in the RCT but living under the same roof with the RCT participants, in order to help readers assess the potential impacts of household transmission among the participants randomized to mask wearing. To address household transmission in future studies, trialists could implement randomization by the unit of household or community (such as counties and cities) instead of individuals.
To conclude this section, DANMASK-19 provides important and insightful information on the effects of recommendation to wear surgical masks on limiting the transmission of SARS-CoV-2 infection in the community settings. It can also inform future trials by reflecting on study limitations and further improving methodological quality. However, results from DANMASK-19 were inconclusive and more RCTs are needed to answer the question.
Due to lack of RCTs on this topic, we also identified some key literature which could provide evidence from non-interventional studies.
- Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis (Chu et al., 2020)
This systematic review included 172 observational studies involving more than 25000 patients in healthcare and non-healthcare settings from 16 countries. Results showed that mask use could significantly reduce the risk of infection with SARS-CoV-2 [N = 2647; adjusted OR: 0.15; 95% CI: 0.07 to 0.34; ARR: 14.3 (10.7 to 15.9 percentage points)]. However, due to the observational study design, the quality of evidence rated using the GRADE approach is low. Additionally, there was a stronger association of protection with N95 or similar respirators, compared to surgical masks or similar reusable cotton masks (12 to 16 layers).
- Trends in County-Level COVID-19 Incidence in Counties With and Without a Mask Mandate — Kansas, June 1–August 23, 2020 (Van Dyke et al., 2020)
This is an ecologic study, in which the observation occurs at the population/group level, not the individual level. In this ecologic study, researchers looked at 24 counties of Kansas which implemented a mask mandate in the public space following the governor’s executive order and 81 counties in the same state without a mask use mandate. The study found that the incidence of COVID-19 (calculated as the 7-day rolling average number of new daily cases per 100,000 population) decreased among counties with a mask mandate [mean decrease of 0.08 cases per 100,000 per day; net decrease of 6%], but kept increasing among counties without a mask mandate [mean increase of 0.11 cases per 100,000 per day; net increase of 100%].
- Community Use Of Face Masks And COVID-19: Evidence From A Natural Experiment Of State Mandates In The US (Lyu et al., 2020)
This study examined the effects of US state government mandates for face mask use in public space issued by 15 US states plus Washington, D.C. and calculated the changes in the daily county-level COVID-19 growth rates. Face mask mandate was associated with a decline in the daily COVID-19 growth rate by 0.9, 1.1, 1.4, 1.7, and 2.0 percentage points in 1 to 5, 6 to 10, 11 to 15, 16 to 20, and 21 or more days after state face mask orders were issued, respectively. As a result of the implementation of these mandates, an estimate of over 200000 COVID-19 cases were averted between March 31 and May 22, 2020.
- Absence of Apparent Transmission of SARS-CoV-2 from Two Stylists After Exposure at a Hair Salon with a Universal Face Covering Policy (Hendrix et al., 2020)
This study described an event in which 139 clients were serviced by and exposed to 2 hair stylists with COVID-19. Both the clients and the 2 stylists wore face masks during the service. None of the 139 clients reported signs or symptoms of COVID-19. All 67 nasopharyngeal swab specimens from 67 clients who agreed to be tested were tested negative using PCR. This event suggested that wearing face masks might mitigate the spread of COVID-19 in community settings.
- Case-Control Study of Use of Personal Protective Measures and Risk for SARS-CoV-2 Infection, Thailand (Doung-ngern et al., 2020)
This is a case-control study including 211 cases of COVID-19 and 839 controls in Thailand. The study found that individuals who reported having worn masks all the time during contact with other people were less likely to contract SARS-CoV-2 infection [adjusted OR: 0.23; 95% CI: 0.09 to 0.60], compared to those who did not wear masks.
- Reduction of secondary transmission of SARS-CoV-2 in households by face mask use, disinfection and social distancing: a cohort study in Beijing, China (Wang et al., 2020)
This retrospective study investigated 335 subjects in 124 families which had at least 1 lab confirmed COVID-19 case. The overall secondary attack rate in these families was 23.0% (77/335); wearing masks at home before the onset of the primary COVID-19 case in the family was associated with a lower risk of secondary transmission of COVID-19 [[adjusted OR: 0.21; 95% CI: 0.06 to 0.79].
Outside of the above mentioned non-interventional studies, there are more studies existing in the literature exploring the association between mask wearing and transmission of SARS-CoV-2 (Payne et al., 2020; Schwartz et al., 2020). Overall, these non-interventional studies, including observational and epidemiological studies, supported that face masks may be effective in reducing transmission of SARS-CoV-2 in community settings. However, due to the nature and limitations of observational and epidemiological studies, these findings only suggest an association between community mask use and reduction of COVID-19 transmission. Further research adopting an RCT design is still needed.
Currently, data from observational studies and epidemiological studies overwhelmingly favor the community use of face masks in order to limit the spread of COVID-19. The evidence from the only published RCT -- DANMASK-19, however, is inconclusive and lacking certainty -- therefore could neither confirm the effectiveness nor prove the efficacy of the wearing-masks-in-public policy. It is still too early for the public, the health care professionals and the policy makers to jump to any conclusion, especially the conclusion which denies the use of face masks to prevent the transmission of COVID-19.
Bundgaard, H., et al. (2020). Effectiveness of Adding a Mask Recommendation to Other Public Health Measures to Prevent SARS-CoV-2 Infection in Danish Mask Wearers : A Randomized Controlled Trial. Ann Intern Med. doi:10.7326/m20-6817
Byambasuren, O., et al. (2020). Estimating the extent of asymptomatic COVID-19 and its potential for community transmission: Systematic review and meta-analysis. Official Journal of the Association of Medical Microbiology and Infectious Disease Canada, Accepted version, e20200030. doi:10.3138/jammi-2020-0030
Chu, D. K., et al. (2020). Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis. The Lancet, 395(10242), 1973-1987. doi:10.1016/S0140-6736(20)31142-9
Doung-ngern, P., et al. (2020). Case-Control Study of Use of Personal Protective Measures and Risk for SARS-CoV 2 Infection, Thailand. Emerging Infectious Diseases, 26(11), 2607-2616. https://dx.doi.org/10.3201/eid2611.203003
Hendrix MJ., et al. Absence of Apparent Transmission of SARS-CoV-2 from Two Stylists After Exposure at a Hair Salon with a Universal Face Covering Policy — Springfield, Missouri, May 2020. MMWR Morb Mortal Wkly Rep 2020;69:930-932. DOI: http://dx.doi.org/10.15585/mmwr.mm6928e2
Lei, H., et al. (2020). Household transmission of COVID-19-a systematic review and meta-analysis. J Infect. doi:10.1016/j.jinf.2020.08.033
Lyu, W., et al. (2020). Community Use Of Face Masks And COVID-19: Evidence From A Natural Experiment Of State Mandates In The US. Health Aff (Millwood), 39(8), 1419-1425. doi:10.1377/hlthaff.2020.00818
Nogrady. (2020). What the data say about asymptomatic COVID infections. Nature, 587, 534-535. doi:https://doi.org/10.1038/d41586-020-03141-3
Payne DC., et al. SARS-CoV-2 Infections and Serologic Responses from a Sample of U.S. Navy Service Members – USS Theodore Roosevelt, April 2020. MMWR Morb Mortal Wkly Rep. 2020;69(23):714-721.10.15585/mmwr.mm6923e4. https://www.ncbi.nlm.nih.gov/pubmed/32525850
Schwartz KL., et al. Lack of COVID-19 transmission on an international flight. Cmaj. 2020;192(15):E410.10.1503/cmaj.75015. https://www.ncbi.nlm.nih.gov/pubmed/32392504
Stirratt, M. J., et al. (2015). Self-report measures of medication adherence behavior: recommendations on optimal use. Transl Behav Med, 5(4), 470-482. doi:10.1007/s13142-015-0315-2
Van Dyke ME., et al. Trends in County-Level COVID-19 Incidence in Counties With and Without a Mask Mandate — Kansas, June 1–August 23, 2020. MMWR Morb Mortal Wkly Rep. ePub: 20 November 2020. DOI: http://dx.doi.org/10.15585/mmwr.mm6947e2
Wang, Y., et al. (2020). Reduction of secondary transmission of SARS-CoV-2 in households by face mask use, disinfection and social distancing: a cohort study in Beijing, China. BMJ Glob Health, 5(5). doi:10.1136/bmjgh-2020-002794