The scientific case against face masks

Description

This winter season, the New York Times, Washington Post, Wall Street Journal and Atlantic, among other outlets, have all published articles on the same theme. According to their advice, we should re-don masks to prevent seasonal spread of influenza, RSV, Covid-19 and run-of-the-mill colds. This seems poised to become a yearly occurrence, as with the accompanying post-holiday mandates in some schools, colleges, and elsewhere that these articles actively encourage.   

However, while these articles are full of quotations from health officials and disease experts, glaringly absent is high-quality data to support claims that masking reduces spread of circulating seasonal viruses. 

The reason for this omission may be that, three years into the pandemic, there are no rigorous studies showing masks to be an effective method of viral infection control. In fact the highest-quality scientific studies, randomised controlled trials (RCTs), show the opposite: that masks make little to no difference in controlling spread of influenza, SARS-CoV-2, or RSV. 

In May 2020, the CDC summarised data from 14 RCTs as failing to show a significant benefit of masks in reducing transmission of influenza. An analysis of nine trials conducted by Cochrane, an organisation that conducts large reviews of health-care interventions, reached similar conclusions in November 2020. Studies of masking to prevent common colds and RSV also had negative results. 

For Covid-19, there are two RCTs evaluating masks’ ability to cut viral spread. One, conducted in Denmark in the spring of 2020, found no statistically significant difference in infection rates between masked and unmasked groups. Another, bigger, RCT — conducted in Bangladesh from late 2020 until the following spring — showed a small but statistically significant reduction of symptomatic Covid-19 cases in villages using surgical masks. Yet even this small benefit was lost upon reanalysis using different statistical parameters. An additional finding that mask colour made a difference in effectiveness further suggests that this positive data were skewed in some way. 

Another rigorous study, though not an RCT, was conducted in schools in Catalonia, Spain. This showed that unmasked five-year-olds had similar Covid-19 case rates to masked six-year-olds, the age at which masking was mandated. 

These high-quality studies are noticeably not linked in any “mask up” articles. Instead, links (if any are provided) reference low-quality observational studies such as a CDC analysis from 2021 that found higher paediatric Covid-19 rates in unmasked schools compared to masked ones. This study was subsequently debunked upon reanalysis using more districts and a longer time period. 

A recent analysis conducted in Massachusetts schools, reporting that masks reduced Covid-19 case rates, is another popular source used to support media exhortations to mask up. However, this study was also riddled with issues, such as changes in testing practices after masking was dropped (impacting Covid-19 detection) and differences in levels of natural immunity between masked and unmasked schools before and during the study period. In addition, many of the schools cited as requiring masks during the study period had already dropped their mandate. 

Many media outlets have also repeated that “high-quality masks” are the solution to viral spread. But, again, we lack randomised data providing evidence that these masks do a significantly better job than other types. While fitted N95s can be effective at protecting against bacterial and droplet transmission, several RCTs have found them not to be significantly better than surgical masks at protecting against influenza or SARS-CoV-2 infection.  

We all want masks to work, but thus far high-quality data indicates that they don’t, at least not on a measurable population level. The same is true of rapid testing. Considering the limited accuracy of a single asymptomatic swab at detecting SARS-CoV-2, the advice to “test before seeing Grandma” is not supported by evidence that this would provide meaningful protection. In fact, a few studies have shown that large asymptomatic Covid-19 testing programs have had limited utility in reducing transmission. 

Insisting upon mask effectiveness or the reliability of antigen testing may actually be emboldening vulnerable people, or those who interact closely with vulnerable people, to take risks by eliciting a false sense of security. This phenomenon, known as the Peltzman effect, where people act more carelessly when they perceive risk is lower due to the presence of a guardrail, has been shown to influence a range of behaviours including driving, drug use and sexual activity. 

In the case of the coronavirus, this might involve forgoing other precautions that would actually protect themselves and others, such as getting vaccinated, avoiding crowds, making sure rooms are ventilated well, and staying home when sick. Even the White House Covid-19 response coordinator, Ashish Jha, recently appeared to lament that an overreliance on masking has replaced necessary investments in improving indoor air quality.

The fact that many news outlets continue to promote narratives that are not supported by high-quality scientific evidence may be contributing to historically low levels of trust in the media. Hyping unproven mitigation measures was, and continues to be, a serious mistake.

Dr Leslie Bienen works in health care policy. 
Dr Jeanne Noble is an emergency physician and director of Covid Response at the UCSF Parnassus Emergency Department. 
Dr Margery Smelkinson is an infectious-disease scientist whose research has focused on influenza and SARS-CoV-2.

Discussion