A woman wearing a surgical mask and gloves during the coronavirus outbreak walks past a mural in the … [+]
Let’s get one thing out of the way right at the start: if you’ve come to this article wondering if you should wear a mask to protect society at large from coronavirus infection, this article won’t answer that question for you. The CDC is considering the question because that’s what responsible public health agencies do as evidence and circumstances change, and I’m awaiting their guidance like everyone else.
What this article will do is discuss why the question is so complicated—yes, it really is—, what factors are at play and what you will want to consider in deciding whether to wear a homemade mask. (It remains true that severe shortages in surgical masks mean any mask wearing should be something you already possess or will make, and it remains true that most people do not properly wear N95 respirators.)
I wrote my previous article on masks on Feb. 29, which, in CoronaTime, is approximately a millennium ago. It would be another week and a half before the WHO would even declare COVID-19 a pandemic. Both the science and the pandemic itself have shifted and are continuing to shift. That article’s information was true at the time, and most of it remains true now. It addressed one main question: should you wear a mask to protect yourself from coronavirus?
“The question a month ago was will they protect you, the wearer, and the answer is still, they probably won’t protect you,” Eli Perencevich, the University of Iowa infection prevention specialist I spoke to for the last article, said when I spoke to him today about the topic. One of the biggest reasons they won’t protect the average wearer is that most don’t wear them correctly—even when trained—and unconsciously engage in counterproductive behaviors, such as touching the mask frequently.
So why, then, do healthcare workers wear them? Aside from the fact that they (usually) wear them correctly (and have people telling them when they don’t), there are other reasons I’ll get to—hang tight. Further, it’s still possible future evidence could change what we know now and reveal that wearing masks does protect wearers. Currently, that evidence still doesn’t exist.
But’s there a more important question to address right now—one which was not addressed in my previous article, which is the hot question right now, and which, even if I had addressed it before, would be outdated based on new evidence and circumstances. (In fact, as an update on March 1 noted in that article, “it remains an open scientific question whether (and which) droplet-based respiratory viruses are transmitted this way. So far, all documented transmission for COVID cases has involved droplets.”)
The big question is: Should everyone wear a mask to protect other people from infection? It’s an important question, a question that could not be adequately answered a month ago (I’ll get to why), a question that’s being intensely and appropriately considered and debated now, and a question all of us desperately want an answer to.
The best answer for the moment: Quite possibly, yes, universal mask wearing might decrease asymptomatic and presymptomatic transmission of the disease. The evidence isn’t strong, solid, or crystal clear (it rarely is), but it might be better to err on the side of trying it. It won’t surprise me if the CDC shortly comes out with a mask recommendation to reduce community spread.
Circumstances and Science Change
Since the article a month ago, we have much more evidence on COVID-19 transmission, and the prevalence of the disease itself is far more widespread.
“Even the science and recommendations of a month ago suggested that people who are sick should wear a mask to protect others,” Perencevich said. But a month ago, fewer than 100 Americans had confirmed COVID-19. Even though many likely had it who hadn’t been tested, well over 99.9% of the nation didn’t have it, compared to more than 200,000 cases today.
“The critical thing is that we’re in a totally different part of the epidemic, where the shift is that we have to assume many of us are infected, so we would wear masks for a totally different purpose,” Perencevich said. “Now, with social distancing, we’re assuming everyone is sick, so it makes sense for the time being that the CDC is considering broadening mask recommendations, not to protect the wearer, but to protect the family members and others in the community.”
The US is already engaged in a massive social distancing experiment with high levels of community spread, so wearing a mask would be another aspect of social distancing that could protect others, Perencevich said.
Some will argue that recommending that everyone wear masks a month ago would have prevented the explosion in cases in the US. There is no evidence to support that, as I’ll discuss. Further, since a severe mask shortage already existed, a universal recommendation would likely have significantly depleted mask supplies even more, which would have been even more dangerous for the nation. If healthcare workers can’t get enough masks, they get sick, and there are fewer of them to take care of people, so fewer people receive care, more disease is spread, and more people die. Simply telling people to make their own masks rather than buying them would not have prevented a run on masks.
But preserving masks for healthcare workers was not the only reason to recommend against universal mask wearing. At the time, evidence that it would reduce transmission did not exist. Let’s review that evidence.
The Theoretical Evidence
It is, and always has been, theoretically possible that putting a mask on people who are infected but don’t know it yet could potentially lower the likelihood they would infect others. But to make that recommendation would have required more knowledge and peer-reviewed evidence than was available a month ago.
What we didn’t know a month ago:
- How much asymptomatic and presymptomatic transmission was occurring
- How that transmission was occurring — whether it was solely through cough and sneezes, whether singing or forceful talking or heavy breathing (such as during exercise) also increased transmission risk, or whether simply breathing or talking normally did.
- How far droplets containing the virus traveled before falling out of the air, evaporating and settling on surfaces.
- How long it took for the droplets to evaporate.
- How long a potent virus capable of infecting people remained on different surfaces.
- How much transmission occurs through direct respiratory droplets in the air versus fomites, objects that are infected from recently evaporated droplets containing the virus.
The historical peer-reviewed evidence on universal mask wearing to reduce community spread remains inconclusive. It is difficult—probably impossible—to conduct an ethical randomized controlled trial that could ever conclusively tell us whether an entire region wearing masks will substantially reduce transmission, especially since such an intervention should never be implemented without other evidence-based measures that we know work, such as testing, contact tracing, mask-wearing for the symptomatic, and social distancing.
Here’s a short summary of what the peer-reviewed evidence does show, all of which must be considered through the lens of other considerations discussed further down and the fact that COVID-19 is a new disease caused by a new pathogen:
- Homemade masks were not as effective as surgical masks in preventing wearers from expelling droplets, but they did reduce droplets and were better than no protection, according to a 2013 experiment.
- The combination of wearing a mask and hand-washing—but not either one by itself—reduced household transmission of influenza by 35-51% in a 2010 study.
- Flu-like illnesses occurred 13 times more often in healthcare workers wearing cloth masks compared to surgical masks, and 97% of particles penetrated cloth masks, compared to 44% penetrating surgical masks, according to a 2015 study.
- Wearing a mask decreased infection risk by 60-80% when a parent was caring directly for a sick child in the same household, but mask adherence was well below 50%, leading the authors of a 2009 study to conclude that masks were “ineffective in controlling seasonal influenza-like illness” but might work better with better adherence. Further, the authors urged “caution in extrapolating our results to school, workplace, or community contexts, or where multiple, repeated exposures may occur, such as in healthcare settings.”
- A 2015 systematic review of 9 randomized controlled trials consistently found that real-life use of medical masks overall did not reduce infections compared to no masks except in the 2009 study above when adherence was high. Five studies found small reductions in risk with a combination of mask-wearing and hand-hygiene, while the others found no benefits.
- Five separate studies in the 2015 review above found that N95 respirators significantly reduced infections compared to surgical masks.
- Masks blocked live influenza particles in a 2013 simulation experiment with a dummy, but studies showing that masks block droplets or even infectious particles cannot be assumed automatically to prevent infections.
- Wearing masks appeared to reduce SARS transmission risk, with approximately one infection prevented for every 6 people wearing a mask, according to a 2008 systematic review.
- Face masks were not helpful in reducing transmission of pandemic influenza, according to a 2017 systematic review and meta-analysis.
This is an incomplete list, but it covers a substantial breadth of the evidence to show how complex the mechanics of mask-wearing are and how little evidence exists regarding wide-scale community wearing to reduce community spread. But for the latter, it’s important to distinguish between the “absence of evidence and evidence of absence,” as a March 20 editorial in The Lancet Respiratory Medicine pointed out, adding that “universal use of face masks could be considered if supplies permit” (emphasis mine). Still, as a March 5 editorial in The Lancet Infectious Diseases noted, “Although surgical masks are in widespread use by the general population, there is no evidence that these masks prevent the acquisition of COVID-19, although they might slightly reduce the spread from an infected patient.” Existing evidence is based on experience with pathogens before COVID-19. Any guidance on masks must be based on what we learn about COVID-19—and that evidence is quickly accumulating. Hence the CDC’s reconsideration of mask policy.
What many have found to be most persuasive regarding the benefits of universal mask wearing is a popular—but very misleading—graph circulating from #Masks4All that shows cases across many countries. A red circle around the high case numbers of many European and North American nations is labeled “No masks,” and a blue circle around the low case numbers in South Korea, Singapore, Japan, and Hong Kong is labeled “Masks.” The absolute best use of this graph is to teach the fallacy of mistaking correlation for causation, and even then it’s not entirely accurate.
First, the country with the highest proportion of residents consistently wearing masks is China, where the disease clearly was not contained quickly enough. If mask-wearing could very substantially reduce community transmission, the epidemic would not have spread so far so quickly in China.
Second, contrary to what the graph shows, Singapore did not universally recommend masks; the country limited masks to healthcare workers. Third, Japan is beginning to see an uptick in cases, and it remains to be seen where the numbers end up months from now.
Last and most importantly, all the countries showing low case numbers instituted very intense, widespread testing, contact tracing and quarantining— strongly evidence-based strategies for containing an outbreak—and they did so immediately after the first case was identified. As Australian virologist Ian Mackay, PhD, pointed out on Twitter, those countries all had “experienced professionals” in “preparing for and dealing with an emerging respiratory virus,” primarily due to SARS.
In South Korea, drive-through testing centers were available before almost anyone in the U.S. could get tested at all. In Singapore, citizens downloaded smartphone apps that gave the government knowledge of their whereabouts and self-reported symptoms at all times, something that American citizens would never have consented to on a widespread scale.
All these measures significantly reduced COVID-19 transmission in those countries, and it’s impossible to know how much, if at all, mask-wearing made a difference in South Korea, Japan, or Hong Kong, or even in China.
Aside from that graph, many have pointed to the SARS epidemic as evidence that mask-wearing reduces transmission. Even if solid evidence existed to show that wearing masks substantially reduced SARS cases—which is lacking—the fact remains that SARS is a different disease than COVID-19, just as MERS and influenza are different, and it’s not possible to extrapolate findings related to those diseases directly to COVID-19.
With any new disease, scientists are flying half- or totally blind for a while as they scramble to learn about the disease at the same time that public health and government officials are scrambling to contain it. As Harvard epidemiologist Bill Hanage told Ed Yong at The Atlantic, “We’re trying to build the plane while we’re flying it. We’re having to make decisions with quite massive consequences in the absence of secure data.”
Finally, even if it could be shown that mask-wearing is strongly correlated with reduced transmission after controlling for factors such as testing and contact tracing, we don’t know the mechanism. Would it be the physical barrier of the mask itself that’s preventing transmission, or could it be, as Gregory Poland, MD, an infectious disease expert at the Mayo Clinic, suggested to me, that the mask is an indicator of a constellation of behaviors that collectively reduce transmission? That is, does wearing a mask lead someone to touch their face less often (there isn’t evidence of that yet, and self-reported anecdotal evidence is unreliable), and is wearing a mask a constant psychological reminder to do other known infection prevention behaviors, such as frequent, rigorous hand-washing and keeping a good distance from other people?
If the mask is mainly a proxy for other good behaviors, then telling everyone to wear one is only helpful if everyone also follows those other behaviors as well. In countries where mask-wearing, even not during a pandemic, is common, that’s less of a concern. But in countries where mask-wearing is extremely uncommon, such as the US, recommending at the very beginning of the epidemic that everyone wear a mask could have risked providing a false sense of security: People may have been less conscientious about extremely important, effective behaviors that are more mundane, such as hand-washing, or very difficult, such as social distancing.
And that brings us to important cultural differences between different countries.
Cultural Considerations of Mask-Wearing
There are cultural reasons—that do not imply anything negative or positive about any particular country—that mask-wearing recommendations might have different effects in different regions.
As medical anthropologist Christos Lynteris wrote in the New York Times, wearing masks is a cultural symbol of solidarity in China and has been for years. When I visited China ahead of the 2008 Olympics, I wasted no time after arriving in finding, buying and wearing a mask myself for two reasons: the pollution was so awful that it made breathing easier, and I came down with a cold that quickly turned into bronchitis, so I was coughing constantly. After all, “when in Rome”—er, uh, China—as the saying goes.
In China, mask-wearing is common and culturally accepted, both because of pollution and because the SARS epidemic in 2003 caused a nationwide case of PTSD that led Chinese people to take outbreaks extremely seriously as soon as they occurred. And because of that, nearly every household in China already has masks on hand—just like many households in Hong Kong, South Korea, and Japan.
The nearly 350 million people in the US do not have masks conveniently available in their homes, however. It’s easier and less risky—in terms of unintended consequences like mask shortages—to tell everyone to wear a mask to reduce asymptomatic transmission “just in case” in countries where every home already has masks on hand or knows how to make them.
Further, even had the US recommended mask-wearing sooner, compliance likely would have been very poor. Rates of compliance for CPAP masks, which can literally prevent people with obstructive sleep apnea from dying in their sleep, are notoriously low—below 40%. The US has not experienced a traumatic, widespread, deadly respiratory disease epidemic since 1918, so they lack the fear that might motivated mask-wearing that many East Asian people justifiably have.
(Plenty of people from Hong Kong told me on Twitter that America should listen to them because they had been through SARS. They weren’t wrong about that, but unfortunately, the American mindset doesn’t respond to that kind of advice, even if it’s good or well-intended advice. In fact, Americans, and Westerners in general, have an inexcusably xenophobic history of discounting the wisdom and advice of people from other cultures.)
Finally, residents of many countries in the East have years of experience in wearing a mask correctly. As much as everyone wants to believe they will wear them correctly, they actually underestimate how difficult it is to do so: you must wash your hands before and after putting it on, never touch it, wash your hands immediately if you accidentally touch it, and keep it clean and sanitized in between wearing. We all have heard repeatedly how often we touch our faces. If you wear a mask, you might touch your face less—but you might touch the mask itself just as often.
There is still plenty of incorrect mask wearing, and, as my previous article noted, improper mask-wearing does risk spreading infection further, but at least the learning curve is far flatter in places with a longer history of mask-wearing.
What We Know Now
That brings us to the knowledge and evidence we know today, the reason the CDC is re-evaluating its stance on masks. In the past month, we’ve learned the following:
- Presymptomatic transmission likely accounted for up to 6-13% of new cases in a new MMWR report published by the CDC just today.
- Transmission of the disease may occur for up to 3 days before symptoms appear, according to the same MMWR report.
- It may not require coughs and sneezes alone to transmit the disease: “Presymptomatic transmission might occur through generation of respiratory droplets or possibly through indirect transmission,” the MMWR report found. “Speech and other vocal activities such as singing have been shown to generate air particles, with the rate of emission corresponding to voice loudness.”
- Transmission of the disease from infected people who don’t know they are infected could have been as high as 86% in the early days of the epidemic, according to a Science study.
- The SARS-CoV-2 coronavirus that causes COVID-19 can remain present and infectious in aerosols and on cardboard, plastic, stainless steel and even copper anywhere from several hours to several days, according to a March 17 NEJM study.
- Depending on a person’s physiology and the environmental conditions (such as humidity and temperature), droplets containing pathogens may travel as far as 23-27 feet (7-8 meters) from a person sneezing or coughing, according to a March 26 paper in JAMA.
What We Still Need to Know
One of the most challenging aspects of understanding COVID-19 is transmission through droplets and aerosolization. Ed Yong of The Atlantic provides an excellent explainer on droplets and aerosolization vs airborne transmission in his piece today, “Everyone Thinks They’re Right About Masks.” The concept of aerosolization vs airborne transmission is very confusing and complicated, and I strongly recommend everyone read his article to understand it better and understand why we still lack crucial data about COVID-19’s transmission.
Yong’s piece is also one of the only recent ones I’ve read that appropriately addresses the complexities of the question of universal mask recommendations. As is true in science in general, the more certain someone is of something—especially non-experts—the more skeptical you should be of their claims and the more evidence their claims demand. If anyone tells you that the mask question is “simple” or that everyone “obviously, unquestionably” should wear masks, that person is not coming from a thoughtful, conscientious exploration of the evidence. And the simple and attractive mantra that “Masks Saves Lives” is irresponsible when it implies that masks alone—without hand-washing and continued social distancing—can make any difference at all. The fact remains that if every person stayed at home for the next month, mitigation and containment would radically increase.
Practical Concerns of Universal Mask-Wearing
The question now, of course, is if the CDC does recommend masks, or if you decide you want to wear them, where do you get them? You shouldn’t—and probably aren’t even able—to buy surgical masks or N95 respirators because healthcare workers still desperately need them.
Making your own mask, or getting one from someone else who makes them, is the best answer. There are plenty of resources online for doing so, and Forbes contributor Bruce Lee has an excellent primer on DIY mask-making. I won’t link to specific instructions because I have not vetted them, and the quality of materials and methods will vary greatly. It’s important to keep in mind that some materials may not be safe to use, but any fabrics designed for clothing should be safe.
Watch the WHO videos on proper mask usage. Wash your hands before and after putting on the mask. Do not remove a mask to eat and then immediately put it back on without washing your hands or, ideally, changing or washing the mask.
Why Does a Mask Protect Healthcare Workers But Not Me?
That answer hasn’t changed. There isn’t reliable peer-reviewed evidence showing that wearing a mask will protect you from getting infected during everyday errands in the public. Many asked after my previous article why, then, medical staff wore masks (aside from protecting their patients from any other pathogens they might unwittingly be carrying).
First, they are caring directly for symptomatic COVID-19 patients in close proximity who are coughing and sneezing. The risk of droplets hitting their faces wasn’t theoretical—it was happening. When in extremely close contact with a sick person who might cough in your face, any barrier might help, which is why the previous article also recommended that you wear a mask if caring for a person with COVID-19 in your household.
Further, healthcare workers must often do medical procedures that cause droplets containing the virus to aerosolize— temporarily turn into vapor—such as bronchoscopy or intubation. Even doing a throat swab could cause vomiting or severe coughing in someone with an intense gag reflex.
Even when patients wear masks outside of these procedures, the masks likely cannot able to stop all the infectious particles coming from the symptomatic patients. Multiple studies have found N95 respirators, properly fitted, to be far more effective in protecting healthcare workers than medical masks, but something is better than nothing.
Healthcare workers are also at higher risk for transmitting the disease to their coworkers before they realize they are sick, and the risk is far higher in medical settings than outside the hospital.
“A lot of hospitals are going into universal masking in recognition that you should wear maximum protection around COVID patients but also that when you’re crowded into a nursing station or team room, people siting next to you could be infected, Perencevich said.
Finally, if a resource is scarce—as masks are—it is essential to reserve that resource for those who need it most—healthcare workers and symptomatic patients—even if its effectiveness is dubious or extremely small because the risk of not getting 5% protection is so much greater than simply walking through a grocery store.
Did the CDC or WHO Mislead People?
No. Both the CDC and the WHO have difficult jobs in finding and interpreting incomplete, complex and sometimes contradictory evidence about a disease in real time as it’s raging across the globe. Despite criticism of them, both organizations have the smartest people in infectious disease in the world working hard to do that job in good faith, and they have many factors to consider in their guidance. The messaging up until now regarding masks was not necessarily the “wrong” messaging, and it’s highly irresponsible and inaccurate to claim that the WHO “misinformed” people.
Spreading rumors that the CDC or WHO lied or misled people on masks makes that job harder and harms people. The CDC and WHO are working from the best evidence they have access to at the time, and interpreting that evidence—while gathering new evidence for a newly emerged disease that has different characteristics than any previous disease—is not a simple or fast task, even in the best of times. And these are not the best of times.
Where to Read More About Masks
The following articles are excellent sources of more reading on masks:
Note: This article was written with the generous research help of Beth Drummond, MPH, of Seattle, Washington.