Lessons from the Choir: The Science Behind Cloth Face Masks

Something caught my eye in the news[1] recently, about how 45 out of 60 choir members developed symptoms of the Covid-19 viral disease after their March 10 practice session.

What’s staggering about that (at least to a medical school professor in a different field) is not the 75 percent but the setting. I don’t mean its beauty, up in the Skagit Valley in the tulip-growing country an hour north of Seattle, but its resemblance to a clean, well-designed experiment that has an unambiguous answer. Nature seldom provides one that eliminates most alternative explanations. But the Skagit super-spreader episode was in spite of the victims taking all the reasonable-at-the-time precaut­ions for spreading viruses by touch, focusing attention on the airborne route.

Most case clusters have something like a funeral or family reunion at their core, where everyone has a lot of physical contact. Six people just died in South Carolina who had attended the same funeral; more would have probably tested positive for the virus. Kissing is by far the easiest way to get a big load of viruses in a short time. Handshakes, followed by wiping away tears, is high on the list as well.

“In February, a man in Chicago brought food to and hugged two friends who had recently lost a family member. The next day, the man went to the funeral, where he comforted other mourners and shared a potluck meal. A few days later, he attended a family birthday party. The man had symptoms of a mild respiratory illness. Later he’d learn he had COVID-19. His acts of condolence and celebration set off a chain reaction that sickened at least 16 people, three of whom died.” — Science News

The more typical transmission path is where Alice contaminates her hand when swiping a drippy nose, and then smears some of her mucus onto an elevator button or a doorknob — after which some of Alice’s mucus transfers to Betty’s hands, and eventually to Betty’s nose or eyes.

Washing hands often helps to break this virus transmission path, as does the admonition to avoid touching your face. Each time that you wipe your nose, wash your hands immediately. Twice a day is no longer sufficient. A soap scrub does a better job than alcohol-gel hand rubs.

In most case clusters, one cannot distinguish between spreading via touch and spreading via shared room air. Airborne transmission has not been thought to be an important route for delivering enough virus to matter. But there have been except­ional cases which have puzzled epidemiologists, such as how one sick airplane passenger in 1977 could have infected many others who were seated far away.

Officials in Washington State first began to plead with people to keep their distance from one another at the end of February. But on March 10, schools were still open[2]; prohib­it­ions of large gatherings had yet to be announced; and no cases had yet been reported in Skagit County — even though, 60 miles (100 km) to the south, the Seattle area had a serious nursing-facility outbreak, the first cluster of secondary-spread cases in the U.S.

Half the 121 community choir members, average age 67, showed up for the rehearsal, which lasted 2.5 hours. Nobody was sick; members with even mild symptoms were told to stay home. A greeter offered hand sanitizer, and members refrained from the usual hugs and handshakes. In a church hall about the size of a volleyball court, they gathered in rows facing a piano and a choir director. They were not standing shoulder to shoulder on risers but sat in three rows of individual chairs and had space to keep an empty seat on each side.

The first Covid-19 cases began three days after the rehearsal. At seven days, a choir member had heard of enough cases to phone Skagit County Public Health about the apparent cluster. Using the choir’s membership list, a dozen health officers worked for three days to contain the outbreak. For each person with symptoms, they obtained a list of their close contacts during the 24 hours before illness set in. They told anyone who felt sick to quarantine themselves.

Three-quarters of those who gathered to rehearse soon came down with the Covid-19 symptoms. All the sick ones who were tested turned out to have the virus. Two died.

I became curious about what could have amplified the ordinary means of viral spread in the physical situation of a choir. I sang in a choir in my youth and became aware that some singers occasionally have a spray coming out of their mouths when trying to project their voice. Thanks to excellent stage lighting, many of us have probably noticed the spray coming from a rock star’s mouth; the spray can extend far past the handheld microphone.

Figure 1. Loud speech emits about four times as much as speaking softly. Wearing a bandana would catch much of the virus load.

The spray doesn’t go as far as when people sneeze or cough, but one can often see small droplets from loud speech, sometimes strung together. They contain air bubbles which burst, and those can release aerosols at some distance from the speaker. Lydia Bourouiba at MIT has found forceful sneezes create gaseous clouds that can travel 27 feet[3].

A cluster bomb pops open just above its target to release a lot of little hand gernade equivalents over the area of a square block. A sneeze or cough (and sometimes merely loud speech or singing) projects droplets of mucus or saliva ten paces ahead; they evaporate water en route, popping bubbles containing aerosols containing viruses. But instead of falling, the aerosols hang around in the air as an invisible cloud.

The viruses carried by the aerosols (about a thousand times smaller than any visible droplets) are the serious problem, as they do not sink quickly to the floor in the manner of larger droplets. Instead, aerosols hang around in the air in front of the spreader. No ordinary face mask, surgical or otherwise, can filter out incoming aerosols but they may filter the small droplets before they are inhaled.

An invisible cloud forms during quiet breathing in healthy people. Usually air circul­ation dilutes it, moving viruses to places where they will do no harm before deteriorating naturally.

If the singer is in the first row of the choir, with no one out in front, the danger of contagion is not as bad as when the singer is standing on a riser with several rows of choir members out in front. Because I was tall, I always sang from the back row. The aerosol spray zone might extend for six feet; aided by a loud sneeze, aerosols can make it across the room.

Suppose that the one-fourth of the Skagit choir who escaped the virus was on the back row of the choir, in no-one’s invisible projection zone? But that one singer in the back row was slowly coming down with the disease and, for several hours, was creating a cloud of aerosol hanging over the singers in the other rows? As aerosols containing the virus slowly settled out, some would accumulate on their upturned faces or in their hair, where their hand could then transfer it to their eyes or nose.

Bear in mind that this is not the usual casual encounter with someone’s exhaled breath, where acquiring sufficient virus numbers is far less likely. This is a choir practice lasting several hours where people do not move out of position very much and, these days, are constantly sipping from their water bottles to wet their vocal cords.

Figure 2. The vibration of the vocal cords creates small air bubbles in the mucus, which contain aerosol particles, some of which are viruses. As the mucus travels forward in the airstream, water evaporates and the bubbles pop, releasing viruses.

You would think, from the usual models of gradual disease spread, that the number of new cases should rise exponentially at each distance from the source, redoubling daily case numbers every few days. But analyzing patterns in the spread of malaria, HIV, and other diseases, it is said that 20 percent of infected people spread 80 percent of communicable disease.

Super-spreaders mean new cases can grow even more suddenly, and in ways that are quite unpredictable. They may leapfrog ahead of the gradual wave of disease spread. That makes it difficult to predict disease spread, just as extreme weather makes it difficult to predict climate impacts merely from the gradual overheating[4].

Now consider the ‘super-spreaders’ of communicable diseases[5], who seem to spread infection more than ten times more easily than normal people. Since it takes weeks to even suspect a super-spreader episode, they are difficult to study. Super-spreading plays a role in many diseases, from Ebola to tuber­culosis, not just viral diseases.

If trying to project your voice across a large room, you may spray four times your soft speech amounts. In 1968, it was found that six times more is emitted during singing, approx­imately equivalent to that released by coughing.

Finnish researchers [7] used a supercomputer to model cough aerosol spread in typical supermarket air flow.

It turns out that there are also ‘super-emitters,’ people who normally spray out ten times more than the usual number of tiny aerosols — even in ordinary speech in ordinary times. Speech super-emitters[6] are not necess­arily breathing super-emitters as well; perhaps they just have much better lubrication of their vocal cords, whose vibrations help to produce droplets containing air bubbles that are loaded with tiny aerosols. It is not yet known what role super-emitters play in super-spreader episodes; at the moment, it’s just an interesting research lead to follow up.

When anyone becomes sub-clinically ill (no symptoms yet) and starts to project a cloud of virus when singing or speaking forcefully, that’s when one must stand back and hope there is good air flow to dilute the cloud. On an elevator, that’s difficult; fortunately, most elevator rides do not last as long as choir practice or an airplane trip.

Getting a few virus particles is generally thought not to pose a problem, but attention must be paid to special situations such as super-emitters, kissing, doorknobs, elevator buttons, hand­rails, or spending several hours down­wind of someone’s exhales. Such situations are more likely to deliver enough viruses to start a clinical-strength infection.

That’s one setup for infecting the choir in front of an infected singer, but to infect 75 percent of those in the room suggests multiple spreaders or enough aerosol hanging around in the room air to start new infections, something that better ventilation might have prevented. If the room air seems stuffy, always turn on fans and create some cross ventilation. This is a little difficult in airplanes and modern office construction; bringing fresh air in from the outside means heating or cooling it, which costs fuel; managers would much rather recirculate the air — and the exhaled aerosols.

To quote the CDC, “A cloth face covering is not intended to protect the wearer, but may prevent the spread of virus from the wearer to others.” Face masks are usually worn, as in an operating room, to protect others by limiting the projection of infectious material by the wearer. Even a face scarf can trap projected saliva and most of its viral load. And it will remind you not to rub your eyes, as will wearing gloves.

The aerosol super-emitters are not yet well-studied, and physicians cannot yet ask the lab to do a test that measures aerosol volumes during loud speech — but that[8] is coming soon, I should think, giving us better data on the percentage of normal super-emitters. In the meantime, the obvious precaution is to wear a cloth face mask in protect others, when closer than six feet away, as when riding elevators.

William H. Calvin, Ph.D., is a professor emeritus at the University of Washington School of Medicine in Seattle, and the president of CO2Foundation.org. Forthcoming is his seventeenth book, “Extreme Weather and What to Do About It.”

wcalvin@uw.edu, WilliamCalvin.org

[1] https://www.latimes.com/world-nation/story/2020-03-29/coronavirus-choir-outbreak

[2] https://covid19.healthdata.org/projections

[3] Lydia Bourouiba (2016). A Sneeze. N Engl J Med 375:e15
DOI: 10.1056/NEJMicm1501197

[4] Calvin, W. H. (2020). Extreme weather and what to do about it. Seattle, CO2Foundation.org/bk17.

[5] https://doi.org/10.1016/j.chom.2015.09.013

[6] Asadi, S., et al (2019). Aerosol emission and superemission during human speech increase with voice loudness. Sci Rep 9, 2348. doi.org/10.1038/s41598–019–38808-z

[7] The video is from the work of Ville Vuorinen, et al (2020). A 3D model of a person coughing in an indoor environment — how an aerosol cloud travels in the air. Text: https://www.aalto.fi/en/news

[8] Aerosols do not travel very far out of the mouth on their own momentum, but droplets of saliva can make it across the room in the airstream. The droplets contain bubbles that themselves contain aerosols, their viral load soon to be released as the water content evaporates, leaving viruses entangled with salts and proteins, still large enough particles for a good face mask to filter from incoming air.

That makes projected saliva a useful proxy for projected aerosol amounts. Simply weighing a coffee filter held in front of the mouth during a minute of loud speech or singing might be a quick way to spot super-emitters in the general population. It’s imperfect, but we are looking for real standouts, individuals who project ten times more droplets than others do.

See this article for general advice.