Editor’s note: The Salt Lake Tribune is providing free access to critical stories about the coronavirus. Sign up for our Top Stories newsletter, sent to your inbox every weekday morning. To support journalism like this, please donate or become a subscriber.
It hasn’t even been five months since health officials in Wuhan, China, reported unusual pneumonia cases to the World Health Organization.
But those five months have been the most active in the history of epidemiology. Since that report, we’ve learned so much about the coronavirus. One of the most important lessons? How the disease is spread.
In particular, so-called superspreading events seem to be a major cause of infections. One London School of Hygiene analysis
suggested that 80% of the secondary transmissions were caused by just 10% of infected people. In other words, if you want to avoid getting COVID-19, one of your major focuses should be avoiding a superspreading event.
So as Utahns leave their homes and reengage with society, we thought now would be a good time to scour the research to note where these events have been documented and where they haven’t. We can also learn about the circumstances that led to each superspreading event, and do our best to avoid them.
The result is the following compendium, in alphabetical order, of public places and the lessons we can learn.
A 29-year-old South Korean man who didn’t know he had the virus went barhopping on April 30, after the country loosened restrictions. He and a friend went to five bars and clubs
in Seoul. To date, 187 positive cases
have been tied to this night on the town, including 93 people who visited the same nightspots as the pair. At these clubs, it is pretty clear they did not maintain social distancing restrictions
Because of South Korea’s frankly incredible contact tracing operation, we know a lot about these cases. Researchers say 5,517 people
attended the five spots the man visited during the timeline of possible spread. But because these were gay clubs — and homosexuality isn’t always accepted in Korea — many wrote down fake names upon entry. As a result, Korean health officials hadn’t been able to contact about 700 of them as of May 13.
We also know a lot about the secondary cases. For example, a 25-year-old teacher got the virus at one of those clubs and later infected 11 people, including five students at the school. As a result, 1,000 members of the church that those students attend have been told to stay home and monitor for symptoms.
In all, over 35,000 people
have been tested in relation to this cluster. 35,000! And 8,500 police officers
were deployed to stop the outbreak. In reaction, South Korea pushed back
opening schools one week. But about a month later, Korea’s efforts to quarantine those associated with the cluster seem to have worked.
• Dance floors are probably incompatible with social distancing.
• High-capacity bars and clubs are going to be potential locations of superspreading events.
Japanese public broadcaster NHK
, in consultation with local health experts, got 10 people to dine at a buffet. They put fluorescent paint on the hands of one person to simulate the virus after a sneeze or cough. After 30 minutes, they turned the black light on to see where the virus had gotten.
The answer: everywhere. People’s hands, faces, plates, napkins, and more were covered with the virus-simulating paint.
Salad bar restaurants like Souplantation and Sweet Tomatoes have decided to permanently close
. In Utah, during the moderate and low stages of virus risk, buffets must be served by a restaurant employee; they can’t be self-serve.
• Self-serve buffets probably create an unmanageable amount of viral spread.
In January, an asymptomatic woman who had recently been to Wuhan took a bus
with 66 other passengers to a worship event at a Buddhist temple. She felt sick after returning home.
The ride was 50 minutes each way and passengers stayed in the same seats on both legs of the trip. Of the 66 passengers, 24 were infected. There was no statistically significant correlation between how close the passengers sat to the woman — labeled IP below — and whether or not they got the disease. Many further than 6 feet away caught it.
The initial infected passenger in row 8B seemed to spread the virus to many others on her bus, designated by the pink, green, and yellow shading. (https://www.researchgate.net/publication/340418430_Airborne_transmission_of_COVID-19_epidemiologic_evidence_from_two_outbreak_investigations)
The bus’ air-conditioning was set to recirculation mode. The study’s authors note that all passengers who sat next to a window on the left side of the bus — where the air-conditioning vents were — remained healthy other than the one person who sat next to the sick woman, but they don’t know if that’s by chance or due to air circulation patterns.
A United Kingdom study found that male bus drivers
were more likely to die from COVID-19 than the average citizen, at the rate of 26.4 deaths per 100,000 males.
• In closed, small environments, virus transmission can occur throughout a room.
• Keeping outside ventilation high on buses seems to be a worthwhile goal.
• Bus drivers should receive protective gear to prevent infection.
It may be the single most famous outbreak in the U.S.: the Skagit County, Wash., choir practice. Last week, the Centers for Disease Control and Prevention compiled the results
of its contact tracing.
The choir met every Tuesday evening until March 10. At that last meeting, 61 members were present and chairs were arranged close together in six rows of 20 with many empty chairs. They practiced for 40 minutes together, for 50 minutes separated into two smaller groups, and then for 45 minutes sang together again. There was a 15-minute break between the second and third session for oranges and cookies, but many didn’t eat.
No one reported physical contact between members and most everyone left immediately after practice. Hand sanitizer was distributed.
But, in the end, 53 of the 61 contracted the coronavirus. Three were hospitalized, two died. Interestingly, the CDC did find someone who had coldlike symptoms in the choir beginning March 7 — that person is the most likely Patient Zero in this case, but it’s not certain that person was responsible for the spread.
This seems to happen repeatedly. The Amsterdam Mixed Choir gave a performance March 8, then 102 out of 130 singers
tested positive. Fifty members of the Berlin Cathedral Choir tested positive as well.
• Singing appears to significantly raise the likelihood of transmission.
On Tuesday, the CDC released a report
about a rural Arkansas church. The pastor and his wife began feeling sick March 10 and 11 after hosting a three-day children’s event a few days earlier.
The event included 1.5 to 2 hours of indoor church services on each day. The children also competed each day to collect offerings from adults. There was some singing, in which most children and some adults took part. On one day, the church hosted a buffet.
After an investigation, another husband and wife said they were symptomatic during the children’s event. On March 11, the church hosted a Bible study, though the pastor said people followed social distancing procedures.
In all, 92 people attended the church between March 6 and 11, and of the 45 tested, 35 became infected, seven were hospitalized, three died.
The Washington Post has a good review
of what happened in a church in South Korea on a larger scale. Patient No. 31 in South Korea went to her Shincheonji church while sick, and a month later, the cluster had accounted for 5,080 cases, more than half that nation’s total. Wow.
• Churches can be the site of community-changing superspreading events.
• High-risk activities like singing and buffets may make church gatherings more dangerous.
This story is heartbreaking.
In February, a man who had recently traveled out of state attended a close family friend’s funeral
in Chicago. He was experiencing minor symptoms at the time. The night before, he attended a potluck-style meal with the family of the deceased, and reported hugging some of them. Four of those he hugged would contract the virus. One died.
Three days after the funeral, the same condolence-giving, out-of-state-traveling man went to a birthday party for a member of his own family, which nine people attended. The man embraced others and shared food. Seven of the nine became infected; two of them died.
Some of the people this man infected passed the virus on to others, including a fellow churchgoer and health care professionals.
This one poor guy went to a funeral and birthday party and it led to the infection of 16 people and three deaths. We know more about the disease than we did in February. Don’t be that guy.
• Avoid hugging and sharing food, especially while sick.
One manager concealed a report
of coronavirus in his store in the Shandong province of China. Later, 17 people tested positive.
The United Food and Commercial Workers union released a report Wednesday
saying that 68 grocery workers have died of the virus.
But there’s relatively little evidence that says grocery stores are the site of superspreading events, and given that they have stayed open throughout the crisis, you’d think we’d know by now. There was one study that used a computer model for how airborne particles would spread in a grocery store
, but a computer model isn’t reality.
• Constant movement in spacious buildings probably helps prevent exposure to coronavirus droplets.
In Cheonan, South Korea, a workshop for Zumba fitness instructors — Zumba is surprisingly big in Korea — was held Feb. 15. Of the 27 instructors in attendance, eight eventually tested positive.
These instructors then went to their communities and taught two 50-minute Zumba classes each week. By March 9, less than a month later, there were 112 cases tied to this Zumba workshop; 57 were students of the positive instructors, the rest secondary infections from those students. In all, those classes had 217 students attend, for an infection rate of about 26%.
Of the eight positive instructors, there were two
who led most of the classes where people got sick. One held a class in a gym that had five students — but three got infected. Later, another COVID-19 positive instructor taught Pilates and yoga to 25 people in that same room. None got the virus.
Two hypotheses: Either the Pilates instructor wasn’t as contagious as the Zumba instructor, or high-intensity dance classes like Zumba mean more heavy breathing — and therefore more virus-laden droplets — than low-intensity Pilates.
In general, studies that have looked at infection
in gyms have found them to be high-transmission areas, especially when there is insufficient ventilation.
• Gyms can quickly spread the coronavirus, especially when instructors become infected.
• High-intensity workouts may be more dangerous than low-intensity workouts, though that’s unclear.
In Wenzhou, China, 34 people were infected
in a shopping mall. The cluster’s first recorded case involved a person who worked on the mall’s seventh floor, where it quickly spread to other nearby employees.
A few days later, staff who worked on lower floors started getting the disease, and a few days after that, customers started getting infected.
Spread of coronavirus in a Wenzhou mall in China. https://wwwnc.cdc.gov/eid/article/26/6/20-0412_article
However, the staff with the virus on the floors below No. 7 and the customers who tested positive all denied knowingly coming into contact with one another. It’s possible they all brought the disease to the mall separately, but given that they all developed symptoms at nearly the same time and the high number of cases associated with this mall, it’s unlikely. Everyone did share restrooms and elevators, which could explain spread from floor to floor or store to store.
In another case in Tianjin, China, 21 people were infected
in a mall department store: six staff, and 15 customers over the course of five days. The store was especially crowded for Chinese New Year, so reports were that up to 10,000 people would need to be quarantined because they visited during the outbreak period. The study’s authors struggled to figure out the disease’s spread through the store staff and customers.
• Thanks to high traffic numbers, one mall or large store can infect many people, even though the likelihood of any individual customer being infected is low.
In early March, the South Korean version of the CDC was informed about an outbreak in a call center in a 19-story building. The first 11 floors are commercial offices — the call center is on floors 7-9 and 11. Residents in apartments live on the floors above that.
In response, South Korea just started testing everyone. That included 811 people at the call center, 111 people who work in the floors below, 203 people who live in the apartments above, and 20 visitors.
Of those 1,145 people, 97 tested positive and 94 of them worked on the 11th floor in the call center. The other three worked on the 10th or ninth floor. In fact, 89 of the 97 worked on the same side of the same floor of the call center. Positive cases are in blue.
A coronavirus outbreak in a call center in South Korea. Persons at desks shaded blue contracted the virus. https://wwwnc.cdc.gov/eid/article/26/8/20-1274_article
As the study notes, “residents and employees in the building had frequent contact in the lobby or elevators,” but as you can see, widespread transmission didn’t occur there. The vast majority of it happened in the 11th floor on one side of the call center.
• The most dangerous method of transmission in an office is spending a long time near an infected person.
• Using shared facilities like restrooms with an infected person appears to be less dangerous.
• Spending a short amount of time with an infected person, like on an elevator ride, is not especially dangerous.
• Jobs with frequent talking
, like call centers, do appear to have elevated risk for superspreading.
We don’t have any direct studies on COVID-19’s transmission on airplanes, but we do have some on a related coronavirus: SARS-CoV-1.
On one plane
, a 72-year-old infected man with a cough in seat 14-E of a flight from Hong Kong to Beijing ended up passing SARS to 18 passengers and two flight attendants. The flight was about three hours long. While those seated at the very front and very back of the plane were safe, those infected included those within several rows of the man, including on the other side of the aisle.
A plane with a SARS-infected person in 2003. The black seat represents the index case, while those shaded in gray were infected on the plane. (https://www.sciencedirect.com/science/article/pii/S2214999616307421#bib13)
Other planes studied, however, did not have that level of transmission. As researchers studied the introduction
of SARS to France, a nearly 15-hour flight from Hanoi to Paris infected only two of the 69 people aboard. On two other flights with an infected SARS passenger, only one person out of 561 people was infected. Both of those flights were shorter, though, just 90 minutes long.
A recent opinion piece
in The Washington Post from Joseph Allen, a professor of exposure assessment science at Harvard, asserted that “You don’t get sick on airplanes any more than anywhere else,” because the ventilation system requirements for airplanes meet or exceed CDC regulations for COVID-19 rooms in hospitals.
• Don’t go on a plane with a cough. Everyone should wear a mask.
• If there is someone with a cough, those viral particles can likely travel more than 6 feet, but probably not throughout the plane.
• The odds of an outbreak on any individual plane trip are low.
In Wisconsin, 400,000 people went to various polling locations to vote in the April 7 primary election. Since then, people have wanted to know whether the in-person voting led to a spike in coronavirus cases.
At the end of April, the Wisconsin Department of Health announced that 52 voters or poll workers had tested positive
so far. However, it wasn’t clear if that was more than the normal number of cases you’d expect.
(Rick Wood | Milwaukee Journal-Sentinel via AP file) Waukesha settles on one polling place for in-person voting for Election Day due to the coronavirus Tuesday, March 31, 2020 in Waukesha, Wisconsin.
Three studies have looked at the issue since. One, released April 28
by a team from a Milwaukee hospital, found that coronavirus rates actually declined in the 10 days after the voting when compared to the 10 days before it.
A second study released a day later
from researchers at Stanford looked at hospitalization rates as well, and also didn’t find any increase after the election.
However, a third study released May 11
found that counties with higher numbers of in-person voting per location did have higher coronavirus rates after the election. Researchers estimated that when the number of voters per location increased by 100, a 3.5% increase in coronavirus rates was found.
Regardless, the election clearly did not cause a huge spike, like some doomsayers predicted. This could be due to the precautions polling places and voters took, or it could be that quick interactions like voting are unlikely drivers of coronavirus transmission.
• In-person voting either has a small or neutral impact on coronavirus spread when precautions are taken.
The New South Wales district of Australia looked at 15 schools
that were open between March 5 and April 3. That included 10 high schools and five primary schools.
There were 18 coronavirus cases split evenly, nine students and nine staff. But out of 863 close contacts with those people — of which 384 were tested — only two secondary cases were discovered: one in a high school, and one in a primary school. That’s a tiny transmission rate.
However, the schools weren’t necessarily operating as normal. Many kids chose to stay home, which may have made social distancing more feasible. Large assemblies were avoided, and more outdoor classes were held.
Another study looked at a 9-year-old from Britain
who attended school while symptomatic. The 73 students he came into close contact with all tested negative. For whatever reason, coronavirus transmission to and from kids
just doesn’t happen as often as you’d think.
Still, because schools have been closed nearly everywhere, we’re not certain that the virus can’t effectively spread in classrooms. And there’s still a lot of mystery here: Serology studies show that kids are getting the virus
somehow, and viral load studies
are showing kids have just as much virus in their throats as adults, so why haven’t we witnessed spread in schools?
• Schools are a significant source of spread for other diseases like influenza, but so far, not the coronavirus. We don’t know why.
• A individual who has contact with an infected child in a school is unlikely to be infected.
Sports venues and stadiums
It was a great day for Atalanta. On Feb. 19, the Italian soccer team from Bergamo defeated Spain’s Valencia 4-1 in the Champions League, the pinnacle of European soccer. About 40,000 people from Bergamo went to watch the match in the San Siro stadium in Milan.
It also led to a virus explosion in Bergamo a week later.
“It’s very probable that 40,000 Bergamo citizens in the stands of San Siro, all together, exchanged the virus between them,” Bergamo Mayor Giorgio Gori said on Facebook
. “So many Bergamaschi that night got together in houses, bars to watch the match and did the same.”
(Antonio Calanni | AP file photo) The Champions League round of 16, first leg, soccer match between Atalanta and Valencia at the San Siro stadium in Milan, Italy, on Feb. 19, 2020, was the biggest soccer game in Atalanta’s history and a third of Bergamo’s population made the short trip to Milan’s famed San Siro Stadium to witness it. Nearly 2,500 fans of visiting Spanish club Valencia also traveled to the Champions League match. More than a month later, experts are pointing to the Feb. 19 game as one of the biggest reasons why Bergamo has become one of the epicenters of the coronavirus pandemic — a “biological bomb” was the way one respiratory specialist put it — and why 35% of Valencia’s team became infected.
A journalist who covered the match was Bergamo’s second positive case. And then 35% of Atalanta’s staff
tested positive. More than a third of Valencia’s players would eventually test positive. The director of health there called the match a “biological bomb.”
This can happen on a smaller scale, too. On March 6, 2,800 people showed up
for sectional basketball semifinals at a high school in Lawrence, Ind. While testing wasn’t widespread, five of the attendees would later die, and “at least a dozen others” tested positive or had virus symptoms. Decatur County, Indiana had a per-capita death rate that was twice as high
as any other Indiana county — they blamed their spread on another high school basketball game.
• Getting thousands or tens of thousands of people together in one building can result in community-changing “biological bombs."
• Sports celebrations (singing, hugging, cheering) could potentially mean more spread.