SARS-CoV-2, the coronavirus that causes the disease known as COVID-19, is a novel virus. It is undeniably true that science has made many discoveries about the virus and disease in record time (remember, the first cases were only reported only late last year in China), including isolating the offending coronavirus and sequencing its genome in record time, developing tests for it, and even finding potential treatments that appear to decrease mortality, such as dexamethasone and remdesivir. However, even with the rapid pace of discovery, some of it based on what we know about other coronaviruses, including SARS and MERS, and at times showing science at its messiest, thus leaving the door open to grifters and conspiracy theorists, there still remain many mysteries about COVID-19, how it is spread, how its spread can be prevented and slowed, and why a minority of people become ill with life-threatening disease while the majority do not, with a third or more having no symptoms at all, or symptoms so mild that they don’t think much of them.
We’ve written before about the recent controversy over whether COVID-19 can spread through aerosols (answer: probably, but it’s complicated) and whether masks work to slow the infection (answer: yes, contrary to the misinformation being spread about masks). These are not just academic questions. Given how COVID-19 is spreading rapidly in many parts of the US, perhaps the most urgent question remains, amazingly enough; How does COVID-19 most efficiently spread? If we can target how COVID-19 spreads, we have the best chance of having the most impact in slowing down its spread. This brings us to the topic of so-called “super-spreaders” and “super-spreading events”. It’s a great example of how confusing science at the “bleeding edge” can be and, more importantly, how hard it is to make decisions based on science when the science is still uncertain and in flux.
An article yesterday in the Washington Post by Ariana Eunjung Cha, unfortunately, starts with just such an event in my state:
It wasn’t until Day 7 of her team’s coronavirus investigation when it dawned on Linda Vail, the health officer for Michigan’s Ingham County, that this was going to be a big one. It had started with just two infections at the college bar on June 18, not long after the state began reopening. But the numbers quickly jumped to 12, then 18, then 34.
As of Friday, she was staring at a spreadsheet with 187 infected at Harper’s Restaurant and Brew Pub.
“The tables were six feet apart, but no one stayed there,” she said. “The DJ was playing music so people were shouting, the dance floor started to get crowded. We had flattened the curve and then boom.”
The East Lansing case is what’s known as a super-spreading event — possibly the largest so far in the United States among the general public. Many scientists say such infection bursts — probably sparked by a single, highly infectious individual who may show no signs of illness and unwittingly shares an enclosed space with many others — are driving the pandemic. They worry these cases, rather than routine transmission between one infected person and, say, two or three close contacts, are propelling case counts out of control.
More than 1,000 suspected clusters — ranging from the single digits to thousands — have been logged in a database compiled by a coder in the Netherlands. A megachurch in South Korea. A political rally in Madrid. An engagement party in Rio de Janeiro. Nearly all took place indoors, or in indoor-outdoor spaces.
These “super-spreader” events all share certain characteristics. Specifically, nearly all of them took place indoors or in indoor-outdoor spaces. Unfortunately, although so-called “COVID parties“, parties in which people intentionally try to be infected with COVID-19, are probably far more urban legend than reality, there are lots of events still occurring in which people don’t wear masks, don’t social distance, and are in general very cavalier and careless about the possibility of contracting COVID-19, such as underground parties being held in New York City.
While we know that COVID-19 is mostly spread by larger respiratory droplets, there’s emerging evidence that it can also be spread by aerosols, smaller respiratory droplets that can hang in the air a lot longer. (Again, I discussed this issue in detail a couple of weeks ago.) Now, here’s the strange part:
Why, for instance, didn’t the earliest infections in the United States, or the infamous Lake of the Ozarks party, spur lots of cases, while a much smaller gathering at a Michigan bar produced nearly 200? Why out of countless large gatherings held — church services, soccer games, choir rehearsals, and Zumba classes — did only a fraction ignite significant infections?
Part of the uneven spread of the coronavirus — and the phenomenon of superspreading — can be explained by extreme individual variation in infectivity, researchers say.
Some people will not transmit the virus to anyone, contact tracing has shown, while others appear to spread the virus with great efficiency. Overall, researchers have estimated in recent studies that some 10 to 20 percent of the infected may be responsible for 80 percent of all cases.
Scientists are only starting to understand the different factors — physiological, behavioral, environmental — that play a role in amplifying transmission.
There’s a term used by infectious disease experts known as R0 (pronounced “R-not”), which is an estimate of the average number of people a single individual with an infectious disease will infect. Measles, a disease we used to discuss a lot in the context of the antivaccine movement’s depressing vaccination coverage with the MMR vaccine and the measles outbreaks that have resulted, is one of the most easily transmissible viruses there is, with an R0 of around 12-18, while Ebola, which I’ve also discussed before in the context of comparing its transmissibility with measles, has an R0 of around 2, but that number is an uncertain estimate because it depends on a lot of factors.
The new coronavirus turned out to have a reproductive number somewhere between two and three. It’s impossible to pin down an exact figure, since people’s behavior can make it easier or harder for the virus to spread. By going into lockdown, for instance, Massachusetts drove its reproductive number down from 2.2 at the beginning of March to 1 by the end of the month; it’s now at .74.
This averaged figure can also be misleading because it masks the variability of spread from one person to the next. If nine out of 10 people don’t pass on a virus at all, while the 10th passes it to 20 people, the average would still be two.
One critical factor is that COVID-19 can spread from people who are either presymptomatic (they don’t have symptoms yet but go on to develop symptoms) or asymptomatic (they never go on to develop symptoms), meaning that people who can transmit the disease often interact with people while releasing the virus. One prominent common trope spread by COVID-19 conspiracy theorists and deniers of the severity of the pandemic is that asymptomatic COVID-19 patients do not spread the virus, but the evidence for asymptomatic spread has become much stronger over the last couple of months, along with the evidence that facemasks work to slow the spread of coronavirus.
There also appear to be large differences in how effectively a given infected person spreads COVID-19. For example, in a recent study out of Hong Kong that is currently available as a preprint—as always, remember that this means it hasn’t undergone peer review yet—researchers investigated several COVID-19 clusters and looked at the results of contact tracing to determine the chain of transmission. What they found in these “super-spreading events” is that about 20% of those infected were responsible for about 80% of viral transmission. Another group, consisting of around 10% of the total, infected one or two others, consistent with an R0 of around 1.5, while the remaining 70% didn’t infect anyone at all. The reasons for this difference were not entirely clear:
Superspreading is considered a function of both variations in individual transmissibility and individual susceptibility or exposure. Our results show that the number of individual secondary cases was significantly higher within social settings such as bars and restaurants compared to family or work exposures (p<0.001). This is certainly due to the greater numbers of contacts expected in such settings. Social exposures are therefore at an increased risk for SARS-CoV-2 transmission and likely constitute the core behavioural risk factor for SSEs. Targeted interventions should therefore focus on reducing extreme numbers of social contacts at high-risk venues such as bars, nightclubs and restaurants, which also appear at increased risk of SSE (22), either via closures or physical distancing policies, both of which currently remain implemented in Hong Kong (17).
There are other studies that come to similar conclusions. For example, in this one from Israel, currently in preprint, investigators sequenced 212 SARS-CoV-2 nucleotide sequences and used the information to perform a comprehensive analysis to trace the origins and spread of the virus, estimating an R0 of around 2.0 to 2.6 and noting large differences in transmission of SARS-CoV-2, with around 1-10% of infected individuals resulting in 80% of secondary infections. Another study, this one from Georgia and also in preprint, estimates that 2% of individuals seeded 20% of cases. There’s also another parameter, the k-value, which estimates how much viral infections tend to cluster, which has been estimated to be around 0.1, indicating that 10% of infected people might be responsible for 80% of secondary spread.
These observations have significant implications, if they are borne out in further research. As noted in May in Science:
That could explain some puzzling aspects of this pandemic, including why the virus did not take off around the world sooner after it emerged in China, and why some very early cases elsewhere—such as one in France in late December 2019, reported on 3 May—apparently failed to ignite a wider outbreak. If k is really 0.1, then most chains of infection die out by themselves and SARS-CoV-2 needs to be introduced undetected into a new country at least four times to have an even chance of establishing itself, Kucharski says. If the Chinese epidemic was a big fire that sent sparks flying around the world, most of the sparks simply fizzled out.
Especially compared with other viruses:
The lower k is, the more transmission comes from a small number of people. In a seminal 2005 Nature paper, Lloyd-Smith and co-authors estimated that SARS—in which superspreading played a major role—had a k of 0.16. The estimated k for MERS, which emerged in 2012, is about 0.25. In the flu pandemic of 1918, in contrast, the value was about one, indicating that clusters played less of a role.
So we know that COVID-19 tends to spread in clusters. What could be the factors that determine who is and is not a “super-spreader”? We already know that enclosed indoor spaces are conducive to the spread of coronavirus. It’s been speculated that the situation matters more than the person infected. Yes, there could be biological differences among people regarding how much coronavirus reproduces and how much virus is in their respiratory droplets, but scientists suspect it’s the situation that matters more:
Some people also have more opportunity to get sick, and to then make other people sick. A bus driver or a nursing home worker may sit at a hub in the social network, while most people are less likely to come into contact with others — especially in a lockdown.
Dr. Nelson suspects the biological differences between people are less significant. “I think the circumstances are a lot more important,” she said. Dr. Lloyd-Smith agreed. “I think it’s more centered on the events.”
A lot of transmission seems to happen in a narrow window of time starting a couple days after infection, even before symptoms emerge. If people aren’t around a lot of people during that window, they can’t pass it along.
And certain places seem to lend themselves to superspreading. A busy bar, for example, is full of people talking loudly. Any one of them could spew out viruses without ever coughing. And without good ventilation, the viruses can linger in the air for hours.
Still, scientists don’t discount the possibility of biological differences as a cause for “super-spreading”:
While it’s often impossible to identify the person who triggered an outbreak, there have been some commonalities among those who have been pinpointed as the likely source in studies. They tend to be young. Asymptomatic. Social.
Scientists suspect these “super-emitters” may have much higher levels of the virus in their bodies than others, or may release them by talking, shouting or singing in a different way from most people. Research based on the flu, which involved college students blowing into a tube, showed that a small percentage tended to emit smaller particles known as aerosols more than others. These particles tend to hang or float, and move with the flow of air — and therefore can go much farther and last longer than larger droplets.
In a study published in Emerging Infectious Diseases by Japan’s Hitoshi Oshitani at Tohoku University of 22 superspreading individuals with the coronavirus, about half were under the age of 40, and 41 percent were experiencing no symptoms.
This is mainly speculation, though. There is a paucity of evidence supporting even the existence of especially infectious people with COVID-19.
Personally, I’m more on the side of concluding that it’s more the situation, than the biology of the infected, that determines who can be a “super-spreader”. However, I also did several PubMed searches, and I must admit that the literature on this phenomenon is very confusing right now. It also doesn’t help that “super-spreader” doesn’t have a very clear epidemiological definition, as discussed in depth in this article by Emma Cave. She notes that this is not a new term and is defined (usually in retrospect) as an individual who has a greater than average propensity to infect a larger number of people. (Perhaps the most famous example of a “super-spreader” was Typhoid Mary.)
However, the terms “super-spreader” and “super-spreading” are problematic for a number of reasons, as Cave argues. Again, its precise meaning in epidemiology remains rather vaguely defined. More importantly, though, it is a term that can easily take on a moral judgment and be used to blame certain individuals:
The term is particularly problematic when applied to individual ‘super-spreaders’, as it can mean different things to different groups. Media interest in super-spreaders focuses on the early stages of the epidemic when efforts are being made to contain, trace and delay. Used in this way, a ‘super-spreader’ will generally have interacted with a larger than average number of people, making tracing difficult or impossible. At the other end of the spectrum, scientific interest can focus on the heterogeneity of populations in the transmission of infectious disease. Used in this sense, ‘super-spreading’ is connected to the scientific nature of the virus and the way it manifests in some humans. There is speculation that some people with COVID-19 are especially infectious (Boseley and Belam 2020).
Super-spreading is therefore a product of biological, behavioural and environmental factors. It can be used to describe decisions, policies, events, settings and individuals—in fact, anything that contributes to increased rates of infection can be seen (by some groups) as super-spreading. The wide and varied use of the term ‘super-spreader’ is problematic for two reasons considered in subsequent sections: it can lead to apportionment of moral blame to alleged super-spreaders and it could detract from scientific investigation into heterogeneity of COVID-19 if misunderstanding leads to diminished public support and trust.
Cave concludes by proposing that a different, more neutral term be used. She makes a good argument.
The COVID-19 pandemic has, unfortunately, provided a perfect storm of information mixed with misinformation and rapidly changing science that confuses not just the general public but the public policy response. The entire issue of “super-spreading” is one such issue. On the other hand, if “super-spreading events” really are the main driver of the COVID-19 pandemic, it could be possible to tailor a public health response that isn’t so blunt, that doesn’t limit so much of what we can do. For example, in Japan health officials have noted that “many COVID-19 clusters were associated with heavy breathing in close proximity, such as singing at karaoke parties, cheering at clubs, having conversations in bars, and exercising in gymnasiums”, and Japan’s Prime Minister’s Office and the Ministry of Health, Labor and Welfare announced three situations that could increase the risk for COVID-19 cases and advised the population to avoid the “Three Cs”: closed spaces with poor ventilation, crowded places, and close-contact settings.
And wear a mask!
I’ll conclude by warning that the tsunami of misinformation mixed with rapidly evolving science is likely to continue to get worse before it gets better. There’s an election coming up, and it’s likely that science will be politicized even more than it’s been thus far, particularly as various candidates for vaccines go into clinical trials and preliminary results are breathlessly reported to the press. That, however, is a topic for another day.