Hours after a federal judge struck down the federal mask mandate covering air travel and other public transportation last month, Delta Airlines celebrated the move in a statement saying that Covid-19 “has transitioned to an ordinary seasonal virus.” By the next day, after an intense backlash from public health experts, Delta had taken the offending language down.
“‘Ordinary viruses don’t cause 1 million deaths in one country in just 2 years,” tweeted epidemiologist Jessica Malaty Rivera, a senior advisor at the Rockefeller Foundation’s Pandemic Prevention Institute.
SARS-CoV-2 remains a long way from being ordinary. It has not yet found seasonal cadence — take the recent surge in Europe and the U.K., which comes just weeks after the initial Omicron wave subsided — and it’s still capable of inflicting mass death and disability (see Hong Kong’s lethal last few months).
But there are signs that the virus — and our relationship to it — is shifting in subtle ways that make it more like seasonal flu than it was at the start of the pandemic.
When everyone’s a superspreader, no one is
One of the most intriguing shifts involves how Covid now spreads from person to person.
Early on, a hallmark of SARS-CoV-2 transmission was that the majority of infections hit a dead end. A 2020 study from Hong Kong found that 80% of new infections were caused by just 10% to 20% of cases, often in indoor superspreading events. That meant most people didn’t spread the virus to anyone else.
Scientists call this phenomenon of patchy transmission “overdispersion.” Dispersion is a measure of how uniformly a pathogen spreads — does it steadily chug along or break out randomly in big bursts? And though its importance has long been overlooked, understanding dispersion is critical for developing effective infection prevention strategies.
“It’s a reason why some infections are more controllable than others, even for the same reproductive number,” said Benjamin Cowling, an infectious disease epidemiologist at the University of Hong Kong who led the 2020 study on overdispersion. It’s a lesson some countries learned right away.
When Hitoshi Oshitani, a virologist and infectious disease specialist at Tohoku University Graduate School of Medicine saw data from Japan’s retrospective contact tracing teams, he realized that most transmission was being caused by a few infectious people gathering in poorly ventilated indoor spaces like gyms and restaurants. If they could cut the occurrence of those potential superspreading events, they had a chance at containing the virus. Oshitani, who advises the Japanese government, suggested a simple mantra that became key to the country’s Covid-19 success: avoid closed spaces, crowded places and close-contact settings — later known as the Three C’s.
But as more infectious variants have emerged, Oshitani’s team has observed changing transmission patterns in Japan. Rural areas that avoided SARS-CoV-2 surges in earlier waves have been inundated with Omicron, he told STAT via email. Clusters are also showing up more in schools and nursing homes. And some data indicate that secondary attack rates in households are higher for Omicron — meaning if someone brings the virus home, more people they share a roof with are likely to contract it.
Studies in Norway and in the U.S. have also shown that Omicron spreads much more easily in households, suggesting that superspreading events may be becoming less important as primary drivers of contagion chains.
It’s not that superspreaders have become less super in their spreading; it’s that with Omicron, everyone else may be catching up with them.
“Epidemiologically the Omicron variant is quite different from previous strains,” Oshitani wrote. “The level of overdispersion is also probably different. But we need more data to adjust our public health responses.”
Cowling is also trying to understand how Omicron and other variants are changing the degree to which the virus spreads through clusters. But it’s become much harder to study as contact tracing programs in Hong Kong have buckled under the deluge of new cases.
“They had trouble keeping up with 100 cases a day in late 2020,” Cowling said. At the peak of Hong Kong’s fifth wave, earlier this year, between 50,000 and 100,000 new cases were being reported daily. “The system basically broke down,” he said. “So I’m not sure what exactly we’re going to be able to show beyond what was happening in the very early stages of that wave.”
To Seema Lakdawala, a microbiologist at the University of Pittsburgh School of Medicine, these steeper, shorter waves of infection caused by Delta, Omicron, and now BA.2 look like signs of a potentially important shift in the virus’s behavior.
“It could be that more individuals are now forward-transmitting and we’re seeing a move away from cluster transmission to one that is more linear like you would expect for flu,” she said. The original strain of SARS-CoV-2 was estimated to have a dispersion parameter, k, of around 0.1, meaning that fewer than 20% of people infected passed on the virus to someone else. Pandemic influenza is less sparing. With a k of around 1, it reliably hops hosts more than 60% of the time, based on estimates from the 1918 pandemic.
A recent modeling study led by Lidia Morawska at Queensland University of Technology found that the Delta variant is less reliant on superspreading events, with a k of 0.49. Her team hasn’t yet repeated the work for Omicron, but she expects that its preference for the upper respiratory tract, where it replicates at astonishing rates, probably results not just in more transmission, but more uniformity in who transmits to others.
“Even a very short time is sufficient to inhale enough of this virus to be infected,” Morawska said. “Short enough that ventilation may not have had a chance to remove the virus from the air.”
That’s why she and others are now pushing for the use of germicidal ultraviolet light, which can zap infectious viral particles in the air, killing them in an instant. This technology could have prevented the Gridiron superspreader event last month in Washington, University of Maryland aerobiologist Don Milton argued in a recent New York Times op-ed. Disinfecting UV light “should become the norm for large indoor gatherings where meals are served and masks cannot be worn,” he wrote.
Less susceptible targets
But the virus is just one component of what makes for a superspreading event. The other is the host network where it lands — which is a function of the current levels of population immunity and how many contacts people are making.
“Based on everything we’ve seen throughout the pandemic, the underlying population susceptibility seems to be the primary driver of spread,” said Emily Gurley, an epidemiologist at Johns Hopkins University. “I think that’s more important than changes in the virus itself.”
That means that even as SARS-CoV-2 has evolved to be more contagious, it is encountering a small and ever-shrinking proportion of the population whose bodies have never seen some version of it before. Most people, through prior infection, vaccination, or a combination of the two now have immune systems capable of fending off the deadliest outcomes of contracting SARS-CoV-2. And that’s starting to look a lot like what happened when pandemic flu transitioned to seasonal flu.
We’re not at an immunological détente yet. Over the last decade, seasonal flu killed about 30,000 people each year, on average. Covid-19 killed 148,000 people in the first four months of 2022 alone. Last year it was the third leading cause of death in the U.S., after heart disease and cancer.
But we’re not impossibly far away. And the availability of effective early treatments like Paxlovid in addition to vaccines will also continue to tilt the scales away from an encounter with SARS-CoV-2 turning into a lethal or disabling one, for most people. (Although millions of immunocompromised Americans remain at risk of those worst outcomes.)
“A lot of it comes down to immunological diversity,” said William Hanage, an epidemiologist at Harvard’s T.H. Chan School of Public Health.
“Gatherings are less likely to be as significant a component of spread at this point, but superspreading events will continue to be possible,” he said. “That’s what we’re seeing in D.C. at the moment. People who have avoided Covid thus far are making contacts they weren’t making six months ago, and so the virus is just making hay with all those contacts.”
It drives home another thing about dispersion — it can change based on people’s behavior too. In a study that is currently under review, Cowling’s team found that as Hong Kong got more strict about limiting large gatherings and requiring masks in public places during initial waves of the virus, its overdispersion actually went up, completely contradicting what they expected.
“We thought if we stop the superspreader events then the secondary case numbers will be ones or twos not tens or twenties,” said Cowling. “And that happened some. But what also happened is there were a lot more people who didn’t transmit to anyone. We think of overdispersion as just the big numbers, but it’s also the zeroes.”
And the zeroes are a lot harder to keep track of. So it’s possible that superspreading events have been overestimated for Covid-19 because they’re big dramatic events—like the Skagit Valley Chorale or the Gridiron gala.
It’s also possible that flu might actually be more like Covid-19 than we appreciate. “I suspect that there’s actually a lot of superspreading with flu, we just haven’t studied it in the same way that we have for this coronavirus,” said Cowling.
‘Flu-like’ genetic drift
Flu and SARS-CoV-2 are starting to resemble each other at a more basic level too — how they evolve.
During the first year and a half of the pandemic, new variants of concern arose from distantly related branches of the SARS-CoV-2 family tree. Delta didn’t arise from Beta, which didn’t arise from Alpha. The constellation of mutations each new strain acquired that gave it a competitive advantage evolved independently. Omicron was an even more extreme example of this.
The variant popped onto the scene in South Africa at the end of 2021 looking like a version of the virus that hadn’t been seen in someone since mid-2020, leading experts to speculate it went underground either in an immunocompromised person’s body or into a different species entirely.
But since Omicron has spread around the world, the new variants that have emerged and outpaced it — BA.2, BA.1, BA.4, and others — have all splintered off from the same starting point. This sort of ladder-like accumulation of mutations is something much more characteristic of how influenza evolves. In general, there’s one major lineage that “drifts” year to year, inching toward more immune escape and higher transmissibility, rather than leaping out of nowhere.
This could be good news, because more stable, predictable evolution would make it easier to develop meaningful Covid-19 vaccines and boosters, better tuned to handle whichever version of the virus will be circulating six to 12 months from now, as Trevor Bedford, a computational biologist at the Fred Hutchinson Cancer Research Center pointed out in a recent Twitter thread.
But, as he noted, it’s hard to say how long this “flu-like drift” will last before another huge evolutionary jump, like Omicron, might take place.
“Nobody knows what this virus is going to do next,” said Hanage. That’s why he thinks the viral videos of flight attendants collecting masks sends the wrong message. “This is not a thing that ends in a way that people understand. The pandemic will be done but not in the way that most people think of as done.”