For the past two months, the earliest known community spread of Covid-19—passed between people here in the United States, rather than a traveler bringing it home from abroad—was a case identified at the UC Davis medical center on February 26. Now we have confirmation the virus was circulating here far earlier than that. Last week, health officials in Santa Clara County, California, announced that a 57-year-old woman died of Covid-19 at her home in San Jose on February 6. Two others followed her in mid-February and early March. None of those people had recently traveled abroad. Their diagnoses were all confirmed posthumously from tissue samples taken by the county coroner and sent to the Centers for Disease Control for testing.
Health officials are likely to turn up more cases like those. “They’re really like iceberg tips,” said Sara Cody, the public health officer for Santa Clara County, at a press conference last week. “When you have an outcome like death … that means there’s an iceberg of cases of unknown size.”
The findings have underlined the early failures of the US’s pandemic response. As federal officials, including President Donald Trump, downplayed the crisis and cast it as a foreign problem, the virus was already here, slipping through border checks and spreading undetected due to a lack of tests. They also left people wondering once again whether they might have already had the disease and hoping they are now immune. (The answer, still, is probably not.)
For scientists who have studied the early spread of Covid-19 in the US, the virus’s timeline did not need to be rewritten. “There’s nothing even vaguely surprising about this finding,” says Arthur Reingold, an epidemiologist at the University of California, Berkeley. “It’s easy to be smart in hindsight, but all the evidence we have has pointed to this.”
“This is in line with our projections,” says Kate Coronges, director of the Network Science Institute at Northeastern University. Using travel patterns and data about the virus’s transmission, a team there recently estimated that tens of thousands of invisible cases were already occurring in US cities prior to March 1. As the new cases confirm, fatalities from the virus were invisible, too, going unexplained or labeled with the wrong cause of death. But even with the new information, their models still point to a likely mid-January introduction in states like California and New York.
Even if the experts expected this result, the new cases highlight some of the challenges of precisely understanding the early stages of the virus’s spread in the US—a process that could have lessons for containment if the virus roars back later this year after social distancing measures are relaxed. That starts with what researchers have learned about how the virus arrived here in the first place.
Much of what we know about the early spread of Covid-19 has been learned from sleuthing through the virus’s genetic code. As a virus spreads, hijacking the cellular machinery of its host’s cells to replicate its RNA, it picks up mutations at a relatively predictable pace. That allows researchers to walk backwards through the virus’s evolutionary history, tracing when mutations occur and where the virus splits off into new strains. Researchers have used those genetic techniques to trace the origins of the virus back to strains circulating in China, and to draw various paths that brought it to the US.
In the US, an analysis by Trevor Bedford, a computational biologist at the Fred Hutchinson Research Center, put the virus’s introduction at some point in the middle of January. But the virus took multiple routes. Some strains came via travelers arriving directly from China. Other strains, especially on the East Coast, arrived by way of Europe. From there, the various strains began to bounce around the country, as Americans continued to jet about prior to the enactment of shelter-in-place rules.
Those early introductions are tricky to model. The most common epidemiological graphs are those that show the size of an outbreak growing on a smooth curve. But that’s an oversimplification, says Kenah Eben, an Ohio State University epidemiologist. While those models assume anyone can give the disease to anyone, viral spread is chancier than that. People have different webs of interactions—they form knotted groups determined by family and friends, their commutes and workplaces.
Early on, before an outbreak is widespread, there’s a good chance that the virus won’t escape those clusters. That kind of clustering is difficult to model, but it’s critical for understanding the initial stages of an outbreak, Eben says. “It makes it harder for epidemics to break out,” he says. He compares these various introductions to “little embers”—some that fizzle out, and others that help spark and then fuel a broader conflagration, as seen in places like New York City.
It’s too soon to tell the full story behind these three early California deaths, says Lauren Ancel Meyers, an epidemiologist at the University of Texas at Austin. While it’s possible they’re the offshoots of imported cases that failed to take root, they could also be part of a low-level circulation of the virus that eventually bloomed into the outbreak we have now. “It’s consistent with both,” she says. “Those things were happening simultaneously.”
Genetic analysis of the samples sent to the CDC could allow researchers to provide an answer. But the CDC hasn’t recovered enough genetic material to do that type of work just yet.
Another route to understand the full scale of early Covid-19 infections would be to try to disentangle some of the early Covid-19 deaths from flu season data. But that would be difficult, says UC Berkeley’s Reingold, without having more confirmed early cases and deaths. California governor Gavin Newsom has announced that state officials plan to investigate more autopsy samples, going back to December. It’s possible more data could be coming from that state-wide autopsy effort. It’s also possible we’ll never know the exact answer. And that’s probably fine. “Once everything is on fire, the embers don’t matter as much,” Eben says.
But it is important that the US heed the lessons contained in the unknowns of how the virus arrived here—the way that it slipped through cracks in the border, then spread unseen—especially as health officials try to prevent future waves of infection. In the absence of proper screening and testing, the virus “spreads quickly and silently,” Meyers says. Her team’s models have suggested that, by the time a single case was unearthed in a given county, there was a greater than 50 percent chance that a sustained epidemic was already underway. That means government officials will need to be especially nimble with reinstating social distancing policies, if necessary. Even places thought to be well-equipped for testing and contract tracing, like Singapore and China, have had fresh outbreaks of the virus seeded by cases imported from abroad.
The new data points don’t change one basic fact about the pandemic’s spread: Only a small fraction of people have had the disease, and an even tinier fraction were infected before March, during its period of stealthy circulation. Early serological studies, testing for antibodies in the blood that indicate past infection, seem to back that up. So far, with the exception of very hard-hit places, like New York City and its immediate suburbs, the numbers are consistently in the low single digits. That suggests it’s very unlikely that the weird flu-like thing you had in January was Covid-19, and many more people could still catch it. As Cody put it last week: “We have a long, long, long, way to go with this virus.”
- What does Covid-19 do to your brain?
- An oral history of the pandemic warnings Trump ignored
- We need a vaccine—let’s get it right the first time
- Un-miracle drugs could help tame the pandemic
- WIRED Q&A: We are in the midst of the outbreak. Now what?
- Read all of our coronavirus coverage here
