The Premonition: A Pandemic Story

by Michael Lewis

He’d been promoted, basically overnight, from the world’s most ignored pandemic modeler to the world’s most important pandemic modeler.

Twenty-five million rode a bus to school. “I thought, Holy crap, half the kids in the U.S. hop on a school bus.”

He watched the way they horsed around and jumped on each other’s backs and behaved in ways that he no longer did. Look! They’re so different. They’re not little adults. They have a different sense of space.

As the teacher spoke, he reached out his arms in either direction. Look! he said to himself. It’s three feet. I can touch the person next to me.

It had taken the Glasses’ model to point it out. Why? Carter wondered. Why hadn’t they seen it? Then it struck him. They saw it all with adult eyes. They forgot the world that their kids live in, and that they once lived in.

At some point Richard and Carter realized that they’d need to change the minds of everyone working in and around public health, and that this meant first changing people’s minds inside the Centers for Disease Control.

The CDC felt more like an academic institution than a government bureaucracy.

The next day he sent, not just to Rajeev but also to Richard and Lisa, a single-spaced, thirteen-page memo, “Analysis of Philadelphia Outbreak 1918.”

In the end he plotted both the deaths and the restrictions imposed to prevent them, and saw that the earlier the restrictions imposed in any given outbreak, the fewer the deaths. In the case of Philadelphia, he wrote, “the closing of schools and churches, banning of public meetings, and banning of large public gatherings occurred relatively late into the epidemic”—nearly one month after the outbreak began and just a week before its peak.

The death rate in St. Louis was half that of Philadelphia because St. Louis’s leaders used the cover provided by the federal government to distance its citizens from one another.

It was seven months before the United States public-health system fully bought into the power of social distancing.

The official newly in charge of him, Heidi Avery, who came from some deep place in the intelligence community and was now called deputy assistant to the president for homeland security, told him that the Obama administration had decided to dissolve the Biodefense Directorate and fold it into something called the Resilience Directorate.

In terms of pandemic response more generally, he thought it was a big mistake: henceforth disease would be managed alongside all the other existential threats to American life, and, along with other, less sexy threats, be shoved into the background. George Bush’s moment of terror after reading John Barry’s book had led his administration to break with tradition and create an office that did nothing but worry about pandemics. Now Obama was about to do away with it.

A couple of months into the swine flu pandemic, Richard sensed that his journal might be a valuable historical document. Once it became clear that swine flu would come and go, like a massive hurricane that dissipated before making landfall, it became something else. A message in a bottle. A premonition. A warning.

The Mexicans, interestingly, had taken the new pandemic strategy of the United States and run with it. They’d closed schools, and socially distanced the population in other ways that, studies would later show, shut down disease transmission. The CDC, by contrast, sent the message that each American school should make its own decision, which was a bit like telling a bunch of sixth graders that the homework was optional.

But one set of experts, inside the White House, thought they should close schools; and the other, at the CDC in Atlanta, thought they shouldn’t. When Obama had asked Carter what he should do, Carter had told him that he ought to close schools until they knew what they were dealing with. Obama had instead listened to the CDC.

Carter agreed to stay until the end of Obama’s first term. Whenever some urgent problem arose that needed thinking through, Heidi Avery would bring it to Carter and ask him to think about it. Deepwater Horizon. Fukushima. The earthquake in Haiti. “She’d say, ‘Run ahead and see what you can find, and then come back and tell me,’ ” recalled Carter. She gave him a nickname: Scout.

She’d figured out that Carter Mecher saw things that other people didn’t.

Then, on the second line, she made a prediction. “It Has Started,” she wrote.

We would like you to come to DC and talk to the Jasons, it began.

He began by telling them about his new technology. Until now, he explained, the identification of a new virus has been a tedious, slow-moving process. The trouble with viruses is that they’re so very small. You need a lot of virus before you can see it,

The Virochip, as it was called, was actually a glass microscope slide. Its surface held genetic sequences from every known virus. These sequences, along with the genetic information of living creatures, were stored in a federally funded database called GenBank, inside the National Institutes of Health.

To the Jasons, Joe explained how the Virochip had taken the guesswork out of the investigation. When some new pathogen infected human beings, you didn’t need some expert virologist to make guesses about what it might be. You could now approach the pathogen without any knowledge or preconception, and allow its genes to tell you what it was. The biology would reveal itself, if you gave it a chance.

If the unidentified virus is a new virus, why would it stick to anyplace on the Virochip? People always asked this. All viruses on earth are genetically related, Joe explained, because they’d evolved from common ancestors. If a virus is new, and thus doesn’t match up perfectly with the DNA on the chip, the chip can still lead you to its family. Its grandparents or, at least, its distant cousins. The chip, in other words, could be used not just to diagnose an existing virus but to discover a new one, as it had with SARS.

It was after that first outbreak of a new coronavirus that Joe’s phone got named the Red Phone. “We had gained some fame during the SARS thing, and we began to get cold-called,” he said. “If you’ve tried everything else and you don’t know what it is, you can pick up the phone and call us.”

“The only way to prove that a virus causes a disease is to isolate the virus and inject it into a healthy animal,” Joe explained. “If the injection causes the disease, everyone agrees that you’ve proved your case.”

To prove that the virus they’d isolated was killing boas and pythons, they’d first need to grow this ancient ancestor of Ebola in the lab, find some healthy boas and pythons, and then infect them with the virus. To inject a virus into an African python took some trouble. Snakes don’t have injectable veins. They do, perhaps surprisingly, have hearts, and that’s where the virus must be injected.

On top of it all, the DeRisi Lab had stumbled into a surprising discovery. When they’d injected pythons with the same virus, this ancestor of the Ebola virus that killed boas, the pythons had survived. “Pythons are an old-world snake and boas are a new-world snake,” said Joe. “This thing that didn’t bother old-world snakes caused havoc inside new-world snakes.”

On the first of January 2020, Joe DeRisi was passing through the airport in Guangdong on his way to Cambodia. In addition to his lab at UCSF, he now ran a peculiar new institution called the Chan Zuckerberg Biohub. Created with a $600 million gift from the Facebook founder and his wife, the pediatrician Priscilla Chan, it had set itself a preposterous goal: to eliminate all disease on earth by the end of the twenty-first century.

Just then Joe was flying to Cambodia to install a node in what he hoped would one day be a global network to detect disease. “Early-warning radar for emerging pathogens,” he called it. It was an idea that had once interested the U.S. government. The original pandemic plan, conceived by the Bush White House, spawned a program called Predict, which set out to test animals around the world to determine which contained viruses that might jump into people. The Trump administration had zeroed out the program’s funding and so Predict, in the end, failed to predict anything. That didn’t trouble Joe so much as the thought that the Biohub, using genomic technology, had found a simpler, more practical way to achieve the same goals: by catching any new virus the moment it appeared in humans.

The natural home of such an ambitious global project was the U.S. government or the World Health Organization. The absence of pandemic prevention was another example of a deficiency in the system. Compared to its expected benefits, its costs were trivial, but no one company or person had the incentive to attack the problem. “We went to the CDC and pitched them and we were given a fairly cold reception,” said Joe. “And we were like, ‘We’ll pay for it!’ And they were like, ‘Huh, that’s weird.’ They basically didn’t care. Walking out, I kind of knew in my gut we’d have to go it alone.” In the end, the Biohub had partnered with another nonprofit, the Gates Foundation, to build, in effect, a global network of infectious-disease trip wires.

Joe reckoned they’d have the system in place by 2022. It wouldn’t be seamless: China would remain a black hole, as the Chinese had declined to participate. But Joe thought they could see what they needed to see inside China by creating trip wires in surrounding countries. That was one reason Cambodia was important, and why he was on his way there. It was close to China, and a draw for Chinese tourists. If a new virus ever took a trip out of China, Cambodia would likely be an early stop.

He spent ten days in and around Phnom Penh and left feeling good about his new friends’ ability to work the gene machine. It was the flight home, on January 10, 2020, that he found unsettling. He changed planes again in Guangdong—the province from which the Chinese doctor, the superspreader of SARS, had come. The airport was transformed. There were now lots of security people wearing masks. Passengers were required to step, one by one, inside an acrylic stall and be scanned for fever. “They weren’t fooling around,” said Joe. “I thought, What the hell is going on?” He...

Outside the VA, at least a few people from his White House days had not lost track of Carter Mecher. Tom Bossert, for example. As deputy homeland security adviser to George W. Bush, Bossert had watched Carter and Richard reinvent pandemic planning, reinterpret the greatest pandemic in human history, resurrect the idea that a society could control a new disease by using social distancing in its various forms, and then somehow lead the CDC to the conclusion that the whole thing had been their idea. Donald Trump had shunned most anyone associated with any former president but had made an exception of Bossert, whom he’d named his first homeland security adviser. “The job I had was chief risk officer for the country,” said Bossert. “And it should have been named that.” After his appointment, Bossert built a team of people to deal with biological risks, and instantly called Richard Hatchett and Carter Mecher. I just want you to know that if the shit ever hits the fan, the first phone call I’m making is to you, he’d said. He considered clearing Carter and Richard into the White House in the first days of the Trump administration so that in the event of some disease outbreak, they could fly in and get right to work. But then, on April 9, 2018, Trump hired John Bolton as his national security adviser, and the next day, Bolton fired Tom Bossert, and demoted or fired everyone on the biological threat team. From that moment on, the Trump White House lived by the tacit rule last observ...

For more than a decade the seven doctors had come together each time a biological threat presented itself. MERS, Ebola, Zika: they’d all been involved in each of those outbreaks, one way or another, behind the scenes. In flurries of phone calls and emails, they’d seek to figure out what was going on, and what each might do to influence the situation and save lives. They might be mistaken for a secret society—if the person at the center were not so insistently happy to share his thoughts with anyone who asked. They’d even been given a name: Wolverines. A former Bush White House colleague had come up with that, and somehow it had stuck.

“Saw unconfirmed reports today of 17 more cases in Wuhan,” he wrote to the others. “So I got to thinking and pulled a Carter. Is it likely that this thing is much larger than what we are seeing?”

The total number of infected people had risen from forty-five to sixty-two, most of them in Wuhan. Outside of China, two cases had already been identified in people who had traveled from Wuhan—one in Thailand, the other in Japan. Lawler pointed to those two cases and asked: What were the odds that there were fewer than one hundred cases in all of China if there were already two infected international travelers from Wuhan? Then he launched into a Carter-like back-of-the-envelope calculation, which is to say that it was both academically laughable and fantastically insightful. Redneck epidemiology.

The next day, the first American tested positive for the new virus. He was a man in his thirties who had traveled a week before from Wuhan to Seattle. Still the United States government was showing no sign of alarm; the only action the CDC had taken was to issue a travel alert, and to screen travelers entering the United States from China for fever. “It’s one person coming in from China, and we have it under control,” said President Trump. “It’s going to be just fine.”

Back at his desk, Carter collected the official statistics from China: cases, hospitalizations, deaths. He compared them to whatever he could find in Chinese blogs and newspapers. It was slow going, as much of the information he found was in Chinese. “I had no idea what the stuff was,” he said. “It was all gobbledygook to me. My computer kept lighting up: ‘This site is Unsafe!’ ” He’d cut and paste what he found into Google Translate just to see what it was. Some were death notices. He saw that the dates on deaths reported by the authorities every day at midnight were later than the dates in the local reports—that is, the thirty-seven people in Wuhan who had reportedly died by January 23 had died earlier than the official stats suggested. The timing was important. “I was trying to get a sense of how old the starlight was,” said Carter. The average time from infection to death was maybe two weeks, and so the deaths gave you a sense of how widespread the disease had been maybe two weeks earlier. If, on top of that, the Chinese were slow to report the deaths, the lag, and the multiple, was even greater. He noticed that, while the Chinese government had been reporting only a handful of cases two weeks earlier, they were now behaving as if there had been many more. “Reading tea leaves,” he wrote to his fellow Wolverines. “I see that China is building a 1,000 bed quarantine hospital in Wuhan—in 5 days. They also called in the military to assist . . . Reminds me of the military c...

What other virus does this virus most closely resemble? The obvious first answer was the new virus’s closest known genetic relative, the original SARS of 2003.

This new SARS had similar official stats, but he saw signs that the stats were misleading. The new SARS was spreading much more quickly from country to country than the original, for example. It was also eliciting very different behavior from the Chinese government. On January 23, the authorities closed Wuhan and forbade anyone from entering or leaving the city. “It underscores the point that Richard has made many times about how hard it is to get a feel for how bad or mild an outbreak is while it is happening around you,” Carter wrote. “I could be wrong, but this just doesn’t feel mild.” Then he nodded in the direction of the 1918 pandemic. “Wuhan = Philadelphia. Hopefully, we are watching and learning like St Louis . . . We are soon going to be running uphill from a brush fire.”

What Carter couldn’t understand was why the United States government lacked the same urgency.

The Mann Gulch fire captured the difficulty people had imagining exponential growth, even when their lives depended on it.

Still, they could hardly ignore the federal government. They watched the U.S. government repatriate Americans from Wuhan on January 29. The first group went to March Air Reserve Base, in Riverside County, California; the second, in early February, to four different places, one of them a National Guard base just outside Omaha, where they remained quarantined for fourteen days. The Omaha National Guard base was a short drive from the Global Center for Health Security, the place charged with treating Americans infected with mysterious new pathogens, and run by James Lawler. Lawler discovered—and could not quite believe—that the CDC didn’t plan to test any of the new arrivals unless they had a fever. All of the foreigners being shipped home from Wuhan were being tested before they got on the plane, and the CDC felt that was adequate.

He and his staff had already created their own test, based on the test created by the World Health Organization, and so they didn’t require the CDC’s help, merely its approval. The CDC sent one of its epidemiologists to visit James Lawler. At the end of the meeting, the guy said he needed to check with Atlanta. “The next day I get this panicked call from him,” said Lawler. “It’s gone all the way up to [CDC director Robert] Redfield. He said, ‘You can’t do it!’ I said, ‘Why?’ He said I would be ‘doing research on imprisoned persons.’ ” Never mind that every single one of the fifty-seven Americans in quarantine wanted to be tested: the CDC forbade it.

A disease projected to kill fewer than 90,000 Americans was a “Category 1” and called for nothing more than home confinement of the obviously ill. A Cat 5 (more than 1.8 million Americans) or a Cat 4 (900,000 or more dead Americans) required the CDC to call for all available measures: isolate the ill, cancel all public gatherings, encourage telework, enforce social distancing, and close schools for up to twelve weeks. After his back-of-the-envelope calculation, Carter concluded that, if the society failed to intervene, the virus would kill between 900,000 and 1.8 million Americans. “The projected size of the outbreak is hard to believe,” he wrote.

According to the pandemic plan, the federal government should at least have been preparing the country for the full suite of interventions. It wasn’t. So far as Carter could tell, it wasn’t even working all that hard to keep track of the virus.

To find the one hundred to two hundred that Carter imagined were already roaming around the United States, the CDC would need to be testing seven times that many people, or between seven hundred and fourteen hundred. “Right now we are in containment,” wrote Carter. “Think of the cases that popped up across the US like embers that are capable of starting a fire.

On January 31, the United States government finally acted, sort of. It restricted travel by foreigners into the country and required any Americans returning from China to quarantine for fourteen days. “We pretty much shut it down from coming in from China,” said President Trump. By then, thought Carter, the virus was likely already so widespread inside the United States that the focus on foreign travelers was a pointless distraction. “It’s a waste of time,” he wrote after Trump’s announcement. “You’re protecting your front door from intruders and they’re taking your stuff out the back door.”

What does one do, in the Mann Gulch fire, at 5:45? How does one respond to the sight of a thirty-foot-high wall of flame racing in one’s direction? State and local health officers still had no ability to test, as they were waiting for a test being created by the CDC. The CDC itself was testing only sparingly. With so little testing capability, Carter argued, you had to be smart about how you tested. He hit upon an idea: hospitals in the five biggest American cities should test anyone who showed up with flu-like symptoms. “I thought, Let’s go on a fishing expedition,” he said. “I thought, Why don’t we focus on where we think it is.” He began to track hospital reports of such cases and map them against numbers from previous years—and soon saw odd blips in Seattle and New York City. Those blips, he suspected, were cases being misdiagnosed as flu. The blip in Seattle would prove misleading; but the blip in New York, it became clear later, would have uncovered cases, likely a lot of them, had anyone bothered to test the patients.