by Sanjay Gupta
Gao finally realized the enormity of the situation one night on another private call with Redfield. Gao broke down, audibly and tearfully distraught after finding “a lot of cases” in the community who had never visited the wet market. He knew the situation was not only out of his control, but people were dying, and the crisis was being directed by the higher-ups in government and the military, and that likely had been going on awhile. The initial mortality rates in China were somewhere between “5 and 10 percent,” Redfield told me. “I’d probably be cryin’ too,” he added. (To this day, we don’t know how many Chinese citizens were infected or died; the numbers could be grossly undercounted.)
During my postmortem conversation with Redfield, it became apparent that he was very concerned about Gao’s safety and was protective of him. At times, Redfield leaned forward conspiratorially and told me that he was worried about George Gao’s security, and he wanted to say nothing that could incriminate him in the eyes of the Chinese government, which he doesn’t trust. It was arresting to hear a chief scientist so distressed that his friend and Chinese counterpart might be physically harmed just for revealing the scientific evidence he was uncovering. When Gao and Redfield spoke in early January, it was clear that while China’s CDC was far out of the loop, the country’s central government knew what was going on and was secretly preparing for the spreading disaster: It was at least a month ahead of the rest of the world in terms of securing N95 masks and other PPE, reagents for testing, and the development of vaccines—the essentials they’d need to manage a pandemic. They were buying up these supplies before alerting the rest of the world.
There was other evidence the Chinese knew and were not telling. Toward the end of January, as we all watched the Chinese hastily build two massive coronavirus hospitals in just over a week, people like Redfield and Fauci thought, Wait a minute. Why are you building hospitals overnight if you are not that worried? Thousands of miles away, Debbie Birx was watching the international news in dismay one night over dinner in South Africa, where she was acting in her role as global AIDS coordinator for the President’s Emergency Plan for AIDS Relief program (PEPFAR). The mere sight of sick people overwhelming hospitals and the need for new ones to be built rapidly was enough of a signal that broad-based community spread had been happening, and probably for some time. She and her colleagues found themselves yelling at the TV: This is going to be a pandemic!
Soon after, Redfield was handed a report from his own CDC’s internal modeling for the pandemic: The United States would have 2.2 million people dead by September. It made Redfield pause in shock, and later that night, his wife shuddered to think it meant one of them would probably be dead by the fall.
“Just a Bad Flu Season”7
The one factor Crimson Contagion didn’t account for was the nature of COVID. The covert experiment modeled the pandemic response after influenza—not a coronavirus like COVID that can have a long incubation period during which a person is infectious or, worse, asymptomatic throughout the active infection. That is precisely what set COVID apart from other pandemics and partly explains our chaotic, slipshod reaction to it. We were groping in the dark for weeks. And in that darkness, our missteps and oversights started to metastasize into one of the worst responses in the world. According to Redfield, “Early on the focus was symptomatic cases—case identification, isolation, and contact tracing. But it became very clear to us in late February that, unfortunately, the major mode of transmission of this virus was not symptomatic transmission. And that changed the whole ballgame.”
We eventually learned that COVID was much deadlier than the flu and much more easily transmissible than either of its close cousins, SARS and MERS. We came to terms with the reality that aerosolized particles and asymptomatic carriers are significant drivers of its relentless spread. Everyone I spoke to who was part of the initial task force shared with me that their “oh no” moment was when it became suddenly clear and unmistakable that asymptomatic spread was happening with this virus. COVID created millions of modern-day Typhoid Marys, unwitting silent carriers of a deadly disease. Dr. Birx was especially unnerved by the realization this bug had secret superpowers that nobody acknowledged in its flight out of the gates. But she quickly saw the parallels to her decades-long experience in sub-Saharan Africa combating the AIDS epidemic.
Human immunodeficiency virus (HIV), the virus that causes AIDS, has surprising similarities to COVID even though they are very different. Both have an asymptomatic phase, although for HIV, it can be eight to ten years. With COVID, the asymptomatic phase may be eight to ten days. So if you rely on people entering emergency rooms or coming to the hospital, you’re already far behind in finding and stopping the community spread. Without proactive testing, by the time the first person develops severe illness, there is an avalanche of cases aggressively spreading the virus. Birx remembers the rapid spread of the infection on the cruise ships as particularly alarming: When nearly half of the passengers and crew end up positive for COVID and were asymptomatic when tested, that’s a major clue the virus is an aggressive and stealthy predator.
That’s exactly what happened on the Diamond Princess that docked for quarantine in Yokohama, Japan, on February 4, after a passenger fell ill and disembarked on January 25 in Hong Kong.8 That passenger, an eighty-year-old man, is thought to have been the only carrier—patient zero—of the virus onboard, whose infection wound up spreading to 712 others, 14 of whom died. A staggering 50 percent of people aboard got the virus from a single source. When the CDC mapped out the spread of the infections onboard, they discovered that lines of cabins sharing plumbing created a vehicle for the virus’s propagation: the toilets aerosolized the virus. People who had never shared physical space with infected passengers were nevertheless exposed. Between March 1 and July 10, the CDC discovered nearly 3,000 cases of COVID or suspected COVID and 34 deaths across 123 ships.9
The virus’s silent spread meant it was allowed to circulate a lot earlier before being detected. By many accounts, the virus started transmitting somewhere in early fall 2019, and local health officials in China had miscalculated their ability to contain it. This would be the second factor that worked against our pandemic response. First, the deliberate misinformation coming from China; and then came the cover-up.
Yale-trained epidemiologist Dr. John Brownstein, a professor of biomedical informatics at Harvard Medical School and chief innovation officer at Boston Children’s Hospital, has some of the most captivating proof that the virus was sickening people as early as autumn 2019, months before the rest of the world was made aware. Canadian-born Brownstein, as ebullient and youthful as your favorite tenth-grade biology teacher, is a pioneer of digital epidemiology—leveraging diverse digital data sources to understand population health. He has advised the WHO, the Institute of Medicine, the US Departments of Health and Human Services and of Homeland Security, and the White House on real-time public health surveillance data and has authored over one hundred articles in the area of disease surveillance.
Brownstein enlisted the power of both microsatellite technology and Internet search trends to “see” the first ripples of illness hit Wuhan before others took notice. Satellite images showed increasingly filled hospital parking lots in the late summer that didn’t resemble the same lots in years past. There was also an uptick in searches of keywords associated with infectious disease on China’s Baidu search engine (Baidu is Google’s rival in China; because Google is essentially banned in China, Baidu is the chosen search engine). Satellite data like the ones Brownstein used have historically been employed not only by intelligence agencies but also the private sector. Day traders, for example, track traffic patterns in parking lots at places like Walmart and Home Depot so they can more easily get an idea of the goings-on and capitalize on their buys and sells. The data help inform their trades. These photographs can be taken every hour to show when volume in the stores is high versus low. Such technology has also been used to track respiratory diseases. Brownstein hi
mself published a piece years ago that showed that hospitals in Latin America got superbusy during flu season. “You could predict flu season just by looking at the parking lots,” he told me.
Using images from October 2018, Brownstein’s group counted 171 cars in the parking lots at Tianyou Hospital, one of Wuhan’s largest. Satellite data a year later showed 285 vehicles in the same lots, an increase of 67 percent. And there was as much as a 90 percent increase in traffic during the same time period at other Wuhan hospitals. In his paper, posted on Harvard’s DASH server, his team writes, “Between September and October 2019, five of the six hospitals show their highest relative daily volume of the analyzed series, coinciding with elevated levels of Baidu search queries for the terms ‘diarrhea’ and ‘cough.’ ”10 While searches for “cough” typically increased at the beginning of yearly flu season, “diarrhea” was more closely linked to this pandemic. It’s a twenty-first-century way of predicting the beginning and trajectory of an outbreak based on the behavior of large populations.
Not seeing those first cases to help us realize the nature of COVID cost us. Our learning curve grew steeper every day we didn’t know, and we finally woke up to the reality after the nightmare began—which brings me to the third strike against us in our ability to manage the pandemic: testing failures. “Don’t start preparing when you’re in the middle of a pandemic,” Dr. Brett Giroir said to me in my postmortem with him. Giroir, a pediatrician by training, is a former four-star admiral in the US Public Health Service Commissioned Corps and was the sixteenth assistant secretary for health from 2018 to 2021. He was named the testing czar early in the pandemic. “You can’t create something out of nothing,” he said. He mentioned a case in point: “We’ve been investing in vaccines for 20 years and we’ve reaped the benefit of that investment. But for testing, it wasn’t ever mapped out or gamed beforehand so we couldn’t achieve those goals on the fly.… We didn’t have a resilient public, private, commercial, and academic infrastructure that could coordinate and work cohesively.”
Testing, Testing, Failures 1-2-3
Testing too little, too late, was our original sin in the response. When COVID got stuck in a flu model early on, testing simultaneously suffered. “We don’t really diagnose flu,” Birx pointed out. “We treat flu by the symptoms during flu season. When you come down with a flu-like ailment during flu season and you call your doctor, you’ll likely be prescribed flu treatment without a flu test. With COVID, the situation was not like flu—it was like HIV, with a large volume of asymptomatic people perpetuating the virus’s replication and spread. And to deal with that, you needed to have testing.”
Within the first ten days of knowing the virus’s genetic sequence, the CDC sent out test kits it had developed. But they didn’t work. Although the WHO had developed a test before the CDC, one that many countries were using, the United States chose not to use it and instead waited for its own testing system to become established. But that never really happened, at least not to the extent necessary to shove back against the virus’s proliferation. As Birx put it to me, “We let the perfect be the enemy of the good.” Instead of pushing to an impossible “perfect,” and therefore getting nowhere, we should have accepted “good enough” and at least gotten somewhere. Many things worth doing are worth doing badly—even, and especially, in a pandemic.
Redfield was reticent to shoulder any of the blame for his agency’s testing failure, even saying at one point that the CDC should have been congratulated on at least attempting to create the tests. The specifics to the flawed CDC tests are complicated and have been detailed in many news reports; suffice it to say they originally worked at the CDC but not in most public and academic labs where they produced inaccurate and inconclusive results. The CDC’s recall of these tests caused a significant delay in testing—five weeks, The Lost Month, and then some. During that time, the virus was burning through our population as other countries were successfully deploying tests of their own. The mistake in these first-generation kits marred the entire testing enterprise from the start and put testing perpetually behind. Nobody wanted to talk about the misfire; even when I probed Redfield about it, he praised the CDC’s record-time test development but grumbled at the harsh scolding his center received for testing failures on the ground. People unrealistically expected the CDC to test hundreds of millions of swabs every week and produce those hundreds of millions of test kits. But the center was not equipped for such a monumental task; it simply didn’t have the manufacturing capacity to provide that number of kits, let alone perform the kind of testing we all knew was needed for mitigation—especially when it involved an epidemic bolstered by asymptomatic spread.
“We needed a Manhattan Project for testing,” Redfield now notes in retrospect. He can identify the holes easily that were not about to be filled instantly when the pandemic hit, and surely not by a single organization or individual. These are holes that take decades to plug and then build on: A public health infrastructure at the ready for a pandemic, robust data analytics and reliable predictive data analysis, laboratory resilience, and a public health workforce in every health department in the country that’s ready to respond to a deluge of cases. Although the CDC, which is based in Atlanta, Georgia, and has employees in more than sixty countries and forty American states, is tasked with protecting the country from infectious disease threats, it’s surprisingly limited in its ability to order actions. It provides funding for most of our state, local, and tribal public health departments, as well as information and guidelines, but those do not translate to mandates. (An interesting aside: The CDC was founded in 1946 to prevent the spread of malaria across the country as veterans carried it home after World War II.)
Birx stressed the importance of rigorous proactive testing in the first set of gating criteria—the benchmarks on the way to reopening after the nation paused in March. These benchmarks had to be data driven. Who is infected? Who is sick? Who needs treatment? You have to routinely test staff in nursing homes, county health workers, and so on in order to see the epidemic and your imminent surge before that first person is hospitalized. “And if you work in a hair salon,” Birx told me, “you have to be regularly tested—not because we believe that you’re a huge risk to the clientele, but because you are in the public.” These individuals are like our sentinel surveillance people who act as beacons on where the virus is lurking and how we should respond. Unfortunately, the value of aggressive testing never sunk in at the White House.
Another huge misfire around tests occurred over the summer when the CDC issued guidance on its website telling people that they didn’t need to be tested if they were asymptomatic. This was around the same time Trump and his advisers, including Dr. Scott Atlas, a radiologist by training who didn’t believe in testing for asymptomatic cases, seemed to be pushing for a slowdown in testing. People in the White House believed that testing was driving cases rather than slowing them down. By slowing testing, they could make the COVID numbers look better—which is like wearing a turtleneck over a massive lump in your neck to avoid acknowledging what could be serious. At some point, you have to look squarely at the lump and deal with it. Or you’re playing dumb. Redfield says he was never explicitly told to slow testing. Trump’s comment at a Tulsa campaign rally, where he said to the audience he’d asked his people to “slow the testing down please” because “when you do more testing you find more cases,” was later explained away as “semi–tongue in cheek.” But all this muddled messaging would end up being unhelpful, and deadly. The CDC’s guidance was revised within twenty-four hours due to its risk of being “misinterpreted,” to use Redfield’s word. Mixed messaging was yet another strike against us.
Mixed Messaging in a Political Mire
For Redfield, the most egregious move by the White House in the pandemic response came when he says he was aggressively pressured to tamper with the CDC’s most important and prestigious publication, the Morbidity and Mortality Weekly Report (MMWR), a weekly epidemiological d
igest for the United States. Authored by career scientists and approved by the director, the MMWR is the main vehicle for publishing public health information and recommendations that the CDC has received from state health departments. It has been a fixture in our public health landscape for decades and is considered one of the most revered publications health care professionals use to make important decisions, some of them in life-or-death scenarios. When you’re in a pandemic, such a scientific report is critical because it informs doctors, researchers, and the general public about how a pathogen like COVID is spreading and who is at risk.
Claiming that the MMWR’s COVID reporting was aimed at hurting the president’s bid for reelection, Redfield told me the HHS Secretary Alex Azar and his staff, perhaps under the direction of the White House, asked that the reports be modified and, in some instances, delayed their publication. This was an absurd, if not unethical, proposed intervention. “Now he may deny that, but it’s true,” Redfield told me in reference to Azar’s coercion.11 The CDC director was not going to surrender to pressure—not when his name was going on these scientific reports. The MMWR was “sacrosanct” on his watch. On a drive home one evening, after pushing back on the Office of the Secretary in a heated dialogue that lasted for at least an hour, the call came in again. This time, it was Alex Azar’s lawyer and his chief of staff. They wanted certain details changed in the MMWR.
