Uncontrolled Spread: Why COVID-19 Crushed Us and How We Can Defeat the Next Pandemic

by Scott Gottlieb

In New York City, fourth-year medical students at New York University Grossman School of Medicine were given the option of graduating early if they agreed to work on the hospital’s COVID wards.1 It was an eerie throwback to events that unfolded one hundred years earlier, during the Spanish flu, when Philadelphia—critically short of healthcare workers—pressed medical students into service.

By the time the first cases of community spread were diagnosed in late February, SARS-CoV-2 had already rooted itself in our communities. It had been here for a while, at least since January, replicating, spreading, and doubling its numbers every two to three days.7 Then, in March, after thousands of cases had accumulated, the virus abruptly burst into public view.

This wasn’t because the United States had never imagined it might fall victim to a deadly pandemic. We certainly had imagined the possibility. In some respects, we had been preparing for this moment through three presidential administrations, starting with George W. Bush, who warned in a 2005 speech, following the outbreak of Severe Acute Respiratory Syndrome, or SARS-1, and then avian flu, that “scientists and doctors cannot tell us where or when the next pandemic will strike, or how severe it will be, but most agree: At some point, we are likely to face another pandemic. . . . Our country has been given fair warning of this danger to our homeland and time to prepare.”8 We had a pandemic playbook on the shelf, ran exercises simulating the threat countless times, and developed the Strategic National Stockpile to store the medical countermeasures that the top experts thought the country would need. But when the pandemic we long feared finally arrived, we weren’t ready. Many of the plans and preparations turned out to be a technocratic illusion. The stockpile lacked key essentials. A lot of what it contained didn’t work. It was a metaphor for our fragile response.

Even when the shortcomings became obvious, the Centers for Disease Control and Prevention (CDC) continued to rely on its systems for monitoring and responding to influenza, insisting even into 2021 that its flu-based interventions were the right tools in the fight against COVID.

In the future, we cannot rely on global manufacturing for key components. We’ll have to build domestic capacity for making commodity products like gloves and masks and greater resiliency for making high-technology products like biological drugs and vaccines.

When COVID struck, all eyes turned to the CDC, which, it was assumed, would quarterback our response. However, the agency doesn’t have an operational capability to manage a crisis of this scale.

China ignored commitments it had made for sharing information in a public health crisis. It grudgingly revealed incomplete material only after the risks were largely apparent. China never shared samples of the early strains of SARS-CoV-2.20 These behaviors proved we cannot rely on global cooperation alone to alert us to emerging threats. Going forward, we need to cultivate more effective ways to monitor potential hot zones and gather information on these risks.

As for the political response, President Donald Trump and his advisers grew increasingly weary of the economic impact of the measures recommended by public health authorities, and they were wrongly convinced that uncontrolled spread was inevitable regardless of what steps were taken. This partly underpinned an ambivalence by the president on how forcefully to embrace mitigation. That doubt was reflected in public statements, and actions, that were contradictory and harmful. At the same time that the president was berating Georgia Governor Brian Kemp for opening tattoo parlors and bowling alleys too soon, he was tweeting dangerously to his followers to “Liberate Michigan.”21 At the same time that many of the president’s top health officials were urging people to wear masks at the start of the fall surge of infection, the president stood on the balcony of the South Portico of the White House and, in a theatrical gesture, took off his mask while he was still contagious with the virus.

Operation Warp Speed was a bright spot: one of the greatest public health achievements in modern times. The unprecedented research effort helped to deliver safe and effective vaccines and to secure new efficiencies in manufacturing.23 The Trump administration deserves credit for helping to facilitate that accomplishment, which will eventually end the pandemic. The success of Operation Warp Speed proved what government can accomplish when it functions well, to improve our preparedness and protect the nation. Other parts of the system were unable to deliver a more synchronized and effective response.24 We wouldn’t have been able to prevent a US epidemic, but with better preparation, better alignment, better leadership, and better engagement of our tools of public health, we could have delayed its onset and reduced its scope and severity.

The US Centers for Disease Control and Prevention already had evidence, in the form of clusters of cases that occurred in the same families, to suggest that the virus was spreading between people. Yet several days earlier, on January 14, the World Health Organization (WHO), relaying China’s assurances, tweeted that there was “no clear evidence” of human-to-human transmission.

For all the talk of a pandemic playbook that was purportedly on the shelf, initiated in the George W. Bush administration and refined under the administration of President Barack Obama, it was written for a pandemic involving a novel strain of influenza. The Trump health team would use this blueprint nonetheless, and exacerbate its shortcomings by making some fateful mistakes in how it was implemented.

On January 21, the White House started to actively plan the legal, policy, and logistical requirements for implementing an airport screening program to detain travelers from China and subject them to temperature checks and extra surveillance for symptoms.15 The next day, President Trump got his first question on the virus during an interview on CNBC from Davos, where he was attending the annual gathering of the World Economic Forum. Asked by Squawk Box anchor Joe Kernen if he was worried about the outbreak, Trump replied, “No. Not at all. And we have it totally under control. It’s one person coming in from China. . . . It’s going to be just fine.”

The US travel ban that was ultimately announced, which went into effect on February 2, would deny entry to foreign nationals who had visited China in the fourteen days prior to their arrival to the US.17 The World Health Organization would largely endorse China’s lockdown of Hubei Province (where Wuhan is located), one of the most sweeping travel restrictions in modern times by any major economy, and yet the WHO would be critical of the restrictions imposed by the US, and other countries, on foreign nationals traveling from China.18 In other words, Chinese nationals couldn’t move around inside China, and the WHO approved of that policy. But if these same individuals somehow managed to get on a plane and travel to the US, and were subsequently not permitted entry into America, the WHO found that policy objectionable.

Testing would be a major gap in our response and the most visible symptom of the capabilities that we lacked in confronting a public health crisis of this magnitude. We just didn’t have the resiliency in our system to quickly step up the production of test kits and testing equipment at the scale that was required. We couldn’t organize a collective effort to get key industries all working in a coordinated way to plug these gaps in time. Our health officials were overwhelmed early on with repatriating Americans from China and focused too little attention on preparing the homeland for the inevitable spread of the virus. Many publicly clung to a tragic view that SARS-CoV-2 wouldn’t become epidemic here. By March, when the evidence that it would spark a global pandemic could no longer be discounted, it was too late.

Viruses are not technically alive. But they’re not dead, either. They operate in a pseudo corporeal state, where they hijack our cellular machinery to carry out their cycle of replication and spread.13 In doing this, they engage in most of the activities that give an organism the features of life. But viruses cannot do it alone. They need a host to model the gestures of life. They need a specific species that they can invade, pirating its cells and using them to invade and propagate. Once inside a host, often a virus will have a specific “tissue tropism,” meaning it has a fondness for infecting and replicating inside a certain part of our body. Some viruses can infect many types of cells. Other viruses live primarily inside a certain organ.14 The rabies virus travels along our neurons. Hepatitis hides in our livers. SARS-CoV-2 likes to live in the lining of our respiratory tract.

For many patients, the most severe consequences of COVID weren’t the direct effect of the coronavirus on their lungs, but rather the exaggerated immune response their bodies unleashed to kill the virus. Because the SARS-CoV-2 virus was novel, meaning that human beings had never encountered it before and thus had no established means to combat it, the human immune system tended to overreact. In the most severe cases, it could trigger a “cytokine storm,” a cascading activation of immune cells, which in turn excrete small proteins called cytokines. In the normal sequence of our body’s functions, cytokines help coordinate our response to viral invaders and recruit other immune cells to the site of an infection. But if all the cytokines get dumped into our blood simultaneously, too many immune cells can get recruited to the site of the infection, creating an abrupt and severe hyperinflammation that can damage organs.16 It was a cytokine storm that ended up killing many people from COVID.

The new virus had nearly 29,000 nucleotide bases that held its genetic instruction set, and it looked frighteningly similar to SARS-1, which killed nearly eight hundred people globally in 2002 and 2003. But the virus wasn’t SARS-1.19 It was something completely novel. That information alone should have been terrifying. A new respiratory virus capable of making people severely ill was spreading in Wuhan. Yet local officials initially withheld information from the central government in Beijing about the outbreak.20 An employee of one Chinese genomics company said that on January 1 an official at the Hubei Provincial Health Commission instructed the company to stop testing samples related to the Wuhan outbreak and destroy all of the lab’s remaining specimens. The impulse to contain the troubling results eventually spread to officials in Beijing. Two days later, on January 3, the country’s top health authority, the National Health Commission, ordered the local genomics labs not to publish any sequencing information related to the novel coronavirus.21 News of the outbreak would be shared through the postings of anxious doctors who took to social media to exchange the sequencing information.

At the same time, the WHO continued to publicly praise China for its purported openness. To take one of many examples, at a January 30 press conference, Tedros celebrated “China’s commitment to transparency and to supporting other countries. . . . In many ways, China is actually setting a new standard for outbreak response, and it’s not an exaggeration.”49 He would make similarly effusive statements on other occasions, despite the wrangling that continued behind the scenes. A year later, at a media event held a week before the end of President Trump’s term, Secretary Azar reflected on the opening days of the pandemic, saying that his team was “pressing for the Chinese government to send us viral isolates from patients there. And China has still, one year later, failed to provide the first-generation viral isolates.”50 Access to those samples at the outset could have helped the world prepare. And without the source strains, it would be impossible to determine with any certainty the virus’s origin, and whether the market was the focal point of the initial outbreak or, as seems more likely, one stop along its advance.

We can no longer depend largely on global cooperation and the competency and transparency of other nations. COVID wasn’t caused deliberately, but it was enabled and nurtured by the intentional quashing of information.

During the SARS-1 outbreak in 2003, local Chinese authorities had removed seventy sick patients from two hospitals in order to conceal these SARS cases from a delegation that was visiting from the WHO.13 The Chinese officials put the patients into vans while WHO officials toured the hospitals, and they drove the infected individuals around the city to pass time. Some patients were checked into hotels until the WHO inspectors had left the country.

It was only after Robert Redfield had suggested to George Gao that the Chinese CDC broaden its criteria that Gao revealed that the outbreak was probably out of control. Now, nearly two months later, the US CDC’s requirements seriously narrowed who could qualify to be tested in America.35 On March 1, the CDC’s official tally of COVID cases rose from fifteen to seventy-five. Models developed later by researchers at Northeastern University show that by this date, the US probably had 28,000 infections.

But a big part of the challenge was that officials at the CDC were applying a flu model to coronavirus. “People think that because you prepared for flu you were prepared for coronavirus, and it took eight to twelve weeks to realize that the two viruses behaved very differently,” one senior White House official said to me in the spring. When it came to a novel coronavirus, we also lacked a foundational infrastructure for testing and developing countermeasures: we can make new flu vaccines by changing the strain we use in existing manufacturing processes. Similarly, the country has a lot of flu tests that can be adapted to a novel strain of influenza. We lacked any similar foundation to build on for a new coronavirus.

This is what happened with COVID. The federal government lost trust and credibility early, by its inability to accurately convey the true scope of the hazard. The lack of reliable information on COVID’s spread, and the inability of people to access testing, degraded the integrity of the response.

They simply didn’t have the information they needed to conclude the risks were low. By the end of February, in the US, there were plenty of reasons for Americans to be concerned. Some White House officials also started to grow more skeptical of the reassuring reports from HHS. Among them were Grogan and Pottinger. Both didn’t understand how the CDC could be so sure that coronavirus wasn’t already circulating in America given the events unfolding around the world, where the pathogen had spread furtively only to explode after cities had become saturated with it.

These shortcomings were aggravated by the way the coronavirus spread, through asymptomatic transmission. As many as 40 percent of infected patients didn’t develop noticeable symptoms, and thus never show up on the symptom monitoring.

During the spring 2020 COVID wave, it’s likely that we were diagnosing only about one out of every ten or fifteen actual infections. With more testing in place, by the summer, that proportion improved to about one in eight infections, and by the fall, one in four. Thus, even thousands of cases lurking in a large city might not be enough infection to generate an identifiable signal of disease, especially because it was late winter, and the virus was able to hide in the shadow of flu season.

Two of the biggest failures was the CDC’s overestimation of the risk of spread from contaminated surfaces, and its underestimation of the risk from asymptomatic transmission. The two errors were mutually dependent.

Despite the fact that influenza can also spread asymptomatically, the CDC didn’t assume that the same would apply for the novel coronavirus. On January 28, Anthony Fauci said during a Coronavirus Task Force press conference that in “all the history of respiratory-borne viruses of any type, asymptomatic transmission has never been the driver of outbreaks. . . . Even if there’s a rare asymptomatic person that might transmit, an epidemic is not driven by asymptomatic carriers.”

Picking up that thread, Frost asked Fauci: “Some people in the last couple of days in particular have started to elevate their concerns as to whether this disease is already present in the US and spreading across people who are otherwise apparently healthy. Is that something that is impossible to know either way for sure and how worried are you about that possibility?”12 Fauci replied, “Well, certainly it’s a possibility but it is extraordinarily unlikely and let me explain why. The reason is if there were people who were actually spreading it, you would not have them identified, isolated, and contact tracing which means you would have almost an exponential spread of an infection of which we are all looking out for. We have not seen that, so it is extremely unlikely that it is happening.

In the weeks before and after the press conference, Robert Redfield would also affirm the message that the threat to the nation remained low. “The American public needs to go on with their normal lives,” he said.17 In fact, SARS-CoV-2 was spreading widely. The point here isn’t that the federal health officials were wrong. The point is that they were working with faulty tools, and from faulty data sets.

We would learn later that by February, the West Coast was heavily seeded with cases imported from China that circumvented our travel restrictions.20 It turns out that the February 2 ban on air travel from China was quite leaky. An estimated 40,000 American residents had returned to the US from China through major airports after the ban took effect.21 And in the month before the travel restrictions, nearly 300,000 people had traveled into the US from China.22 Airport screening for symptoms was also porous, and there was no way to identify asymptomatic carriers. Meanwhile, the East Coast was being seeded heavily with cases coming from Italy. Wuhan and Italy had close ties. Italy was the first G-7 nation to join China’s investment and infrastructure project, the Belt and Road Initiative, and many Chinese textile workers had immigrated to Prato, the historical capital of Italy’s fashion business.23 Garments would be manufactured in Wuhan, and shipped to Prato, where the final stitches were “made in Italy.”

A review of CDC records by the news organization ProPublica found that of the more than 750,000 travelers screened at US airports by mid-September 2020, officials found only 24 cases of COVID.25 We now know that many hundreds of cases got through. We closed down most, but not all travel from China early on. But we let travelers continue to arrive from Europe even as the epidemic raged there. Matthew Pottinger and others had urged restricting travel from Europe much earlier than our March 11 ban that affected travelers from twenty-six European countries, but the proposal met resistance from Secretary of the Treasury Steven Mnuchin and others on the economic team, who worried about the financial impact of such a move.

In retrospect, we could never stop the virus at our borders. At best, we could have delayed its fuller entry. Screening programs and travel bans were long known to have limited effect in preventing or even slowing epidemics.28 One widely cited modeling study, which examined the impact of travel bans in the setting of a pandemic flu, found that implementing strict travel bans that reduced infections by 99.9 percent would delay the peak of a hypothetical US pandemic by just six weeks. This was if the restrictions were implemented early, on day thirty of a hypothetical foreign epidemic that had not yet spread discernibly to America, and the estimate was modeled from influenza, which is less contagious than SARS-CoV-2.29 In other words, even draconian bans on travel, implemented to stop a virus that was less contagious than SARS-CoV-2, would, at best, be temporizing. It would buy some time; it wouldn’t prevent the epidemic.

The problems with the CDC’s COVID test seem to have started inside the agency’s Respiratory Virology Lab in Atlanta. They had sprung from a contaminated component that was meant to make the test more precise but also made the process more complicated and prone to failure. There were three main parts to test, referred to as N1, N2, and N3. Each component was a different set of primers and probes that would bind to a specific segment of the coronavirus’s RNA. The first two reagents, N1 and N2, were specific to SARS-CoV-2. They targeted two regions in the virus’s RNA that code for the production of a structural protein that plays a key role in the coronavirus’s replication and assembly.1 The third component, N3, was unique to the CDC test and was not included in other countries’ tests, including the one developed in Germany and adopted by the WHO. The N3 component was designed to bind to any SARS-like coronavirus, including some strains that were found in bats but had never been known to have infected humans.

But on February 6 came another red flag. A final quality check showed that the test kit could fail 33 percent of the time. For reasons that have not been fully explained, the CDC decided to ship the kits anyway.5 They were initially sent to labs in thirty-three states.6 On the same day, the World Health Organization directed the shipment of about 250,000 COVID test kits (manufactured by a German firm) to more than 70 labs around the world.7 Once the public health labs got the CDC kits, they first tried to validate them by testing sterile water, a common practice to confirm whether a test kit worked. Sterile water should have generated a negative result since it contains no coronavirus RNA. The labs were distraught to discover the samples testing positive as if the water contained live virus.8 This was not an isolated problem, either. Twenty-four of the first twenty-six labs that tried to run the tests couldn’t get it to work. Nobody knew why.

It was becoming clear that the trouble with the CDC’s manufacturing process might be the result of a more systemic failure, perhaps related to contamination at the CDC site that was making the kits, and this was now affecting many parts of the agency’s diagnostic test. The CDC wasn’t a manufacturer. In an ideal system, the CDC would have contracted with commercial manufacturers from the outset and wouldn’t have tried to make test kits on its own, even for the public health labs. The CDC eventually reached out to contract manufacturers around February 10, but agreements to enlist the help of commercial companies to make the testing components weren’t completed for about ten days, due to complex issues that would arise in the contracting process. Among other contracting matters, the CDC was trying to assure protection of the agency’s intellectual property around its test design.

While this was going on, one of the CDC’s contract manufacturers, Integrated DNA Technologies, Inc., or IDT as it was known, decided to start making its own primer and probe kits using the blueprints that the CDC had already published on January 27. This was the key component that labs needed in order to assemble their own test kits and no longer have to rely on the CDC. Initially, IDT began accepting orders for a version of that panel that could be used for research use only, and not for testing and reporting results to patients.15 However, using the primers and probes, sophisticated labs could now assemble their own tests using the CDC’s original protocol. And IDT had made a reliable set of these components that could be mass-produced. With the IDT kits, labs would now have all of the components they needed to start testing patient samples.

The labs learned much later that the third and extra reagent that CDC included, the N3, was contaminated with coronavirus RNA. But N1 was also contaminated, although at a lower level. At some point in the manufacturing process, someone had probably walked from a dirty room, where scientists used RNA, into a clean one where they assembled the CDC kits, and carried the sticky strands of genomic material with them. Then some pieces of the RNA had gotten into both the N1 and the N3 components. This meant that two of the three primers that were designed to tell whether the test was detecting SARS-CoV-2 were themselves contaminated with strands of coronavirus genomic material. As a consequence, when the test registered a positive result, it wasn’t detecting COVID. Instead, it was cross-reacting with itself, detecting the bits of RNA that had accidentally seeped into the system’s reagents.

Manufacturing the synthetic SARS-CoV-2 RNA that would serve as a “positive control” should have been outsourced to a lab that was disconnected from the CDC facility that was assembling the completed kits. It was critical that the two processes be kept far apart. RNA can be hard to work with. It attaches to surfaces and spreads easily.19 The risks of doing all these things in the same facility are well known in the lab industry and a common source of problems.

The CDC’s single biggest breakdown in its response to COVID was the abortive rollout of its COVID test. In their defense, officials at the CDC pointed out to me that developing tests for commercial and academic labs was never their job. It wasn’t what they had ever done before. Their experience was in working directly with state public health labs. They had done it successfully in 2009 in response to H1N1 swine flu. But once the agency ran into challenges with their development of that initial test for SARS-CoV-2, for weeks after, there still wasn’t another test being pursued in the US other than the effort that was under way at the CDC.

The stockpile was a running sore throughout the crisis. In April, the federal government started distributing ventilators from the stockpile to backstop hard-hit cities. When the machines arrived, they didn’t work. The New York Times reported that HHS had assured governors that the stockpile was holding about ten thousand ventilators in reserve and had another ten thousand on order.24 But the government had allowed a key contract for maintaining the machines to lapse, and more than two thousand of the ventilators on hand were not functional.25 Similarly, when masks were shipped from the stockpile and arrived at hospitals, there were reports that healthcare personnel found them covered in mold, and the rubber straps had become dried and brittle.

Once COVID struck, demand surged. For one ninety-day stretch in the spring of 2020, we used about 140 million N95 respirators.34 In March, some US hospitals were using 1,700 percent more N95 respirators, 500 percent more isolation gowns, 860 percent more face masks, and 300 percent more surgical masks, according to one estimate.35 A lot of money was spent trying to quickly establish domestic manufacturing of all these items. In the Coronavirus Aid, Relief, and Economic Security Act, or CARES Act, $1 billion was set aside for this effort. But the lingering question is how to maintain this excess capacity once the pandemic is over. In past crises, or in preparation for them, we built the domestic capacities we thought we’d need, only to see them atrophy once the crises failed to arrive, and attention shifted elsewhere. This was the case with our preparations in the early 2000s for a feared pandemic involving a bird flu. The need for the domestic manufacture of critical supplies, and the need for policy approaches that provide for sustained capacities, will be a recurring theme in how we prepare for the next pandemic.

state had many free testing sites, and people were encouraged to use them.40 Massachusetts even pioneered the testing of sewage wastewater as a tripwire to give policymakers a way to identify early when and where community spread was getting under way in a certain area. (People shed the virus in their feces.) The data gleaned from testing raw sewage allowed the state’s public health officials to measure how prevalent the virus was in a community and better target interventions.

Data on the genetic sequences of the H5N1 infections that had been identified in 2005 showed that this particular strain was mutating in ways that could make it more menacing to people.10 It was believed that the H5N1 virus had first jumped from birds to humans in Hong Kong in 1997, in an outbreak that infected eighteen people, six of whom would die.11 The infection was fearsome, causing severe inflammation of the lungs, a condition known as pneumonitis.12 When it first emerged, people were stricken only after close contact with infected chickens; there was no evidence of person-to-person spread. The outbreak was ultimately stopped. But at a big cost. Among other measures, Hong Kong slaughtered more than 1.5 million chickens.

Modeling suggested that timing matters: the interventions had their greatest impact if schools were closed before 1 percent of a local population was infected.41 More systematic studies had found that in the setting of flu epidemics, closing schools for long stretches of time reduced the total number of community cases.42 These steps could also reduce peak attack rates by up to 45 percent among a community, according to the research (and by as much as about 50 percent among children).43 Real-world studies, including surveys done after the 2009 pandemic that analyzed the influence of school closure on transmission, supported these conclusions.

On March 13, walking around the West Wing of the White House, you wouldn’t know we were about two weeks away from a synchronized, national shutdown of nonessential businesses. I was at the White House that day for a scheduled meeting to see President Trump to share my views on the unfolding crisis. We were three days away from the lockdown of San Francisco, seven days away from the stay-at-home order in New York. Our models would later show that the epidemic was already spreading uncontrollably across the nation.1 And yet a lenience about the personal dangers pervaded the building among the White House staff that I met with that morning. Vice President Pence insisted on shaking my hand and seemed nonplussed when I immediately dabbed my palms with Purell.

President Trump had called me the previous day and asked me to come to Washington to meet with him. He wanted to discuss the unfolding crisis and get my take on the outlook. Trump had taken an interest in my work at the FDA and had supported my efforts to rein in teen use of tobacco products, to advance policies that promoted generic drug competition, and other public health initiatives. I had stayed in touch with senior White House staff and was grateful to the president for the active support he showed me when I served as his FDA commissioner.

The president seemed to agree on the need for stronger action. He was well briefed before my meeting. Even though some of his public statements over that time period would have suggested he didn’t see the looming calamity, or the need for strong action, his private demeanor that day left me with the clear impression that he recognized the grave risks, he was more solemn than previous meetings, and when it came to the question of mitigation, he was mostly sold on the ideas before I had arrived. It would be revealed seven months later that the president was sitting down for a series of taped interviews with Washington Post journalist Bob Woodward around the same time. Trump revealed his fears to Woodward and would face understandable criticism for, at times, publicly downplaying the risks while privately he was worrying about the dangers. Six days later, Trump told Woodward he was expressly downplaying the dangers to avoid creating a panic.7 The president, Woodward would write, understood the gravity of the threat we faced.

Underlying all this, the CDC believed that the coronavirus was behaving like influenza, and so the most plausible explanation for these mysterious chains of transmission must be some contaminated surface that patients had touched, where they picked up the virus through respiratory droplets left behind on a subway pole or a door handle. In reality, a lot of those puzzling chains of transmission weren’t the result of fomites, but rather, asymptomatic carriers who had gone on to unknowingly infect those around them. Since the CDC believed that SARS-CoV-2 spread like flu, however, they discounted the role of asymptomatic transmission.

The CDC’s preoccupation with fomites probably contributed to the agency misjudging the contribution of aerosol transmission and discounting the use of masks early in the pandemic. It became increasingly clear that a lot of spread wasn’t merely through droplets but also through aerosol transmission.29 Droplets are larger particles we secrete when we talk, cough, or sneeze.30 They are heavier and can’t travel long distances. Aerosols are smaller particles that emerge after droplets start to evaporate. They can spread through the air over longer distances and stay suspended for prolonged periods of time. This becomes especially true when the aerosols get into sealed, indoor spaces. If the right patient was in the right confined space at the right time in their illness, typically a day or two before symptoms set in and when they were maximally contagious, then the pattern of spread they could spark could look more like measles (which spreads through aerosols) than flu (which spreads primarily through droplets). Yet, in considering COVID, the CDC had flu in mind.