Uncontrolled spread, p.27
Uncontrolled Spread, page 27
The agency started to track these new variants, but it declined to recognize the sequencing data that it wasn’t doing itself, or even to include it on the agency’s website. This included the sequencing work being done by the public health labs, which the CDC also bypassed. From the CDC’s standpoint, the problem was that not all of the sequencing was being done according to the same protocols, and if outside labs weren’t following the same approach as the CDC, then the agency didn’t believe the data could be reliably pooled. So the CDC simply set it aside and didn’t include it in the agency’s analyses or national estimates on spread.
Looking at the CDC’s website, you would see only a fraction of the sequencing work being done across the country. Worse still, by the spring, the CDC’s website only updated data on a monthly basis. So the sequencing information they publicly reported reflected analytical work that had been conducted weeks prior to its posting. The CDC’s data were as much as a month out of date. As a consequence, to get an accurate snapshot on the spread of the new variants, you’d have to go to the websites maintained by different states and cities. New York and Los Angeles were posting regular updates on the sequencing work being done by each of those cities. There were also some researchers who were posting to Twitter the sequencing they were doing. The CDC wasn’t only doing less sequencing, it was reporting only a fraction of the sequencing that was under way. In contrast to the US, the UK was able to detect the more contagious B.1.1.7 variant because it was sequencing about 10 percent of all patient samples. By the time the delta variant, B.1.617, emerged in May 2021, the UK was sequencing 60 percent of all patient samples.33 By comparison, the CDC had set as its initial goal getting five sequences from each state per week. When B.1.1.7 arrived in the US, the CDC was sequencing only 0.3 percent of patient samples. This figure probably undercounted the total amount of sequencing going on in the US because it was based on what was being reported into central repositories (like GISAID, a global initiative set up in 2008 to provide open access to genomic data on influenza strains). Since there was no organized reporting for COVID sequencing data in the US, we weren’t even capturing what was getting done.34 The Washington Post estimated that the US ranked forty-third globally in its percentage of cases sequenced.35
The CDC’s historical role is to provide careful analysis and shape our insight into more fundamental questions of public health. It’s a retrospective mind-set. In a crisis, what’s needed is rapid sharing of information and quick analyses that can inform real-time decision making. It’s a much more prospectively focused challenge and requires a forward-looking mind-set.
An analysis that suggests, at the outset of a crisis, that spread is primarily through aerosols can be far more impactful at informing our decisions and response than waiting for a report to provide a more definitive answer to the same question after you’re twelve months into the pandemic. By then, a lot of mistakes might have been made based on faulty assumptions. The CDC’s carefully crafted analyses take time to develop, time that may not be available to policymakers in a crisis, where even partial reporting is better than making the same decisions in an information vacuum.
Ultimately, the mandate to collect and report clinical information in a fast-moving crisis may need to reside with a different agency, one that has a national security mind-set. The CDC can take on some of these functions, but its mission and capabilities would need to be profoundly reformed. Epidemiologists Dr. Caitlin Rivers and Dr. Dylan George advocated the creation of a centralized system for disease forecasting. The idea is to have the epidemiological equivalent of the National Weather Service.36 Such an agency would be tasked with developing sophisticated disease modeling to help guide public health policy. It could provide the base on which to rebuild the nation’s data infrastructure for pandemic monitoring and preparedness. Congress ultimately set out to create such an effort inside a newly constituted component of the CDC that will be properly staffed, resourced, and more importantly clearly programmed by Congress to execute this new mission.
Wired magazine referred to this concept (in analogy to Philip K. Dick’s novel Minority Report) as “federal PreCrime for pandemics. Precognitive epidemiology. Make up whatever sci-fi words for it you want; the fact is, one thing the COVID-19 pandemic proved is that pandemics can happen and certainly will again. Building a place to develop the sophisticated models and simulations that can give a hint of when and where an outbreak will hit, and give guidance on how to stop it . . . well, that sounds like a pretty good idea.” The center, Wired observed, “would also become a central place to gather all that data, via public health surveillance and lab work—the equivalent of ocean buoys and satellites—and for dissemination of that information to local public health workers. Right now, lots of the most important data is siloed among different researchers and labs. The result is, disease modelers have to wheel and deal to get access to data, and production of models responsive to emergent problems is ad hoc. Meanwhile, if you want to know whether you need to double up your masks today, there’s no National Epidemic Center web page where you can check the forecast.”37
During COVID, the lack of reliable, real-time information on what measures were working to help contain the spread meant that many of our efforts remained subject to debate long after careful evidence could have substantiated their value or disproved their merits. Worse still, in the absence of definitive evidence on what worked, it left a void where critics were able to put forward their own cherry-picked evidence, or try to discredit the value of certain mitigation steps that they rejected and there wasn’t enough definitive data to fully contest these intrigues.
In the absence of good information, bad information was able to guide the debate. Perhaps nowhere did this play out with more conflict and misfortune than in the debate over masks.
I was involved in some of the early efforts to coax the coronavirus Task Force to endorse the use of masks. Other nations, including South Korea, had mandated the wearing of masks. Initially, task force members were reluctant to embrace a similar recommendation. One of their early concerns was that guidance to wear masks would send a mixed message about the benefits of social distancing. At the time, the task force was encouraging people to stay at home. Some members told me that they were concerned that issuing a call to wear masks would be interpreted by people as an indication that it was safe to go out, so long as you wore a mask.
It could, they worried, confuse the public.
There was also long-standing ambivalence at the CDC toward the use of masks. One task force member told me that the CDC raised concerns that masks would end up encouraging consumers who wore them to touch their face more, and in turn make them more likely to spread infection through fomites. One instance: the CDC initially told a major airline that their flight attendants couldn’t wear masks because the CDC was concerned that the flight personnel did not know how to properly fit the masks, and it would lead to more touching of their faces and could ultimately increase spread. The concerns were based, in part, on the CDC’s flawed premise that more of the early transmission was being driven by droplets and contaminated surfaces rather than aerosolization. One senior airline executive told me that their internal data showed that after their flight personnel started to wear masks, the incidence of coronavirus infections among staff fell sharply.
Masks had never been viewed as a standard part of the response to a pandemic. In fact, the much-discussed pandemic plan that the Obama administration had provided to the incoming Trump team didn’t take up the issue of widespread masking as a potential approach to containing a pandemic, or even mention masks a single time.38 I worked with a group of experts at Johns Hopkins University to craft a report we issued through the American Enterprise Institute (AEI) on March 29, 2020, on how to safely reopen the economy. (The report was titled “National Coronavirus Response: A Road Map to Reopening.”)39 We recommended universal masking. A lot of the heavy lifting in writing the report was done by Caitlin Rivers. She is a skilled scientist, with the rare gift for being able to translate scientific goals into prose and condense complicated public health objectives into policy goals that had enough coherence and unfussiness to be easily adopted. In my experience, this kind of practicality, this skill of being able to convert policy into interpretable narrative, is what often separates good policymakers from great ones. In crafting our proposal, we wanted to make sure there was enough supply of masks to support the objective. So we talked to the chief executives of the major medical product distributors. They told us that the supplies were still severely limited, and hospitals were struggling to maintain enough masks for their medical personnel. Ultimately, we issued a proposal for high-quality cloth masks.
President Trump was asked about our recommendation at a March 30 press conference.40 “Scott Gottlieb, your former FDA commissioner, wrote a roadmap for recovery after coronavirus. . . . The roadmap suggests that everybody wear a mask in public. Is that something that the Task Force thinks is a good idea?” a reporter asked Trump. “I saw his suggestion on that,” the president replied. I was told by Marc Thiessen, a Washington Post columnist and colleague of mine at AEI, that Thiessen had sent Trump a copy of the report, through the president’s personal assistant, to be printed off and shared with the president. “So, we’ll take a look at it. For a period of time, not forever,” Trump said. “I mean, you know, we want our country back. We’re not going to be wearing masks forever, but it could be for a short period of time. After we get back into gear, people could—I could see something like that happening for a period of time, but I would hope it would be a very limited period of time. Doctors—they’ll come back and say, ‘for the rest of our lives, we have to wear masks.’”
The president took the idea back to his advisers. The task force had already been debating the measure and was on the cusp of issuing a recommendation. Four days later, on April 3, the CDC issued a new recommendation that Americans wear cloth face coverings in situations when they were in public and couldn’t socially distance. President Trump announced the directive at a task force briefing, emphasizing that the guidance was “voluntary” and saying that he wouldn’t be wearing a mask himself, undercutting the message while unveiling it.41
The data showed that masks were not a panacea, but they could help reduce spread.42 One study later showed that weekly increases in per capita mortality were four times lower in places where masks were the norm or recommended by the government, compared with other regions.43 Another study published around the same time looked at the effects of mandates for mask use issued by governors in April and May. The authors estimated that these state policies reduced the number of new COVID cases by up to two percentage points per day.44 Other studies would go on to report similar findings. Masks were not a solution, but a higher quality mask worn properly could reduce risks.
The central premise behind our call for masking in the AEI report that laid out a road map to reopening the economy was our view that widespread adoption of maks could reduce asymptomatic spread. If people were infected, asymptomatic, and in settings where they might transmit the virus but they had a mask on, the mask would reduce the chance that they could spread the infection. Wearing a mask was in many respects an act of civic virtue, a way to protect your friends and neighbors if you were one of those individuals who was unknowingly infected and at risk of becoming a superspreader. Many cloth masks weren’t intended to protect you from getting infected, although a high-quality cloth mask would provide a person with some protection. To secure more robust protection, a person would need a higher quality N95 mask. The idea of cloth masks, simply put, was to protect others from you, not you from others, a distinction I made often.45
But the whole premise of mask wearing quickly became a faux protest, an expression of the disdain that some people felt for government directives that brought us shutdowns, closed schools, and other restrictions. Sensing that political zeitgeist, the president and his staff couldn’t resist the political urge to support these impulses and undermine their own guidance. People who wore masks inside the White House were mocked by senior staff. Members of the task force also didn’t model the behavior early on, even the doctors. They would get tested when they entered the White House, but then wouldn’t wear masks once they were inside. It was one of our greatest missed opportunities; especially our inability to get higher-risk individuals N95 respirators that could offer them better protection. Like the dispute over other forms of mitigation, the debate over masks had echoes in the tensions that surfaced during the 1918 Spanish flu, when antimaskers had gone as far as to fabricate a telegram from the US surgeon general, falsely claiming that he didn’t believe laws requiring masks were effective at reducing the pandemic flu.46
However, in the setting of COVID, it was unfortunately the president who did the most to turn face masks into a political flashpoint, creating an ideological movement that ultimately cornered the administration into a dogmatic position that proved harmful to itself and harmful to the nation.47 It may have been a misreading of the politics: antimasking didn’t become a political movement until many political leaders made it so. In fairness to the White House, the CDC gave political leaders mixed guidance on masks during the early months of the pandemic and was itself initially resistant to their widespread use.
The president could have found a middle ground on masks. His message could have been: We don’t need mandates. We’re adults. We control our government; our government doesn’t control us. However, we’re going to act responsibly and wear masks. We can do this ourselves. Personal responsibility; not government control. He could have couched it in a way that appealed to his political supporters while appealing to the nation to take more public health precautions. He could have encouraged people to wear masks. I had continued to urge the White House staff to support the use of masks, and I asked myself many times, why didn’t the president?
Most thought his resistance was to satisfy his political base, some of whom saw mask mandates as a breach of their liberties. Others said it was the president’s innate disdain for government control, his contempt for the “nanny state.” It was, in my estimation, some combination of all of those things. However, I heard another theory from those close to the president: Trump just thought the masks made people look funny.
When people insisted on wearing masks around him, the president would ask them to stand outside of the camera shot. As one person close to Trump told me, “the president thought that masks made you look weak.” White House staff had been prodding the president to wear a mask in public as a way to reinforce their use and sought the optimal occasion where it would be hard for the president to refuse. They found one in early July when Trump would wear a mask for a trip to Walter Reed National Military Medical Center to visit with wounded service members.48 However, he had agreed to do so only after finding a mask that he believed looked sharp: dark blue and emblazoned with the presidential seal.49
Our challenges in collecting, analyzing, and reporting information reinforced the shortcomings in our pandemic planning. As we learned, the tactics designed to mitigate spread of flu didn’t work as well against a novel coronavirus. Flu was not a bad starting point for our efforts, considering how little we knew about COVID. However, we didn’t learn fast enough about what was and wasn’t effective. We were caught in the fog of a viral war. We lacked the data to reveal how the virus spread and how our tactics were working against it. Then, we lacked the information and analyses to tell us how to adapt our response.
All of these problems showed just how poor our systems were for confronting a pandemic: our information systems, our analytical tools, and our ability to modify our policy response.
The emergence of HIV had proven that viruses with characteristics that are both highly novel and highly lethal can arise unexpectedly.50 When it came to new and deadly forms of coronavirus, MERS and SARS-1 should have prompted us to plan differently. The appearance of these pathogens triggered efforts to develop vaccines and therapeutics that targeted this class of viruses. However, that work was largely shelved when the immediate risk seemed to recede. A group was formed inside the CDC to work on coronaviruses, but its efforts were dwarfed by the much larger group dedicated to influenza, and much of the coronavirus effort dissolved.
All of our focus was on flu. The next pandemic, it was firmly believed, would be triggered by a novel influenza. It’s likely that the next pandemic will indeed be from a new strain of influenza, and it will be a lot worse than COVID. However, we need to be prepared for the unknowns. That starts with the analytical systems to identify and characterize a new pathogen. It could be a new strain of influenza with features that are irregular. Or another coronavirus that’s even deadlier and more contagious than SARS-CoV-2. Or, instead, something else that we never anticipated.
Chapter 14
Hardened Sites
The family had just returned to South Korea after a short trip to China, where an unidentified respiratory disease had started to spread, and they brought the pathogen home with them. One of the parents and a child soon became seriously ill with pneumonia. Before they developed symptoms, they were in close contact with friends and other relatives. A spark of transmission had been lit. The mysterious disease began to spread uncontrollably inside South Korea.
A team from the South Korean Centers for Disease Control and Prevention activated a plan to isolate the respiratory virus, identify its source, and deploy a diagnostic test to screen other people for the novel pathogen. They managed to get control of the outbreak before it exploded.
The entire incident was just a drill, a virtual scenario that South Korea ran as a tabletop exercise in December 2019. It was part of preparedness efforts first implemented after the country’s dangerous experience with MERS in 2015. A month later, South Korea would confirm its first case of COVID-19.1 “It was blind luck—we were speechless to see the scenario become a reality,” said Dr. Lee Sang-won, one of the South Korean CDC’s experts who led the drill. “But the exercise helped us save much time developing testing methodology and identifying cases.”2
