On August 8, a few days after President Trump tore into me on that phone call, I walked into the Oval Office for a major vaccine briefing with him. Since April 21, I’d been to the Oval only a handful of times, either to serve as a prop for a governors’ meeting or for a brief huddle prior to a press conference on vaccines. I saw this as my first opportunity to engage in discussion of real, substantive policy with the president in the Oval in 109 days. Unfortunately, it only confirmed a fear that had been steadily building in me: not only were Scott Atlas and his dangerous ideas now controlling Trump’s message, and could impact policy on the pandemic, but in Atlas, Trump had found a scientific messenger willing to tout his views, regardless of how many lives they cost. This meeting made it disturbingly apparent that when it came to the pandemic response, the Oval Office had become an echo chamber.
The White House was on its back foot at this time, trying to recover from a revelatory interview. On August 3, President Trump had appeared on HBO’s Axios series. Journalist Jonathan Swan conducted an interview in which the president floundered. He repeated his claim that there were those who said (including him) you can test too much. When pressed for who these people were, Trump claimed there were “manuals” and “books” where this point was substantiated. When pressed again, he couldn’t name any of them. The president also again said that we had higher cases of the coronavirus “because of the testing.” This didn’t get much attention given that this was the same interview in which he claimed that he had done more for the Black community than any president, with the possible exception of Abraham Lincoln. As President Trump made his way through a series of questions, he held up “supporting” graphics (which I had not made) to defend his position. He handed one over to Jonathan Swan, believing that it supported his contention that the United States had the lowest rates in the world in several categories.
Though his base may have seen the interview as the left-leaning media attacking the president, with it, President Trump once again laid bare the issues that existed in the White House. Who had made the charts and graphs? Who was briefing the president and what were they telling him? It wasn’t me. It wasn’t the task force. The figures he was citing weren’t from the data sources we’d vetted as accurate.
In the same interview, the president said of people dying of the virus, “And it is what it is.” Of the lag time between tests and results: “There’s nothing you can do about that.” Worse, he continued to make the claim that the virus was now under control.
I’d made clear dozens of times that the virus wasn’t under control, that we could do better, that we had solutions to the testing issues he’d said were not solvable. He wasn’t listening to the task force. He wasn’t listening to Bob Redfield or Tony Fauci. He was listening to someone, or several someones, people were manipulating data and giving it to the president without review or discussion. I was reminded of the disinfectant moment at that press briefing.
The next day, Stephen Miller told me that President Trump needed to know exactly where things were improving. I wasn’t sure why he needed to know that now, but I suspected he was looking to support the claims he’d made in the Axios interview. Miller said that he and the other speechwriters believed the “vigilance” message would be more effective if they leavened it with a dash of hope. I’d chafed against this kind of snapshot thinking before. Yes, there were counties where case and fatality rates were falling—parts of Texas and Florida came to mind—but cases were far worse in Oklahoma, Missouri, and Tennessee. The virus was relentlessly on the move through the heartland and would reach the Northern Plains just as cooler weather settled in. Highlighting improvement in one place would give a falsely positive picture of the scope of the pandemic. It wouldn’t account for rising and falling patterns in regions, for how it improved in one spot and deteriorated in another. Instead of supplying the rosy picture Miller wanted—presumably, for campaign rally speeches—I sent him the now-familiar U.S. map showing the status of every state. The map was overwhelmingly red and yellow, with few pockets of green. That was the state of things as the summer surge ebbed slightly and the fall surge approached. Knowing what a trusted advisor Miller was, I also sent him the Arizona data, to show how effective proper mitigation could be.
Whenever anyone—Mark Meadows, Marc Short, and Stephen Miller being the most prominent—came to me with a request for good news, I complied with specific, honest, detailed figures. I always pointed out that such snapshots were distortions. Yes, these specific instances of improvement are occurring, but each state was in a different place in the cascade. Even if smaller areas are showing improvement, others are showing an early increase in positivity. Here’s the projection that shows that those areas are going to worsen. I knew the winter was coming and we needed to prepare, and part of that preparation was to make sure both the president and the public understood the risks.
The Trump reelection campaign was rolling along. That week, they claimed to have knocked on one million doors. How many of those canvassing workers and respondents socially distanced or wore masks is unknown, but if they did not, it showed a continued disregard for best public health practices. President Trump spoke to one hundred supporters in Cleveland, attended two fund-raisers in the Hamptons, and launched two campaign tours (one from Pennsylvania, the other from Florida), with buses traveling across the country until the general election, going to rally after rally.
Though most large White House events were moved outdoors, into the Rose Garden—a step in the right direction—indoors, people gathered unmasked, a failure that resulted in one superspreader event after another. There was always someone in their midst who was silently infected, sometimes within hours of a negative test, with the virus exploding from the cell factories in their noses and spreading to those around them invisibly in the air. No masks. No physical distancing. No mitigation. No worries. In the months to come, the West Wing became a hot zone over and over. My assistant, Tyler Ann, and I masked consistently. To avoid risk of exposure, other task force doctors called into work rather than coming into the White House. I didn’t have that luxury. Mitigation works. Tyler Ann and I are uninfected proof of that truth. In the end, in this White House, it became easier to count who was not infected than who was.
THE OVAL OFFICE MEETING of August 8 was ostensibly to discuss Operation Warp Speed and the progress on vaccine development, and because of this, I walked in hopeful. Dr. Moncef Slaoui delivered the news that the vaccine trials were on track. They would be conducted according to established rigorous scientific protocols. I’d heard a number of times from Steve Hahn that the president had repeatedly questioned whether that process could be sped up. The president peppered Dr. Slaoui and others with questions. Dr. Slaoui and Steve endured. They pushed hard, but always with safety as the primary aim, resisting the president’s urging for “sooner” and “faster” and his incessant “Are you sure?” regarding the projected approval time line. Unstated, but very clear, was the president’s desire to have the vaccines available before the election. Moncef, Steve, and Alex Azar stayed the course. They’d cut no corners, but would very quickly deliver safe and effective vaccines—not in time for the election, but in record time nonetheless.
I was motivated by a different set of numbers, numbers that reflected what was happening in the immediate—test positivity cases, hospitalizations, and deaths. As a result, as the meeting was drawing to a close, I decided that, though I wasn’t on the agenda, and despite months of the president’s completely ignoring me, I would reintroduce him to the reality that, as vital as a vaccine was, we had other aspects of the pandemic response still to deal with.
“We are currently at just under fifty thousand cases a day,” I said. “We are trending downward from the July peak, but I am very, very concerned about the fall. That could be an incredibly deadly period. The way to prevent that is to mitigate—expand testing, masking, and reduce indoor unmasked gatherings. We’ve got to fight the virus that’s here now and coming to new communities in the fall.”
I noted that the president didn’t do what he had so frequently done in the past when I gave him unpalatable news. He used to turn his head to the side and put his hands up, as though shielding himself from me. This time, his reaction was more muted. He sat back and folded his arms, nodded, and then shook his head, as if he’d heard all this before and wished I would come up with something new. When I pressed on, he waved me off and turned toward Scott Atlas: “Is that how you see it?” He didn’t so much ask the question as offer up a prepared introduction to what he knew was coming next.
“I couldn’t disagree more strongly,” Scott Atlas said, leaning forward, eyeing me without turning toward me—both of us were seated facing the president. “No matter what we do, the outcome is going to be the same. The same! You don’t need to mitigate. You don’t need to test. All we need to do, and we’ve already done this, is to protect the most vulnerable.” He shrugged and held his hands out, palms up, as if it were all that simple and straightforward.
Ignoring the satisfied look on the president’s face, I challenged this grossly inaccurate and inadequate statement: “We can’t protect the vulnerable if the virus is already in the community. If we—”
“That’s just wrong!” Atlas said, cutting me off.
For the next five to ten minutes, Scott Atlas and I argued. Back and forth we went: I’d make my case based on the data, and he’d cut me off, his rebuttals increasing in frequency and intensity, each one beginning with either “You’re wrong!” or “That’s just wrong!” It wasn’t a case of differing data; it wasn’t about diverging points of view. It was black and white: I was wrong, and he was right.
I refuted each of his dangerous assertions:
That schools could open everywhere without any precautions (neither masking nor testing), regardless of the status of the spread in the community.
That children did not transmit the virus.
That children didn’t get ill.
That there was no risk to anyone young.
That long Covid-19 was being overplayed.
That heart-damage findings were incidental.
That comorbidities did not play a critical role in communities, especially among teachers.
That merely employing some physical distance overcame the virus’s ill effects.
That masks were overrated and not needed.
That the Coronavirus Task Force had gotten the country into this situation by promoting testing.
That testing falsely increased case counts in the United States in comparison with other countries.
That targeted testing and isolation constituted a lockdown, plain and simple, and weren’t needed.
Point by point, as I dismantled his case, Atlas grew angrier, cut me off more aggressively. To further emphasize his points, he’d push himself back in his chair and splay his arms in the air while nearly shouting his new mantra, That’s just wrong! Using alternately defensive and offensive language to bolster his position, he became animated and boisterous, his thinking neither scientific nor logical.
I should have expected as much. He’d started his campaign against the facts (and, by extension, me) by editing or refuting my daily reports. For weeks since that inauspicious start, he had been discrediting me and the science through emails and in person during Covid Huddles and in task force meetings. He never offered clear evidence to support his position. He bullied instead of debated using facts.
Now, as I sat there, alternately eyeing him and then the president, Atlas rose up, leaned forward, and said, apparently out of desperation, “You’re going against the president’s policies!”
I flinched, surprised and concerned. To this point, the president hadn’t articulated any policies beyond we will never shutdown this country again. All along, whether it was through the CDC, the White House, the task force, or other agencies, the federal government had consistently been developing guidance, not specific policy, guidance that was approved by the Staff Secretary and OMB. We’d been providing the states with options for how to implement effective public health recommendations. But those recommendations weren’t policies. They weren’t hard-and-fast rules or regulations—a synonym for policy. Nothing we had developed and distributed had used that word unless it went through the Staff Secretary. Nowhere had we stated that these were actionable, enforceable measures or that failure to abide by them would result in consequences. There would have been a firestorm of protest if we had. Most governors, and the American public, would have seen this as the federal government overstepping its authority. Now here was Atlas attempting to speak for the president and set policy from in front of the Resolute Desk and not through data-driven discussions within the task force.
Since he’d told me “never again” with regard to shutdowns, the president had given neither me nor any other person on the task force a policy objective. As for the rest—masks, testing, and other elements of the response—the president hadn’t explicitly elucidated (his tweets notwithstanding) any measures or approaches of the kind Scott Atlas was now insinuating existed. If Atlas was overstating the approval the president had for his, Atlas’s, opinions, that was one thing; if Atlas was speaking on behalf of the president, that was another, far more dangerous thing for the country.
If Scott Atlas was now the voice of President Trump’s policy, it was worse than his saying I was wrong. The test of whether this was Scott Atlas overstepping his bounds or whether it indicated that he had the president’s approval to speak on his behalf like this was right there in the room with us. In that moment, the president neither endorsed nor refuted Atlas’s positions. He was silent. He didn’t say “Scott, that’s right. Those are my policies,” and he didn’t say they weren’t. This left the field of play completely open, where it was before I entered the room.
I countered Atlas’s points with the same message I’d delivered consistently. The summer figures were harbingers of more dangerous and deadly times ahead. I continued this drumbeat: the time to more proactively and aggressively mitigate was now. I’d been saying the same to Kellyanne Conway and Hope Hicks, to Alyssa Farah, Morgan Ortagus, and Kayleigh McEnany, and to every senior advisor in the West Wing. I delivered the same message now in front of the president: Scott Atlas’s words weren’t just irresponsible. They weren’t backed by accompanying evidence to support them.
For every unsupported “That’s just wrong,” I offered a substantiated “Here is the proof.” What he offered by way of a rejoinder was a reiteration of the herd immunity theory without directly calling it herd immunity.
I pointed out that now that we had an even more accurate and efficient data collection system, with 93 percent of hospitals reporting, there was even greater weight behind my assertions. We knew the majority of hospitalizations and deaths were from the community, not just nursing homes.
I used Arizona’s turnaround to demonstrate the efficacy of the UPenn/CHOP model.
Atlas shook his head and entire torso while shouting, “That’s just wrong!”
Meanwhile, President Trump sat expressionless, his arms folded across his chest. The posture said: I don’t want to hear this. I don’t care what you’re saying.
As the confrontation drew to a close, a bemused but detached expression animated the president’s face. It wasn’t what we were saying that had grabbed his attention as much as the volume at which it was being said. Atlas was assertive, aggressive, perhaps hoping that his passion would carry the day. If he could speak loudly enough and long enough, perhaps he’d demonstrate that he was right—at least to someone who didn’t understand the data and enough to earn the status of senior Covid-19 advisor. Bombast versus substance makes for interesting spectacle when there is no legitimate reality at stake. But in a room representing the highest level of American government, with a viral pandemic raging and countless lives on the line, the he-who-speaks-loudest and he-who-speaks-what-I-want-to-hear nonetheless seemed to carry the day.
It was tragic. It betrayed people’s trust. It cost lives.
I knew that if I rose to match Scott Atlas’s fever pitch, I’d be characterized as shrill or hysterical, the doom-and-gloom lady out of control. The adjectives used to describe women are often laced with hidden meanings that trivialize or undermine. The proverbial rock and hard place—a man’s passionate stance versus a woman’s letting her emotions get the better of her.
As with so many meetings and crisis points, this one ended with no resolution, no dramatic shifts. I knew that making a final push, imploring the president not to listen to Atlas or others, was going to fall on deaf ears. The president had enjoyed the show, and just as he’d done in the very first meeting I had with him, he tuned to another station.
“Very good. We’re done.”
I had a sense that the first two words were directed only at Scott Atlas and the last two were for me.
As I got up and walked out, the feeling was mutual. I was done. I was done with this type of spectacle in front of the president. I was done with Scott Atlas. I was done with beating my head against the most stubborn and unrelenting wall I’d ever come up against in my professional career and my life. I was angry with Scott Atlas for his handling of science. I was furious with whoever had handed this man the keys not just to the White House but to the vehicle that could steer the response. Whoever that person was, they were accountable for whom they brought on board, how they managed them, and what they allowed them to do.
Though the president didn’t utter the catchphrase, “You’re fired,” I very much had the sense that I had been.
I’d long suspected that the president was being influenced by someone like Scott Atlas. That was obvious. But to sit in a meeting and see that person, to experience viscerally how the shift in influence was affecting the president, left me feeling lightheaded, disembodied. As the meeting ended, I felt no sense of relief, only overwhelming sadness and frustration. I knew that while continuing to fight the virus, I would still be fighting Scott Atlas, too.
FOLLOWING THE OVAL OFFICE meeting, Scott Atlas did two things to establish his authority and take over the direction of the federal response to the Covid-19 pandemic:
First, when rebutting anyone’s point of view, he used the president’s name at every opportunity—not data or analysis; just the president’s name, to give himself credibility. I’d seen this move many times before: someone with a tenuous position in an organization, or an untenable position or approach, will attempt to firm up both by citing the name of the person with the highest level of authority. Basically, Scott Atlas was saying, I know what the president wants. I have insider knowledge that you lack. I have access to him you don’t. Consequently, if you disagree with me, you are disagreeing with the president. This ploy had never worked with me.
Second, Atlas kept repeating his “policy” comment. When, in task force meetings, we discussed making some modification to guidance, he’d frequently say, “Those are not the president’s policies.” But the president hadn’t defined any new policy positions, in the Oval meeting or elsewhere. His sole policy position that we never shutdown the country again remained in place. Nothing had changed. Atlas may have been using the president’s name, but until the president or vice president specifically told me about a new policy, what Atlas espoused would be irrelevant. When it came to this president, and in military terms, I wasn’t violating a direct order. Perhaps I was going against the president’s wishes, but wishes are not the same as orders or a stated policy.
To be clear, I was never told by anyone in the White House to stop issuing my daily report, the critical four-pager I created for the Covid Huddles, which clearly showed day by day where we were in the pandemic and that charted actions and local communications. I was never barred from taking my messages to the states and counties by phone or in person—indeed, the White House and Vice President Pence facilitated my trips. I was never prevented from doing local press to get the message out locally. Though the president might have been doing the opposite of what I was recommending and saying the opposite of what I was saying, his White House supported my actions, and I was never stopped.
Despite Atlas’s saying that policies existed, they didn’t. But because I didn’t demand that such policies be formally spelled out and confirmed, I could continue to operate on the razor’s edge of Scott Atlas’s opinion and the president’s failure to make clear what his policy was and issue direct orders to implement it. As a result, I could still go to state and local governments with clear science and data-driven recommendations to mitigate community spread without defying the commander in chief’s orders. If he didn’t specifically lay out what was permissible, then I was free to do and say what I knew was necessary.
EARLY ON THE MORNING of August 11, I wrote to Bob, Tony, and Seema, laying out in detail the seven points of contention I had with Atlas’s position that I’d gleaned from my “debate” with him and from his previous statements and op-ed pieces. In the aftermath of my confrontation with Atlas in front of the president, I’d committed myself to working on two parallel but related tracks—continuing to coordinate all aspects of the pandemic response, and managing Atlas in the task force and ensuring he wasn’t viewed within the White House as having any real influence or impact.
I pointed out that Atlas had created an alternate reality, a parallel universe where the virus could spread unmitigated among younger people without infecting any of the most vulnerable, the thirty-five million Americans over seventy, a group that had suffered and died at the greatest rate since the pandemic began. We had seen this in the spring and in the summer: there was no way to lock thirty-five million Americans away in an impenetrable bubble. What Atlas had offered wasn’t a plan to mitigate the spread; it was a plan to speed it up. What looked like something logical on paper had the fundamental flaw of being impossible to implement. There was no way to separate the vulnerable from the community in which they lived. We needed to expand testing, not decrease testing.
Normally, theories, models, and approaches in science are tested with the use of well-designed experiments or at a population level. In this case, Atlas’s and my positions weren’t mere laboratory experiments. They were being conducted in the real world, in the lived experience of the American people. From the beginning of the outbreak through the end of 2020, 180,000 of the 250,000 recorded Covid-19 deaths were vulnerable Americans aged seventy and over. And that was with inconsistent mitigation efforts in place. We can’t calculate how many would have died if Atlas and other proponents of herd immunity had had their way, but we know that figure would have been so much higher.
Even with our mitigation plan in place, the virus was moving among younger people and reaching the vulnerable. Seema Verma worked around the clock with nursing homes and nursing home associations to protect their residents. Despite those best efforts, time after time, the virus penetrated our defenses, and by June 2021, at least 185,000 Covid-19 deaths were found to have occurred in nursing homes. It simply wasn’t possible to protect the elderly without taking measures to stop community spread.
Even once the vaccines were available, we knew that not every vulnerable person would have a fully effective immune response to them. As we age, or for those with underlying immune deficiencies, or those who undergo treatments that suppress our immune system, our immune responses to vaccines are blunted. If SARS-CoV-2 was actively circulating in communities, vaccination alone would never be enough to protect everyone. Therefore testing and masking would need to be continued in the presence of vulnerable family members (the elderly and those immunosuppressed) whose immunity status was uncertain despite vaccination. Full stop.
Thanks to the UPenn/CHOP model, we knew which measures were effective: proactive testing, masking, and decreasing gatherings. What had begun as a theory, and then became a computer-generated model, had now been put to a real-world test. As Arizona and other states that followed the model had shown, we could mitigate against community spread; the numbers would decrease.
Scott Atlas didn’t offer a plan or proof—but he did offer something: the distraction the White House needed and wanted in the moment. The administration wanted to campaign, to raise money, to hold indoor gatherings—and they wanted someone to tell them it was all going to be fine.
On August 13, Seema Verma shared with me an email she had received on March 21 from Scott Atlas. My warning emails to the task force had prompted her to recall this past message from him. In it, Atlas claimed that “the total lockdown is a massive overreaction [his emphasis] and super harmful to our entire society, destroying the economy, inciting irrational fear, and even diverting medical care away from sick people.” At the time he wrote this, we were only five days into the initial Slow the Spread campaign. By no reasonable person’s estimation were we in “total lockdown.”
What astounded me about this email was what it revealed about Atlas’s thinking, which had not evolved from the end of March to the beginning of August. Over those four months, we had learned a lot about the virus and had spent months working to refine all aspects of the response. Atlas believed that we needed to protect nursing home residents, 1.5 million people. But he didn’t seem to understand, or didn’t care, that there were far more elderly Americans who didn’t live in long-term-care facilities but who had comorbidities that put them at great risk—35 million-plus of them. He didn’t understand, or didn’t care, that those who died in long-term-care facilities were infected by members of the community (nurses, orderlies, janitors, visiting family), not by other residents.
Atlas still believed that, despite the mitigation efforts we had put in place, this virus had a natural course to run. He attributed the success of the forty-five days of Slow the Spread not to all the measures governors had enacted, but to the virus doing what it was going to do. Even if we did nothing, his thinking went, we’d still see the same numbers of cases and deaths. I knew this wasn’t true, as we could see direct, temporally related consequences of mitigation. We could see it in differences in the deaths-per-one-hundred-thousand between states with mask mandates and those without.
As much as I blamed Scott Atlas for his deeply flawed approach to pandemics and epidemiology, ultimately, the responsibility for this turn of events lay on the president’s desk. Someone in his administration had brought Atlas in. They had allowed him to make reckless, scientifically unsound recommendations. I had serious doubts about whether the president, beyond watching him on Fox News, had ever actually had Atlas vetted. And I felt very strongly that our contentious Oval Office meeting represented the first time Atlas had ever been seriously challenged. I don’t know whether it was a case of the administration lacking a nuanced understanding of what Atlas was advocating, or of their lacking the will to challenge someone who clearly had the ear of the president. It really didn’t matter; the end result was the same: Atlas was inside the building. It was my job to make sure that was where his views and his influence were contained.
Of course, Atlas couldn’t just leave things at that. He wanted and needed to substantiate his claims further, to bring more people within the administration and around the country over to his side. So, on August 13, shortly before I departed for another visit to the states, while the participants of a Covid Huddle were sitting around the conference table in the Roosevelt Room, he announced that he was convening a panel of infectious disease experts for a roundtable discussion. (He would later call it a Medical Experts Roundtable.) He would invite Dr. Martin Kulldorff, PhD, from Harvard; Dr. Joseph Ladapo from UCLA; Dr. Jay Bhattacharya from Stanford; and Dr. Cody Meissner from Tufts—essentially, the same group that had been championing herd immunity from the start. He wanted the president and vice president to attend, and he wanted the event to be open to the press. I believe Atlas hoped that the intellectual heft of the participants’ academic positions would bend public opinion his way and give a White House public stamp of approval for his theories and his approach.
As Atlas outlined his plans for the event, I sat there living the nightmare of the moment while worrying about the one to come. The people he wanted to include in his roundtable were academics who lacked on-the-ground, commonsense experience in public health and infectious disease epidemiology. As he ran through his plans for the roundtable, Kellyanne Conway, who was seated near me, slid a sheet of paper over to me on which she’d written, “GOOD LORD!” I nodded, envisioning dark lords of misrule entertaining the president while the vice president sat there helpless. I hoped Trump would see the dog-and-pony show the event was sure to be as merely another example of unreality TV being captured inside the White House.
Once the Huddle broke up, I immediately contacted Kellyanne with this message: Scott Atlas was a dangerous man, and his roundtable would be even more dangerous. Allowing him access to the press with the president and vice president in attendance would give him and herd immunity the executive branch’s seal of approval. We couldn’t let that happen.
I did everything I could to block the planned roundtable event from happening. I went to the vice president’s staff, the White House communications team, and anyone who would listen—letting them all know what a mistake it would be to validate Atlas’s pseudoscience in public. The event shouldn’t take place at all, but especially not at the White House, and not in front of the press. We couldn’t be seen as endorsing this theory in public. I wrote to the other doctors with my concerns.
We were heading into the deadliest time with this virus. We needed to be talking about testing and mitigation, not about a recklessly damaging approach that the vast majority of scientists had rejected. I had to continue to try to prevent the Atlas event from taking place. Any public perception outside or within the White House that his views were impacting the direction of the response would further erode the public’s already tenuous faith in the gains we were making.
Later, I could see how undermining Atlas’s authority and the integrity of his plan took me places I’d hoped never to have to go. It wasn’t enough to keep him and his cohorts out of the public eye, and it wasn’t enough to keep his plan from receiving presidential approval. I also couldn’t let him believe he could make headway with anyone on the task force, beginning with the doctors. If Atlas found a foothold there, who knew how high he’d climb or whom he’d get to accompany him.
Vice President Pence wanted all the doctors to sit down with Atlas to try to come to consensus. After a phone call with Tony in which I reiterated my position that Scott Atlas was a clear and present danger, Tony wrote to me and the other two docs. He agreed with me, shared my concerns. Like the vice president, he hoped that we could all sit down and have a “non-confrontative discussion to go over in detail the basis of his claims.” But Tony hoped we could wait a week to do this, until he could resume speaking after a procedure that prevented him from doing so.
I was beyond working it out or reaching consensus. That wasn’t going to work. My position and Atlas’s were too far apart. Based on my experiences with him, I knew he wasn’t going to listen to reason. He wouldn’t engage with evidence. He was wedded to his beliefs and theories, and that was that.
I informed the docs that I was done with trying to reason with him. I just couldn’t do it. I couldn’t invest one more ounce of energy refuting his claims. I had told myself that I would have to fight both the virus and Atlas. His opinions needed to be contained within the White House. My priority was on mitigating the virus.
Fortunately, Tony reconsidered his position on a face-to-face meeting. He wrote to me: “I know what I’m going to do. I am going to keep saying what we have been saying all along, which contradicts each of his 7 points listed below. If the press asks me whether what I say differs from his, I will merely say that I respectfully disagree with him.”
I responded to them all: “Perfect—will do the same.”
I don’t say this lightly: affording Atlas any professional respect after his actions in the White House would have been a mistake. Some doctors on the task force believed that because of Atlas’s position as a medical doctor, he should be afforded professional respect. I had started from that place. I had sat down with him and gone through all the evidence. But he had chosen to say nothing and to delete me and refute my data behind my back. He didn’t engage in dialogue; instead, he bullied. In the end, I believe all the doctors came to the same place, but it was a rough several weeks getting everyone there.
I understood from past experience that any pandemic response will be tied into politics. For decades in other countries I’d worked to get leaders to act against their own political self-interest in order to get needed public health measures enacted. But I hadn’t understood that with this administration, the order of political magnitude was going to be so much higher. I also didn’t anticipate that some physicians would use their MD to bring credibility to their deadly theories.
Over the days and months since that Oval Office meeting, with Scott Atlas’s “You’re wrong!” still ringing in my ears, I had to face this harsh reality: The president and some of his senior advisors wanted someone who could match their cynicism and their perceptions of the pandemic, which were driven by personal opinion and perceptions rather than data. Now more than ever, I knew I couldn’t quit. I could still make progress—by coordinating the task force, by working with the vice president, by working through Jared’s Huddle to ensure we were meeting the needs of the state and local governments, and by speaking with state and local officials. I would continue learning what was and wasn’t working and get that information across the country through our reports and trips.
Scott Atlas could continue to believe he was speaking for the president. I couldn’t change his mind, but I could make sure his influence didn’t extend beyond the Oval Office.
THREE DAYS BEFORE I met with the president and confronted Scott Atlas, I had received a draft document titled, “Considerations for Covid Testing.” As I suspected it would, it didn’t talk about the clear importance of testing—the need to test both symptomatic and younger asymptomatic individuals to stop community spread. Proactive testing would (as I hoped and wanted) drive up the number of cases, alerting asymptomatic people to their infected status and preventing superspreader events from occurring. As was being shown by those routinely testing, doing so rapidly drove down spread, significantly blunting the outbreak. In a task force meeting, Atlas expressed agreement with the president on our needing new testing guidance posted to the CDC’s website and said he would be the person to make this happen. I knew what he would advocate for. In his March email to Seema, he’d come right out and said it: “Fear of exposure, or people who are without symptoms who are outside priority groups, do not need urgent testing and should not seek it [his emphasis].”
As I read the draft of his proposed testing guidelines, I was dismayed to see they reflected a lack of strategic implementation that would be effective at preventing community spread. I added a bullet point about the need for people in high-transmission zones who were unmasked for the duration of any indoor gathering of more than ten people to consider getting tested. Such gatherings constituted superspreader events. Nurses I’d spoken with in hospital after hospital reported that the patients on ventilators had been infected at family gatherings—birthday parties, weddings, funerals. But Atlas’s proposed guidance supported prioritizing testing only the symptomatic, not the asymptomatic, too. We were still getting hung up on this issue.
I received pushback on my bullet point addition from Brett Giroir. I respected Brett’s position as testing czar, and up to that point, he’d agreed with me on the role of asymptomatic spread and the need to detect it early through widespread testing. Though the language I had used in my bullet point was intentionally mild, Brett objected to it. I had known that any strong signaling of the need to test asymptomatically wouldn’t get past the staff at the CDC, given their continued resistance to widespread testing independent of symptoms when test positivity was rising in the community. But now it hadn’t passed the sniff test with Brett Giroir, who’d been supportive of my position up until now. I suspected that Brett had either come under the influence of Scott Atlas, whose strong opposition to testing was so essential to his championing of herd immunity, or been instructed by HHS to find a middle ground. That would mean supporting Atlas while still trying to maintain as much testing as possible.
In response to my questioning about his shifting view, Brett wrote to me on August 8, saying that “people know that you can’t fix this by public health guidance with testing everyone on the street.” The language was right out of the Book of Atlas. This about-face by the country’s testing czar couldn’t have come at a worse time: testing needed to expand in preparation for the fall to blunt community spread and to ensure kids stayed in schools.
We discussed the new CDC testing guidelines in a task force meeting that same week. Bob Redfield and his people were the ones in charge of what was released as CDC-endorsed actions and guidance. And in task force meetings, it was Bob who usually brought up CDC-related matters. This time, though, it was Brett Giroir who went over the proposed guidance that had clearly been influenced by Scott Atlas. Essentially, it was another case of “Let’s test less.” In part, the revised guidelines said that—with the exception of people who had developed symptoms, were elderly, or had a medical condition that made them vulnerable—people who knew they’d been exposed to a test-positive infected person didn’t need to be tested. These revised guidelines were an eerie echo of what Atlas had spent months calling for publicly.
If Atlas got the CDC to make a definitive statement like this, my job would be made much harder. The CDC and I continued to engage in back-and-forth on the degree to which silent spread among young adults contributed to community spread. This new guidance was exactly the opposite of what I’d been advocating, which was more testing of young asymptomatic individuals. Sports teams were using regular testing to find the infected early to prevent spread. If approved, the Atlas-influenced new guidance would not only reduce the number of tests being conducted, but would also make contact tracing impossible and prevent those without symptoms but engaged socially from being tested, effectively cutting two legs off the platform on which a pandemic response must be built.
I voiced my opposition. We had to stop discouraging people who thought they’d been exposed in indoor gatherings from getting tested. I believed that by proactively testing people who were in high-risk environments like bars and indoor gatherings, we could prevent spread to parents and grandparents. Tony wasn’t present, but the other docs chimed in to support me.
So, on August 26, when Brett presented the task force with the final draft, I spoke up again, saying clearly, “I don’t approve this. I can’t.” Scott Atlas stepped in and again went after me, saying that I was wrong about testing, wrong about asymptomatic spread. He concluded with the statement he’d made in our heated Oval Office exchange: that his views represented the president’s position and policies.
Angry but under control, I said again, “I can’t approve this.”
The vice president stepped in to say, “I really want us to have consensus on this.”
“I can’t approve this,” I said. “I can’t keep the CDC from issuing this. I don’t have oversight of them. I do have some oversight, though, over what gets put out from the White House. If the CDC wants to do this, that’s one thing. But it can’t go out from the White House as something the task force approved. It can’t go out on the White House website as something the task force endorsed.” There were limits to what I could do, but I had to strongly speak my mind no matter what.
Bob Redfield was the head of the CDC. If he disagreed with the document’s points about asymptomatic spread—as I believed he did—then it was up to him to take on this fight. I looked to Bob. He wore the expression of a condemned man resigned to his fate, and it became clear to me that this document, which undercut so much of my belief about strategic testing, was being forced down his throat.
In the end, the revised testing guidance, which aligned so closely with what Atlas had called for publicly, went out, despite its flaws. I was flabbergasted when, in announcing the new guidelines through the national media, Giroir said the White House Coronavirus Task Force had approved them. I immediately contacted him, asking him how in the world he could have done this. He offered the flimsiest of excuses, saying that he thought I had said it could go out. But the point I’d made hadn’t been at all subtle; I reminded him of what I’d said: the CDC could issue what it wanted to—I couldn’t control that—but what came out of the task force, and therefore the White House, was under my control. And I’d make it clear that this document hadn’t received White House approval, either directly or tacitly. The task force unanimously had not approved the CDC’s recommended testing guidelines. Period.
Lowering testing levels was exactly what we’d been fighting against for months. Initially testing rates had been low thanks to the CDC’s early errors in scaling up testing. We’d been gaining ground and the new point of care/rapid antigen tests were helping us do that, but testing remained a long-standing problem. Then the president had made the illogical (and public) connection between the role testing played in the increase in cases. Now, with Scott Atlas having influenced the guidance, he was putting into practice the push for lower testing President Trump had spent months calling for publicly. In the history of the president’s rhetorical assaults on testing, this was the first time there had actually been a guidance shift from the CDC. I didn’t know if he had directed Atlas to do this or if Atlas was taking it upon himself to influence one of the president’s “policies.” Either way, Atlas was now the de facto testing czar.
That all this was happening right as the political campaigning was shifting into high gear was hard to ignore. At a crucial time in the pandemic response and the election cycle, testing and campaigning were now intertwined. The administration knew the two key issues most prominent in Americans’ minds: the economy and the pandemic. The two were inextricable, of course, but in the world of politics, simplification works best when it distorts most. Even if there was a false sense that Covid-19 cases were decreasing, the Trump people could tip the scales in their favor. Scott Atlas and his approach to testing was the now-visible thumb on that scale. If we didn’t test as much, the number of cases being reported would decrease, and it would appear as though this administration’s handling of the pandemic was producing results. The reduced case numbers could be used to prove a false claim: that the United States was triumphing over the virus. We weren’t.
I was surprised that Brett Giroir had been part of this change in guidance, and I wondered what kind of pressure he’d been under. From comments Brett made to me later on, I gathered he thought the posting of the new testing guidance was inevitable. He believed he was trying to salvage as much of the public health elements of the document as possible. But his reversal on testing was a disturbing sign of how quickly testing had been lost to politics. In the past, the administration had boasted that our ever-increasing testing capacity was a signal that we were on top of things—when we weren’t. Now this boast had been turned on its head: With this new Atlas-driven guidance, Trump had put in place a formalized message through both the White House and the CDC that would reduce testing at a crucial moment for both the election and the pandemic. This had to be countered.
The case numbers pointed to another dire period. “I am at a loss [as to] what we should do,” I wrote to Bob, Seema, and Tony on August 13. “We need to stop these infections, or there will be 300K [dead] by Dec[ember].” I was wrong. By November 30, there were nearly 260,000 deaths. By December 15, we would surpass my 300,000 estimate. By Christmas, we were on the verge of 350,000 deaths.
I knew what to do to stem the incoming tide of infections and deaths. I felt handcuffed, but not helpless. I’d figure out a way to manage the crisis and keep the numbers from rising even higher. The UPenn/CHOP model had worked well in Arizona and elsewhere. That was the message I would carry to the governors and public health officials.
IRUM AND I SET out on another round of state visits—this time to the heartland. Iowa, Ohio, Kentucky, Tennessee, and Virginia were on the schedule. In each place, cases were either rising or had briefly plateaued at a high level. Also, many states did not yet have a plan in place for returning college students. We had to push nationwide for weekly mandatory testing of those students, regardless of whether they lived on campus or off.
The testing issue followed me to the states. The new CDC guidance equivocated on who should be tested after being exposed: “You do not necessarily need a test unless you are a vulnerable individual or your health care provider or state or local public health officials recommend you take one.” In other words, the CDC wasn’t directly saying “Get tested.” The media cut through the flimsy language and called it what it was: most outlets reported that the CDC had come out against asymptomatic testing. After all, it was right there on the CDC website—what the Trump administration had wanted for months: fewer tests and fewer cases. But these weren’t the only words in the new guidance that could potentially cause confusion. A few lines down, the guidance had retained the phrase “you can be infected and spread the virus but feel well and have no symptoms.” Obviously, the CDC’s contradictory remarks created confusion.
Meanwhile, the scientific community was in an uproar. The CDC’s most recent word on testing was roundly blasted as wrong. Reducing testing was not a solution; it was a problem. But the guidance worked as the Trump administration had wanted it to work: testing across the United States plummeted in most states by 5 to 10 percent and in some by 20 percent, and at the worst possible time. Of course, for some in the administration, it was the best possible time: they were out campaigning. And unless Americans were really paying attention to everything in the news, they wouldn’t have fully understood why the case numbers were dropping. Those plummeting numbers falsely signaled a trend that the administration, busy out on the campaign trail, heralded as a triumph.
I’d been in the field enough to know that there were governors who would embrace this new guidance, so I had to do everything I could to refute it, to make clear—whether it was internally, within the White House, or externally, in the governor’s reports, or while out on the road—that they should ignore it. I would continue to challenge the administration’s stated and unstated faulty positions. The fall was coming. We needed to continue to warn people about what the future held. Testing needed to be expanding, access to testing needed to be expanded.
Given everything that had been going on with Scott Atlas, and the chaotic nature of the White House, I thought it best to put those sentiments in writing: “If one is threatened: we are all threatened. I trust we have each other’s backs as[,] united[,] we can weather this to ensure we continue to make progress to save American lives.” We needed this unified approach. I’d been out there in the states, proselytizing for the UPenn/CHOP model; advocating for mask wearing, expanding testing, and ending in-person, maskless get-togethers. We had to counteract the administration’s anti-mask, anti-testing messaging. President Trump, his family, his aides, Vice President Pence, and many others were campaigning in mask-free venues; fortunately, many were held outside. But campaigning with packed rallies drove the perception that all was fine and that if you truly wanted to “Make America Great Again,” you wouldn’t wear a mask, either. I tried to convince the vice president innumerable times to wear a mask. He said he would. I was very concerned that he would get infected. I spoke with Marc Short, reminding him that it was his responsibility to protect the vice president from getting infected, including at the Atlas event.
On August 25, I wrote a sharply worded email to Marc Short, informing him that I would not be a part of Atlas’s roundtable, especially not with a fringe group who championed herd immunity and who believed that the United States could be like Sweden, which had followed the same path of minimal mitigation—but with population far, far healthier than our own. I offered to be out of DC, on another visit to the states, to give the vice president cover for my absence. And again ask him to stop the roundtable. I again cited my three hundred thousand death figure, adding that half a million people would likely die before a vaccine was in full use.
In the end, with the help of many others, I was able to stop the big circus Scott Atlas had planned for his open roundtable with major press coverage. He and the other doctors came to the White House to meet in private with the president, some advisors, and with Secretary Azar, but the press wasn’t present. In requesting that no press be allowed, I had been able to diminish some of the public perception that Scott Atlas and his theory had taken control of the response. As much as I would have liked for him to have no voice in the public discussions, and to engage in constructive dialogue with the docs on the task force, it was more important that he stay a talking head.
I had to do more to achieve that aim. I would be out of the White House for a period, on state and university visits. Somehow, I had to cut Atlas off from any degree of influence he might try to exert on anyone short of the president, whether during my absence or while I was still there. Individually, I contacted Marc Short, Mark Meadows, and Jared Kushner. I communicated clearly to everyone: “I won’t be in any meetings any longer if Scott Atlas is present at them. If that means a meeting in the Oval Office[,] it doesn’t matter. If it’s at the task force, it doesn’t matter. If it’s at the Covid Huddle, I don’t want him there.”
Not only did they indicate that they understood, but they agreed. Scott Atlas no longer attended task force meetings. I would be out on the road for much of the coming months, but Tyler Ann was present for all the Covid Huddles. I would phone in from the road to participate in these meetings, and she’d confirm for me that Atlas wasn’t present.
After sending my email to Marc Short, I was finally able to meet with Vice President Pence alone. For days, I’d been asking for ten minutes of his time, but he had been out campaigning. I knew that what I had to say would put him in a difficult position. If we met privately, I’d be able to give him plausible deniability.
I immediately reiterated the points I’d been making to the docs and to Marc Short, Mark Meadows, and Jared: Scott Atlas was a danger. The position he and the president were advocating wasn’t working and would worsen the fall surge that was rapidly approaching. When I told the vice president that Scott Atlas was effectively persona non grata in the task force and in the Covid Huddles, he nodded in agreement.
Emboldened, I took things a step farther: “You know that I’ve been out in the field, meeting with governors, doing press, and meeting with the community. I want you to know that what I have been saying and what I will continue to say in every state contradicts what the president is saying publicly. I want to be clear: I can’t support what Scott Atlas is saying, and I will need to say exactly the opposite of what the president is saying.”
The vice president’s eyes narrowed for a moment. He considered his response for another few seconds. Finally, he looked me in the eye, his voice steady, and said, “You need to do what you need to do.”
I read this as permission for me to do what was right. It meant he understood and tacitly supported my position over Atlas’s. The vice president was already getting constant pressure internally for what I was saying and doing, but in that moment, I saw he believed that taking the fight to the governors was the right thing to do. From that point, I would continue to infer that Vice President Pence had my back. The administration could always say I’d gone rogue, that I was part of the “deep state,” bent on undermining the president and disrupting the election by distracting voters from the “real” issues. I wanted to believe—and this is my optimistic nature rising to the fore—that the vice president knew I was doing what was right and needed. I couldn’t give up hope that our arguments had been persuasive, that data and facts and science had prevailed at least to some degree.
When I returned to my office, I received a draft of an op-ed Scott Atlas had written. It looked like I had more contradicting to do. Now that he was part of the White House staff, any of Atlas’s communications would be subject to review by the Office of the Staff Secretary. Jared Kushner and a long list of others in the Executive Office of the President were on the distribution list. I cc’d them all in my response, which, once again, was a cogent refutation of Atlas’s arguments. I paid special attention to his false claims about lockdowns, reminding everyone that our mitigation efforts—wearing a mask indoors, testing to find silent spread, decreasing social interactions—could not be characterized as part of a “lockdown.” Keeping retail spaces and schools open were the opposite of a lockdown. I spelled out again the simple measures that had been enacted to produce very positive results. I reminded them that we’d been balancing the economy with public health. I also referenced the role silent spread was playing.
Hoping to prevent the damage that another Atlas op-ed might do, I wrote a separate email to Jared: Why were they letting Atlas be a distraction with this op-ed during the Republican National Convention? In another email to him, I again refuted the notion of herd immunity, pointing out why it was impossible to say it was responsible for some of the improvements we were seeing. At the end, I wrote that if Dr. Atlas now spoke for the president and his policies, Jared should let me know, and I would “take that under advisement.”
Jared didn’t respond. As was the case when his father-in-law was similarly noncommittal about the extent of Atlas’s influence, I presumed that Jared’s silence on the matter meant that Atlas didn’t speak for the president. Even if Jared had said that Atlas did, I would have stayed on and fought harder against him. Fundamentally, I was constituted to always do the right thing, even if others didn’t see it that way. I couldn’t change my stripes; nor did I see a need to.
Being on high alert has always been a part of my nature and my profession. With Covid-19, I was always eager for anyone in higher-risk age groups or those with underlying medical conditions to be as vigilant as possible. In the White House, I spoke with all forms of staff, from housekeeping to senior advisors to the vice president. I was preaching the gospel of vigilance. My concern wasn’t bound by any strictures, political or otherwise. Consequently, my concern extended to the members of the Biden campaign and the candidate himself.
On August 29, I saw an opportunity to put that worry into action. I received a text from my friend Chip Lyons. He was contacting me on behalf of David Kessler, the former FDA commissioner, who later served as the dean of the Yale School of Medicine and then in the same capacity at the University of California, San Francisco Medical School. Chip stated up front that Kessler was then informally serving as an advisor to former vice president Biden and was in communication with him regularly. Kessler wanted to be in touch with me, but he understood that I might be reluctant. I let Chip know that I was willing to speak with Kessler.
At about this time, I’d been reading the Biden pandemic plan, which had been made widely available. In short, I was impressed by it, particularly because its emphasis on the importance of testing aligned with my own beliefs on that crucial mitigation tool. I wasn’t sure what role, if any, Kessler had played in developing the Biden plan, but now knowing that he was an insider on the Biden team, I was eager to learn why he wanted to be in touch with me. I suspected that it had something to do with vaccine development. I’d learned that the Biden camp was in contact with the various manufacturers directly, so, in my mind, I believed they knew as much as I did. Perhaps they hoped I’d have some additional insight.
I wanted them to know what data we were collecting—so if they had additional data requests or additional thoughts we could incorporate them. I was always open to learning from others and doing things better. Kessler came to my house shortly after the twenty-ninth and I provided him with samples of the daily reports—both my daily pandemic analysis and the four pager we created for the Covid Huddle and governor’s reports we’d been producing. I went through the data streams we had created to be able to get eyes on the pandemic across the country—from test positivity, to cases, to hospitalizations, to deaths, down to the most granular level possible—counties and metro areas. Pictures and graphics told the story, some of which are still updated in the HHS Community Profiles we quietly posted in the late fall. He thanked me and said that he’d review them and get back to me with any questions or requests for additional data he thought would be useful to have integrated into the data stream we had created. I was glad to know that he and others on the Biden team were thinking ahead, and that my usual desire to have a Plan B and Plan C was being met. November’s election was still months away, but preparing for either result would ensure continuity of care so essential to proactively managing the next surge.
I felt comfortable agreeing to speak with Kessler for several reasons. The first was that I had always acted apolitically throughout my career. If I could discuss pandemic matters with another medical professional in any small way, I was open to the idea, regardless of party affiliation. Second, because I didn’t know anything more about vaccines than the Biden people, I wouldn’t risk leaking any information. Speculation about the Trump administration’s hope to spring a preelection vaccine surprise was rampant in the media, but I had no secrets to share. And my third reason for agreeing to speak with Kessler was my concern for the health and safety of the former vice president. Biden wasn’t actively doing personal campaign appearances, but many of his staffers were out on the campaign trail (along with, critically, the uniformed Secret Service protection), who could be exposed and asymptomatically infected. I wanted to make sure they knew how critical silent spread was, how important testing and quality masks were across the board. I also wanted them to know what data we were collecting so if they had additional data or additional thoughts we could incorporate them. I was always open to learning from others and doing things better.