President Trump’s accusatory tweet barreled in at 7:49 a.m. a week ago Saturday: The “deep state” at the Food and Drug Administration was trying to sandbag his election prospects by slowing progress on coronavirus treatments and vaccines until after Nov. 3.

Shocked and upset, FDA Commissioner Stephen Hahn, who was tagged in the tweet, immediately began calling his contacts at the White House to find out why the president was angry. During his conversations, he mentioned the FDA was on the verge of granting emergency authorization to convalescent plasma as a treatment for covid-19. The agency planned to issue a news release.

The White House would upend those plans, turning a preliminary finding of modest efficacy into something much bigger — a presidential announcement of a “major therapeutic breakthrough on the China Virus,” as White House press secretary Kayleigh McEnany previewed in a tweet late that Saturday night.

At a news conference on the eve of the Republican National Convention, Trump lauded an emergency authorization for convalescent plasma as a “very historic breakthrough.” Hahn, who had rushed back to Washington from a family home in Colorado, was initially restrained but then doubled down on Trump’s talking points. He said that 35 of 100 people with covid-19 “would have been saved because of the administration of plasma”— a gross overstatement denounced by scientists and public health experts.

The misrepresentations became a stunning debacle for the FDA, shaking its professional staff to the core and undermining its credibility as it approaches one of the most important and fraught decisions in its history amid a divisive presidential election — deciding when a coronavirus vaccine is safe and effective. Yet again, the president had harnessed the machinery of government to advance his political agenda — with potentially corrosive effects on public trust in government scientists’ handling of the pandemic.

Hahn apologized the following day for misspeaking, saying on Twitter, “The criticism is entirely justified.” But demoralized employees felt he had allowed the agency to become a prop in the president’s reelection campaign — a bit player in a reality TV show scripted by political operatives, not scientists, according to several people familiar with their thinking who spoke on the condition of anonymity to protect the employees from retaliation.

It wasn’t so much the inaccurate use of 35 percent — anyone can make an error. What rankled agency insiders was the way a defensible FDA decision to authorize an incremental advance for a disease with few treatments was being described as a huge leap forward in an over-the-top White House rollout. How, they wondered, would the FDA have any credibility on a vaccine decision if it bungled something much simpler?

“There is this weird Stockholm syndrome where FDA higher-ups are starting to identify with their captors,” said Gregg Gonsalves, an assistant professor at the Yale School of Public Health. “. . . It’s not clear they’re going to do the right thing.”

The incident was just the latest misstep for the FDA, following flip-flops earlier this year on authorizing and then revoking clearance for the antimalarial drug hydroxychloroquine and its initial decision not to regulate coronavirus antibody tests.

It also feeds a long-running narrative of a White House repeatedly undermining its health and science experts, not just at the FDA but also at the Centers for Disease Control and Prevention. Last week, the CDC came under fire from a host of medical and public health groups and infectious-disease experts for an abrupt change to its guidelines, which no longer recommend testing for asymptomatic people even if they had contact with an infected individual — a shift that coincides with the president’s stated desire to reduce testing.

“I’ve been following health regulatory decisions for decades and have never seen this amount of White House arm-twisting to force agencies like FDA and CDC to make decisions based on political pressure, rather than the best science,” said Jerome Avorn, a professor of medicine at Harvard Medical School, who decried the “routine policy distortions we now see nearly every week.”

The FDA’s situation is further complicated by an inexperienced commissioner who former agency leaders say failed at a critical task: to clearly explain the complicated “risk-benefit” calculation that goes into every drug authorization or approval. Hahn, a radiation oncologist with no government experience before joining the FDA last December, declined to comment for this report. He has stressed in recent days that he is committed to the independence of FDA scientists and is determined to shore up public trust.

The White House said in a statement that the president has pushed for the best possible science because he believes all Americans should have access to proven, safe and affordable treatments.

“Every decision the FDA has made under the Trump Administration has maintained the agency’s gold standard for safety and been data-driven to save lives, and this false narrative that the media and Democrats have created that politics is influencing approvals is not only false but is a danger to the American public,” White House spokesman Judd Deere said.

The behind-the-scenes look that follows of the decisions that led up to the briefing and the ensuing fallout is the result of interviews with more than 30 people, including current and former administration officials, their advisers and other experts who spoke on the condition of anonymity to disclose internal deliberations.

'Less and less autonomy'

Leading up to the event were infighting among some of the nation’s top scientists, frustrating delays and the last-minute arrival, just hours before the news conference, of a trove of new data, previously undisclosed, that was rough and difficult to interpret but that some thought raised questions about convalescent plasma’s benefit.

Weeks before the briefing, FDA scientists had concluded that plasma, an old therapy in which the yellowish liquid portion of blood is taken from recovered patients and transfused into ill people, was safe and might offer benefit to some hospitalized covid-19 patients. They reviewed published data, as well as a national, early-access plasma program being overseen by the Mayo Clinic, and were confident the treatment met the relatively low standard for an emergency use authorization. Such clearance, a temporary approval used during public health emergencies, requires only that a treatment “may be effective” and its “known and potential benefits” outweighed its risks.

But the FDA would soon learn scientists at the National Institutes of Health disagreed with that assessment. Officials’ willingness to hear NIH’s objections slowed the process and frustrated several White House officials.

Hahn already had critics in the West Wing. His refusal to reauthorize hydroxychloroquine as a covid-19 treatment had earned him the enmity of White House trade adviser Peter Navarro, who is close to Trump and remains a strong advocate of the drug. The FDA cleared the drug earlier this year but then revoked that when studies found cardiac risks and no benefit — a move Trump complained about as recently as the Republican convention.

Another sign the White House was exerting more political control was its installation about two weeks ago of a new chief FDA spokesperson — Emily Miller, a staunch gun rights advocate with no health or science expertise. She previously worked as a reporter for One America News, a conservative cable channel that frequently espouses conspiracy theories and is allied with Trump.

Meanwhile, the agency’s relationship with its parent, HHS, is increasingly tense as HHS Secretary Alex Azar and Hahn disagree on policy matters. Over Hahn’s vociferous objections, HHS recently blocked FDA from regulating laboratory-developed tests, including some coronavirus tests, a move that raised concerns about an influx of faulty tests.

Early last week, HHS’s political leadership was angry that Hahn had publicly apologized for misstating plasma’s potential benefits without clearing that with HHS communications staff. Michael Caputo, assistant secretary for public affairs at HHS, argued that it muddled Trump’s and the administration’s message on the treatment, according to one former and one current official.

HHS officials are “making clear what their expectations are in a way that is not traditional,” said another individual familiar with the situation. “It doesn’t mean it’s illegal or wrong. . . . Generally speaking, you don’t have HHS leadership weigh in on specific drug approvals, an [emergency use authorization], or those sorts of things.” The FDA does have “less and less autonomy.”

HHS spokeswoman Caitlin Oakley denied conflicts between Hahn and Azar. “Your characterization is inaccurate,” she said in a statement. “HHS and FDA are working hand in hand to combat covid-19 and that’s where our focus remains.”

The growing political pressure on the FDA comes as some administration officials say they are trying to design a PR campaign to convince the public that scientists are in charge of developing vaccines and treatments and that anything the government approves will be safe and effective — an effort repeatedly upended by Trump.

“How can we get people to trust a vaccine if the president is saying the deep state is behind it?” one senior administration official said.

'People thought this was probably going to work'

For more than a century, doctors have used the plasma from patients who have recovered from infectious diseases to treat sick people; the theory is that antibodies in the plasma bolster the immune system’s war against the infection. The FDA, swamped by individual requests to use the approach on dying patients, shifted gears and approved plans by the Mayo Clinic to set up a nationwide program to give patients access to the therapy.

Ideally, an unproven treatment should be tested in a clinical trial where patients are randomly assigned to receive either the therapy or a placebo. But Michael Joyner, the Mayo anesthesiologist who led the plasma program, said there were concerns that such trials would only be offered at a limited number of hospitals.

“People thought this was probably going to work — it had worked in the past, it was biologically plausible,” Joyner said. “People wanted to make sure it was widely available.”

The program was initially set up to treat 5,000 patients, but it grew to more than 75,000. Plasma donations can vary widely in the dose of antibodies they contain, so Mayo researchers began testing samples to see if the patients who had received higher doses of antibodies fared better than those who got lower doses. Such a finding wouldn’t prove it worked without a control group for comparison, but it would be suggestive.

By late July, Mayo researchers were finding that patients who received high-dose plasma were less likely to die than those who received low-dose plasma, if they received it early. NIH officials asked to see the data the FDA was compiling, and on July 29, agency officials briefed them on a Zoom call.

Peter Marks, director of the FDA’s Center for Biologics Evaluation and Research, which regulates blood products and was overseeing the authorization process, believed that convalescent plasma met the relatively low bar for an emergency authorization. He thought the evidence — which included several different analyses of the Mayo data, use of the plasma to treat other respiratory illnesses, animal studies and more than a dozen clinical reports — showed it was likely to provide a modest benefit for certain patients.

But the agency’s effort to authorize plasma got bogged down after that Zoom meeting. NIH Director Francis Collins was skeptical and immediately expressed his opposition to the emergency authorization, saying the evidence was too weak, according to two individuals familiar with the situation. Anthony S. Fauci, director of the National Institute of Allergy and Infectious Diseases, also was critical.

Hoping to bring NIH on board, the FDA delayed a decision on the authorization and asked Mayo for more data, including measuring the levels of antibodies in thousands of additional samples. Such an analysis would potentially help build confidence that the modest efficacy signal was real.

How much data is enough?

The White House, meanwhile, was increasingly impatient about getting clearance for what Trump called “the beautiful ingredient.”

Aides including Adam Boehler, the chief executive of the U.S. International Development Finance Corporation and a close associate of Jared Kushner, the president’s son-in-law, repeatedly called the FDA to check its progress. Sometimes, the officials would call several times a day, according to an individual familiar with the situation.

Hahn brought Collins’s and Fauci’s objections to the White House, believing all sides should be heard. Heated disagreements went on for several days, with Collins fiercely opposing an authorization without more concrete data, according to a senior administration official. White House aides felt Hahn should have handled the issue himself.

“Francis thinks he’s a master drug developer and he has an opinion on everything,” said one person familiar with the situation. “But it’s up to Hahn. It’s FDA’s decision.”

On Aug. 19, the New York Times reported the emergency authorization was on hold because of objections from top NIH officials. The FDA, furious about NIH’s public airing of concerns, issued a statement saying NIH did not dictate FDA decisions and vowed to press forward.

In a frank phone conversation earlier this month, some White House officials — including Boehler and John Fleming, a senior adviser to White House Chief of Staff Mark Meadows — scolded Collins, saying the New York Times article about the dispute created public confusion and undermined the process.

Collins declined to comment for this report, as did Fauci.

As Trump was briefed on the disagreement between the FDA and NIH, he was increasingly annoyed, White House officials said. That helped prompt the Aug. 22 tweet, which turned a debate over clinical data into a politically charged drama.

Some White House aides shared Trump’s frustration but worried the angry tweet could erode public trust in critical FDA vaccine decisions. They did not view the delay as Hahn’s fault. Still others believed Trump was trying to send a message to NIH but confused the two agencies.

'Off the vaccine itch'

Even as the debate swirled, FDA officials made the decision to authorize plasma by Aug. 24. The Mayo program had become far larger than anyone expected and was becoming a burden for the medical center.

But while the FDA plan was to issue a news release, and perhaps brief reporters on a call, HHS and the White House saw a political opportunity.

“HHS is always looking for good news that the president can be involved in,” said an individual familiar with the situation. “By talking about plasma, it got him off the vaccine itch for a couple of minutes. It scratched that itch.”

At 9 p.m. on Aug. 22, the additional data that had been requested weeks before by NIH and FDA landed in the inboxes of FDA officials. In a mad dash, government statisticians tried to determine what it showed. But they found the data hard to interpret and full of errors. There was some disagreement about whether the data indicated the plasma did not provide any benefit at all, but FDA statisticians concluded they needed more time to “clean” and scrutinize the information before reaching any conclusions, according to several people familiar with the discussion.

FDA officials had already decided they would not stop the authorization based on new information unless it showed a safety problem, which this did not. They are continuing to review it, confident it will confirm their view of a modest benefit. Hahn has publicly said the agency would reconsider its decision if new data shows the treatment is ineffective.

“An ambiguous finding derived from one data set that the FDA received that was not properly validated (and subsequently found to be flawed) was not a reason to delay taking a regulatory action that could potentially bring benefit to additional patients in need,” the FDA said in a statement.

It noted the agency “makes its regulatory decisions based on the totality of the available evidence.”

Joyner of Mayo acknowledged such data “takes time to curate and validate. It is unrealistic to expect insights in essentially real time.”

Continuing fallout

FDA now faces a crisis of confidence inside and outside the agency in the run-up to sensitive decisions on vaccines that would be used by millions of people. If, as some predict, it clears the first vaccine under the lower bar of emergency authorization to make it available to health-care workers, for instance, the action may be viewed as politically driven — even if it is scientifically sound. Hahn has said several times that such an authorization is on the table, and that any decision will be based on the data.

On Friday, Miller, the agency’s newly arrived spokeswoman, was removed in what many agency watchers saw as an effort to restore credibility. In addition, a longtime communications consultant who had advised Hahn to correct his misstatements about convalescent plasma had his FDA contract terminated.

“I advised the commissioner, as I would do so again, to correct the record to clarify points about the efficacy of plasma,” the consultant, Wayne Pines, said in an interview. “I’m an advocate, as I always have been, for transparency at the agency. The public needs to have confidence in the FDA.”

HHS Chief of Staff Brian Harrison said in a statement the contract cancellation was “routine” and “the timing was 100-percent coincidence.”

Those familiar with Hahn’s thinking say he intends to try to restore confidence by taking steps to make the FDA’s decision-making as transparent as possible, including explaining more clearly the evidence the agency uses in reaching its conclusions.

But some Democrats say that’s not enough. They’re demanding that Hahn assure them he has put in place safeguards to avoid pressure from Trump as the agency pivots to vaccines.

“The president just refuses to stop,” House Energy and Commerce Committee Chairman Frank Pallone Jr. (D-N.J.) said in an interview. “All these things sow all kinds of confusion in the public’s mind, so the stage is already set where a lot of people are likely to think this process is not scientific, not transparent.”

Some FDA staffers acknowledge concerns. “For the first time,” said one official, “the agency is involved in something that is existential to the president. It’s our turn in the barrel.”

A senior administration official said the agency must avoid the fate of some other agencies. “CDC is not a public health agency anymore,” the official said. “And I would like to try to make sure that the FDA doesn’t end up being a disaster area also.”