Skip to Main Content

There are serious signs the Food and Drug Administration is getting cold feet over the notion of issuing emergency use authorizations to allow for the widespread early deployment of Covid-19 vaccines.

Instead, it appears the agency may be exploring the idea of using expanded access — a more limited program that is typically used for investigational drugs — in the early days of Covid vaccine rollouts.

Whereas a few weeks ago the agency’s concern was to protect against the possibility that unproven vaccines would be pushed out prematurely due to pressure from President Trump, now the fear is that early authorization of vaccines could squander a one-time chance to determine how well the various vaccines work and which work best in whom.


Marion Gruber, director of the FDA’s office of vaccines research and review, put the issue on the table when members of the Vaccines and Related Biological Products Advisory Committee began to discuss a series of questions FDA staff posed at the end of a grueling day-long virtual meeting Thursday.

“We are concerned about the risk that use of a vaccine under an EUA would interfere with long-term assessment of safety and efficacy in ongoing trials and potentially even jeopardize product approval,” Gruber said. “And not only the first vaccine, but maybe even follow-on vaccines.”


The acting chair of the committee, Arnold Monto, from the University of Michigan, who has decades of experience studying the efficacy of vaccines, put it in more dire terms during the discussion. The maker of a Covid-19 vaccine that is given an EUA might not be able to generate enough additional data to ever successfully apply for a full license, Monto said.

The problem stems from thorny ethical questions about whether — once a vaccine has been cleared for use by the FDA — the people who were randomly assigned to receive a placebo in its clinical trial must be informed and offered vaccine. Vaccinating the people who received placebo injections — the trial’s control arm — would end the ability to continue to compare the two groups after what would have been a short trial.

Pfizer and BioNTech, the collaboration expected to be first to apply for an emergency use authorization — sometime in mid-November — have indicated they plan to unblind their trial and offer people in the placebo arm vaccine. (When a trial is blinded, participants don’t know if they received vaccine or a placebo injection.)

Early unblinding of these trials actually runs counter to the FDA’s advice. The agency is urging vaccine manufacturers to keep their trials blinded as long as possible, to collect as much data as they can.

The EUA could trigger another related problem. People in clinical trials might choose to pull out and try to get the vaccine that has been authorized for emergency use, especially if they are in a high-risk group that is likely going to be at the front of the line when vaccines begin to become available. Enrollment in the vaccine trials for other Covid vaccines might slow as people decide they don’t want to risk being randomized to receive a placebo, and instead wait for their turn to get vaccine cleared under an EUA.

If they are stopped too early, the trials, which were structured to come up with quick answers as to whether Covid vaccines prevent symptomatic Covid-19 infections, might fail to answer additional, important questions needed to figure how to best use the various vaccines that have been produced, if — as expected — multiple vaccines prove to work.

For most of the vaccine trials, the “primary endpoint” is showing they prevent symptomatic Covid-19 disease in at least 50% of vaccinees. But there are also “secondary endpoints.” Secondary endpoints include whether the vaccines reduce the number of severe Covid cases, and how well they work in important subsets of the population such as the elderly or people of color, who have been disproportionally hit by the pandemic. Trials that end after reaching their primary endpoint will leave gaping knowledge gaps, a number of experts warned.

“We may have limited and in some instances no information about some of the secondary endpoints,” said Stephanie Schrag, an epidemiologist from the Centers for Disease Control and Prevention who made a presentation at the meeting. “This would be particularly true in the instance of an early EUA because many of these secondary endpoints require longer time than the primary to accrue events.’’

Jesse Goodman, a former FDA chief scientist who listened to the meeting, concurs with the agency’s concerns. Answers about how well these vaccines work, how long they work, and which work best for which segment of the population will always be clearer if they are generated by randomized controlled trials — the gold standard of clinical trials.

“Let’s say one of these vaccines has, you know, 60% efficacy and another one has 80%. Or one of them, the efficacy waned after four months, and the other lasts a last year. It’s going to benefit people to find that out now rather than three years later from crummy observational data,” Goodman told STAT.

Using expanded access rather than emergency use authorizations would be a more cumbersome process and create some challenges. People who were to be vaccinated would have to sign informed consent forms, which requires a discussion of risks and benefits during vaccine administration; safety data from vaccinees also would have to be gathered. But Goodman, who actually suggested expanded access as an option in a commentary he co-wrote in JAMA in July, said that this route might better ensure the continuation of the clinical trials.

Members of FDA’s advisory committee appeared to share the concerns of the agency staff who were asking for their guidance.

Sheldon Toubman, a lawyer from New Haven, Conn., who is the consumer representative on the panel, said it would be his preference that the vaccines not be deployed under emergency use authorizations. Toubman said the public fears politics, not science, is driving the approvals process and EUAs won’t dispel those views.

A number of polls, including one published Monday by STAT and the Harris Poll, have shown that the public is cooling to the idea of Covid vaccines. The declining percentage of Americans who say they are willing to be vaccinated is thought to be linked to the politicization of the vaccine approvals process, which Trump has attempted to fast-track in the weeks leading up to the election.

Toubman’s position was echoed by representatives of the HIV Medical Association and the Infectious Diseases Society of America during a public comment section of the hearing.

Emily Martin, an assistant professor of infectious diseases epidemiology from the University of Michigan who studies vaccines effectiveness, urged the committee to advise the FDA that EUAs should not allow companies to halt their clinical trials early.

“Without complete and full randomized trial data, we will lack the evidence base needed to monitor and adapt vaccination strategies as needed over the many years these vaccines will be in use,” Martin said. “Ending these trials early will irrevocably hamper our ability to optimize the effective use of the vaccines going forward.”

Goodman agreed that an early EUA could actually leave a manufacturer with too little data to persuade the FDA to issue a full license, though he thought that was not the likeliest of scenarios.

“I think more likely than that, but really, as concerning from a public health point of view would be if … we don’t get adequate enough information to understand how these vaccines compare with each other and perform in terms of their safety and efficacy, we don’t know how to use them in a public health context,” he said. “Which could end up hurting far more people at the end of the day.”

  • When is the COVID 19 vaccine going to be issued to the public. Russia and china have their own vaccines, and here we are with none ready to go.

    Reminds me of when George Bush (2nd) said how inferior Canadian drugs were to the United States pharmaceuticals. Next thing you heard, George Bush requested Canadian Flu vaccines made available to the U.S. population. Maybe we can learn from all this and get the Russian COVID 19 vaccine sent to the U.S. and start innoculating our country. Might speed up our own vaccines to get out to the U.S. population.

  • I watched the full 8+ hrs of the meeting. No real mention that thousands of Americans are dying every week. There was plenty of concern expressed about the risk of deploying the vaccine and not enough regarding the risks of NOT doing so. This is different than other vaccines that the FDA and the vaccines advisory committee typically review. We have an infection that is actively lethal for a huge portion of the population right now. It really requires a different mindset. “Abundance of caution” is not necessarily the right response.

    Observational studies are very do-able. The study patients will make themselves available for the full two years if the placebo patients are crossed over to active agent, but not if they drop out and are lost to follow-up.

    • Engineer, not health sciences person here — I agree with you. 10-20-30 million people get vaccine A, another 10-20-30 get vaccine B. Sure, it’s not randomized, double blind, placebo controlled, bla bla bla, but I refuse to believe that out of such numbers one cannot get useful information on effectiveness and side effects. Maybe the medical community should make sure it’s up to date on machine learning techniques.

  • My personal opinion is that EUAs are given for a reason…thus what it stands for EMERGENCY USE AUTHORIZATION. This is a public health crisis and an emergency in every sense of the word. This is a novel coronavirus which has killed over 200,000 people in this country and shutdown the US economy. These people who are continually moving the goal post are doing so for pure political reasons. They just don’t want Trump to get credit for streamlining the process in order to SAVE AMERICAN LIVES !!!!

  • I listened to much of the FDA Ad Com yesterday, frankly it was a disappointment and the Ad Com members themselves were underwhelming.

    FDA made a decision to consider Expanded Access at the very end of a long day of deliberation; it is neither practical nor ethical.

    If granted EUA, Pfizer (and likely Moderna ) planned to stop enrollment, unblind and give active vaccine to placebo patients, since it would be unethical to continue placebo if the vaccine is shown to be effective… the first efficacy threshold for EUA is high (>=74%) ….why would high risk placebo patients remain in the trial for a year, knowing that they could be protected with an effective vaccine? In addition, courageous African American volunteers (and the AA community) would be turned off to the vaccine if many on placebo get infected when they could have been vaccinated (google the Tuskegee Study) . They should be rewarded for their courageous contribution to the study, and not kept at risk for a year.

    The FDA further threatened to not grant full licensure if companies unblind, citing this need for long term safety data with placebo control, which is frankly BS. Most vaccine AES occur within days or weeks of administration and are transient…the mandatory 2 mos of post dose follow up is adequate. Plus how does placebo adjustment help with a rare long term safety event? The FDA has always endorsed post approval prospective registries where one could use matched referent groups and complex stats to compare. Trial patients on vaccine could thus be followed prospectively via a safety registry. This would also provide important data of the durability of the immune response.

    FDA agreed to and endorsed the event driven design with early looks for EUA and they should find a way to live with it.

  • Somebody has to explain it…I understand comparing a placebo to a vaccine to see if the vaccine really is making a difference. But then you know the vaccine works, so you give it to people. Why does it not make sense to then find out more data based on statistics of what happens when you have it?
    You get vaccine > you get covid > you check when you were vaccinated and look for patterns.

    There is a big difference between what doctors want to have happen and what the public will accept. “Hey guys! I know people are dying but we’re not going to give you this shot that we’re 99% sure works like a charm, because, hey, we need more data that we…well, we want it. Hang on for 6 more months. Maybe longer. We’ll let you know.”

    • good point Hank. Perhaps they are worried not so much about efficacy but lethality of the vaccine. If a vaccine comes out and kills one in a thousand, it could scare the public about getting vaccinated regardless of a future, safer version coming out. Most people have a low chance of dying from Covid-19, so if the vaccine runs the same per-individual lethality as the disease, why get the vaccine? Just a thought.

    • Yes, you can. But that kind of trial (called a non-inferiority or superiority trial) takes a lot longer to do. Each new vaccine has to prove it’s better than — or at least not less protective than — the first vaccine. The bar is higher, the studies are bigger. That would prolong the time to market of other vaccines, some of which may be more effective or easier to administer than the first vaccine.

  • Here we go again! The FDA is moving the goalposts once again. Every time they come up with the requirements for an EUA, they change it to delay the vacine, again and again. Based on their new argument of waiting to see which vaccine is most effective, it will be YEARS before we have that data because there is an enormous amount of vaccines being tested. Do we wait for the results of all of them? Do we move the goalpost again…..and again…. and again? Enough FDA! Why shouldn’t people go overseas to get a vaccine? It’s better than what the FDA offers, which is…nothing!

  • Vulnerable populations should not be sacrificed in order to add additional data forever and to what end. This is a fatal disease in the elderly and they need vaccine NOW!!!! Control freaks don’t have a right to withhold information from the vulnerable or placebo groups

  • Vulnerable populations should not be sacrificed in order to add additional data forever and to what end. This is a fatal disease in the elderly and they need vaccine NOW!!!! Control freaks don’t have a right to withhold information from the vulnerable or placebo groups

Comments are closed.