The Department of Health and Human Services last week announced a new standard for testing the safety of vaccines, a “radical departure from past practices.”
All new vaccines will be evaluated against a placebo, an inert look-alike that serves as a point of comparison, the department said. Health Secretary Robert F. Kennedy Jr., as well as many anti-vaccine groups, has long argued that placebo-controlled trials were the only way to fully understand vaccine side effects.
To scientists who have spent their careers evaluating vaccines, the plan did not seem so radical. New vaccines are often tested against a placebo in clinical trials. One researcher has created a crowdsourced spreadsheet of more than a hundred examples.
But it also concerned vaccine experts that Mr. Kennedy seemed not to recognize the circumstances when placebo groups are neither ethical nor practical. The idea is widely accepted by scientists and enshrined in ethics frameworks for medical research.
“He’s asking for something that’s not ethical,” said Arthur Caplan, a leading bioethicist at the New York University Grossman School of Medicine.
Why Are Placebos Used in Vaccine Trials?
Randomized placebo-controlled trials are often described as the “gold standard” of research: they allow scientists to tease out whether the effects they observe result from the drug itself or some other factor, such as the expectation of treatment.
Patients in trials of drugs for depression or pain, for example, often report feeling better even if they haven’t received the treatment — the so-called placebo effect.
In vaccine trials, researchers also use placebos, often shots of a saline mixture, to help quantify behavioral and psychological changes that might arise from participants knowing whether or not they received the real vaccine.
Had some participants during the Covid-19 vaccine trials known they didn’t get the actual vaccine, for example, they might have avoided the virus by staying home, potentially skewing data about infection rates.
Similarly, simply believing that they have gotten the real shot can make trial subjects more attuned to possible side effects and more likely to report them.
One study found that nearly a third of participants who received a harmless placebo during Covid vaccine trials reported side effects, most commonly headaches and fatigue.
Why Don’t Some Trials Include Placebos?
Placebo-controlled trials, which may last years and cost tens of millions of dollars, are not always feasible or necessary, especially for vaccines that are updated every year, like the flu shot.
“By the time you’re done with that trial, it would be two years later and your vaccine is already outdated,” said Dr. Daniel Salmon, director of the Institute for Vaccine Safety at the Johns Hopkins Bloomberg School of Public Health. “Pragmatically, it’s just not possible.”
Instead, updated flu shots — which are only slightly tweaked each year — are tested in animals to see whether they produce an immune response similar to earlier versions of the vaccine, which have extensive safety and efficacy data.
There are also ethical reasons why using a placebo may not be appropriate. Perhaps the clearest example comes from the polio vaccine trial in 1954.
Hundreds of thousands of children had volunteered, eager for a chance at protecting themselves against a devastating disease that was sweeping through classrooms and communities, causing irreversible paralysis and death.
More than 200,000 of those children were randomly assigned to the placebo group, receiving doses of saline water instead of the vaccine. By the end of the trial, 16 of those children had died of polio.
All of the children who received the vaccine survived.
Bioethicists and scientists believe that the saline shots were ultimately necessary to prove the vaccine’s effectiveness. And since there was no alternative vaccine or treatment at the time, the children assigned to the placebo were not at higher risk than the average American primary schooler.
But institutional review boards, which are charged with evaluating the ethics of medical studies before they start, carefully weigh the experimental benefits of using a placebo against their responsibility to protect the participants in the trial.
“It is better to do a randomized, placebo-controlled trial,” said Dr. Paul Offit, director of the Vaccine Education Center at Children’s Hospital of Philadelphia. “But there is a certain cruelty to that.”
How Are Vaccines Tested Without a Placebo?
In general, review boards abide by the principle that placebos are not appropriate if there is a “proven intervention” that already exists that has an established record of safety. In those cases, researchers will compare the new vaccine with the standard of care.
For example, during the development of the first vaccine for HPV, a sexually transmitted virus that causes nearly all cases of cervical cancer, researchers compared its efficacy and safety with that of a placebo.
The vaccine was a huge success. The shot reduced the risk of cervical cancer by 70 percent.
Years later, when researchers developed a vaccine that they believed would protect against five additional types of the virus, they could not ethically test it against a placebo. Doing so would have meant withholding a safe and very effective vaccine from half the participants.
Instead, scientists measured how much more effective the new HPV shot was than the original vaccine.
Michael Osterholm, an epidemiologist at the University of Minnesota, said offering a saline shot in such a scenario would be akin to running an experiment in which “half the parachutes opened and half did not.”
The same logic often applies to clinical trials for new cancer medications, which commonly enroll patients undergoing chemotherapy and radiation as a control group, rather than a placebo.
“It’s going to be a little harder to see the effect because you’re getting some benefit from the control group,” said Dr. Caplan, the bioethicist.
“But it’s wrong to say, ‘Well, sorry, you have to die faster if you’re assigned to the placebo.’”
Apoorva Mandavilli contributed reporting.
