This week, the U.S. Department of Health and Human Services. The National Institutes of Health (NIH) has a new clinic to study and care for people with health problems after vaccination. The Centers for Medicare & Medicaid Services (CMS) has proposed paying physicians for the time they spend filing reports to the Vaccine Adverse Event Reporting System (VAERS). Electronic health records will be required to capture vaccine-related adverse events, and VAERS itself is being modernized. "We must listen to patients and families who report vaccine injuries—not dismiss or gaslight them," Secretary Robert F. Kennedy Jr. said in the announcement.
On paper, some of this sounds reasonable. People who have a real, rare reaction to a vaccine deserve good care, and better data is always welcome. Vaccine safety experts told much the same thing. Johns Hopkins vaccine-safety researcher Daniel Salmon said more research "would be very helpful," then added the condition that matters most: "It needs to be rigorous science."
That condition is where this initiative runs into trouble. To see why, it helps to start with what we already know about how often vaccines cause serious harm, and how that compares with the harm caused by the diseases they prevent.
How Often Vaccines Cause Serious Harm
Vaccines are medicines, and like all medicines, they can cause side effects. Most are mild and short-lived: a sore arm, a low fever, a day of feeling run down. Serious reactions do happen, but they are rare, and we know about them because the existing safety system found them.
Here is what the best evidence shows for some of the best-documented serious reactions:
- Severe allergic reaction (anaphylaxis). A of more than 25 million doses found 33 confirmed cases, a rate of about 1.3 per million doses. Anaphylaxis is why people are asked to wait a few minutes after a shot. It can be treated on the spot.
- Febrile seizures after MMR. The measles, mumps, and rubella vaccine causes about. These seizures are frightening for parents, but CDC notes they do not appear to cause any long-term harm.
- Low platelet counts (ITP) after MMR. A large U.S. study estimated, usually mild and gone within about a week. Wild rubella infection causes this problem far more often, at about 1 in 3,000 infections.
- Guillain-Barré syndrome after flu vaccines. In seasons when flu vaccines have been linked to this nerve disorder, the extra risk has.
- Intussusception after rotavirus vaccine. This type of bowel blockage occurs in, mostly in the week after the first or second dose.
- Myocarditis after mRNA COVID-19 vaccines. The highest reported rate was in males aged 16 to 17 after their second Pfizer-BioNTech dose, at. Nearly all of those patients were hospitalized, but 98% had been discharged at the time of review; most had symptoms resolved by discharge, and there were no confirmed myocarditis deaths among people under 30 who had no other identifiable cause.
Large independent reviews point the same way. The Institute of Medicine's 2011 review of 158 possible vaccine–adverse event pairs found that, and that the reactions convincingly linked to vaccines were few and mostly well known, such as anaphylaxis, fainting, and febrile seizures. A 2021 federally funded review of 338 studies found and found high-strength evidence of no increased autism risk after MMR.
The federal compensation program offers another useful yardstick. From 2006 through 2024, more than 5.6 billion doses of covered vaccines were distributed in the United States, and. That works out to roughly one person compensated per million doses. Even that figure overstates proven harm: about 60% of compensation comes through negotiated settlements in which HHS did not conclude that the vaccine caused the injury.
What Life Looked Like Before Vaccines
Now compare those numbers with the diseases themselves. A landmark by CDC researchers estimated how many Americans got sick and died each year from vaccine-preventable diseases before vaccines were widely used. sets those historical figures against cases reported in 2025.
Behind each row are individual risks that are hundreds or thousands of times larger than the risks from the matching vaccines:
- Measles. Before the vaccine arrived in 1963, measles infected an estimated 3 to 4 million Americans a year, putting. Today, CDC says.
- Diphtheria. In the 1920s, diphtheria caused, and it still kills 5% to 10% of the people it infects. By comparison, moderate or severe reactions to the DTaP vaccine occur in fewer than 1 in 10,000 doses.
- Polio. More than. Paralytic polio kills 2% to 5% of children and 15% to 30% of adults who develop it. The injected polio vaccine used in the United States contains no live virus and cannot cause polio.
- Hib. Before the vaccine,. Hib meningitis killed 3% to 6% of children even with treatment, and 15% to 30% of survivors were left with hearing loss or other lasting brain damage.
- Rotavirus. Before vaccination, rotavirus sent. The vaccine's small intussusception risk is real, and it is far outweighed by the hospitalizations it prevents.
- Chickenpox. The pre-vaccine era saw about. Since the vaccination program began in 1995, chickenpox hospitalizations and deaths have fallen 93% and 94%.
- Hepatitis B. As many as, and CDC estimates that.
Measles Side by Side
Measles shows the comparison most clearly, because we have good numbers for both the disease and the vaccine. (Please note the chart's log scale.)
Even when the chart gives the vaccine every disadvantage, using the low end of disease estimates and the high end of vaccine estimates, the gap is large. For every million children who catch measles, we should expect at least 1,000 cases of encephalitis and at least 1,000 deaths. For every million MMR doses, we should expect a few hundred febrile seizures that leave no lasting harm, about 25 cases of temporary low platelets, and no deaths that these figures attribute to the vaccine. A brain-inflammation condition called ADEM follows about. Because children get two MMR doses, you can double the vaccine numbers, and the comparison still isn't close.
Nor is this just a historical exercise. As of October 1, the United States had recorded, the most since 1991, according to. CDC currently lists two confirmed measles-associated deaths this year and notes the count is subject to change. five measles-associated deaths nationwide and notes that Pennsylvania has seen more measles infections than any state in 35 years. Over the long run, CDC estimates that routine childhood vaccination of children born from 1994 through 2023 will prevent.
Why This Looks Like Another Anti-Vaccine Move
If the goal were simply to care for the rare person with a true vaccine reaction, we already have tools for that. CDC's, established in 2001, gives clinicians expert consultations on complex vaccine safety questions about individual patients. The, running since 1990, analyzes electronic health records every week and compares vaccinated people with comparison groups to catch rare problems. Those systems are how we learned about the risks listed above. Expanding and funding them would be a straightforward way to strengthen vaccine safety.
The new initiative takes a different path, and its details, along with what's missing from them, are telling.
It boosts the system that can't show cause and effect. VAERS is an early-warning system that anyone can report to. CDC says plainly that, and FDA has warned that some people. Paying doctors to file more reports, and making reporting easier through mobile tools and QR codes, will almost certainly raise the number of reports. Without denominators and comparison groups, that rise in reports can easily be presented as a rise in injuries, even if the true rate of injury hasn't changed at all.
The clinic's science is undefined. According to, HHS has not released a budget, a funding source, enrollment criteria, a research protocol, plans for comparison groups, a method for confirming diagnoses, or who will review the findings. It's also unclear whether an ethics board has approved a study protocol. The clinic was described as doing observational research, which can describe patients' symptoms but generally cannot, on its own, show that a vaccine caused them. The launch included a roundtable with parents who believe vaccines harmed their children. Listening to those families matters, but a clinic that starts from the assumption of injury and has no comparison group is set up to confirm a belief, not test one.
The people behind it have already told us what they want to find. Kennedy previously led Children's Health Defense, which describes as an anti-vaccine group, and he told a Children's Health Defense event last month that his role now lets him put federal resources behind the questions he once asked from outside government. that Kennedy asked NIH early this year to spend $5 billion studying a vaccine-autism link, and that his top adviser, who has no scientific or public-health background, repeatedly sought access to FDA's Sentinel safety database to compare autism rates in vaccinated and unvaccinated children. FDA staff initially refused over concerns about a "fishing expedition." Experts told Reuters such comparisons are invalid, because vaccinated children see doctors more often and so tend to receive more diagnoses of every kind.
It fits a clear pattern. Since 2025, Kennedy has, to say that the claim "vaccines do not cause autism" is "not an evidence-based claim," ended the universal hepatitis B birth-dose recommendation, and cut the routine childhood schedule from 17 diseases to 11. In March, a federal judge, writing that "a committee of non-experts cannot be said to embody 'fairly balanced... points of view' within the relevant scientific community." The administration has appealed only the part of the ruling about the advisers, and. Meanwhile, a Kennedy adviser has said HHS was working to, and Kennedy's former personal attorney has. Public health and legal scholars responded that the petition and warned it could scare parents and clog the program for families with legitimate claims.
Taken together, the pieces fit one another: more unverified reports, an injury-focused clinic without a published protocol, efforts to widen what counts as an injury, and advisers who have already reached their conclusions. A real vaccine-safety initiative would publish its protocol, include comparison groups, build on the systems that have already found real risks, and report what it finds even when the answer is "no link."
What Real Vaccine Safety Looks Like
Taking vaccine injuries seriously and supporting vaccination don't conflict. The same safety systems now being sidelined are the ones that found the rotavirus-intussusception link, the MMRV febrile-seizure signal, and myocarditis after mRNA vaccines, and that changed recommendations as a result. That is what "following the evidence" looks like.
The evidence is clear on the central question. Serious vaccine injuries are measured in cases per million doses. The diseases those vaccines prevent are measured in cases per thousand infections, and before vaccines they killed thousands of Americans every year. Any honest look at vaccine risk has to put both sides of that comparison in front of the public. An initiative that counts only one side, while measles spreads at levels not seen in 35 years, isn't making vaccines safer. It's making the diseases more likely to come back.