A new analysis of data from Florida found a higher risk of mortality in those receiving the Pfizer versus the Moderna COVID-19 vaccine in the year after a first dose. It does not establish that the Pfizer vaccine was dangerous or caused deaths, contrary to suggestions by the state’s surgeon general or statements by online purveyors of vaccine misinformation.
The study did not compare outcomes to unvaccinated people, so it cannot say whether vaccination with either shot was associated with higher or lower mortality overall. The paper acknowledged that one possible reason for Pfizer’s higher relative mortality risk is that the Moderna vaccine was more effective against COVID-19 than the Pfizer shot.
The study was published on Nov. 4 in the journal Global Epidemiology and was mostly conducted by members of the Florida Department of Health using state data. It found that compared with a Moderna dose, adult Pfizer recipients had a 40% higher risk of dying from any cause within one year of a first dose. Relative to Moderna, the Pfizer vaccine was also associated with a 91% higher risk of dying from COVID-19, a 57% higher risk of dying from cardiovascular disease and a 37% higher risk of dying from something other than COVID-19, as recorded on a death certificate.
Previous research has not found that either vaccine increases mortality. Both vaccines have saved many lives by preventing COVID-19 deaths.
Experts told us the study itself, which was upfront about its limitations, is not especially problematic. But some interpretations of it — including by some of the paper’s own authors — are.
“As I said in January 2024, no one should be taking mRNA COVID-19 vaccines because of their terrible safety profile,” Dr. Joseph Ladapo, a co-author and Florida’s surgeon general, wrote on X the day after publication. “But for those who had to take one, Florida’s public health data strongly implies that thousands of Floridians died as a result of receiving Pfizer instead of Moderna. The number of lives lost in the U.S. and internationally is staggering, based on this analysis.”
Jeffrey S. Morris, the director of biostatistics at the University of Pennsylvania who wrote a detailed critique on X of what the paper can and cannot say, told us that Ladapo’s comment goes beyond the evidence.
“That is like orders of magnitude beyond what his own paper showed,” he said.
Ladapo, who has a history of promoting flawed studies that argue the mRNA vaccines are dangerous, including one from the Florida Department of Health that claimed to identify a higher risk of cardiac death after vaccination in young men, cited unsubstantiated concerns when calling for a halt in the use of all mRNA COVID-19 vaccines in January 2024. That fall, the Florida Department of Health began advising all people — even older people at high risk of severe COVID-19 — to not get vaccinated with an mRNA vaccine. At the time, and still today, there is only one non-mRNA COVID-19 vaccine available in the U.S., the protein-based Novavax vaccine. No other state recommends against mRNA COVID-19 vaccination.
Speaking to Children’s Health Defense, the nonprofit opposed to vaccines that Health and Human Services Secretary Robert F. Kennedy Jr. previously led, Ladapo framed the results as indicative of a safety concern, saying the new study helps show that “our system for evaluating vaccine safety is broken.”
“The findings of this study should not be interpreted as establishing the overall safety of the Moderna vaccine,” he said, citing prior research that he claimed also raised safety concerns about the Moderna vaccine (we’ve written about that study, too).
“We reiterate our recommendation that, based on what is currently known about the safety of mRNA COVID-19 vaccines, no one should receive them,” he added.
The Florida Department of Health did not respond to our inquiries asking about his interpretation of the study and whether the department shared his views.
Online, others made even more explicit claims about the new paper.
“This is an extremely important paper. It shows that the Pfizer vaccine was not safe. It killed people. There is no other way to interpret this data,” Steve Kirsch, a former tech entrepreneur who has long spread inaccurate information about the vaccines, said on X on Nov. 5.
“An official government study just confirmed that COVID shots resulted in a CATASTROPHIC loss of life,” Nicolas Hulscher, another prominent spreader of vaccine misinformation, wrote on X, sharing a video in which he further claimed that extrapolating the Florida findings to the U.S. would mean “you get a minimum of 330,000 American deaths from Pfizer alone.”
These claims are false.
Natalie Dean, an associate professor of biostatistics and bioinformatics at Emory University, told us the study isn’t evidence that either of the vaccines is killing people. “I don’t see how this is an indictment of the safety of these vaccines since it’s just a comparison of two different vaccines,” she said.
“The way the paper is getting talked about does not align with the actual evidence in the paper,” Morris, who is also a distinguished research fellow with our parent organization, the Annenberg Public Policy Center, said.
“The study legitimately demonstrates an association between vaccine brand and subsequent mortality in a selected matched Florida cohort,” he said in an X post. “It does not establish that these differences are causal, much less that they reflect vaccine-related harm.”
When asked for comment, a Pfizer spokesperson told us that medical experts and regulatory authorities around the globe have reviewed and approved or recommended the vaccine, and that it continues to be monitored for safety. “With billions of doses of the vaccine administered around the world, the safety profile for the vaccine for all authorized groups continues to be favorable,” the spokesperson said.
Study Limitations
In the Global Epidemiology paper, the authors attempted to identify similar individuals in the state who received a Pfizer versus a Moderna vaccine in the first eight months of the vaccine rollout. They then looked at death outcomes within a year of the first vaccine dose to estimate relative mortality risk.
To try to control for factors other than vaccine brand, the authors matched about 1 million individuals who received a Pfizer vaccine with 1 million who received a Moderna dose on seven different variables. These were age group; sex; race; ethnicity; calendar month of vaccination; type of vaccination site, such as private pharmacy or hospital; and census tract.
The authors also did other checks to try to show that this matching was largely successful, finding no statistical difference, for example, when the same pairings were assessed for cancer deaths or suicide or homicide.
The problem is that since this is an observational study, there is still concern that the mortality differences observed between the vaccines were not due to the vaccines themselves, but to other factors.
A key issue Morris and Dean raised is that the vaccine recipients were not matched for medical frailty or underlying health status. If the Pfizer recipients were in poorer overall health or had more risk factors for COVID-19 or other diseases, that could explain a large part of why the Moderna recipients appear to have lower mortality, rather than suggesting anything about the vaccine itself.
Then, even if the mortality difference is due to the vaccines, Morris explained in an article on X, it does not necessarily mean that the Pfizer vaccine is causing harm. It may simply be that the Moderna vaccine is better at preventing harm, making the Pfizer vaccine look bad in comparison.
“You can see the better effectiveness of the Moderna vaccine in there,” Dean said of the study’s findings.
Several studies have found that the Moderna vaccine was slightly more effective at preventing severe COVID-19 outcomes than Pfizer’s. Although quite similar, the original Moderna vaccine contained significantly more mRNA per dose than the Pfizer shot and the second dose was given an extra week apart.
Higher protection against COVID-19 could also spill over into the other mortality categories as well, particularly during Florida’s delta variant wave in 2021, Morris said. “Any deaths that were in fact due to COVID, but they didn’t formally get tested would not have been documented as COVID, so they would be in the non-COVID deaths,” he told us.
Studies have found that severe COVID-19 infections raise the risk of cardiovascular events and death well after the infection.
In the paper, the authors were careful to note many of these caveats, and to say that the findings could be explained by a more protective Moderna vaccine, other factors the researchers were not able to control for, or what they term “non-specific effects” of the vaccines (or some combination thereof).
It is nevertheless clear that in public, some of the authors preferred to promote the latter idea, which could include the notion that the Pfizer vaccine has some kind of harmful effect.
In a post on X, Retsef Levi, the corresponding author, wrote, “This study adds to a growing evidence that the mRNA COVID vaccines of Moderna and Pfizer may have different non-specific effects.”
Levi is a professor at Massachusetts Institute of Technology’s business school who has long said that mRNA COVID-19 vaccines should not be used. He was one of Kennedy’s appointees to the Centers for Disease Control and Prevention’s vaccine advisory committee in June 2025 before a judge paused the appointments and temporarily blocked the group’s decisions in March.
In his X post, Levi also pointed to a recent nanoscience study that he said “showed significant differences in systemic persistence of nano-lipids and spike mRNA between Moderna and Pfizer mRNA COVID vaccines!”
The senior author of the nanoscience paper, immunologist Stephen Kent of the University of Melbourne, told us in an email that it is “inappropriate to refer to our paper supporting death from Pfizer” and that “it would not be scientifically appropriate to use our findings to draw conclusions about vaccine safety or mortality risks.” He, too, said the observed outcome in the Florida analysis could be related to Moderna protecting against COVID-19 better than Pfizer.
Putting the Study in Fuller Context
Another argument against jumping to the conclusion that the Pfizer vaccine is actively harming people through an unspecified mechanism is the larger body of research conducted on the COVID-19 vaccines.
In particular, there is a Veterans Affairs study published in December 2021 in the New England Journal of Medicine that also compared the Pfizer and Moderna vaccines to one another using a matching approach similar to the one in the Florida analysis. (This study also did not match for health status or comorbidities, but unlike the Florida analysis, it did provide a table showing that the matched pairs did not substantially differ on those parameters.)
The study was smaller and over a shorter time period, but it found that the Moderna vaccine recipients had a lower risk of a variety of adverse COVID-19 outcomes. Although not statistically significant, Pfizer recipients were 11% more likely to die from COVID-19 than Moderna recipients. The study did not, however, find any other mortality differences between the vaccines. As a check on their main findings, the authors looked at non-COVID-19 deaths, finding virtually identical patterns of death in matched recipients of both vaccines (see figure S3).
“They found absolutely no difference in non-COVID deaths, like not even a trend,” Morris said.
A few other studies have identified a slightly higher all-cause mortality after vaccination with various Pfizer shots compared with Moderna, but nothing approaching 40% — and some results have favored the Pfizer vaccine instead.
If the Florida study results “were really true and really driven by vaccine harm, we would expect them to be replicated elsewhere, but it doesn’t seem to be,” Morris said.
Notably, the studies that have looked at all-cause mortality among vaccinated and unvaccinated people have consistently found much lower rates of death in vaccinated people.
A 2025 French study, for example, compared mortality over four years, finding about a 25% lower risk of death in vaccinated versus unvaccinated individuals (with a lower risk for Pfizer than Moderna).
Scientists don’t think the vaccines actually lowered death risk by that much because of something called the healthy vaccinee effect. People who choose to get vaccinated are generally healthier than those who don’t — especially closer to the vaccination date, since people near the end of their lives aren’t getting vaccinated. While studies try to adjust for this issue, and longer-running ones are less affected by it, it remains a concern when interpreting the results. Still, these findings do not support the notion that the vaccines are increasing the risk of death. The authors of the French study concluded that “a causal link between mRNA vaccination and excess long-term mortality appears highly unlikely.”
For both vaccines to have some kind of toxic effect, as some claim — and for the Pfizer vaccine to have a nearly 40% higher risk of death than Moderna, but not be detected in these sorts of studies — the healthy vaccinee effect would have to be “really enormous,” Morris said.
Dean noted that even if the Moderna vaccine is better at protecting against COVID-19, it was still important to have both vaccines as options. When the shots were first rolled out, not only was it not yet known how the two vaccines compare, but there weren’t enough shots of any single brand for all Americans.
Some data suggest the Moderna vaccine caused more intense expected side effects, such as a fever, than the Pfizer shot. “For some people, Pfizer has been a better option” because of the milder side effects, she said. “I just don’t understand indicting things backwards.”
It’s worth noting that the original versions of the Florida analysis, which were posted to preprint servers, made more sweeping statements more akin to how some authors have presented the paper online, with one preprint concluding that the results are “suggestive of differential non-specific effects” of the two vaccines “and potential concerning adverse effects on all-cause and cardiovascular mortality.” The published version lacks these statements.
Individuals can now point to a published document “that is going to be hard for people to understand because it’s very complex and there’s a lot of tables,” Dean said, and it can be used to “make more extreme claims.”
Editor’s note: FactCheck.org does not accept advertising. We rely on grants and individual donations from people like you. Please consider a donation. Credit card donations may be made through our “Donate” page. If you prefer to give by check, send to: FactCheck.org, Annenberg Public Policy Center, P.O. Box 58100, Philadelphia, PA 19102.
