A Food and Drug Administration (FDA) expert found indications of safety problems with COVID-19 vaccines several months after they were authorized, according to newly released emails.
Dr. Ana Szarfman, an FDA vaccine safety analytic expert at the time, met with Dr. Peter Marks, the FDA’s then-top vaccine official, and others at the agency on March 1, 2021, to discuss issues with the approach employed by the FDA to analyze the safety of vaccines following the emergency authorization of COVID-19 shots.
Szarfman had advocated for the FDA in the 2000s to adopt the approach, an algorithm called the multi-item gamma Poisson shrinker developed by William DuMouchel, the chief statistician at Oracle Health Sciences. Szarfman helped develop the FDA’s data mining system, which fully implemented the algorithm by January 2021.
In the March 2021 meeting, Szarfman told her superiors that limitations of the algorithm included masking, or difficulty detecting safety signals, and that DuMouchel had developed a new method called the regression-adjusted gamma Poisson shrinker that she had examined. The updated algorithm was “state of the art” because it “incorporates more information,” which “leads to a lower rate of missed signals and less false alerts,” she told them.


Szarfman, in subsequent emails, shared additional analyses that identified more safety signals, such as sudden death for the Moderna and Pfizer-BioNTech COVID-19 vaccines, as well as more details about her concerns that the method the FDA was using was insufficient in light of the updated approach. FDA leaders ultimately directed her to stop performing the analyses.
“Please hold off on creating and sending data mining reports and analyses using COVID-19 vaccine [adverse event] data,” one wrote.
FDA officials responded by writing in emails that they were worried about spreading the findings from the new algorithm.
Marks said in a Sept. 15, 2021, missive to one of Szarfman’s superiors, Dr. Patrizia Cavazzoni, that Szarfman “has decided on her own to do vaccine analysis” using data from the Vaccine Adverse Event Reporting System but without assistance from other FDA statisticians.
Peter Stein, another FDA official, replied to Marks, “We’ve made it clear to her that she should not be discussing or providing internal analyses externally, and needs to focus on her assigned work.”


Szarfman, who has retired, and DuMouchel did not respond to requests for comment.
An email to Menschik, who is still employed by the FDA, returned an automatic message saying he is on extended leave.
Contact information for Stein, who is not listed in the Health and Human Services Department employee directory, could not be located.
“I think this is a bombshell,” Johnson said. “I think this is a huge scandal.”
Johnson has a hearing on the records scheduled for the afternoon of April 29.
Karl Jablonowski, director of science and research at Children’s Health Defense, a group that scrutinizes vaccines and was cofounded by Kennedy before he took office, was among the experts slated to testify during the hearing.
Jablonowski told The Epoch Times that Johnson’s report on the emails “is really heartbreaking because it’s looking back at a time where had they done something different—and they were encouraged to do something different—the end result would have been so much better for the country.”