FDA officials knew COVID vaccine safety system was flawed, suppressed whistleblower, investigations reveal

The architecture of public health oversight in the United States has come under intense scrutiny following the release of internal government documents that reveal a profound disconnect between the private technical assessments of federal regulators and their public-facing safety mandates. Two separate investigative reports published in November 2026, supported by disclosures from the U.S. Senate and Freedom of Information Law (FOIL) requests spearheaded by attorney Jimmy Wagner, have documented that officials at the Food and Drug Administration (FDA) and the Centers for Disease Control and Prevention (CDC) were aware of critical flaws in their vaccine safety surveillance infrastructure as early as 2021. Despite these internal warnings, the agencies continued to utilize compromised data-mining protocols, effectively blinding their safety monitoring systems to potential adverse events including myocarditis, pericarditis, Bell’s palsy, and tinnitus.
The revelations center on the Empirical Bayesian (EB) data-mining algorithm, a tool designed to flag statistically significant spikes in adverse event reports within the Vaccine Adverse Event Reporting System (VAERS). Documents spanning from October 2020 through July 2024 indicate that this algorithm was rendered mathematically ineffective due to the unprecedented volume of COVID-19 vaccine reports. Because the vast majority of VAERS entries during the initial rollout were related to COVID-19 vaccines, the “background” frequency used for comparison became inflated. When both Pfizer and Moderna vaccines triggered similar rates of adverse events, the algorithm’s “masking” effect prevented it from identifying these occurrences as statistical anomalies.
A Chronology of Suppressed Warnings
The failure of the surveillance system was not an oversight, according to internal correspondence, but a documented technical limitation that career officials attempted to rectify. Dr. Ana Szarfman, a medical officer with extensive expertise in pharmacovigilance, identified the “masking” phenomenon early in the rollout. By early 2021, Dr. Szarfman had developed, in conjunction with renowned statistician William DuMouchel, a modified algorithm capable of bypassing the statistical noise to isolate genuine safety signals.
Internal emails reveal that when Dr. Szarfman sought to integrate these fixes into the FDA’s oversight process, she was met with administrative resistance. Sources indicate she was instructed to “cease and desist” her efforts to modify the surveillance protocols. The decision to suppress these improvements ensured that the FDA’s public assertions—which frequently cited a lack of “safety signals” as evidence of the vaccines’ profile—remained unchallenged by their own automated systems.
By May 2021, the institutional dissonance became apparent. CDC Director Rochelle Walensky publicly stated that the agency had “not seen a signal” for myocarditis. However, internal communications from May 13, 2021, show CDC Deputy Director Tom Shimabukuro writing to colleagues, “I’m perplexed that myocarditis isn’t alerting for either of the mRNA vaccines.” This email was sent just one day before the agency publicly categorized myocarditis as an “extremely rare” side effect, suggesting a discrepancy between private confusion and public certainty.
The Medical Examiner’s Dilemma
The investigation extends beyond federal agencies into the offices of regional medical examiners, where forensic pathologists were forced to reconcile the reality of post-vaccination deaths with the prevailing public health narrative. Documents from the New York City Office of Chief Medical Examiner and similar entities in Connecticut reveal a culture of internal apprehension.
On June 26, 2021, Dr. James Gill, the Chief Medical Examiner for Connecticut, initiated a discussion among a national forensic pathology group regarding a 16-year-old boy who had been found dead in his bed four days after his second mRNA vaccination. Dr. Gill noted that the teenager’s heart exhibited “striking ischemic changes.” Weeks prior, Dr. Barbara Sampson of New York City had consulted with Harvard-affiliated pathologists regarding a case of focal lymphocytic myocarditis occurring two months post-vaccination, noting in her private records that she “doubted” an association while simultaneously questioning whether such pathology had been previously documented.
These exchanges underscore a broader pattern of professional conflict. Within the National Association of Medical Examiners, debates over how to categorize these deaths became increasingly polarized. When one member raised questions regarding cardiac incidents in athletes, Dr. Jay Stahl-Herz of the New York City Medical Examiner’s office dismissed the inquiry as “misinformation from a non-medical source.” This prompted a rebuke from Dr. William Oliver, who argued that it was premature for the medical community to adopt a dogmatic stance on vaccine safety, urging colleagues to examine the original reports rather than attacking the credibility of those raising concerns.
Statistical Failure and the VAERS Oversight Gap
The reliance on a broken algorithm created a "black hole" in public health monitoring. When the CDC eventually conducted Proper Reporting Ratio (PRR) analyses in 2022—separate from the flawed Bayesian system—the results were stark. The analyses identified hundreds of adverse events meeting signal criteria for myocarditis, pericarditis, Bell’s palsy, and tinnitus across 13 consecutive weekly reports. Despite this, the FDA’s primary surveillance mechanism triggered zero alerts.
Statistician William DuMouchel, who had offered his expertise to correct the system, later expressed profound regret over the agencies’ refusal to act. “If they had paid attention to Ana [Szarfman], they would have done better,” DuMouchel stated in an interview with The BMJ. The failure to pivot during the critical early months of the vaccine rollout represents a significant departure from standard pharmacovigilance practices, which typically err on the side of caution when data-mining tools indicate potential instability.
Implications for Institutional Trust
The implications of these findings are far-reaching. Critics argue that by prioritizing “vaccine confidence” over transparent safety surveillance, federal agencies inadvertently undermined the very trust they sought to maintain. The decision to suppress internal warnings and dismiss legitimate medical inquiries created a vacuum that allowed speculation to flourish, as the public was largely kept in the dark regarding the technical limitations of the safety monitoring apparatus.
Attorney Sujata Gibson, who conducted an extensive review of the medical examiner communications, suggests that the documents paint a picture of a system that prioritized the maintenance of a specific narrative over the investigative rigor required of public health institutions. “The city had the warnings,” Gibson noted. “Deaths within a day of vaccination. A teenager found dead days after his second dose. Evidence that didn’t fit the narrative was swept under the rug.”
The Path Forward
The convergence of these investigations, spanning the FDA, CDC, and the nation’s largest medical examiner offices, suggests a systemic failure that demands a comprehensive re-evaluation of how public health agencies conduct safety surveillance during national emergencies. The transition from the pandemic era to the current day has left behind a legacy of eroded institutional credibility.
Whether these revelations will lead to structural reform remains a subject of intense debate in Washington. Senate investigators continue to press for additional documentation, and legal challenges regarding the transparency of these agencies are ongoing. As the medical community grapples with the fallout, the central question remains: how to balance the urgent need for mass immunization campaigns with the fundamental responsibility to provide the public with an honest, unvarnished assessment of risk. The events of 2020 through 2024, as documented in these files, serve as a stark reminder of the dangers of institutional groupthink and the critical importance of maintaining independent, transparent oversight in the interest of public safety.







