The Pfizer Document That Went Viral — And What FDA-CBER-2021-5683 Actually Says
Released under a US court order and hosted on the FDA's own servers, the Pfizer 5.3.6 post-authorization adverse event report is a real public record. Here is what it actually contains — and why the claim it proves the vaccine causes hantavirus is false.
May 9, 2026 · By Justin Plosz · Ottawa, Canada · Community · 14 min read
What Is This Document And Where Did It Come From?
The document circulating on social media — variously photographed from phone screens, screenshot in Instagram Stories, and shared with captions asserting the COVID-19 vaccine causes hantavirus — is real. It is not fabricated. It is not a leak. It is hosted, in full, on the United States Food and Drug Administration's own FOIA reading room at downloads.regulations.gov under docket number FDA-CBER-2021-5683.
Its full title is: "5.3.6 Cumulative Analysis of Post-Authorization Adverse Event Reports of PF-07302048 (BNT162B2) Received Through 28-Feb-2021." PF-07302048 is the internal Pfizer compound number for BNT162b2 — the Pfizer-BioNTech mRNA COVID-19 vaccine commercially distributed as Comirnaty. The "5.3.6" designation refers to a section of the International Council for Harmonisation (ICH) E3 regulatory submission format, which is the standardised structure used for clinical trial and post-authorization safety submissions across all major regulatory jurisdictions, including Canada's Health Canada.
The document covers adverse events reported to Pfizer from the date of initial authorization through February 28, 2021 — approximately three months of post-authorization monitoring. It is dated April 30, 2021 and was submitted as part of Pfizer's Biologics License Application data package.
How it became public is its own story, and it is a legitimate one. In September 2021, a group called the Public Health and Medical Professionals for Transparency (PHMPT), represented by attorney Aaron Siri of Siri and Glimstad LLP, filed a Freedom of Information Act request for all data and information submitted to the FDA in support of the Pfizer-BioNTech COVID-19 vaccine's authorization. The FDA initially sought 55 years to complete the release. A federal district court judge, Mark Pittman of the Northern District of Texas, disagreed. He ordered accelerated rolling production of the documents — a decision upheld on appeal. The first tranche of documents, including the 5.3.6 report, was released in November 2021. The documents have been continuously available since then at downloads.regulations.gov and, in a compiled archive, at phmpt.org.
What The Document Actually Contains: Key Statistics
The 5.3.6 report is 38 pages, plus a nine-page appendix. It is an integrated summary of adverse events reported to Pfizer's worldwide pharmacovigilance system in the three months following emergency use authorization. Here is what the document actually reports.
In the period covered (December 1, 2020 through February 28, 2021), Pfizer received 42,086 case reports describing one or more adverse events following vaccination. Of these, 25,957 were medically confirmed, and 16,129 were non-medically confirmed (reports from patients or non-medical professionals rather than from clinicians). The document notes that 1,223 of the cases resulted in a fatal outcome. This is the figure most frequently cited by critics of the vaccine — and it is also the figure most frequently misrepresented.
What the document actually says about those 1,223 deaths is: they are spontaneous reports of deaths that occurred following vaccination and were reported to Pfizer's safety database during the monitoring window. The document does not conclude, and is not designed to determine, that these deaths were caused by the vaccine. Standard post-authorization pharmacovigilance collects all reports of adverse events following vaccination without pre-adjudicating causality — that is the point of the monitoring system. The document explicitly states that the benefit-risk profile of BNT162b2 remains favourable based on the data reviewed.
The most frequently reported adverse events in the confirmed case set were: headache, fatigue, pyrexia (fever), injection site pain, and pain. These are the expected reactogenicity profile for an mRNA vaccine — responses consistent with the immune activation the vaccine is designed to trigger.
The Nine-Page Appendix: What 'Adverse Events Of Special Interest' Actually Means
The page that generated the viral claims is Appendix 1 of the 5.3.6 report: a list titled "Adverse Events of Special Interest" — commonly abbreviated AESI. This is the list that contains, among hundreds of other conditions arranged alphabetically, "Hantavirus pulmonary infection."
Understanding what this list is requires understanding what an AESI is in the context of drug safety regulation. The concept is defined by the Council for International Organizations of Medical Sciences (CIOMS), the body that develops international pharmacovigilance standards, and is used by regulatory agencies worldwide including Health Canada, the European Medicines Agency, and the US FDA.
An Adverse Event of Special Interest is a condition identified in advance — during the design of a pharmacovigilance plan, before any data is collected — as warranting enhanced surveillance. The criteria for including a condition on an AESI list are not that the drug is known or suspected to cause it. The criteria are that the condition: (a) is biologically plausible as a potential immune-mediated or otherwise mechanism-relevant response; (b) has been associated with similar vaccines or drug classes in the past; or (c) is serious enough that any unexpected signal would be of public health significance.
In plain language: the AESI list is the set of conditions that pharmacovigilance teams are told to watch for and report immediately if they see a pattern. Its purpose is precisely to catch signals the manufacturer did not anticipate. Including a condition on the list is not a statement that the vaccine causes it. It is the opposite — it is a commitment to actively look for evidence that it might, so that any genuine signal can be detected and acted on early.
The 5.3.6 report's AESI appendix lists over 1,200 specific conditions across nine pages. These include every major organ system, every major infectious disease category, every serious immune-mediated condition known to medicine, and yes, several rare infectious diseases including Hantavirus pulmonary infection. The list was developed consistent with WHO and CIOMS guidance. Hantavirus appeared on it because serious respiratory infections were considered important to monitor given the respiratory context of the pandemic period — not because any evidence existed that BNT162b2 caused or could cause hantavirus infection.
The Viral Claim, And Why It Is Incorrect
The claim circulating on social media — that the Pfizer 5.3.6 document "states the COVID vaccine can give you hantavirus" — has been reviewed by Snopes, Lead Stories, and Science Feedback, all of which have rated it false or misleading. The specific claim appears to have re-emerged in early May 2026 following renewed circulation of the original screenshot.
The factual errors in the claim are two.
First, hantavirus pulmonary infection is not listed as a confirmed, identified, or suspected adverse effect of BNT162b2 anywhere in the 5.3.6 document or in any post-authorization safety communication from Pfizer, the FDA, Health Canada, or the European Medicines Agency. It appears in the pre-specified AESI monitoring list, where it is categorised as a condition to watch for, not as a condition the vaccine is known to cause.
Second, the 5.3.6 document is a passive surveillance summary — it reports what was submitted to Pfizer's safety database. When hantavirus pulmonary infection did not appear in the 42,086 cases reviewed, the monitoring continued. A condition appearing on a pharmacovigilance watchlist that does not subsequently generate a meaningful signal is a good outcome — it is how the system is supposed to work.
The document is genuine. The viral interpretation of it is not.
This distinction matters not because the questions surrounding COVID-19 vaccine safety are trivial — they are not — but because misrepresenting a real public document makes it harder for the public to evaluate genuine safety signals when they exist. The AESI framework caught the rare myocarditis signal in young male recipients within months of vaccine deployment — precisely because monitoring teams were watching for cardiac inflammation on the AESI list. That is the system working correctly. Misreading the AESI list as a confirmed side-effects roster undermines public confidence in the same monitoring infrastructure that identified that real signal.
The Court Order That Made The Document Public — And What That Actually Means
A recurring element of the social media coverage of the Pfizer documents is the claim that their release proves the FDA was trying to hide something. The reality is procedurally different, and worth understanding accurately.
The original FDA position — seeking 55 years to process the FOIA request — was not based on a claim that the documents were secret in the substantive sense. It was based on the administrative workload of processing approximately 450,000 pages of documents at the then-available rate of 500 pages per month. Judge Pittman's order accelerated the timeline to 55,000 pages per month, and the documents have been released on that schedule. Pfizer and the FDA did not assert that the documents were exempt from disclosure. The dispute was about speed, not about whether the public was entitled to see them.
This matters because the documents, read in full, do not contain evidence of suppressed harm data. They contain the routine internal post-authorization reporting that is standard practice for every approved biologic. The external controversy arose not from the content of the documents but from the decision to seek an administrative delay — a decision that, fairly or not, reinforced existing public skepticism about the transparency of the authorization process.
The proper response to that skepticism is not to misrepresent what the documents say, but to read them. They are available, in full, at downloads.regulations.gov (docket FDA-CBER-2021-5683) and in a searchable compiled archive at phmpt.org. Every Canadian who wants to evaluate the primary record can do so directly.
The Canadian Picture: What Health Canada Saw And How It Monitors
The Pfizer 5.3.6 document covers worldwide post-authorization data, including adverse events reported through the Canadian pharmacovigilance system. Canada's parallel monitoring infrastructure is the Canadian Adverse Events Following Immunization Surveillance System (CAEFISS), operated jointly by the Public Health Agency of Canada and provincial and territorial health authorities.
Health Canada conducted its own review of Pfizer's safety submission data as part of its Authorization with Conditions (AuC) process for BNT162b2, which resulted in authorization on December 9, 2020. Health Canada's post-authorization safety commitments mirror the international pharmacovigilance plan described in the 5.3.6 document, including the same AESI monitoring categories.
Canada's National Advisory Committee on Immunization (NACI) released periodic vaccine safety updates throughout the rollout, including guidance on the myocarditis signal in younger male recipients — information that came directly from the AESI monitoring system working as designed. The myocarditis guidance was updated in June 2021 for mRNA vaccines; the signal was detected, communicated, and factored into dosing guidance for that age group.
The CAEFISS database, Health Canada's Summary Basis of Decision documents, and NACI's advisory statements are all publicly available at canada.ca. They represent Canada's parallel record — one that, like the US FDA record, is accessible to any Canadian who wishes to evaluate it.
What To Make Of This — And What Questions Are Actually Worth Asking
PRC publishes this article because the Pfizer 5.3.6 document is a genuine public record that Canadians have a right to understand accurately, and because the current viral framing of it — on both sides of the debate — is frequently wrong in ways that obscure the real questions.
The real questions the FOIA release process raised are these: Was the FDA's initial request for 55 years to process a FOIA request a reasonable administrative position, or an institutional reflex toward delay that undermined public trust at a critical moment? Does the passive pharmacovigilance system that generated the 5.3.6 data — which relies on voluntary reporting by healthcare providers and individuals — adequately capture the full adverse event profile of a vaccine deployed at unprecedented speed and scale? Were the benefit-risk conclusions of the FDA, Health Canada, and other regulatory bodies derived from a sufficiently independent assessment of Pfizer's own submitted data?
These are serious questions. They have been raised seriously by credentialed researchers, including the authors of a 2022 peer-reviewed paper in Vaccine (Fraiman et al., doi:10.1016/j.vaccine.2022.08.036) that reanalysed the randomized controlled trial data and found the absolute risk of serious adverse events in the Pfizer trial group exceeded the absolute risk reduction in COVID-19 hospitalisations — a paper published, peer-reviewed, and available to read.
Those questions are worth engaging. The claim that a pharmacovigilance AESI monitoring list is a list of confirmed vaccine side effects — including hantavirus — is not one of them. The former is a legitimate scientific and policy debate. The latter is a misreading of a regulatory document. Public discourse is better served by being able to tell the difference.
The document is available at downloads.regulations.gov. Read it.
Key takeaways
- The Pfizer 5.3.6 document (FDA-CBER-2021-5683) is a real, publicly available FDA FOIA document hosted at downloads.regulations.gov. It is not a leak, and it was not hidden — it was released under a court order following a successful FOIA lawsuit.
- The document covers 42,086 adverse event case reports received by Pfizer in the first three months of BNT162b2 deployment (through February 28, 2021). The most frequently reported events were headache, fatigue, fever, injection-site pain, and pain.
- The '1,223 deaths' cited from the document are spontaneous reports of deaths occurring after vaccination that were submitted to Pfizer's pharmacovigilance database. The document does not conclude these deaths were caused by the vaccine.
- The nine-page 'Adverse Events of Special Interest' (AESI) appendix is a pre-specified pharmacovigilance monitoring list — conditions regulators are told to watch for and report. It is not a list of confirmed or suspected vaccine side effects.
- Hantavirus pulmonary infection's presence on the AESI monitoring list does not mean the vaccine causes hantavirus. No causal signal between BNT162b2 and hantavirus was identified in the 5.3.6 data or in any subsequent pharmacovigilance record. The claim is rated false by Snopes, Lead Stories, and Science Feedback.
- The AESI monitoring framework successfully identified the rare myocarditis signal in young male mRNA vaccine recipients in 2021 — an example of the system working as designed and resulting in updated dosing guidance from NACI and other regulators.
- Health Canada conducted an independent safety review of Pfizer's submission data and operates Canada's parallel monitoring system (CAEFISS). Health Canada's public safety communications are available at canada.ca.
- There are legitimate peer-reviewed scientific debates about the benefit-risk analysis of COVID-19 mRNA vaccines — including the Fraiman et al. 2022 paper in the journal Vaccine. These debates are worth engaging. Misreading a pharmacovigilance monitoring list is not part of that debate.
Frequently asked questions
- Is the Pfizer 5.3.6 document real or fabricated?
- The document is real. The Pfizer BNT162b2 '5.3.6 Cumulative Analysis of Post-Authorization Adverse Event Reports' (FDA-CBER-2021-5683) is a genuine FDA FOIA document. It is hosted in full at downloads.regulations.gov (docket FDA-CBER-2021-5683) and in a searchable compiled archive at phmpt.org. It was released as part of a rolling court-ordered FOIA production following a lawsuit by the Public Health and Medical Professionals for Transparency (PHMPT) represented by attorney Aaron Siri.
- Does the document say the COVID-19 vaccine causes hantavirus?
- No. Hantavirus pulmonary infection appears on the document's pre-specified 'Adverse Events of Special Interest' (AESI) monitoring list — a standard pharmacovigilance tool listing conditions that monitoring teams are told to watch for and report if they observe a pattern. The presence of a condition on an AESI list is not a statement that the vaccine causes it. Hantavirus is not listed anywhere in the 5.3.6 document as a confirmed, identified, or suspected adverse effect of BNT162b2. This claim has been rated false or misleading by Snopes, Lead Stories, and Science Feedback.
- What does 'Adverse Events of Special Interest' actually mean?
- An Adverse Event of Special Interest (AESI) is a condition identified in advance — before data collection begins — as warranting enhanced surveillance. The criteria for inclusion are not that the drug causes the condition, but that the condition is biologically plausible as a potential response, has been associated with similar vaccines historically, or is serious enough that any unexpected signal would be of public health significance. The AESI framework is used by all major regulatory agencies globally, including Health Canada, the FDA, and the European Medicines Agency. It is the same framework that successfully identified the rare myocarditis signal in young male mRNA vaccine recipients in 2021.
- How many adverse events were reported in the document?
- The 5.3.6 report covers 42,086 case reports received by Pfizer from initial authorization through February 28, 2021 — approximately the first three months of deployment. Of these, 25,957 were medically confirmed. 1,223 cases resulted in a fatal outcome as reported. These are spontaneous adverse event reports, not confirmed causal attributions — standard post-authorization pharmacovigilance collects all reported events without pre-adjudicating causality. The document concluded that the benefit-risk profile of BNT162b2 remained favourable.
- Why did the FDA want 55 years to release the documents?
- The FDA's initial request for 55 years to complete the FOIA production was based on an administrative workload calculation — approximately 450,000 pages of submitted documents at the then-available processing rate of 500 pages per month. The FDA did not argue the documents were exempt from disclosure. Federal Judge Mark Pittman of the Northern District of Texas ordered accelerated production at 55,000 pages per month; that order was upheld on appeal. The documents have been released on that schedule and are publicly available.
- Who obtained the documents and how?
- The documents were obtained through a Freedom of Information Act (FOIA) lawsuit filed by the Public Health and Medical Professionals for Transparency (PHMPT), represented by attorney Aaron Siri of Siri and Glimstad LLP, in September 2021. Federal Judge Mark Pittman ordered the FDA to produce the documents on an accelerated schedule beginning November 2021. The full archive is available at downloads.regulations.gov (docket FDA-CBER-2021-5683) and at phmpt.org.
- What does Canada's Health Canada say about the Pfizer vaccine safety data?
- Health Canada conducted an independent review of Pfizer's safety submission data as part of its Authorization with Conditions (AuC) process for BNT162b2, which was authorized on December 9, 2020. Canada's post-authorization monitoring is conducted through the Canadian Adverse Events Following Immunization Surveillance System (CAEFISS), operated jointly by the Public Health Agency of Canada and provincial health authorities. Health Canada's Summary Basis of Decision documents and NACI advisory statements are publicly available at canada.ca.
- Are there legitimate scientific questions about COVID vaccine safety that are worth discussing?
- Yes. A 2022 peer-reviewed analysis published in the journal Vaccine (Fraiman et al., doi:10.1016/j.vaccine.2022.08.036) reanalysed the Pfizer and Moderna randomized controlled trial data and found that the absolute risk of serious adverse events in the vaccine arms exceeded the absolute risk reduction in COVID-19 hospitalisations in those trials. This paper is peer-reviewed, published, and available to read. It represents the kind of legitimate scientific re-evaluation that the research community is supposed to conduct. It is a different category of argument than the claim that a pharmacovigilance monitoring list constitutes a confirmed side-effects roster.
← Back to PRC Newsroom · Public Relations Canada
Enable JavaScript to view the interactive version of this page.