What Pfizer’s safety document actually records

The report is titled 5.3.6 Cumulative Analysis of Post-authorization Adverse Event Reports. It was approved internally on April 30, 2021 and covers reports received through February 28, 2021. The pages carry Pfizer’s confidentiality marking and FDA production numbers.

The document later became public as part of FDA records released after litigation brought by Public Health and Medical Professionals for Transparency. “Confidential” describes the document’s original status; it does not mean the appendix itself was a hidden list of confirmed side effects.

For the period from December 1, 2020 through February 28, 2021, Pfizer estimated that approximately 126.2 million doses had shipped worldwide. Its database contained 42,086 case reports with 158,893 coded events.

A report is not the same as a verified vaccine injury. The database included spontaneous reports, reports from health authorities and literature, marketing programs, non-interventional studies, and serious events from clinical studies regardless of causality assessment.

Appendix 1 runs from report pages 30–38. It alphabetically lists MedDRA terms that Pfizer treated as adverse events of special interest. The report says the list drew from Brighton Collaboration/SPEAC, the ACCESS protocol, a preliminary CDC list for VAERS surveillance, and an MHRA guideline.

AESI means a medical event selected for closer observation and rapid evaluation. A term can be included because it is associated with COVID-19, because it has mattered for vaccines generally, or because regulators want surveillance to detect a possible signal. Inclusion does not show that the event occurred after this vaccine, how often it occurred, or whether the vaccine caused it.

The circulating count is usually 1,291. Text extraction and counting can vary slightly because the appendix contains compound MedDRA terms, line breaks and repeated terminology. The searchable transcription below preserves every row extracted from a published line-by-line reproduction of the appendix rather than silently deleting apparent duplicates.

The report itself warns that spontaneous reporting is affected by publicity, awareness, seriousness and underreporting; it yields reporting proportions rather than incidence rates. Clinical information may be missing, and an accumulation of reports does not necessarily mean the product caused the event.

That does not make the reports meaningless. They are important for signal detection: identifying patterns that warrant controlled studies, medical-record review, comparison with background rates, regulatory action or label changes. The appropriate question is not “was it listed?” but “what did follow-up evidence establish about frequency, causation and risk by population?”