Pfizer-BioNTech delayed reporting vaccine-associated deaths among BNT162b2 clinical trial participants until after the U.S. Food and Drug Administration (FDA) issued an Emergency Use Authorization (EUA) for the product.
The vaccine makers also failed to account for a large number of subjects who dropped out of the trial.
Together, these strategies kept regulators and the public ignorant of a 3.7-fold increase in cardiac deaths among subjects who received the vaccine, according to analysis in the International Journal of Vaccine Theory, Practice, and Research.
The authors of the paper described it as a “forensic analysis,” defined by the U.S. National Institute for Standards and Technology as “the use of scientific methods or expertise to investigate crimes or examine evidence that might be presented in a court of law.”
What the analysis shows
Corinne Michels, Ph.D., retired distinguished professor of biology at Queens College, New York, led the DailyClout Pfizer/BioNTech Documents Investigations Team on what the authors claim was the first independent examination of original data from the Pfizer-BioNTech COVID-19 mRNA vaccine (BNT162b2) clinical trial.
Investigators looked at each of the 38 deaths occurring between July 27, 2020, the start of phase 2/3 of the Pfizer-BioNTech vaccine trial, and March 13, 2020, the end date culminating in Pfizer-BioNTech’s 6-month interim report.
This trial phase involved 44,060 subjects. Half received a dose of BNT162b2, half got a placebo consisting of an inactive sterile salt solution.
The trial was unusual because at week 20 after the FDA issued the EUA for the vaccine, trial subjects in the placebo group were allowed to switch to the vaccinated group and receive their first BNT162b2 shot.
Switching from the placebo to the vaccinated group — or “unblinding” — normally occurs when the benefit of the drug is so great that not treating subjects becomes unethical. For example, investigators might consider unblinding a cancer trial if at some point all untreated patients deteriorated or died but all treated patients improved.
Unblinding conditions may be specified in the study design, but they usually involve input or review from medical ethicists.
Of 20,794 unblinded placebo subjects in the Pfizer trial, 19,685 received at least one dose of BNT162b2.
Normally the decision to unblind a vaccine trial would be based on the product’s safety and effectiveness in reaching certain endpoints or objectives. Endpoints for a drug to prevent viral infections might be a positive test or self-reported COVID-19 illness (the “case” numbers that drove much of COVID-19 policy), illness requiring hospitalization or death.
But, perhaps unexpectedly, after 33 weeks the data revealed no significant difference between deaths in the vaccinated and placebo groups for the initial 20-week placebo-controlled portion of the trial.