Claims online that a database shows 44% of pregnant women who took the Pfizer COVID-19 vaccine had a miscarriage are based on a miscalculation. The claims appear to stem from a blog post that now seems to have been deleted, where the author misinterpreted a dataset to reach the inaccurate figure.
The claim may have originated from a blog published on Aug. 12, 2022, with the headline: “Multiple Pregnancies in a Pfizer Trial Ended in Miscarriages. Pfizer Misleadingly Reclassified Them” (archive.ph/jfLUL).
A spokesperson for the blog said that “DailyClout made a correction on the original report as soon as the error was brought to our attention.” It now appears to have been deleted.
One user who shared the figure on Twitter said: “I think 44% of pregnant women in the Pfizer vaccine trial losing their babies should be a bigger story” (here). The tweet had gathered more than 32,000 likes and over 10,000 retweets at the time of writing.
Other examples of the claim shared by users online can be found (here), (here), (here) and (here).
SOURCE OF THE DATA
The author of the blog post points to a document with more than 3,600 pages that was uploaded by a website titled “Public Health and Medical Professionals for Transparent Documents.” The document purportedly shows Pfizer data released by the U.S. Food and Drug Administration (FDA) in response to a Freedom of Information Act (FOIA) request.
The dataset is linked in the archived blog post, but can also be found by searching for the dataset it references, which was uploaded on June 1, 2022, titled (125742_S1_M5_5351_c4591001-interim-mth6-adverse-events.pdf) via the website (phmpt.org/pfizers-documents/).
Reuters sought comment from the FDA about the authenticity of the dataset but had not received a response at the time of writing. Pfizer did not comment on what the data in the document shows.
WHAT THE DOCUMENT SHOWS
The document seems to show a log of adverse events and serious adverse events recorded as of March 2021. The blog says, “one must dig through the rest of the large document to learn that 22* of the 50* women suffered “Abortion Spontaneous,” “Abortion Spontaneous Complete,” “Abortion Spontaneous Incomplete,” or “Miscarriage.””
The blog points to a table on pages 3,643-3,644 titled: “Listing of Subjects Reporting Pregnancy After Dose 1” where 50 participants do appear. Each participant has a subject identification number.
The blog then lists 20 pages where it says 22 spontaneous abortions/miscarriages were recorded.
The blog author then calculates a miscarriage rate of 44% by dividing the number of mentions of miscarriage/spontaneous abortion (22) found in their search of the document by the number of pregnancies (50) listed in the table on pages 3,643-3,644.
FLAWED CALCULATIONS
The calculation is incorrect for three main reasons: 1) the numerator (22 spontaneous abortions) is not made up of unique cases but includes repeated participant case numbers; 2) the denominator (50 pregnancies) is not a complete list of pregnancies in this dataset; and 3) what the document and dataset show remain unclear.
First, experts said not all the entries the numerator comprises are unique.
According to Tony Panzarella, Program Director, MSc in Biostatistics at the Dalla Lana School of Public Health, University of Toronto, “The reported 44% (22/50) estimated percentage of subjects experiencing a miscarriage is incorrect.” He told Reuters the numerator included multiple duplicates.
For example, case number C4591001 1013 10131255 appears on both pages 219 and 3,519-3,520, and case number C4591001 1231 12313998 appears on both pages 1,806 and 3,551.
Second, among the subjects identified as having had miscarriages or spontaneous abortions, “the majority were not part of the 50 subjects that comprise the denominator,” Panzarella said.
For example, case identifier C4591001 1013 10131255, as mentioned above, is not included in the table on pages 3,643-3,644.
In other words, the numerator of 22 and denominator of 50 are both inaccurate, and therefore the calculated percentage of pregnancies lost is incorrect as well.
Finally, what the data in this document shows more broadly remains unclear and incomplete. This article will be updated if Pfizer or the FDA provide any information on what the data in the document shows.
A Pfizer spokesperson told Reuters the company does not directly comment on claims shared online, but that the vaccine has been shown to be “safe and effective for pregnant women and is recommended by global health organizations and regulatory agencies around the world, including the World Health Organization, the U.S. Centers for Disease Control and Prevention and the U.S. Food and Drug Administration.”
VERDICT
False. The 44% figure purporting to represent the percentage of pregnancies lost among participants in the Pfizer COVID-19 vaccine trial is inaccurate. The blog’s author failed to take into account duplicate listings in the over-3,600-page document when calculating the numerator (number of miscarriages), while also using a denominator taken from a table that did not include all the pregnancies listed in the document.
This article was produced by the Reuters Fact Check team. Read more about our fact-checking work here.
Our Standards: The Thomson Reuters Trust Principles.