A number is circulating that a pregnant reader could be forgiven for finding terrifying: that 82 percent of vaccinated pregnant women miscarried. Michael Knowles built a segment around it, titled as a question, "Did The COVID Vaccine Cause An 82% Miscarriage Rate?" [2]. The video's own answer is the first thing worth reporting. Knowles calls the 82 percent figure "a red herring" and "a statistical misreading," and lays it at the feet of politicians pushing it rather than the data [2, 8:37-9:21]. This piece is not a correction of him. It is a correction of the figure he is flagging, written for the readers who have watched the 82 percent travel without his caveat attached.
The number traces to a 2021 misreading of a single federal dataset, the CDC's v-safe Pregnancy Registry, reported by Shimabukuro and colleagues in the New England Journal of Medicine that year. The registry followed pregnant people who had received an mRNA vaccine. The error divides the roughly 104 first- and second-trimester miscarriages in the early data by about 127, a denominator that wrongly drops the roughly 700 enrollees vaccinated in the third trimester whose pregnancies had not yet resolved. 104 over 127 is about 82 percent. The denominator that belongs there is 827, the completed pregnancies. 104 over 827 is about 12.6 percent, which sits inside the normal background rate of miscarriage. The NEJM table is paywalled and we did not fetch it this pass; the arithmetic of the misreading is widely documented, and we flag the primary as unverified here.
The concern underneath the number was not invented, and the strongest version of the worry deserves stating. Texts released this week from Anthony Fauci's government phone show him privately raising, in January 2021, a theoretical mechanism by which a strong immune reaction after a second dose "could be associated with miscarriage in the first trimester," a sequence we reported on August 11. A theoretical first-trimester concern, flagged privately by a regulator, is a real entry in the record. The 82 percent figure is not the measurement of it. It is a division problem with the wrong number on the bottom.
When the mechanism was actually tested, it did not appear. A case-control study run through the CDC's Vaccine Safety Datalink, a surveillance network spanning the agency and more than a dozen health systems, compared vaccination histories across roughly 300 miscarriages and 600 live births from January to October 2021. Yale Medicine's account of it is unambiguous: "There was no association between being vaccinated and having a miscarriage, the researchers found" [1]. The number of doses, the vaccine type, and the timing of the shot were likewise unlinked to miscarriage [1]. Our August 11 review documented the larger companion analysis, 105,446 pregnancies with an adjusted odds ratio of 1.02 and a 95 percent confidence interval of 0.96 to 1.08, an interval that runs straight through no effect.
The shape of this cycle is familiar and worth naming. A real document, the Fauci text, sits under a manufactured number, the 82 percent, and the number is what travels. What is unusual here is that the loudest amplifier, the video with the figure in its title, is also on record calling it a misreading. A pregnant reader who saw the 82 percent and not the correction is the person this is for. The tested rate is about 12.6 percent, the background rate, and the studies designed to find a vaccine signal found none.
