The Pattern
On August 11, on Fox Business, Representative Lisa McClain of Michigan named her source. “The New England Journal of Medicine said eighty-two percent of women who got the vaccine in their first or second trimester had a miscarriage.” The New England Journal of Medicine ran that study. Its own reported finding, from the same data, was 12.6 percent.
This review is about the other 69 points.
The subtraction
The study followed 827 pregnancies to completion. Of those, 712 ended in live birth, 104 in miscarriage, 1 in stillbirth, and 10 in ectopic pregnancy or induced abortion. 104 divided by 827 is 12.6 percent. That is the number the journal published.
82 percent comes from a different, smaller group: 127 people who had been vaccinated early enough, in the first or second trimester, to fall inside a separate follow-up window. Most of them had not given birth yet when the researchers checked back in. They were still pregnant. Dr. Jake Scott, quoted in Roll Call’s review, gave the comparison its plainest form: it is like calculating a school’s dropout rate in October by counting only the students who have already left. The ones still enrolled do not get a vote.
104 is a real number. 827 is a real number. 127 is a real number. 82 percent is what happens when the second one gets divided by the third instead of the first.
Four platforms, one week
McClain said it on Fox Business. Senator Ron Johnson of Wisconsin said a version of it on Fox News the next day: “The safety signals, the red flags, were screaming at these people, and you got Anthony Fauci there lying bold-faced to the American public.” Senator Rand Paul of Kentucky, who released the texts jointly with Johnson on August 10 from Fauci’s own government-issued phone, posted on X the next day: “Once again, Anthony Fauci was saying one thing in public and something completely different in private. He was willing to put women and their unborn children at risk instead of admitting that he was wrong about the vaccines that he tried guilting Americans into taking.” Senator Ted Cruz of Texas and NIH Director Jay Bhattacharya joined the same push. PolitiFact fact-checked Johnson’s claim two days later, on August 12, and found no such finding anywhere in the literature he cited.
Four messengers. One number. Zero of the four sourced it to a page anyone could turn to and check, because the page says something else.
What the texts say instead
The private text exchange is real, and it is one message, not two. In January 2021, Fauci wrote: “Since many people have significant cytokines storm and fever after the 2nd dose, this theoretically could be associated with miscarriage in the 1st trimester.” That is the line Paul and Johnson are quoting. It is also, in full, a scientist naming a theoretical mechanism, not confirming a finding. The same exchange records his conclusion: “No issues have arisen” among the vaccinated pregnant patients being tracked at that point, weighed against the documented danger of COVID-19 infection during pregnancy itself. He raised the concern and answered it, in the same conversation, four and a half years before anyone needed only the first half of it.
Paul’s post reads the first half as a confession and never quotes the second. Johnson’s line goes further: lying, bold-faced, to the public. Neither version survives contact with the full exchange. A scientist who names a theoretical risk and then reports that the risk did not materialize has not been caught. He has done the job.
What still holds
12.6 percent is not the only number in this file. A final CDC analysis of the same v-safe pregnancy registry, published in February 2026 once far more enrollees had completed follow-up, tracked 12,097 vaccinated people and found a 10.8 percent miscarriage rate and no elevated risk relative to the general population. Same registry, five years of additional follow-up, and a number within two points of the original. Nowhere near 82.
Verdict: The claim that COVID-19 vaccination carries an 82 percent miscarriage rate, and that this figure comes from the New England Journal of Medicine, is Distorted. The underlying numbers are real. The denominator is not the one the study used. The journal’s own published rate, from the same cohort, is 12.6 percent. The same registry’s final analysis, five years later and covering 12,097 people, found 10.8 percent. The private text cited as evidence of concealment says the opposite of what it is being used to prove.
Image note: the scratch-to-reveal illustrations above are AI-generated editorial art by Deceit, made for this piece. The figure is a composite; no real person is depicted. Not documentary evidence.
