Recently, a fact-checker responded to our article on the major COVID-19 vaccine study published in Pediatrics, arguing that the article was misleading, anti-scientific, and contrary to the established consensus on vaccine safety in pregnancy.
That would be deeply concerning—if it were true.
Let’s be clear: our article never claimed that COVID-19 vaccines cause birth defects, miscarriage, or infertility. What we did say, and still maintain, is this:
A highly publicized study used to reassure pregnant women about vaccine safety excluded over 20,000 pregnancies that ended in miscarriage or non-live outcomes—the very outcomes it was supposed to measure.
And that’s not just worth questioning. It’s essential.
The Core of Our Critique
The Pediatrics study in question analyzed 78,052 pregnancies that ended in live birth, and excluded 20,341 pregnancies that didn’t.
That’s called live-birth bias, a form of selection bias well known to epidemiologists. It artificially narrows the scope of the data, removing the very cases that could reveal harm. Even the study’s authors admitted that this exclusion “could lead to an underestimation of identified outcomes.”
But they went on to conclude the vaccine was safe.
That’s the flaw. That’s what we called out. And that’s what the fact-checker ignored.
Fact-Check or Framing Exercise?
Rather than address that central issue, the fact-check attempted to reframe our position entirely:
- Claiming we said vaccines cause birth defects. We didn’t.
- Claiming we denied the scientific consensus. We didn’t.
- Claiming we oppose vaccine recommendations. We don’t.
We questioned the integrity of a single study that has been used to justify those recommendations—because the study design was incomplete and the data unverifiable.
That’s not misinformation. That’s scientific scrutiny.
About That Data…
Another major concern in our original article was the use of proprietary, commercial data from the Merative® MarketScan® database.
This data is:
- Not publicly accessible
- Not peer-verifiable
- Not confirmed to be authentic patient records
When two of the study’s authors were asked if they had verified the authenticity of the data, they did not respond.
Imagine making national health policy recommendations using a black box of data you can’t inspect, replicate, or validate. Would we accept that in any other field of science?
Science Is a Conversation
To those who disagree with our conclusions, we say: good. That’s how science moves forward. But disagreement is only meaningful if it’s honest.
If your rebuttal ignores the actual critique and substitutes claims we never made, you’re not fact-checking. You’re reframing.
Here’s the difference:
- Rejecting science is ignoring data.
- Practicing science is interrogating it.
We are not asking the public to reject vaccines. We’re asking researchers and institutions to earn trust by doing better science, with transparency, and with integrity.
And when studies used to shape policy come with glaring holes, someone needs to say so.
That’s what we did.
And we’ll keep doing it.
“It’s like claiming air travel is 100% safe—after excluding every crash, in-flight death, and emergency landing.”
— A reader’s analogy that says it all.
For more on this topic, see our original article: How Can This Be Called Science? The Fatal Flaw in a Major COVID Vaccine Study for Pregnant Women.
Image Acknowledgement
We’re grateful to the talented photographers and designers on Unsplash for providing beautiful, free-to-use images. The image on this page is by Devon Divine. Check out their work here: https://unsplash.com/@lightrisephoto. The image was edited using canva.com
