Pregnant women deserve better than “trust us” science
A major study has been used to reassure pregnant women that Covid-19 vaccines are safe. But the data behind the claim are fatally flawed.
In medicine, few assurances carry more emotional weight—or greater responsibility—than the claim that something is “safe during pregnancy.”
Pregnant women are justifiably cautious about what they expose themselves to during this vulnerable time, and history has given them every reason to be.
The thalidomide disaster, diethylstilboestrol (DES), and other cautionary tales have shown what can happen when scientific rigour is sidelined in favour of commercial interests.
So, when a new study published in Pediatrics - the official journal of the American Academy of Pediatrics - claimed that Covid-19 vaccination in early pregnancy was safe, it came with an air of authority and reassurance.
News headlines followed suit, and public health recommendations continued to promote the vaccine’s safety in pregnancy.
But scratch the surface of this study, and something starts to unravel.
Not only are the data unverifiable and privately sourced, but the study contains a fatal flaw that renders its conclusions virtually meaningless.
The fatal flaw
The study analyzed 78,052 pregnancies that ended in a live birth—but left out 20,341 pregnancies that ended in miscarriage or other non-live outcomes.
That’s not a minor oversight.
The very purpose of studying vaccine safety in pregnancy is to assess whether exposure in utero leads to adverse outcomes—like miscarriage, birth defects, or foetal death. Yet one-fifth of the pregnancies were excluded from the analysis, removing exactly the kind of outcomes the study was supposed to detect.
This introduces what’s known as live-birth bias—a selection bias that arises when research includes only live births, disregarding the possibility that harmful effects may have caused some pregnancies to end prematurely.
Put plainly, if you only study babies who made it to birth, you’re ignoring the ones who didn’t—and any harm that may have played a role.
Even the study’s authors acknowledge this limitation, conceding that the exclusion “could lead to an underestimation of identified outcomes.” Still, they move forward to conclude there’s no association between the vaccine and birth defects.
Omitting over 20,000 pregnancies isn’t just a technicality - it’s a fatal flaw.
If even a small fraction of those pregnancies ended in miscarriage or birth defects linked to vaccination, the entire outcome could tip the other way.
Commercial data with no accountability
Then there’s the source of the data itself—a point entirely overlooked.
Rather than using clinical records from hospitals or national birth registries, the study relied entirely on a commercial database from Merative® MarketScan® Research Databases.
These databases are vast, aggregating de-identified insurance claims, prescriptions, lab results, and hospital records from more than 263 million Americans. But they are also privately owned, and their inner workings are entirely opaque.
Researchers using MarketScan data cannot verify whether the patients are real or theoretical, whether records have been altered, or how the data has been cleaned or processed before delivery.
In essence, they are working with a black box, one that comes with no guarantee of integrity.
Experts have already noted that the data from this unverified source shows signs of being unreliable.
The authors ran 93 separate statistical tests to look for differences in outcomes like birth defects. By chance alone, you’d expect a handful to be statistically significant. But none were.
The probability of that happening randomly is just 0.8%—a sign that the dataset may have been fabricated, or that its integrity is in question.
When two of the study authors - Dr Stacey Rowe and Dr Annette Regan - were asked if they had verified the authenticity of the MarketScan database—that is, if they could confirm these were ‘real’ patient data—they did not respond.
This isn't a hypothetical problem.
The medical literature has already been rocked by the Surgisphere scandal, where fraudulent hospital datasets were used to produce papers in The Lancet and The New England Journal of Medicine.
Those papers were eventually retracted, but only after independent researchers demanded to see the raw data and were denied - the data were likely fabricated.
Reassurance without evidence
Despite these glaring problems, the study’s conclusions are being used to reassure pregnant women.
In Australia, for example, the government’s official guidance recommends Covid-19 vaccination in pregnancy, stating that the “recommendations for pregnant women are the same as the general population.”
This, despite the fact that pregnant women were excluded from the pivotal clinical trials and no randomised studies have ever been completed to assess the vaccine’s safety in early-pregnancy.
The result is a landscape where pregnant women are asked to make a “shared decision” with their doctors—based on scientific literature that’s increasingly built on unverifiable data, flawed assumptions, and little to no independent scrutiny.
We are drifting into a new era where conclusions are based on data that sit behind corporate firewalls. An era where trust is expected, but no longer earned.
The Pediatrics study is a case in point.
It carries the imprimatur of authority, published in the flagship journal of the American Academy of Pediatrics. But, in reality, the analysis was based on commercial datasets that cannot be independently verified, and a methodology that systematically excludes the very outcomes it was supposed to assess.
This isn’t just bad science—it’s misleading by design.
And when it comes to pregnancy, where the stakes are literally life and death, that kind of scientific chicanery is a betrayal.
Pregnant women deserve better than a “trust us” approach to medicine.
They need full access to the data, honest communication about uncertainties, and above all, respect for their right to make informed decisions based on real evidence, not selective reporting.
Until that happens, we should remain sceptical of any study that asks us to believe in the evidence without seeing it.
Author’s conflicts of interests:
"Trust us bro" is Jeffrey Morris's dogma that means nobody can verify any of this data. It is most prevalent in the pregnancy data relating to COVID - I wrote about it in a 2 part series after Viki Male infamously dumped a bunch of these junk studies as evidence of the safety of the COVID vaccines in pregnancy.
Here if anyone wants to look into this further specific to the pregnancy papers (apologies for the crossref:
https://www.arkmedic.info/p/debunking-viki-male-part-1
https://www.arkmedic.info/p/debunking-viki-male-part-2-the-calvert
As an aside "FeesGarden" also covers this synthetic data scandal in her latest cryptic blog which is Australia focused - pertinent given the involvement of Monash and Perth universities in the background of the paper you have discussed.
https://feesgarden.substack.com/p/argus-safety-web-the-australian-story-448
That's a pretty trick experimental design - exclude all the people who might lead to you NOT getting the result you want, do the study, then declare your result is confirmed. Oh wait, that's exactly what Pfizer did, where if my memory is correct they excluded pregnant females from their study, which got the jabs approved.