JAMA provides an expert example. They nearly got away with it. But then we read it.<\/h3>\n\n\n\n
<\/a><\/p>\n\n\n\n By James Lyons-Weiler, PhD<\/p>\n\n\n\n Let me walk you through a masterclass in how to bias a study so thoroughly that even a real risk can disappear. Not with fraud, not with data tampering\u2014just good old-fashioned design flaws, all pointing in one direction. (I will leave it to the reader to decide on this). We\u2019re going to talk about a study recently published in JAMA Network Open<\/em> by Bernard et al. (2025), which claims that getting an mRNA COVID-19 vaccine during the first trimester of pregnancy isn\u2019t associated with birth defects.<\/p>\n\n\n\n You might have heard this result and thought, \u201cWell, that\u2019s reassuring.\u201d<\/p>\n\n\n\n Let me show you why it\u2019s not.<\/p>\n\n\n\n The researchers looked at over 527,000 live-born infants<\/strong> in France, comparing those whose mothers got an mRNA COVID-19 vaccine in the first trimester with those who didn\u2019t. They then looked at the rate of major congenital malformations (MCMs)<\/strong> and said: \u201cNo difference.\u201d<\/p>\n\n\n\n Their conclusion? The vaccine isn\u2019t teratogenic. It doesn\u2019t cause birth defects.<\/p>\n\n\n\n Here\u2019s the problem: They only looked at live births<\/strong>. That means any fetus with a severe enough defect to lead to termination or fetal death<\/strong> is simply not counted<\/strong>. And guess what? The most severe defects\u2014the ones most likely to show up if a vaccine did have a harmful effect\u2014are the ones most likely to lead to termination<\/strong> or stillbirth<\/strong>.<\/p>\n\n\n\n Let me repeat that: They deleted the most relevant cases before the analysis even started.<\/strong><\/p>\n\n\n\n We don\u2019t need to guess how big this omission is. France participates in EUROCAT, a European congenital anomaly surveillance system. From their own data:<\/p>\n\n\n\n That\u2019s 9.46 per 1,000<\/strong> anomaly-affected pregnancies removed<\/strong> by design.<\/p>\n\n\n\n If you want that as a percentage, it\u2019s about 32%<\/strong> of all anomaly-affected pregnancies.<\/p>\n\n\n\n So when Bernard et al. say they found no increased risk in their cohort, they\u2019re only looking at the surviving 68% of anomaly cases<\/strong>. That\u2019s like claiming a ship is safe because most of the survivors didn\u2019t drown\u2014without counting the ones who never made it onto the lifeboats.<\/p>\n\n\n\n Here\u2019s how this works, mathematically. The observed odds ratio in a live-birth-only study (OR_obs) equals the true odds ratio (OR_true) multiplied by a bias factor<\/strong>. If vaccinated pregnancies are even 10% less likely<\/strong> to deliver an anomaly-affected infant alive (due to higher termination or stillbirth), the observed odds ratio gets pulled down.<\/p>\n\n\n\n Quick math:<\/p>\n\n\n\n That\u2019s all it takes to turn a real signal into \u201cnothing to see here.\u201d<\/p>\n\n\n\n And remember, France\u2019s surveillance data already tells us that a third of anomaly-affected pregnancies never show up as live births<\/strong>. So the odds that vaccine-exposed fetuses with defects were underrepresented is not speculative\u2014it\u2019s baked into the design of analysis. Like we have seen so many times before.<\/p>\n\n\n\n Historical precedent makes clear that live-birth-only studies consistently underestimate teratogenic risk. For drugs such as valproate, isotretinoin, and thalidomide, early safety studies that limited analysis to surviving infants failed to detect strong signals of harm. Only when pregnancy registries began systematically including terminations for fetal anomaly (TOPFA) and stillbirths did the full scope of risk become apparent. In the case of isotretinoin, for example, the prospective Motherisk and EUROCAT registry data revealed a striking prevalence of central nervous system and craniofacial defects\u2014patterns that were underrepresented or entirely missed in live-birth analyses. Similarly, valproate\u2019s association with neural tube defects was underestimated until registry-based studies incorporated non-live outcomes and stratified by timing of exposure.<\/p>\n\n\n\n This history isn\u2019t anecdotal\u2014it\u2019s methodological canon. The WHO and EMA both emphasize the necessity of including prenatal losses in teratogenicity surveillance for precisely this reason. The Bernard et al. study, by analyzing only live births and excluding all pregnancies ending in fetal death or termination, repeats the same early-stage design errors that delayed recognition of harm in past pharmacovigilance failures. The claim that \u201cno increased risk\u201d exists should therefore be interpreted not as evidence of absence, but as a predictable artifact of an exclusionary analytic frame\u2014a frame that, in prior contexts, has allowed dangerous exposures to persist unchallenged for years.<\/p>\n\n\n\n Let me show you how every other decision they made also<\/strong> worked to hide any possible risk:<\/p>\n\n\n\n When every source of bias points in the same direction\u2014toward the null or even \u201cprotective\u201d\u2014you don\u2019t have evidence of no risk. You have evidence of bias overpowering signal<\/strong>.<\/p>\n\n\n\n We ran the numbers. If you apply just a 10% selection correction<\/strong> to their published values, here\u2019s what you get:<\/p>\n\n\n\n All of those cross into elevated risk territory. And that\u2019s with only 10%<\/strong> differential selection. A 20% difference pushes those even higher.<\/p>\n\n\n\n If you\u2019re thinking, \u201cCould vaccinated pregnancies really be 10% more likely to terminate anomaly-affected fetuses early?\u201d The answer is yes<\/strong>. And again, France\u2019s EUROCAT data shows 32% of anomaly-affected pregnancies are not live births<\/strong>.<\/p>\n\n\n\n So we\u2019re not imagining the bias. We\u2019re quantifying it.<\/p>\n\n\n\n The authors conduct a massive battery of statistical tests\u201475 individual major congenital malformation (MCM) comparisons, 13 organ system comparisons, and 78 stratified subgroup comparisons\u2014yet report only six statistically significant results, all in the direction of reduced risk<\/em>. They dismiss these as likely type I errors arising from multiple comparisons. But the omission here is twofold: first, they fail to apply any formal correction for multiple testing (such as Bonferroni or Holm), despite the sheer volume of comparisons warranting it; and second, they neglect to explore what the expected distribution of false positives would look like under a truly null association. With over 160 comparisons, even a conservative \u03b1=0.05 threshold implies that 8\u201310 significant results should appear by chance alone. That none<\/em> of these suggest elevated risk\u2014even weakly\u2014is statistically improbable and warrants scrutiny.<\/p>\n\n\n\n This absence of positive findings across so many comparisons is, paradoxically, a signal in itself. If noise alone were at play, we would expect both upward and downward fluctuations\u2014some false harms, some false protections. The asymmetric outcome (only protective results reach nominal significance) suggests not random error, but systematic bias in effect estimation, reporting, or both. It implies either suppression of true positive associations or design artifacts (like live-birth conditioning or exposure misclassification) so powerful that they erase even chance-level appearance of harm. In this light, the authors\u2019 insistence that their null results are \u201creassuring\u201d fails not only in design logic, but also in basic statistical expectations.<\/p>\n\n\n\n This isn\u2019t a smoking gun. It\u2019s a missing body count<\/strong>. When you build a study that excludes the most severe cases, undercounts the ones that remain, and smooths the rest with statistical weighting, you don\u2019t get a \u201csafety study.\u201d You get a safe-looking study<\/strong>.<\/p>\n\n\n\n Methodological bias in a study can either be due to incompetence or fraud. When all of the biases point in the same direction, well I’ll leave it to the reader to decide . Either way, the study is biased\u2014in every direction that would help a null result come out looking \u201creassuring.\u201d<\/p>\n\n\n\n If you\u2019re a policymaker, a journalist, or a pregnant woman trying to make informed choices, understand this: A study that doesn\u2019t count all the outcomes can\u2019t make all the claims.<\/strong><\/p>\n\n\n\n Especially not about safety.<\/p>\n\n\n\n Citations and data sources:<\/strong><\/p>\n\n\n\n Just goes to show you can’t trust the white coats and their “trust the science” lies, and especially the pharma prostitutes who they love to feature on megacorp news. So they murdered many unborn babies with this gene therapy, and they went out of their way to recommend it to pregnant mothers while assuring them […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5,7],"tags":[],"class_list":["post-13908","post","type-post","status-publish","format-standard","hentry","category-health","category-world"],"blocksy_meta":[],"featured_image_src":null,"author_info":{"display_name":"Jason","author_link":"https:\/\/jasonsblog.ddns.net\/index.php\/author\/jturning\/"},"_links":{"self":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/13908","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/comments?post=13908"}],"version-history":[{"count":1,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/13908\/revisions"}],"predecessor-version":[{"id":13909,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/13908\/revisions\/13909"}],"wp:attachment":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/media?parent=13908"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/categories?post=13908"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/tags?post=13908"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}The Setup: A Live-Birth Only Study<\/h3>\n\n\n\n
\n\n\n\nThe Scale of the Omission<\/h3>\n\n\n\n
\n
The Bias Multiplier: A Lesson in Selection Bias<\/h3>\n\n\n\n
\n
Other Ways They Softened the Signal<\/h3>\n\n\n\n
\n
What Happens When You Correct for This?<\/h3>\n\n\n\n
\n
![\"\"\/](\"https:\/\/substackcdn.com\/image\/fetch\/$s_!ktih!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep\/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf84f106-cd8d-4cf4-b2f8-271002522655_1918x1180.jpeg\")
<\/a>Multiple Hypothesis Testing Can Mask<\/h3>\n\n\n\n
The Bottom Line<\/h3>\n\n\n\n
\n