A viral Substack post published March 30, 2026 by “A Midwestern Doctor” — a pseudonymous writer with a large following — claims that hospital births are systematically dangerous, that routine obstetric interventions cause more harm than good, and that home birth is the safer, more natural alternative. The post has been shared widely among the natural birth community, anti-vaccine circles, and MAHA (Make America Healthy Again) advocates.
I read it carefully. Some of what it says is worth taking seriously. Much of it is not. And the parts that are not — including two graphs that have no source, no population denominator, and no causal validity — are exactly the kind of thing that sounds compelling until you look closely.
Let me be specific.
This is from an anaonymous person
"A Midwestern Doctor" writes under a pseudonym to an audience of 334,000 subscribers, offering specific clinical guidance on birth location, newborn vaccination, and prenatal ultrasound — high-stakes decisions with real consequences for real patients — while remaining completely unverifiable. Readers are told "a physician says this" and left to take it entirely on faith.
No specialty, no training, no licensing board, no hospital credentials, no disciplinary history.
In any actual clinical encounter, we call that a problem: informed consent requires knowing who is advising you, and professional accountability requires that a clinician be traceable when advice causes harm.
The anonymity here is not whistleblowing — it is the strategic borrowing of medical authority without any of the accountability that normally accompanies it.
A physician who counsels a patient toward home birth and something goes wrong answers for it.
An anonymous Substack writer does not.
With 334,000 subscribers, many of them pregnant women making time-sensitive decisions, that asymmetry is not a minor inconvenience. It is an ethical failure.
The Graph That Shows a Rising Maternal Mortality Rate
The post includes a line graph showing maternal deaths per 100,000 live births rising steeply from roughly 7 in 1990 to over 30 by 2021. No source is given. No dataset is cited. No methodology is explained.
The actual CDC data does show a rise in reported maternal mortality in the United States over this period. But what the graph does not tell you — because the author does not tell you — is that a significant portion of this rise is an artifact of coding changes. In 2003, a standard pregnancy checkbox was added to U.S. death certificates. States adopted it at different times, over more than a decade. When a state adopted the checkbox, its reported maternal mortality rate jumped — not because more women were dying, but because more deaths were being identified and captured. Studies published in Obstetrics & Gynecology and the American Journal of Obstetrics and Gynecology have shown that this coding change accounts for a substantial portion of the apparent increase.
The U.S. maternal mortality rate is genuinely higher than peer nations, and that is a real problem worth taking seriously. But a graph without a source, presented without any mention of the coding artifact, is not analysis. It is advocacy wearing the clothes of data.
The Autism and Ultrasound Graph
The second graph is worse. It plots autism prevalence alongside ultrasound machine intensity (SPTA values from “Martin 2010”) and shows the two rising together from 1991 to 2010. The implication is clear: ultrasound causes autism.
This is a textbook example of ecological correlation presented as if it were causation. Internet usage, organic food sales, and the number of Nicolas Cage movies released per year also correlate with autism prevalence over the same period. That is not a joke — it is a standard illustration of why correlation without biological plausibility, dose-response specificity, and controlled study design is meaningless.
The author attributes the graph to “Martin (2010)” — a reference that does not appear in any citation list in the piece, because there is no citation list. I cannot verify this source exists. The large systematic reviews on prenatal ultrasound safety — including a 2015 Cochrane review and a 2017 meta-analysis in Ultrasound in Obstetrics and Gynecology — have not found evidence of harm at diagnostic intensity levels. The consensus that diagnostic ultrasound is safe at recommended doses is not a conspiracy. It is the result of decades of population-level surveillance.
Raising questions about unnecessary ultrasound use is reasonable. Presenting a sourceless correlation graph and implying that ultrasound causes autism is not.
What the Post Gets Right
To be fair: some of the clinical observations in this piece are grounded in real evidence. The case for delayed cord clamping is strong — ACOG endorsed it for term infants in 2016, and the benefits in preterm infants are well established. The documented harms of routine episiotomy are real; rates have appropriately fallen from over 60% in 1979 to under 10% today. Skin-to-skin contact after birth does improve maternal bonding and breastfeeding outcomes. Birthing positions affect pelvic outlet dimensions and labor progress. These are not fringe positions.
The author is also right that the U.S. spends more on childbirth than any comparable nation and does not achieve better outcomes. That disparity is real and demands better answers than it has received.
What the Post Gets Wrong — And Dangerously So
The post implies, and at times states directly, that home birth is safer or equivalent to hospital birth for low-risk pregnancies. This is not supported by population-level data. Our own analyses of CDC natality data — covering millions of births over multiple years — consistently show that planned home births in the United States carry higher rates of neonatal mortality and perinatal complications than planned hospital births, even after stratification for risk factors. This finding is reproducible. It is not a product of hospital-system bias in the data.
The post also waves away the hepatitis B vaccine at birth with a string of speculation — that it was added to habituate parents to vaccine appointments, or to protect a specific refugee population, or for liability reasons. Not one of these claims is sourced. The actual evidence base for newborn hepatitis B vaccination is straightforward: it prevents vertical transmission in infants of infected mothers who were not identified prenatally, and it provides protection across childhood in populations where subsequent vaccination cannot be guaranteed. The author is free to disagree with this policy. Presenting unsourced conspiracy framings as the “four answers from insiders” is not disagreement. It is disinformation.
The section on ultrasound mixes legitimate questions about overuse with references to EMF concentration in the uterus from a building biologist whose paper cannot be located, Chinese abortion studies from the 1980s with no quality assessment, and the author’s personal clinical impression that “regularly ultrasounded” babies lacked vitality. This is not evidence. Framing it alongside real citations creates the illusion of an evidence base that does not exist.
My Take
The obstetric system in the United States has real problems. Overintervention is documented. Cesarean rates are too high. Informed consent is often inadequate. Delayed cord clamping is still not universal despite 10 years of ACOG endorsement. These are things I have written about and published on for decades.
What does not help is a 53-page Substack post that mixes legitimate critique with unsourced graphs, ecological correlations dressed as causation, and conspiracy framings of vaccine policy — published anonymously, with no credentials disclosed, no peer review, and no accountability. It reaches hundreds of thousands of people. Some of them will make birth decisions based on it.
If you want to read criticism of obstetric overintervention, read Henci Goer.
Read the Cochrane reviews on continuous fetal monitoring. Read our published analyses of induction trends in the CDC natality data.
The critique of American obstetrics does not need fabricated graphs to make its case.
The real data are damning enough — and they do not require you to believe that ultrasound causes autism.
Evidence matters. That is true even when the conclusion you reach is one the medical establishment would prefer you did not.