In 2003, a Texas court sentenced Robert Roberson to death after he was convicted of murdering his 2‑year-old daughter, Nikki, by shaking her so forcefully that he caused irreversible brain damage and death from so-called shaken baby syndrome, also called abusive head trauma.

Roberson’s appeal is based on new evidence that doctors misdiagnosed the cause of death and scientific doubts on whether shaken baby syndrome even exists.

The government plans to put Roberson to death on Oct. 17 unless he is granted clemency. Last week, a bipartisan majority of legislators in the Texas House asked the state Board of Pardons and Paroles to recommend clemency. The final decision will rest with Gov. Greg Abbott.

Pronouncements from the medical establishment strongly influence politicians and courts, leaving lasting marks on public policy. In less than a month, such pronouncements may cause the first-ever execution of a man based on dubious medical science.

Science of shaken baby syndrome is questionable

In a 1971 British Medical Journal article, pediatric neurosurgeon Norman Guthkelch hypothesized that aggressively shaking a baby can cause what has come to be known as “the triad” of intracranial hemorrhage, brain swelling and bleeding behind the retinas.

His hypothesis gradually became accepted wisdom by the medical establishment. The National Center on Shaken Baby Syndrome claims that law enforcement authorities process at least 1,300 cases of SBS/AHT per year.

The center points to a consensus among elite medical institutions, from the American Academy of Pediatrics to the Centers for Disease Control and Prevention to the World Health Organization, that shaken baby syndrome is genuine and a significant problem. Mayo Clinic and Cleveland Clinic websites explain the syndrome to lay readers.

Yet researchers have challenged the hypothesis since the 1980s. Some studies have concluded that shaking cannot biomechanically cause the injuries Dr. Guthkelch described.

A 2016 systematic review by the Swedish Agency for Health Technology Assessment and Assessment of Social Services concluded that there is “limited scientific evidence that the triad and therefore its components can be associated with traumatic shaking (low quality evidence),” and that there is “insufficient scientific evidence on which to assess the diagnostic accuracy of the triad in identifying traumatic shaking (very low quality evidence).”

A study published last year in Forensic Science International found that a significant number of patients were misdiagnosed with abusive head trauma, citing other conditions that can cause clinical and imaging “findings commonly associated with AHT.”

Shaken baby syndrome ‘too often’ diagnosed

Dr. Guthkelch himself called to rethink the shaken baby syndrome. In 2011, the pediatric neurosurgeon told a National Public Radio reporter that he “worries that it is too often applied by medical examiners and doctors without considering other possible causes for a child’s death or injury.”

Writing in the Houston Journal of Health Law & Policy in 2012, Guthkelch called his hypothesis into question and disparaged the medical community’s “level of emotion and divisiveness on shaken baby syndrome/​abusive head trauma interfered with our commitment to pursue the truth.”

In that same journal, an article by legal and medical scholars concluded: “Over the past decade, we have learned that this hypothesis fits poorly with the anatomy and physiology of the infant brain, that there are many natural and accidental causes for these findings, and that the onset of symptoms does not reliably indicate timing.”

Shortly before he died, Dr. Guthkelch told The Washington Post in 2015 that he was struck by the high proportion of diagnoses of shaken baby syndrome he reviewed that were attributable to natural causes and not abuse: “I was absolutely horrified when I came back 20 years later to hear all this rubbish about incarcerating mothers, et cetera, et cetera.”

A decade later, the medical establishment and the law have yet to catch up with the science. At best, shaken baby syndrome is rare yet remarkably overdiagnosed. However, enough research has been done to place the hypothesis in doubt.

In 2005, Stanford professor John Ioannidis published his landmark study, “Why Most Published Research Findings Are False.” The distinguished scientific journal Nature reported a record 10,000 published papers were retracted in 2023, and that may just be “the tip of the iceberg.”

Nature separately reported on researchers’ propensity to fudge or falsify data, claiming that in some fields, “at least one-quarter of clinical trials might be problematic or even entirely made up.”

Despite these facts, “narrative inertia” can mean years before medical organizations change their positions on diagnoses and treatments.

During the COVID-19 public health emergency, the public learned the hard way how dogma, groupthink and the suppression of scientific debate describe the culture of the modern medical establishment. One year after the CDC declared the public health emergency was over, people still suffer from collateral damage to mental and physical health, education and socioeconomic conditions caused by the pandemic policies.

It will take a long time for the medical establishment to fix its scientific sloppiness and groupthink. In the meantime, prosecutors and courts should avoid cases based on uncertain scientific theories, regardless of pronouncements by members of the medical profession, to prevent the incarceration of innocent people.

Unlike a research paper, the government cannot retract a person’s wrongful execution.