Some researchers of landmark shaken-baby syndrome studies question the diagnosis
Shrouded in today’s debate are the complicated origins of a national criminal-justice issue
By Anna Bisaro
The Medill Justice Project
In the 1960s, neurosurgeon Dr. Ayub Ommaya anesthetized rhesus monkeys, secured them in a contoured chair and then accelerated them along a 20-foot track until they crashed into a wall.
The experiments, supported in part by the U.S. Navy, led Ommaya to conclude whiplash, resulting from rapid acceleration and deceleration, can cause cerebral concussions and brain injury, including bleeding on the surface of the brains of these monkeys who have been used in various scientific research.
Ommaya cautioned against drawing conclusions about humans based on his monkey experiments. But his research and that of medical experts who followed him have been used as part of the foundation for shaken-baby syndrome—a triad of injuries including brain bleeding, brain swelling and bleeding within the eyes—and played a role in the conviction of caregivers throughout the United States and abroad over the past several years.
Sometimes lost in the national conversation about shaken-baby syndrome today is that Ommaya and some other key medical experts who pioneered this diagnosis have questioned its foundation in what has become an emotion-laden debate about child abuse.
The debate isn’t about whether vigorously shaking an infant is a good idea, or that doing so could never result in injury. Rather, some medical experts question whether science has reached the point of firmly establishing that caregivers are guilty of such a grave crime—harming a child—when the only signs of abuse are internal head injuries that may be explained by other causes, such as a child’s preexisting medical condition, a rare disorder or an unaccounted for accidental fall.
In 1946, Dr. John Caffey, an American radiologist, wrote a paper in the American Journal of Roentgenology, noting that infants suffered brain bleeding and injuries to long-bones, such as ribs, arms and legs, at the same time. Caffey observed, “Fractures of the long bones are a common complication of infantile subdural hematoma.” He concluded the “fractures appear to be of traumatic origin” but he was uncertain of their cause. This was one of the first studies to begin to hint at the shaken-baby syndrome diagnosis.
Ten years later, Virginia Jaspers, a nanny from New Haven, Conn., was arrested after an 11-day-old girl in her care for died from “head and body injuries resulting from violent shaking,” according to a Newsweek article in Sept. 1956, citing the findings of the medical examiner. The article said she admitted “maltreating five children, including…three who died in her care.” Shaken-baby syndrome hadn’t emerged yet as a diagnosis but Jasper’s case would later be cited as an example of the devastating impact of shaking on children.
Then, in the 1960s, Ommaya began collision testing with monkeys, and his findings were published in 1968 by the Journal of the American Medical Association with Dr. Fred Faas and Dr. Philip Yarnell.
The physical differences between man and monkey are so substantial that Ommaya wrote in the 1968 paper, “We do not contend that our experiments prove that whiplash injury in man can produce lesions similar to what we have demonstrated in monkeys.”
Ommaya died in 2008 but in an interview for this article, Faas, a professor of endocrinology at the University of Arkansas, said the U.S. Naval Air Systems Command supported the research to see how whiplash affects fighter pilots.
Ommaya published later papers that attempted to apply his monkey research to humans to determine how much acceleration and velocity would be needed to concuss a human. While his “scaling of the model to humans is not validated,” his estimated amount of acceleration and velocity needed to concuss a human were “suggested as possible tolerance levels” for fighter pilots, according to a Naval Biodynamics Laboratory report in December 1989. Ommaya’s research also was later used in automobile safety research.
While Ommaya noted the difference between man and monkey, the coauthors of his study remain uncertain how their research applies to shaken-baby syndrome. Denver neurologist Yarnell, one of his research co-authors, said in an interview for this article, “The baby’s brain is a lot different than an adult monkey. [The study] just shows that you can get brain injuries from acceleration and deceleration.” He added, “I could see where people could have extrapolated. Our research wasn’t done with the specific purpose of proving that shaking babies could lead to subdurals. It was done to find out the effects of acceleration and deceleration on the brain.”
In an interview for this article, Faas, the study’s other co-author, said it would be “hard to prove” the monkey injuries could be similar to those of infants because the experiments were done on monkeys, but “it seems like a reasonable application.”
Three years after the monkey study, Dr. A. Norman Guthkelch, a British pediatric neurosurgeon, wrote a landmark paper in the British Medical Journal in 1971 suggesting that shaking of infants could lead to retinal hemorrhages and brain bleeding.
“I began to see a few cases of a rather unusual form of child abuse,” Guthkelch said in an interview for this article. Children“were being brought into the hospital, not particularly ill. They had two features in common. They all had retinal hemorrhages and they all had subdural fluid.” The infants he observed did not have outward signs of abuse, but Guthkelch wrote that all of the caregivers had admitted to shaking their infants, what he called a socially acceptable form of discipline at the time. “The mere absence of visible injury on the head does not exclude direct violence,” he wrote in his paper. He cautioned parents against using “a good shaking” to quiet their children.
A year later, in 1972, Caffey wrote a paper in which he said, “[S]ubstantial evidence, both manifest and circumstantial, has gradually accumulated which suggests that the whiplash-shaking and jerking of abused infants are common causes of” bone and brain injuries. Caffey’s 1972 paper, in the American Journal of Diseases of Children, cited the Jaspers story in Newsweek and Guthkelch’s 1971 paper. Ommaya told Caffey there was no information then to suggest a human could cause the same damage to an infant by shaking as the monkeys had sustained in his collision test, according to a 2011 Brigham Young University Law Review article. Caffey died in 1978.
Like Guthkelch before him, Caffey asserted that brain bleeding may be enough to identify that a child’s head was subjected to whiplash shaking, even without external signs of impact to the head. In 1974, Caffey wrote a follow-up paper in the journal Pediatrics, calling the phenomenon “whiplash shaken infant syndrome.”
Fast-forward nearly a quarter century later, to the late 1990s, when shaken-baby syndrome had become a well-accepted medical diagnosis. Some pioneering medical experts increasingly expressed skepticism about its validity.
In 1997, Ommaya testified on behalf of Louise Woodward, a British nanny in Newton, Mass., accused of murdering an 8-month-old baby by shaking him violently and throwing him against a flat surface.
Ommaya said then the experiments he had done in the 1960s could not be applied to the case, said Dr. Ronald Uscinski, a neurosurgeon at Georgetown University, who, like Ommaya, testified on behalf of Woodward.
“[T]he Ommaya paper emerges as the sole source of experimental data from which the initial hypothetical shaking mechanism was drawn,” Uscinski wrote in a paper published in a Japanese online medical journal, Neurologia medico-chirurgica, in 2006. “[R]atification within the medical community was based principally on anecdotal reports and case studies.”
In an interview for this article, Uscinski said, “People have jumped onto it [Ommaya’s monkey experiments] without thinking it all the way through. Pistons were used to accelerate the chairs holding the monkeys during the experiments, and humans cannot generate that much force with shaking,” said Uscinski, who met Ommaya in the 1970s and was mentored by him.
“Shaking alone is not sufficient,” Uscinski said.
In 2008, Carrie Sperling, then executive director of the Arizona Justice Project, asked Guthkelch, one of the seminal researchers of shaken-baby syndrome, to review the case of a couple whose son had had a difficult birth, lived through a serious bout of pneumonia and experienced regular seizures. One day, the seizures intensified beyond the parents’ control, and they brought their infant son to the hospital. After their child died, the boy’s father was accused of violently shaking him.
Guthkelch agreed and wrote an affidavit on behalf of the defendant in that case. The state dismissed the charges against the father.
“This wasn’t a one in a million kind of case,” said Sperling, now a supervising attorney for the Wisconsin Innocence Project, in an interview for this article.
As Guthkelch began to review the recent medical literature on shaken-baby syndrome since his research decades earlier, “I realized that what I had described was being made into a completely different disease,” said Guthkelch, now 98 years old. “We’ve assumed the cause of shaken-baby syndrome on the basis of a few cases.”
Guthkelch said the sample size of his observations was too small to make such general conclusions. The science is “greatly premature and sufficiently invalid,” he said.
He added: “If I knew the whip with which innocent mothers would be beaten, I never would have written the damn thing. I stand by every damn word. Social workers would stand by it today. But I truly regretted ever having written it, because people are in jail on the basis of what they claim is my paper, when in fact it is nothing like it.”
Northwestern University undergraduate students Dan Tham and Elisa O’Neil contributed to this report.