Jurisprudence

Finally, a Judge Calls Shaken Baby Diagnosis an “Article of Faith”

And frees Jennifer Del Prete, who was sentenced for murder 10 years ago.

Child's head x-ray
Doctors once believed that bleeding beneath the outer layer of membranes surrounding the brain, bleeding in the retina, and brain swelling always meant that a baby had been shaken. Above, an X-ray of a child’s head.

Photo by Piotr Marcinski/Thinkstock

Almost a decade into a 20-year prison sentence for murdering a baby in her care, 43-year-old Jennifer Del Prete was ordered freed on bond late last week. The ruling is one of a growing number that reflect skepticism on the part of judges, juries, and even prosecutors about criminal convictions based on the medical diagnosis of shaken baby syndrome. The case is also a critical turning point. The certainty that once surrounded shaken baby syndrome, or SBS, has been dissolving for years. The justice system is beginning to acknowledge this shift but should go further to re-examine and perhaps overturn more past convictions. 

Doctors once believed that three neurological symptoms—bleeding beneath the outer layer of membranes surrounding the brain (subdural hemorrhaging), bleeding in the retina, and brain swelling—always meant that a baby had been shaken. Because it was accepted that a baby with these three symptoms would show the effect of brain damage immediately, the “triad,” as it became known, was also used to establish the identity of the abuser—the last person with the baby. SBS was, in essence, a medical diagnosis of murder.

Beginning in the 1990s, hundreds of cases were prosecuted based on this conception of SBS. The evidence of guilt was strikingly similar from case to case. This includes the Illinois prosecution of Jennifer Del Prete. In 2002, Del Prete was working at a small home day care in a Chicago suburb. One day, when she went to feed the 4-month-old baby in her care, she says she discovered the infant limp. Because the baby had the telltale triad of SBS symptoms, doctors were sure that Del Prete had shaken the baby to death. She denied it, and there were no witnesses. But based on the testimony of medical experts—primarily a pediatrician—she was convicted of murder in the first degree.  

For my new book on the legal treatment of SBS, I focused on Del Prete’s case precisely because, among the hundreds of SBS cases that I had collected over the years, hers was so utterly typical. For instance, the case against Del Prete looked much like the case against a Wisconsin caregiver named Audrey Edmunds. In 1996, Edmunds too was found guilty of shaking to death a baby in her care, based on the triad, even though again there were no witnesses and she denied harming the child. She was sentenced to 18 years in prison. 

In 2008, in the first decision of this kind, a Wisconsin appeals court reversed Edmunds’ conviction. According to the court, a “shift in mainstream medical opinion since the time of Edmunds’s trial” substantially undermined the state’s case. New research (continuing since then) had identified other causes of the triad of symptoms that pointed away from abuse or away from the adult with the baby last as the abuser. There was now, the Wisconsin court found, a “legitimate and significant dispute as to how the infant’s injuries arose.” The changing science meant a jury might well entertain a reasonable doubt as to Edmunds’ guilt. After the court vacated her conviction, the state dismissed all the charges against her.

Since then, a few others accused of shaking babies to death have managed to unwind their convictions. Among the exonerated is Drayton Witt, an Arizona father whose conviction for the murder of his 4-month-old son came undone in 2012 partly because the medical examiner recanted his testimony—after Witt had served a full decade in prison. Expert review of the medical records uncovered a plausible cause of the baby’s death. It appeared that the infant, who had experienced seizures throughout his life, died from an ongoing disease, one that led to a kind of stroke.

Shirley Smith, a California grandmother, received clemency from Gov. Jerry Brown in another SBS case. Smith was helping raise her daughter’s 7-week-old infant when he died in the middle of the night. The baby didn’t have the standard triad of symptoms—just  “minimal” subdural bleeding. That makes Smith’s case an outlier. Still, she was convicted in 1996 and spent a decade in prison before she was released by an appeals court. Brown later commuted her sentence to the time she had already served, citing “significant doubts” about her guilt.

Del Prete’s upcoming release from prison is crucially different from the cases of Witt, Smith, and Edmunds. Witt and Smith prevailed, but because of the less typical nature of their cases, no court expressly questioned the SBS diagnosis. In Edmunds’ case, the state court was concerned about what it viewed as a “legitimate and significant dispute” about the cause of the baby’s symptoms. But the judges stopped there. Now for the first time, a federal judge has condemned the standard SBS diagnosis itself.

In a 97-page opinion, U.S. District Judge Matthew Kennelly held in January that Del Prete had established her “actual innocence,” paving the way for a full exoneration. Considering the testimony of doctors on both sides at an eight-day hearing, there was “abundant doubt, not merely reasonable doubt, regarding Del Prete’s guilt.” One reason for this doubt was that the baby’s subdural bleeding began well before she was left in Del Prete’s care. But another, hugely important, rationale for the ruling was that the SBS diagnosis had become, Kennelly wrote, “highly suspect.” Given what we now know, the judge continued, a diagnosis of SBS is arguably “more an article of faith than a proposition of science.”

The prospect that Del Prete was wrongly incarcerated, while her children grew into adulthood, is profoundly troubling. But her case highlights problems that transcend her own tragedy—gross dysfunction in our criminal justice system. While Del Prete’s conviction is evidently unraveling, others just like it remain untouched. That can’t be right. Perhaps hundreds of inmates—people like Beverly Moore and Alma Calderaro—sit in prison because the mere presence of the triad of SBS symptoms was once assumed to prove their guilt. We have no mechanism in place for revisiting a category of flawed convictions. Our criminal justice system is too primed to stay the course.

One hope for these cases lies within prosecutors’ offices, which are beginning to create “post-conviction integrity units” dedicated to righting wrongful convictions. New York County and Dallas County have had such units for years, and now other offices are following suit. This past month alone, district attorneys in Philadelphia and Cleveland each added a section devoted to exonerating the innocent, and Brooklyn District Attorney Kenneth Thompson announced that Harvard law professor Ronald Sullivan would become the chief of another conviction review unit that will work with an independent panel of attorneys.   

To be sure, prosecutors should not have exclusive responsibility for identifying and correcting injustice in SBS cases (or potential wrongful convictions of any kind). A more comprehensive approach would include the formation of statewide innocence commissions, and it would certainly entail a greater role for the courts. But the work of prosecutors is critical in ensuring that injustice is remedied expeditiously and fairly. If one case unravels for want of solid evidence, those whose convictions rest on the same weak foundations should not continue to languish behind bars. Shirley Smith, the grandmother who served a decade, said it well: “Prison is so horrible. It’s just so horrible. It becomes even more horrible when you don’t belong there.”