Shaken baby syndrome’ and the flawed science in our criminal courts

By Radley Balko, Washington Post
Posted 2/21/2014

Jennifer Del Prete has been behind bars for more than a decade for murdering 14-month-old Isabella Zielinski. Del Prete was convicted of first-degree murder after the jurors* in her Illinois trial heard testimony about “shaken baby syndrome,” or SBS.

For years, expert witnesses told jurors that if a deceased infant has a series of symptoms—bleeding at the back of the eye, bleeding in the protective area of the brain and brain swelling—then the child died from vigorous shaking. Typically, they’d state that the victim couldn’t have died from anything else.

Conveniently, the diagnosis provided the cause and manner of death (shaking), identified the killer (the last person alone with the child), and even gave prosecutors mens rea (anger). The diagnosis made national news in 1997 when British au pair Louise Woodward was tried for the murder of Massachusetts infant Matthew Eappen. She was convicted of involuntary manslaughter.

But a determined group of skeptics began to question the diagnosis shortly after that case. As the SBS prosecutions continued, the group of skeptics grew too. The skeptics now include Ronald Uscinski, a former student of Ayub Ommaya, whose research on monkeys in the late 1960s is thought to be the origin of the SBS diagnosis. (See Emily Bazelon’s excellent review of the research in her 2011 piece for the New York Times Magazine, as well as this follow-up.)

New research suggests that most humans aren’t capable of shaking an infant hard enough to produce the symptoms in SBS. It usually takes an accompanying blow to the head. And in about half to two-thirds of the 200 or so SBS cases prosecuted each year in the U.S., there are no outward signs of physical injury. Indeed, this is the reason SBS is such a convenient diagnosis. It allows prosecutors to charge a suspected abuser despite no outward signs of abuse. But we now know that other causes can produce these symptoms, which means that some percentage of the people convicted in SBS cases are going to prison for murders that may have never happened.

Last month, U.S. District Judge Matthew Kennelly ruled that the SBS evidence presented in Del Prete’s case wasn’t scientifically sound, and had jurors heard the evidence against the diagnosis, they would not have convicted her. The problem for Del Prete and those like her is that while the criminal justice system was quick to let in forensic evidence that convicted her, once convicted, the system puts a premium on finality. In order to get a new trial in a case like this, someone like Del Prete must show not only that she was convicted based on faulty scientific evidence, she must also show that she could not have known at the time of her trial that the evidence was flawed. For Del Prete, that will be difficult. She was convicted in 2003, well after doubts began to emerge about SBS.

As the Chicago Tribune reports, her best option now is to argue that her trial attorney was ineffective. In cases where a trial attorney failed to even challenge SBS evidence (and keep in mind, up until the late 1990s, it was considered slam-dunk science), that may be possible. But what about cases where a defense attorney did put on an expert to challenge the state’s SBS claims, but the jury just found the defense witness unconvincing?

In cases where there’s irrefutable proof of a defendant’s innocence, such as DNA testing, the courts have (mostly) allowed for testing, provided that either the testing technology or the DNA sample wasn’t available to the defendant at trial. And when those tests showed defendants to be innocent, the courts have overturned their convictions and freed them. This has been true even when post-conviction DNA testing conflicts with trial testimony of forensic witnesses, who claimed to link the defendant to biological evidence. Indeed, this is why we now know that several fields of forensic study—bite mark analysis, hair and carpet fiber analysis, and some forms of ballistics analysis, to name a few—are either fraudulent, or have been exaggerated in courtrooms. Many of these areas of “expertise” were presented in courtrooms with the veneer of science, even though they didn’t emerge from the fields of science, and hadn’t been subjected to the process of scientific inquiry, with its rigorous standards like peer review and double-blind testing.

DNA testing did emerge from science, and it is usually definitive. The blood or semen or hair either matches the defendant, or it doesn’t. It will show that either the defendant raped or murdered the victim, or that someone else did. Things get murkier when the question isn’t who committed the crime, but if a crime was committed at all. The new research into SBS doesn’t state definitively that without external injuries, a child couldn’t have died from shaking. It suggests only that there are other possibilities—that shaking wasn’t the only possible cause of death. It isn’t an advance in science that will produce dispositive exonorations. It’s an advance that merely calls prior convictions into question.

We haven’t yet found an adequately way to deal with that. DNA testing has forced the courts to admit when the criminal justice system definitely got it wrong. But the courts are much less likely to take action when new research suggests the system might have made an error. Of course, a person wrongly convicted in a case where new science later shows that the evidence used against them was flawed or overstated is no less innocent than someone wrongly convicted in a case where DNA shows a different guy did it. And the same flaws in our system were at work in both cases. It’s just that once a case moves beyond the trial, our system moves from an emphasis on reasonable doubt to an emphasis on finality.

But now that we know that the criminal justice system is prone to error, perhaps it’s time to revisit the post-conviction emphasis on finality, particularly in cases decided on evidence that science later calls into doubt.

*CORRECTION: Del Prete was convicted in a bench trial, not a jury trial. Judge Kennelly’s reference to the “reasonable juror” is a legal standard, not a reference to what actually happened during her trial. Additionally, Louise Woodward’s jury convicted her of second-degree murder. That charge was reduced to manslaughter by her trial judge.