Guest Shot: Once rock-solid child abuse signs now uncertain

The following op-ed was originally published on August 7, 2009 in the Las Vegas Sun.

Once rock-solid child abuse signs now uncertain

By Abigail Goldman

Friday, Aug. 7, 2009

Audrey Edmunds was charged with murdering an infant in her Wisconsin day care center in 1995. At trial, medical experts testified the child’s injuries were pathognomonic — meaning exclusively characteristic — of shaken baby syndrome: retinal bleeding, brain bleeding and brain swelling, conditions sometimes called “the triad.”

Edmunds pleaded not guilty, went to prison and remained there until last year when her conviction was overturned. A court ruled medical and scientific opinions on shaken baby syndrome had changed so much that a new jury might have reasonable doubt about Edmunds’ guilt.

In fact, one of the doctors who initially testified against Edmunds told the judge he was no longer certain she shook the child.

This is a remarkable development, not just for the accused, but for thousands similarly charged. It’s also, some argue, an indication the legal system lags the science. If one-time medical certainties have been undermined, then why are an estimated 95 percent of people prosecuted on the basis of shaken baby syndrome symptoms still being convicted? Was Edmunds an exception to the rule, or has the science expanded such that other wrongly accused people must now be freed, like death row inmates in the wake of DNA evidence?

As is true just about everywhere, the revelation of the Edmunds case has yet to fully play out in Nevada’s courts.

What remains clear, though, is that absent witnesses, shaken baby syndrome is a medical diagnosis of homicide. The victim’s injuries are the evidence, which is complicated when our understanding of those injuries changes. It’s now understood that the triad may indicate something other than child abuse. Some scientists say short falls can cause retinal or brain bleeding. Some doctors say medical disorders, such as blood or infectious diseases, mimic parts of the triad.

And even if it were possible to prove shaking occurred, techniques used to identify who did it may be flawed. Usually, it’s the last person with the baby who is accused of shaking the child. The belief is such violence quickly renders a victim unconscious or worse.

But new studies suggest children who suffer fatal head trauma can appear lucid for up to 72 hours before death. This expanded time line raises a serious question: Did the trauma happen a day or two earlier?

In recent years physicians have replaced “shaken baby syndrome” with other phrases — “abusive head trauma,” for example. The Clark County coroner prefers “blunt force trauma” or “inflicted abusive injuries.” These changes reflect an expanded understanding of the injury, although some note that replacing too-rigid definitions with loose phraseology isn’t really a refinement.

Although the media are raising questions in the wake of a recent study by DePaul University law professor Deborah Tuerkheimer, Clark County Medical Examiner Alane Olson notes that the debate has been ongoing for some time. What is different is Edmunds’ story. Observers such as Tuerkheimer hope it forces the justice system and well-meaning people eager to identify child abusers to recognize their absolutes may be eroding.

Clark County Deputy Public Defender Scott Coffee has been handling “inflicted abusive injury” cases for about eight years, but only gets one about every 18 months. He has never taken one to trial. The defendants usually strike a plea bargain with prosecutors.

Shaken baby syndrome is a loaded issue and the new research is far from perfect. Experts who testify against shaken baby syndrome diagnoses are ostracized and outnumbered by those who defend them, even when there is evidence to the contrary, Coffee says. Defendants would rather spend a few years than risk a lifetime in prison.

Those convictions, Coffee says, even the no contest pleas, where defendants admit nothing, are seen as proof of guilt, when all we really know is that at least one tragedy occurred — and maybe two.