So, what I was hoping would be my biggest, best, most original story of the summer at the Sacramento Bee never ran.
Because I wrote it in the last two weeks, working frantic 14 hours days? Because it’s a contentious issue and the Bee is not known for sticking out its neck? Because it was too long and went through too many edits? Because I was naive and believing and unprepared to fight?
Whatever it is, I’m not going into journalism, it’s not a huge loss for me. And I would have forgotten about it if it weren’t for occasional reminders from Bee interns that I’ve kept in touch with. And if it weren’t for running into it today while looking for graduate school documents.
This is neither the original thing I wrote, nor the final edited version. It was somewhere in between, when one day I decided I would save a copy on my laptop drive and work on it at home.
On Oct. 1, following his sentencing, Carmichael dad Kevin Harper Jr. will take his place quietly behind bars, his payment for a crime that 30 years ago did not exist. The 26-year-old was convicted in Sacramento county last month of shaking his 6-month-old baby son, Jaden Harper, to death. The involuntary manslaughter and deadly assault charges carry a minimum penalty of 25 years.
Shaken Baby Syndrome or, lately, Shaken Impact Syndrome, is believed to be responsible for up to half of all child abuse deaths. But its aggressive criminal prosecution, which has since the 1980s put thousands of parents and caretakers in jail, has been called misguided by some. Shaken baby diagnoses, once on unquestionably solid ground, has in the last ten years come under mounting scrutiny. Now, legal scholars and even some courts have joined the crowd of skeptical scientists in proclaiming the last three decades to be an unfortunate episode of scientific overreaching.
Harper’s is, in many ways, a typical shaken baby case.
A caregiver left alone with an infant arrives at the hospital with the child in his arms. The baby is not breathing, he says, went suddenly limp, or has had a seizure. Despite doctors’ best efforts, the child rapidly deteriorates from unexplained injury. A close medical examination shows blood – blood on the surface of the brain, blood in the retina.
“They found the child had a subdural hematoma and retinal hemorrhages,” said Harper’s attorney Sue Karlton, two bleeding, diagnostic markers thought to be proof of shaking. There’s a third marker – swelling of the brain – that in the most severe cases, kills. Together, in law, they’re called the shaken baby triad.
A diagnosis of child abuse mandates an investigation by law enforcement. But with a shaken baby, said Deborah Tuerkheimer, a Law Professor at DePaul University, the diagnosis is the case. It establishes the crime, the motive, and the identity of the perpetrator – simply the last person with the healthy child.
“Shaken Baby Syndrome is, in essence, a medical diagnosis of murder,” she writes in the Washington Law Review in 2009.
The American Academy of Ophthalmology as recently as 2004 issued guidelines stating if retinal damage takes a certain form, and is found together with bleeding and trauma in the brain, then “shaking injury can be diagnosed with confidence regardless of other circumstances.”
But how confident is confident? Legal experts say studies indicate reasonable doubt. It turns out the set of findings is more common than originally thought. But since the early 1980s, convictions have consistently been obtained and affirmed in the higher courts based on the presence of the triad, sometimes in absense of any other sign of abuse.
The Sacramento Child Abuse Prevention Center reports that since 2005, there’s been an average of one case a year in the county of a shaken baby resulting in death. Nationwide, 1500 or so infants are diagnosed annually with injuries attributed to violent shaking, Tuerkheimer estimates. And the numbers, since 1990, have been on a “sharply upward trajectory,” she writes.
Yet just decades ago, the diagnosis was unheard of. Most accounts of its history point to a seminal 1972 study by pediatric radiologist Dr. John Caffey who noted a mysterious but persistent set of clinical symptoms in a group of children. He proposed that they’d been abused. The rotational accelerations from violent shaking have torn delicate veins in the brain, he said.
Though accepted widely by the medical community, nothing was heard of it for years. Then, in the autumn of 1998 a 19-year-old British au pair made it famous.
Louise Woodward was convicted in Massachusetts of the shaking death of 8-month-old Matthew Eappen and sentenced to 15 years. A judge would reduce her penalty to time served, 279 days, on appeal. But the echoes of her high profile trial lingered, and the pace quickened on shaken baby prosecutions.
“She was a normal, healthy baby,” said Scott Juceam of Hannah Rose, his daughter. “Just learning to walk. Just starting to talk.”
Hannah was found unconscious in her Roseville home on March 11, 2006 while in the care of her nanny Veronica Martinez-Salcedo. She was taken to the hospital where she never woke up.
Dr. Christopher Markus, an emergency room physician at Sutter Roseville Medical Center, was the first to see Hannah. After 35 minutes of examining the baby, Juceam said, Markus approached him and his wife, Lorena, to ask if the child had been shaken. “We’d never heard of it before that,” Juceam said.
A 2007 trial date was set for Martinez-Salcedo, who, as an illegal immigrant charged with murder and child abuse, would be held without bail. But as the family mourned and the prosecution readied its case, elsewhere, something significant was happening.
Physicians who, just a decade earlier, had been instrumental in finding guilt were disowning their own testimony. “New science” has cast doubt on almost every aspect of the shaken baby case, they said. It turns out the shaker’s identity may be masked by a conscious interval in the child, the cause of death is no longer an absolute, and the mechanism of injury is in reality poorly understood. Quietly, and largely without notice, the old scientific consensus had shifted.
Dr. Patrick Barnes, Chief of Pediatric Neuroradiology at Stanford’s Children’s Hospital was the lead expert witness who dealt Louise Woodward her conviction. In early 2007, he was called as a witness in another shaken baby case – this time, in support of Audrey Edmunds’ petition for a new trial. Edmunds, a former Wisconsin day care provider, had by then served 10 years of an 18 year a shaken baby sentence of her own.
He was “embarassed” to recall his old testimony, he said to the judge, “(I and others) told a lie on child abuse based on old diagnostic criteria.”
In the same hearing, Dr. Robert Huntington III, a Professor of Pathology at the University of Wisconsin-Madison, came forward as well. Huntington had testified during Edmunds’ 1996 trial that the timing of the baby’s injuries made her culpable, but later saw something in his practice that would change his mind.
A little girl showed up one day at about 9 a.m. “with bruises and not looking quite right,” he recalled. All day she was a little clingy but not obviously in trouble. But middle of the night she collapsed and by the next morning she was dead. “She’d been in the hospital the whole time,” he said. Her autopsy found all three symptoms belonging to a shaken baby.
Studies have since shown that there can be a delay of more than 72 hours between head injury and onset of symptoms in infants. “Who did it?” was a harder question to answer than originally assumed.
Then, with the fall of a 23-month-old girl, diagnostic certainty crumpled. She slipped off a playset and struck her head some two feet below on carpeted ground. Crying, but alert and talking, she walked with her grandmother to the kitchen. Five minutes later she lost consciousness; 36 hours later she died at the hospital.
There, child abuse experts found the full constellation of head injuries indicative of shaking. But before a prosecution could be initiated, the grandmother presented incontrovertible proof: the accident was caught on tape.
Dr. John Plunkett, a forensic pathologist from Minnesota, published this case study along with 17 other deadly short-distance falls in 2001, the majority with multiple witnesses. Of them, 13 had subdural bleeding, 12 had lucid intervals, and 4 out of 6 eye exams revealed bleeding in both eyes.
In all of the latter cases, there was also subdural bleeding. “At one point the notion was when you saw the triad, you knew with certainty that there was shaking,” Tuerkheimer said. “I don’t know there are any reputable pediatricians saying that any more.”
But most contentious, at the core of the debate, is a crucial omission.
“No child has ever been witnessed to have been injured by shaking alone,” said Dr. Jan Leestma, a Chicago forensic neuropathologist.
This hole in the shaken baby case has permitted questions about its very possibility. “In the biomechanics community,” said Chris Van Ee, an impact biomechanics specialist. “I have not seen much of any support that it’s real.”
Studies show the rotational forces produced by shaking are less than those in a 1 foot fall, he said, you’ll break a baby’s neck shaking before you cause massive brain damage. “I don’t think it exists,” said Leestma.
On this point pediatricians issue their most adamant rebuttal – experience. “There are thousands of physicians in the US that agree that shaking can kill a kid and about 50 who say otherwise,” said Carole Jenny, a Professor of Pediatrics at Brown University Medical School. And 35 years of encounters with caretakers have convinced Dr. Desmond Runyan, Professor of Pediatrics at the University of North Carolina, of the reality of the syndrome.
Still, the American Academy of Ophthalmology clinical guidelines were updated to reflect more nuanced language. The diagnosis “shaken baby” was discouraged in favor of “abusive head trauma”. Other “explanations” were given a nod.
“The difficulty is that practice on these cases is so variable,“ said Dr. Eli Newberger, a Harvard Professor and Founder of the Child Protection Program at Boston Children’s Hospital. Doctors, in general, are not well-trained in separating real cases from the many mimics, he said.
In the courts, the spectre of injustice has prompted second thoughts. For the first time, a Kentucky judge in 2006 banned testimony of shaken injury absent other evidence of abuse. And by 2007, the UK and Canada had launched large-scale reviews of child death cases, including shaken baby convictions, on account of discredited medical testimony.
Against this backdrop of evolving science, and as courts around the world struggled to keep up, in 2007, Veronica Martinez-Salcedo’s Placer County murder trial began. Greg Avella was a juror.
“It was horrible. It was rancorous, it was contentious.”
On the prosecution’s side was a confused, grieving family; on the defense a sad-eyed nanny whose stories didn’t match. But on trial: conventional medical wisdom.
“The first jury went 10-2 for guilty,” the prosecutor Jeff Wood said. “At one point it was 11-1.” Avella, who has five kids of his own, was that one.
“If you don’t have all the evidence, you’re throwing someone’s life away,” he said. “I couldn’t put her away they didn’t prove it to me.”
The jury hung. They would try again.
Meanwhile, in Wisconsin, Audrey Edmunds after 10 years in jail was granted a new trial where, later, all charges would be dropped. The appellate court cited “a shift in mainstream medical opinion”. A contentious debate had replaced a medical consensus, it said. The shaken baby case was weakening.
A year later, Martinez-Salcedo’s second trial would end in another hung jury – but 9-3 the other way.