Are Innocent Parents Being Prosecuted for Killing Their Babies?

The doctor who came up with “shaken-baby syndrome” thinks so.

Drayton Witt and his wife.
Drayton Witt and his wife

Courtesy of the Arizona Justice Project.

In 1971, pediatric neurosurgeon Norman Guthkelch helped come up with the medical diagnosis of shaken-baby syndrome.* Guthkelch and another pediatrician each wrote a paper proposing that unexplained bleeding in the brain of infants could occur because of whiplash—via shaking—without causing a visible neck injury and without direct impact to the head. That diagnosis became the basis for “do not shake” campaigns and, over the years, hundreds of criminal prosecutions for child abuse.

Now Guthkelch is worried that medical examiners and prosecutors have been too quick to turn to the shaken-baby diagnosis—and that innocent people may be in prison as a result. He called me to express that concern after I wrote about some questionable shaken-baby prosecutions for the New York Times Magazine last year. Guthkelch told Joseph Shapiro of NPR about an Arizona case he found particularly troubling after reviewing it for the defense. “I think I used the expression in my report, ‘I wouldn’t hang a cat on the evidence of shaking, as presented,’ ” he said.

Guthkelch, who is 96, was talking about the conviction of Drayton Witt, who was an 18-year-old when he was charged with shaking his 4-month-old son, Steven, to death in 2000. Based on the standard medical theory of the time, the case looked like a classic case of abusive shaking to doctors who treated Steven in the hospital as well as to the medical examiner who conducted the autopsy. Later at trial, these doctors testified against Witt, who was convicted of second-degree murder and sentenced to 20 years.

But now the case doesn’t look so clear-cut. It may, instead, be an example of doctors applying an outdated method of diagnosis—one that ignored Steven’s troubled medical history and thus missed an alternate explanation for his death. It’s not just Guthkelch who is raising the alarm: The medical examiner who testified against Witt, pathologist A.L. Mosley, has recanted his position at trial. Mosley now believes that Steven died of natural causes. Which raises a couple of questions: Is Drayton Witt serving time for a crime he did not commit? Will prosecutors in Arizona continue to argue that his conviction should stand?

Steven Witt’s health problems started at birth. He was born with the umbilical cord wrapped around his neck, and he had trouble breathing. He got better and went home, but a few weeks later, his mother, who lived with Drayton Witt and later married him, brought the baby to the doctor because he’d been feverish and vomiting for three days. A few weeks later, Steven was sick again, with a fever and an upper respiratory infection. From then on, his mother was in frequent touch with the pediatrician.

A few days before Steven turned 4 months old, his mother took him to the emergency room because of what was by then his familiar fever, congestion, and vomiting. The doctors thought he had pneumonia, and they prescribed an antibiotic. After Steven took his first dose, his left eye couldn’t focus, he refused the bottle, and he lay limp. The next morning, he threw up and started shaking. Witt came home from work, and he and Steven’s mother rushed the baby to the hospital.

In the ER, Steven had a grand mal seizure. He spent the next six days in the hospital, with a fever and pupils that weren’t dilating normally. The doctors weren’t sure what was wrong, but tests showed more massive seizures, even though Steven was taking anti-seizure medication. His mother told the doctors that Steven had lost head control and couldn’t track with his eyes as he could before. Still, they sent the baby home with his parents before the week was up, on May 7, with antibiotics and the anti-seizure drug Phenobarbital.

According to Witt’s defense, Steven was never again a normal, healthy baby. Sometimes, his eyes twitched back and forth. He threw up repeatedly. His mother called the pediatrician and the hospital; his grandmother worried about his health; and a neighbor who was a nurse noticed the baby’s recurring problems. Steven’s mother brought him back to the pediatrician on May 26 because he’d again had a fever for days and had been projectile vomiting. She was told to continue the antibiotics and Phenobarbital, and she and the baby were sent home.

On June 1, Witt took care of Steven during the day while his mother went to work. Shortly after 8 p.m., Witt told Steven’s mother over the phone that he thought the baby was having another seizure because his eyes weren’t normal. They agreed that Witt would pick her up at work and they would drive to the hospital. On the way, Steven had a major seizure, and Witt tried to resuscitate his son. At the hospital, it took 32 minutes to get Steven’s heart started again. He was also seriously dehydrated—because of the vomiting, he’d been unable to keep down food or liquids.

And yet despite the baby’s long medical saga, Witt was almost immediately suspected of abuse. The social worker from child protective services who talked to Witt and his wife at 4 a.m. on June 2 wrote in her notes that the couple had no specific explanation for the baby’s “severe head trauma.” After Steven’s death on the afternoon of June 2, Dr. Mosley, who knew that the doctors at the hospital thought he’d been killed by shaken baby syndrome, concluded that the death was a homicide caused by “shaken/impact syndrome.” Mosley noted that Steven had no neck injury, no bruising or fractures, no external signs of abuse. But none of that mattered. At Witt’s trial, Mosley testified against him, along with several other doctors. Because of the state of the science at the time, Witt had just one doctor in his corner, who only partially rebutted the shaken-baby theory.

Twelve years later, lawyers from the Arizona Justice Project have gone to court asking for a hearing for Witt. They want the state courts to reopen the case and order a new trial, based on the way the scientific understanding of shaken-baby syndrome has shifted. The mainstream position still holds that it is possible to shake a baby to death. But many doctors now allow that a history of illness like Steven’s can account for the subdural and retinal bleeding, and brain swelling, that used to be blamed exclusively on shaken-baby syndrome.

It’s Steven’s medical history that distresses Norman Guthkelch. “This was always a sickly baby,” he told me when I reached him over the phone. “I was taught to start by considering the non-criminal explanation, and this little guy was in sufficient trouble without any need to postulate someone had shaken him.” Mosley has joined Guthkelch in asking the Arizona courts to revisit Witt’s conviction. “There is now no longer consensus in the medical community that the findings I reported in my autopsy report are reliable proof of SBS [shaken baby syndrome] or child abuse,” Mosley told the court. “Steven had a complicated medical history, including unexplained neurological problems. He had no outward signs of abuse. If I were to testify today, I would state that I believe Steven’s death was likely the result of a natural disease process, not SBS.”

Drayton Witt reminds me of Shirley Ree Smith, a grandmother accused of shaking her baby to death in California in 1996. Both may well have been convicted based on past medical assumptions that may not stand up to present scrutiny. That may also be true of some people sent to prison more recently. I still wonder about the conviction of Trudy Munoz Rueda, a daycare provider whom I wrote about in the Times Magazine. Another is a former New York daycare provider named Alma Calderaro. When she was convicted of assault and child endangerment in 2009, and sentenced to eight and half years in prison, Calderaro’s lawyer did not call a single medical expert to testify on her behalf. Now on appeal, three doctors are saying that the baby who collapsed in Calderaro’s care had subdural bleeding that was chronic and that this condition, rather than shaking, probably caused her injuries. Prosecutors have until mid-April to respond.

By calling attention to these cases, I don’t mean to suggest that every defendant convicted of harming a baby who has the symptoms associated with shaken-baby syndrome is innocent. I’ve learned from my reporting that these cases are incredibly fact-specific. In a pending prosecution in Queens that has gotten a lot of attention, the baby who died, Annie Li, had serious bone fractures as well as internal bleeding. The lawyer for Li’s father, who has been charged with manslaughter and second-degree murder, says that while the test results aren’t in yet, a congenital bone condition may be to blame for the fractures. We’ll see.

Meanwhile, the district attorney who prosecuted Drayton Witt hasn’t responded to his petition for a hearing and a new trial. As these cases move through the courts, they are testing prosecutors’ and judges’ willingness to undo old verdicts. That’s never an easy step to take. But sometimes it’s the only one that is just.

Correction, March 15, 2012: This article originaly referred to Dr. Guthkelch as a pediatrician. (Return to the corrected sentence.)