scorecardresearch Skip to main content
opinion | Lee Scheier

Martha Coakley, stop lauding bad science

Martha Coakley, right, embraced Deborah Eappen on Oct. 3. Eappen is the mother of Matthew Eappen, who died in 1997 at the age of 8 months.Josh Reynolds for The Boston Globe

After coming under attack in an political ad for not doing enough to protect children, Martha Coakley, the Democratic candidate for governor, defended her record. In a large above-the-fold photograph published in the Globe Oct. 3, Coakley is seen standing next to Deborah Eappen, mother of Matthew Eappen, the baby whom Louise Woodward was charged with shaking to death in 1997.

Coakley, the prosecutor in that infamous trial, set up the photo op ostensibly to remind the public of her commitment to protecting children. If so, Coakley must think Massachusetts voters have short memories.

Although Woodward was found guilty of second degree murder by the jury, trial judge Hiller Zobel reduced the sentence to manslaughter and set Woodward free. Zobel’s skepticism of the justice of Woodward’s murder conviction was prescient: Dr. Patrick Barnes, Coakley’s chief expert witness in the case, later publicly renounced his own trial testimony as based on flawed scientific assumptions.

Coakley’s odd invocation of this case demands that we look at the facts. What cannot be lost in all of this political maneuvering is the truth about the Woodward case and all the thousands of shaken-baby cases before and since Woodward. The truth is that Martha Coakley’s deft misuse of science actually came very close to sending an innocent caretaker to prison for life.

At the heart of the Woodward prosecution — and many others like it — is the pseudoscience of the shaken-baby syndrome diagnosis. Anecdotal studies by pediatric neurosurgeon Norman Guthkelch in 1971 and radiologist John Caffey in 1972 and 1974 asserted that the presence of the so-called “triad” — subdural and retinal bleeding plus brain swelling — in an infant constituted certain proof that the last caretaker to hold a baby had “shaken” that child to death. Under Guthkelch and Caffey’s theory, this could be the only cause of death. Because of these studies — and because nearly all shaking cases have no eyewitness — the presence of the triad was the only evidence necessary to prove murder.


In recent years, however, experts have become increasingly skeptical. “The problem lies with Caffey’s level of certainty that subdural bleeding along with bleeding in the eyes could only be caused by shaking,” says Louis Fogg, a professor at Rush University and the former president of the Chicago chapter of the American Statistical Association.


Fogg, who received his doctorate from the University of Chicago in scientific methodology, is an expert in designing and evaluating scientific research. He has concluded: “[Caffey] is manufacturing a degree of certainty that doesn’t exist from his research. A reasonable person can’t look at this research and say it is proof of guilt beyond a reasonable doubt. There are a lot of doubts here. This is not the kind of stuff you can send people to jail for.”

Fogg goes on to note that all of the many published studies that purport to confirm the validity of shaken baby syndrome are false. “Every study accepts nonscientific evidence as if it is scientific,” he explains.

Doubts about shaken baby syndrome have grown considerably since it has been established that many diseases, medical conditions, short accidental falls, or re-bleeds also manifest the identical “triad” of clinical signs. Indeed, it likely was an injury from a short accidental fall weeks before Matthew Eappen died, which re-bled later — not any supposed shaking by Louise Woodward — that caused his death.

Barnes, the expert Coakley relied on to convict Woodward, is one of the physicians who has questioned the legitimacy of the diagnosis. When asked on PBS’s Frontline in 2011, “If you were called to testify in the Woodward case today what would you say?” Barnes responded, “I would say that you cannot select out, accuse, indict, or convict any particular caretaker based on the medical evidence that we have.


Perhaps most tellingly, Dr. Gulthkelch himself — one of the scientists who first posited the theory — has since voiced concerns about its application in the criminal justice system.

Nonetheless, in the face of rigorous scientific evidence to the contrary, many in the medical community still claim that shaken baby syndrome is a valid diagnosis because so many within the medical community support it. But science is not a popularity contest. After all, Galileo was the only scientist of his day who believed that the earth revolved around the sun.

And this consensus does nothing to alter the fact that, to this day, there has never been an experiment that proves a causal link between shaking a baby and death. “An association alone between shaking and the triad contains so much inherent doubt that no jury in good conscience should convict on that evidence alone,” says Fogg. “You need strong experimental evidence of causation if you’re going to send someone to prison.”

This lack of strong experimental evidence of causation did not stop Coakley from trying to send Woodward to prison for life. And despite the growing chorus of scientists casting doubt on shaken baby syndrome, such prosecutions continue: In Middlesex County, nanny Aisling Brady McCarthy currently stands accused of shaking 1-year old Rehma Sabir to death. Prosecutors should hang their heads in shame for blindly accepting bad science and bringing these cases to court.


Indeed, Coakley nearly sending a young, innocent woman to prison is nothing to be proud of. It is certainly not worthy of a photo shoot in a political election.

Lee Scheier is an investigative journalist who has written extensively for the Chicago Tribune and many other publications. He has spent the last eight years researching a book on shaken-baby syndrome.