In June, NPR aired 'The Child Cases,' a series investigating child death cases in which parents or caregivers were prosecuted. The stories, produced in collaboration with PBS Frontline and ProPublica, analyzed 23 cases that were overturned or dropped, but did not explain why or how they were selected, raising questions about how representative or common the cases were.
Listener Stephen Boos from Springfield, Mass., wrote our office:
"For the NPR piece, the method by which NPR, PBS and ProPublica identified the 23 cases is important. How many trials were these culled from? If you looked at a sequential series of 50 cases and found 23, then things look as bad as with Dr. Smith. [Charles Smith was a Canadian pediatric pathologist who intentionally altered autopsies in child abuse cases to get convictions.] If this is 23 from the records of many years in many states with many hundreds or even thousands of cases, this is much less of a problem, though a problem still, and a big problem for the defendant. Additionally there is the question of how a case is overturned. Much of the 'new science' coming to court to challenge these cases is questionable science, or science questionably presented."
Here is an explanation from The Child Cases reporting team at NPR, ProPublica and Frontline:
We wanted to determine the frequency that cases are overturned, but that wasn't possible. There's no national database of how many child death cases get prosecuted and not even any agreement on how many such deaths there are a year. So there was no way to get that denominator. As part of a larger investigation, we focused on cases in the U.S. and Canada where charges had been dropped before trial, or juries acquitted at trial, or judges overturned convictions after trial chiefly based on forensic pathology evidence. We didn't look at child-death prosecutions that were overturned for other reasons. After weeding out some cases because we couldn't get all the information we wanted, we analyzed the 23 cases that we were able to report out in depth in order to study patterns and common features.
Researchers at ProPublica and NPR searched legal and news databases to identify cases. We interviewed judges, prosecutors, defense attorneys, child advocates, forensic pathologists and other medical experts. A ProPublica reporter attended a conference of prosecutors and talked to prosecutors and other legal officials about cases. And we used the records of the Goudge Commission to identify cases in Canada.
We found other cases, in addition to the 23 in which the court process had reached a conclusion, in which there seemed to be a plausible or even a strong case of innocence. But there had been no successful appeal. Because of our access to an extraordinarily volume of detailed evidence submitted by both the prosecution and the defense, we focused on the potential innocence of Ernie Lopez.
Some of the 23 cases we reported involved allegations of shaken baby syndrome. Some of them didn't. In some cases, (like that of Melonie Ware) SBS was at first given as a cause of death and then later a different cause of death was identified. Dr. Norman Guthkelch, the man credited with "discovering" shaken baby syndrome, told us he worried how the diagnosis gets used because he's recently seen cases of people in prison who he thinks, from reviewing the medical evidence, are likely innocent.
For her 2009 law review article discussing problems with prosecutions involving shaken baby syndrome, DePaul University law professor Deborah Tuerkheimer tried to figure out how many SBS cases there were. But in the end she could only guess. She estimated that about 200 SBS cases get prosecuted a year.
The explanation is convincing that a troubling trend is afoot. But the complaint by Boos underscores that it is often as important in stories to tell what you don't know as what you know, so that we can all have some perspective on just how large or important a claimed trend might be.