ProPublica, in collaboration with Frontline and NPR, recently published "The Hardest Cases," a report about how medical examiners and coroners have mishandled child death investigations. In some cases, their errors helped put innocent people behind bars.

Recently, we interviewed Dr. Carole Jenny, a pediatrician at Hasbro Children's Hospital and a leading figure at the National Center on Shaken Baby Syndrome. She is a high-profile advocate for putting more resources toward investigating suspect child deaths.

According to the National Child Abuse and Neglect Data System, there were 1,700 child maltreatment deaths last year. Yet, according to the FBI, there were 500 child homicides. Why is there such a big discrepancy?

The FBI data comes from medical examiners and state health departments. The NCANS [data] is from child protection agencies. Child death review team data is yet another source. Different agencies have different ways of collecting data. I don't think any of them are all that good. The techniques people use to obtain that data are very, very different from jurisdiction to jurisdiction.

Last week you told the House Ways and Means Committee that progress has been made in tracking pediatric maltreatment, and now the field needs federal money. Could you elaborate on what you meant?

There is now a board-certified sub-specialty called child abuse pediatrics. Physicians who are already board-certified pediatricians take three extra years of training, learning about all aspects of child maltreatment. Most of the folks who go through this—there's 187 in the U.S.—will work at children's hospitals, doing child protection full time. Here we have three physicians and three fellows in training who do child abuse cases. We keep very busy. It's not cheap.

ProPublica's partner NPR recently reported on how scientists are rethinking Shaken Baby Syndrome. Years ago you credited Dr. Norman Guthkelch, the pediatric neurosurgeon, for first observing the condition in children in a 1971 paper for the British Medical Journal. What's your reaction to his recent statements that doctors are over-diagnosing SBS?

I've talked with him. He has the same position as many of us. We do know that kids' heads don't spontaneously explode. But sometimes we don't know if [they've] been shaken, hit or both. The Shaken Baby Syndrome diagnosis presumes a mechanism—major traumatic injury to the head that has absolutely no accidental explanation. Sometimes it's obviously [that]. Sometimes it's not.

Medical examiner Dr. Jon Thogmartin, another expert we interviewed in The Child Cases, says that when a child dies, people presume it's murder. A prosecutorial mindset sets in. Do you agree?

No, that's not true. When a child collapses unexpectedly, child protective services and law enforcement do get involved. Somebody has to do a scene investigation, and it's not going to be the doctors. I'm not going to go to the house to look for toxins or poisons. No one presumes that child injury is abuse. Abuse is a diagnosis we make after we rule out a long, extensive list of physical and accidental ideologies that would explain the degree and severity of the injuries we see.

You testify in child abuse criminal cases monthly and in cases involving child death four or five times a year. Are you more concerned about innocent parents and caregivers being convicted or guilty ones walking free?

Obviously both are not good outcomes. When people who've hurt children aren't held accountable, it puts other children at risk. No one would advocate putting innocent parents in jail. That's why the quality of scientific practice has to be very high.

We work with the medical examiners in Rhode Island, who are board-certified forensic pathologists. None [of the 3] are board-certified in pediatric pathology. It's not a very well-developed field. In general, many jurisdictions underfund it. There's not a lot of people that go into the specialty [of forensic pathology]. It doesn't make a lot of money. It's not like being a plastic surgeon or dermatologist.

Is the "CSI effect" an issue with juries? Will they only convict if they believe the medical science is foolproof, or are they hungry to convict regardless?

I'm not so sure the two poles are opposite. I haven't seen the "CSI-effect." Juries take the material they get and deal with it the best they can.

Instead of arguing these issues in the lab, [doctors] are confronted with an adversarial justice system. We shouldn't do research in the courtroom. We should do research in the lab and at the bedside.

Is there a problem with the standards for those who testify as experts?

People are for sale. Obviously not all. Some who testify are extremely ethical and professionally responsible.

There's no quality control. Some of the experts who give the most outrageous testimony are in very high demand. They are making very large amounts of money.

It's an area where there seems to be no downside to irresponsible expert witness testimony. Most of the folks who do it are retired, doing it as a hobby. State medical societies won't take it on for fear of liability or litigation. State medical boards in most states don't consider expert testimony to be "the practice of medicine," so they don't take it on. And most prosecutors are very hesitant to charge anyone with perjury because this is, in fact, their opinion. If you have an opinion, it may be false. But it's not necessarily a lie.

Is there a solution to this?

I think the medical societies should take a more proactive stance. In the OBGYN societies, they've taken on the issue of malpractice testimony—people going around the country, providing false testimony, causing ethical, competent doctors to be sued.

I know the American Academy of Pediatrics won't touch this. It has lots of missions to promote child health. If they had to tie up millions in child lawsuits, it would decrease their resources for other core missions.

The American Academy of Emergency Medicine has taken an interesting [stance] on this. If their members feel that other members have testified irresponsibly, they will put the transcripts online so that others can read it and comment. It's kind of like a public shaming.

Why did Hasbro Children's Hospital start its Child Protection Program?

It started in 1996, in response to an incident that happened years before. A baby was brutally abused and seen at three different health care institutions. The abuse wasn't recognized. The child went on to be permanently and seriously disabled with a devastating brain injury. A task force formed to look at why this happened. They [concluded] that there was no physician in Rhode Island who had expertise in child abuse. They raised money to endow the first chair in child abuse pediatrics in the country. They went out looking for someone to take it. They ended up recruiting me. In the first year we saw 79 kids. Now we see 1,800 plus per year. If you build it, they will come.

Correction (July 18): This post originally said the American College of Emergency Physicians had posted transcripts of what some of its members see as irresponsible court testimony. It is actually the American Academy of Emergency Medicine that has done this.