Current Issue
(Fall 2005)

Table of Contents

Accident or Abuse?

Bioengineering finds answers to help battered kids

"She fell off her bed..."

"He fell down the basement stairs..."

Pediatrician Mary Clyde Pierce has heard firsthand just about every story there is to tell about how children got their bruises, cuts and broken bones.

Sadly, many of them are just that--stories. "I've seen children brought into the hospital with an unbelievable number of injuries, and being told by a parent or other caregiver that they fell off the bed," Pierce says.

Even when it's suspected, child abuse can be difficult to prove. Often, Pierce says, it comes down to educated guesses or finding inconsistencies in a caregiver's recollection of events.

"If we guess wrong and send a child home, that child could end up dead. Or innocent people could be wrongly accused," she adds. "If we miss a diagnosis of child abuse, there's a 50 percent chance that child will die in that environment. A lot is riding on that single decision."

Pierce says solid scientific ways are needed to quantify just what kinds of injuries are caused by accidents and which are not.

"We need something with enough rigor behind it to stand up in a court of law," she says.

Pierce says the science of biomechanics--which combines the skills of engineering and medicine--offers powerful data and tools that can help doctors in their diagnoses.

biomechanics student
Mechanical engineering master's student Angela Knight adjusts a crash-test dummy before it's dropped as part of biomechanics studies of childhood falls.

Before she moved to Louisville in late 2003, Pierce was part of a team of doctors and engineers working at the University of Pittsburgh and at Children's Hospital of Pittsburgh on biomechanics studies of injuries in children. Pierce is now an associate professor of pediatrics in U of L's School of Medicine and director of research in the pediatric emergency medicine department at Kosair Children's Hospital in Louisville.

Not long after she arrived, she contacted her longtime study colleague in Pittsburgh, Gina Bertocci, and asked her if she might consider coming to Louisville to continue their collaborations.

"I wasn't really looking to leave the University of Pittsburgh, but then this was a unique opportunity and it was important to continue to do our interdisciplinary work," says Bertocci. In June 2004, Bertocci assumed the post of endowed chair of biomechanics in Speed's mechanical engineering department. She also has a joint faculty appointment in the School of Medicine's pediatrics department.

Arriving at U of L with Bertocci was B.C. Deemer, a mechanical engineering doctoral student. Bertocci's research team, housed in the injury risk assessment laboratory at U of L's Health Sciences Center, includes several graduate and undergraduate students in mechanical engineering. Study collaborators also include radiologists, orthopedic surgeons and biostatisticians at the medical school. Pierce is the lab's medical directory.

"We have a great interdisciplinary team studying childhood injuries--particularly long-bone (arm and leg) fractures as well as head injuries and bruising patterns--from many different perspectives," says Bertocci. "The resources we have here in people, facilities and technology have allowed us to keep working without missing a beat."

biomechanics researchers
(Left to right): The interdisciplinary research team includes Mary Clyde Pierce, associate professor of pediatrics, B.C. Deemer, mechanical engineering doctoral student, Gina Bertocci, Speed's endowed chair of bio-mechanics in mechanical engineering and Kim Kaczor, research coordinator for pediatric emergency medicine.

Among those resources are computers wired to crash-test dummies. "We have a family of crash-test dummies, representing a 12-month-old child, a 3-year-old and a 6-year-old," Bertocci says.

The dummies are dropped and pummeled. Dropped from just two feet off the ground, the 3-year-old dummy's head hits a wooden platform with a sickening, loud thud. Forces are measured, the position of landing is noted, and other data is ascertained and compiled in the computer.

"We want to take this information recorded by the computer and compare that data to injury criteria documented in the literature," Bertocci says. "Since the literature on childhood injuries is sparse, we want to build a knowledge base so clinicians will know what type of injuries to expect from specific accidents, to help them delineate between accidents and abuse."

"There are many unknowns when it comes to childhood injuries," Bertocci adds. "For instance, how much does it affect the outcome or the injury risk if a child is chubby or skinny? What are the differences between stair falls and bed falls? What kinds of injuries occur from falls from different heights? What type of injuries result from falls on different kinds of surfaces? There are so many questions to answer."

Bertocci and Pierce say their goal is to establish what happens to children in actual accidents. Then, if a parent says a child was in an accident, doctors can see if the injuries conform to the story. If not, abuse may be assumed.

"When someone says a child fell off a bed, we would like to have data that tells us what really does happen in bed falls," Bertocci explains.

Part of the study team's efforts includes perfecting their investigatory tools. For example, crash-test dummies were originally designed for "high-energy events" such as auto crashes, not relatively low-energy falls, Bertocci adds.

"Crash-test dummies' necks can't twist from side to side like a real child's, for instance," she explains. Deemer and the lab students have modified the dummies to make them more "biofidelic," or more realistic in simulating human movement.

"The ultimate goal is to not only detect child abuse, but to make sure we protect those who are innocent of it as well," Bertocci says. "We're trying to learn the truth."

Shocking Numbers

Speed biomechanics researcher Gina Bertocci provided the following recent statistics on child abuse in the United States:

Return to Top

Table of Contents