New Research on Predictive Models for Pediatric Head Injuries
by Jo Ellen Knott
Differentiating accidental falls from child abuse in young children poses a significant challenge for professionals who work these cases. Child abuse cases are some of the most challenging for prosecutors, law enforcement professionals, and child protection advocates tasked with finding the truth about what happened to the injured or deceased child.
In most cases, no one other than the accused was present to witness the event, and children (especially infants) are too young to communicate what the circumstances were that led to their injuries. As pointed out by a research team from the University of Utah, the task of uncovering the truth “is complicated even more because there are very few scientifically based and validated tools or datasets available to help distinguish between accidental and abusive trauma in infants.”
More than 600,000 children were victims of abuse in 2021, according to the Department of Health and Human Services Administration for Children and Families. Of that number, 1,820 children died due to their injuries, and most were younger than three years old. For more than a decade, the National Institute of Justice (“NIJ”) has awarded grants for research to help physicians and law enforcement distinguish accident from abuse cases when presented with an injured child.
Two recent studies resulting from an NIJ award offer promising advancements in this critical area. In the first study from University of Utah, scientists predict skull fracture patterns. The second study from the University of Louisville provides statistical models for injuries resulting from falls.
The first study, led by University of Utah bioengineer Brittany Coats, focuses on the biomechanics of skull fractures in infants and toddlers. Researchers examined real human skull specimens from deceased children under three years old, subjecting them to various impact forces and stresses. By analyzing the resulting fracture patterns, the team is building a computer model that can predict the type and severity of skull fractures based on the circumstances in which the child suffered a head blow. This tool has the potential to significantly improve case evaluation, reduce diagnostic uncertainties, and aid expert testimony during legal proceedings.
The second study, led by bioengineer Gina Bertocci (University of Louisville) and Dr. Mary Clyde Pierce (Lurie Children’s Hospital), aims to develop a statistical model for predicting head injury risk in young falls. Their research builds on data collected through a previous NIJ grant that monitored children’s falls in a childcare setting using head accelerometers. This data, combined with real-world injury information from emergency room cases, formed the foundation for a comprehensive database used to create the LCAST tool. LCAST stands for Lurie Children’s (name of hospital) Child Injury Plausibility Assessment Support Tool.
LCAST, currently in use at its namesake Chicago hospital, assists medical professionals in identifying potential child abuse cases. While the model offers valuable information, the researchers emphasize its limitations. The LCAST website clearly states that the system is “strictly a screening tool” to aid abuse recognition, not a definitive diagnostic tool.
The research done by the two universities’ bioengineers has made significant progress in differentiating accidental falls from child abuse. By harnessing biomechanical principles and statistical analysis, Coats and Bertocci are equipping medical professionals with new tools to improve child safety and providing lawyers with data to argue for justice.
Source: Forensic
