The clinical tool Predicting Abusive Head Trauma (PredAHT) can support medical professionals in determining the likelihood that a pediatric head trauma case is abusive given its validation in a recent study.
“Given the rarity of abusive head trauma for clinicians who are not child abuse specialists, it is important to be able to explicitly define which combination of clinical features carries a high probability of abuse when a clinical work-up has been conducted,” wrote Laura Elizabeth Cowley of Cardiff (Wales) University and her associates.
“Knowledge of the results of this prediction tool may assist these clinicians in their discussions with child abuse specialists, in addition to facilitating discussions between child abuse specialists and social welfare, law enforcement, or other professionals involved in the child protection process,” they said (Pediatrics 2015 [ doi:10.1542/peds.2014-3993 ]).
The researchers developed the tool using a dataset of 1,053 cases of children with intracranial injury and data relating to six key clinical features: head or neck bruising, seizures, apnea, rib fracture, long-bone fracture, and retinal hemorrhage.
The authors applied the prediction tool to 198 children, all under 3 years old, consecutively admitted to one of two U.K. hospitals for intracranial injury. Of these, 133 children had nonabusive head trauma and 65 children had abusive head trauma, confirmed based on a case conference or court proceedings, admission by a perpetrator, report of an independent witness, or adherence to criteria with a multidisciplinary assessment. All but six children were under 2 years old, and the majority were under 6 months old.
Using the presence of at least three of the six features to predict abuse, the tool correctly predicted 82% of the 133 nonabusive cases and 66% of the abusive cases. In 12 cases predicted as abusive but which actually involved an accident, the existence of apnea and retinal hemorrhaging contributed to the abuse prediction. The tool inaccurately classified 13 abuse cases as nonabusive due to few positive findings, but a thorough child protection investigation had confirmed abuse.
Based on a cutoff of 50%, the tool’s sensitivity was 72% and its specificity was 86%. The tool had a positive-predictive value of 71% and a negative-predictive value of 86%.
“In the era when the very diagnosis of abusive head trauma continues to be contested in both the public and legal domain, adequately validated clinical decision rules provide more scientific evidence to support clinical decision making,” the authors said. “The PredAHT tool is proposed as an assistive clinical prediction tool rather than a clinical decision rule, which would typically recommend a direct course of action. Prediction tools offer diagnostic or prognostic probabilities, through scores or algorithms, to assist with rather than dictate decision making.”
The research did not receive external funding. The authors reported no disclosures.