Raising suspicion of maltreatment from burns: Derivation and validation of the BuRN-Tool

Area of Science:

• Pediatric emergency medicine
• Child abuse and neglect research
• Clinical prediction modeling

Background:

• A significant percentage of childhood burns (10-25%) are attributed to maltreatment.
• Accurate identification of burn injuries resulting from abuse is crucial for child protection.

Purpose of the Study:

• To develop and validate a clinical prediction tool, the BuRN-Tool, to assist in recognizing suspected child maltreatment in pediatric burn cases.
• To provide healthcare professionals with a simple, objective measure for identifying children at risk of maltreatment.

Main Methods:

• Logistic regression analysis of prospectively collected data from 1327 children with burns.
• Development of the BuRN-Tool by converting regression coefficients to integer scores.
• Validation of the tool on a separate cohort of 787 children using receiver operating curve analysis to establish an optimal cut-off score.

Main Results:

• Key variables associated with referral for maltreatment investigation included young age (<5 years), known social care involvement, concerning explanations, full-thickness burns, uncommon burn locations, bilateral patterns, and supervision concerns.
• The BuRN-Tool demonstrated good predictive performance in the validation cohort, with an area under the curve of 0.87 for scalds and 0.85 for non-scalds.
• Optimal sensitivity and specificity were achieved at a cut-off score of 3.

Conclusions:

• The BuRN-Tool serves as a valuable adjunct for clinical decision-making in emergency departments.
• It assists in identifying children who require further investigation for suspected maltreatment.
• The tool's simplicity and ease of use facilitate its integration into routine clinical practice.