Maltreatment-Exposure Screening and Adolescent Mental Health: An Evidence-Based Medicine Perspective on Accuracy and...

Universal trauma-informed screening is increasingly recommended across pediatric settings, yet it remains unclear which forms of adversity most accurately and equitably identify youth at risk for psychopathology. In response, the current study tested whether defining maltreatment by cumulative quality (i.e., number of subtypes) versus cumulative frequency (i.e., number of experiences regardless of subtypes), as well as potential differences within each subtype, yielded more accurate and fair predictions of trauma-related psychological distress. Participants included 839 adolescents from the LONGSCAN study (50.5% identified as female; 55.2% identified as Black) who completed self-reports of maltreatment exposure and psychological distress at ages 12, 16, and 18. Statistical accuracy was assessed via statistical discrimination (area-under-the-curve [AUC]) and calibration (Spiegelhalter's Z; calibration curves), while statistical fairness was evaluated through differences in accuracy across adolescent subgroups (i.e., gender; race/ethnicity). Results across cumulative risk models revealed equivocal results regarding accuracy, while demonstrating important patterns of unfairness. In contrast, analyses by maltreatment subtype identified emotional abuse quality and frequency as both accurate (AUC > 0.64; non-significant Spiegelhalter's Zs; 95% confidence intervals for intercept and slope included 0 and 1, respectively) and statistically fair in identifying concurrent risk for trauma-related distress. Cumulative frequency and emotional abuse frequency also demonstrated accuracy and potential fairness in predicting risk at age 16, whereas no indicators demonstrated clinical utility at age 18. Overall, findings suggest that emotional abuse may be useful in screening for concurrent psychological distress, while adding to the literature questioning the continued reliance on cumulative ACEs scores in clinical and public health contexts.

The online version contains supplementary material available at https://doi.org/10.1007/s40653-025-00767-1.