MP33: A systematic review of the psychometric properties and diagnostic performance of instruments to identify mental health and substance use problems among children in the emergency department

Abstract

Introduction: The objective of this systematic review was to investigate the psychometric properties and diagnostic performance of instruments used in the emergency department to identify pediatric mental health and substance use problems. Methods: A search of seven electronic databases and the grey literature was conducted. Studies assessing any instrument to identify and or diagnose mental illness, emotional or behavioural problems, or substance use disorders in pediatric patients with presentations for mental health or substance use issues were considered eligible for inclusion. Two independent reviewers judged the relevance and study quality of the studies. A descriptive analysis of the outcomes was reported. Results: From 4832 references, 14 studies were included. Eighteen instruments were evaluated for identifying suicide risk, alcohol use disorders, mood disorders, and ED decision-making. The HEADS-ED has good inter-rater reliability (r=0.785) for identifying general mental health problems and modest evidence for ruling out patients requiring hospital admission (positive likelihood ratio, LR⁺=6.30). The internal consistency varied for tools to screen for suicide risk (α=0.46-0.97); no tools have both high sensitivity and high specificity. The Ask Suicide-Screening Questionnaire (ASQ) is highly sensitive (98%) and provides strong evidence to rule out risk (negative likelihood ratio, LR⁻=0.04). Among tools to screen for alcohol use disorders, a two-item tool based on DSM-IV criteria was found to be the most accurate in identifying patients with a disorder (area under the curve: 0.89), and has modest evidence to rule in and rule out risk (LR⁺=8.80, LR⁻=0.13). Conclusion: Reliable, valid, and accurate instruments are available for use with pediatric mental health ED visits. Based on available evidence, emergency care clinicians are recommended to use the HEADS-ED to rule in ED admission, ASQ to rule out suicide risk, and DSM-IV two-item tool to rule in/rule out alcohol use disorders.