Agreement between child- and informant-based mental health assessments: results from a national population-based study in Lebanon
摘要
Given the importance of accurate assessment of child psychopathology for early intervention, this study aims to examine the agreement between informants and child mental health assessments in the first national Psychopathology in Children and Adolescents in Lebanon Study (PALS) and to identify socioeconomic, family composition, health, and child-related factors associated with agreement.
MethodsThe PALS study includes assessments of depression (Mood and Feelings Questionnaire [MFQ]), anxiety (Screen for Child Anxiety Related Emotional Disorders [SCARED]), and emotional and behavioral difficulties (Strengths and Feelings Questionnaire [SDQ]) through children/adolescents and informants reports. We examined agreement using Cohen’s kappa and factors related to agreement/disagreement using adjusted logistic regression.
ResultsKappa for child-informant agreement for any psychopathology, depression, anxiety, and emotional/behavioral assessments were 0.29 [95%CI = 0.23–0.36], 0.24 [95%CI = 0.12–0.36], 0.24 [95%CI = 0.17–0.31], and 0.14 [95%CI = 0.02–0.26], respectively. Important disagreement was found for anxiety and emotional/behavioral difficulties: child-based anxiety prevalence was 16.52%, while informant-based prevalence was 10.91%; informant-based emotional/behavioral difficulties prevalence was 9.48%, whereas child-based prevalence was 5.01%. Higher disagreement was associated with lower parental socioeconomic indicators, poorer academic performance, child physical illness, family psychiatric history, and previous child mental health service use.
ConclusionResults show low agreement between child- and informant-based mental health assessments in Lebanon, with disorder-specific discrepancies. Disagreement was exacerbated with lower parental socioeconomic indicators and familial psychiatric history and tracked with academic difficulties and child health. Results suggest that integrating multiple informants and types of information enhances mental health assessments, particularly in the context of highly needed research, clinical, and policy efforts.