(PS12-B26) The Role of Parent-child Agreement on Mental Health Symptoms in a Transdiagnostic Treatment for Common Mental Health Problems

Background: Caregivers (CGs) often play an integral role in treatment for youth mental health (MH) problems [1], and treatment requires buy-in from youths and CGs [2]. However, children and CGs often disagree about the nature of child psychopathology, and therefore, about what problems therapy should target, thus presenting the therapist with a ‘dilemma’ [3]. Previous research established that cross-informant concordance is low-to-moderate [4], but it remains unknown whether cross-informant discrepancies impact treatment outcomes.

Method: Data come from an effectiveness trial of a transdiagnostic youth psychotherapy [5] in which participants (N=168; ages 6-15) were treated for common MH problems in community clinics. Participants and one of their CGs reported on the youth’s MH symptoms throughout treatment. We constructed CG-youth discordance (i.e., absolute value of the difference) scores for internalizing and externalizing symptoms using the Youth Self-Report [6] and the Child Behavior Checklist (for CGs). To examine the effect of CG–youth discordance on MH symptoms, we used generalized hierarchical linear models with a time*discordance interaction to test whether changes in CG–youth agreement over the course of treatment were associated with symptoms changes.

Results: Increased CG–youth discordance on internalizing symptoms significantly predicted youth-reported internalizing symptoms (OR=1.03; p=.001), and there was a significant time*discordance interaction (OR=1.02; p=.015) such that less discordance was associated with steeper symptom declines. Similarly, CG–youth discordance on internalizing symptoms significantly predicted CG-reported internalizing symptoms (OR=1.02; p=.006), and there was a significant time*discordance interaction (OR=1.01; p=.006), such that less discordance on externalizing symptoms predicted better trajectories for CG-reported externalizing symptoms. The covariate, discordance on externalizing symptoms was also significant (OR=1.02; p< .001). Youth-reported externalizing symptoms were not predicted by discordance on externalizing symptoms (p=.09), but were predicted by the covariate, discordance on internalizing symptoms (OR=1.01; p=.028). CG-reported externalizing symptoms were predicted by discordance on externalizing symptoms (OR=1.04; p< .001); there was a significant interaction between time*externalizing symptom discordance (OR=1.01; p=.018) in the manner described above.

Conclusion: Our results demonstrate the importance of CG–youth concordance on judgments of youth MH symptoms in predicting treatment trajectories. Specifically, CG–youth concordance on internalizing symptoms was associated with better CG- and youth-reported internalizing symptom trajectories. CG–youth concordance on externalizing symptoms predicted improvements in CG-reported externalizing symptoms, although there was no effect on youth-reported externalizing symptoms. Future studies may probe whether this pattern of findings indicates that CG–youth disagreement has an adverse impact on treatment outcomes or is representative of negative aspects of the CG–youth relationship (e.g., poor communication) that lead to poor treatment response.