Parent-adolescent informant discrepancies: Predicting suicide risk and treatment outcomes

PROJECT SUMMARY Suicide is the 2nd leading cause of death among adolescents in the United States. Research suggests that the onset of suicidal ideation usually arises in early adolescence and these individuals are approximately 12 times more likely than their non-suicidal peers to exhibit suicidal behavior by middle adulthood, highlighting adolescence as a key developmental period for intervention. Yet, these precipitants of suicide death (i.e., suicidal thoughts and behaviors) remain difficult to predict and treat. Comprehensive assessments often include reports from parents and self-reports from adolescents to identify those in need of intervention and examine treatment responses. Historically, these assessments have been interpreted based on highly convergent parent-adolescent reports, or the prioritization of one informants' report over another. There remains a critical knowledge gap around how to best interpret and utilize inconsistent informant reports. Decades of research in youth psychopathology suggest multi-informant discrepant reporting patterns (DRPs) are not synonymous with poor assessment, but may indicate clinically meaningful information. In the present study, we draw from Co-Sponsor De Los Reyes' Operations Triad Model, a set of principles that elucidate the function of converging and discrepant reports, to examine multi-informant DRPs across two contexts. In pursuit of Aim 1, we will initially examine a non-treatment context by drawing data from a completed `Assessment Study' featuring baseline reports from parent-adolescent dyads (n=96 dyads) on the Self-Injurious Thoughts and Behaviors Interview-Revised (R15MH113076; PI/Primary Sponsor Cha). We will test whether baseline parent-adolescent DRPs about community-based adolescents' suicidal thoughts and behaviors predict their future suicidal ideation occurring 3- and 6-months later. In pursuit of Aim 2, we will then examine a treatment context by drawing data from an ongoing randomized controlled trial (i.e., `Treatment Study') featuring pre- treatment reports from parent-adolescent dyads (n=100 dyads) on the Columbia-Suicide Severity Rating Scale (R01MH123556; MPI/Co-Sponsor Yen). We will test whether pre-treatment parent-adolescent DRPs about inpatient adolescents' suicidal thoughts and behaviors predict their post-treatment suicidal ideation and behavior, and whether these effects are moderated by treatment assignment (active `Skills to Enhance Positivity' intervention vs. enhanced Treatment as Usual). Polynomial regression and response surface analyses will be used to probe the directionality and degree of parent-adolescent DRPs (Aims 1-2). With the intention of testing whether multi-informant DRPs predict adolescents' future risk across treatment and non- treatment settings, the long-term goal of this work is to maximize the prognostic value of information that is readily collected as a part of standard mental health evaluations for youth.