Development of the Reactive Aggression Assessment (RAGA-16) Part II: Clinical Validation

ObjectiveDisruptive behavior disorders rank among the most common concerns motivating families to seek mental health services. Reactive aggression (RA) is a transdiagnostic symptom associated with more severe illness and worse clinical course. Despite the clinical importance of RA, few validated screening measures or outcome assessments exist for use with children under 12. The 16-item Reactive Aggression Assessment (RAGA-16) is a novel parent-report instrument with promising evidence of psychometric reliability. The current work presents clinical validation evidence for RAGA-16. It is the second of two manuscripts presenting evidence for the RAGA-16.MethodA retrospective chart review examined the records of children seeking services at an outpatient mental health clinic and emergency department associated with a US academic medical center. Parents completed an intake packet including the RAGA-16 along with additional questionnaires assessing depression, hypomania, ADHD symptoms, and aggression. The chart review included item-level data from these assessments along with each child’s clinical diagnoses, demographic information, and use of PRN medication, seclusion/restraints, or readmission within 30 days. The hypothesized RAGA-16 factor structure and reliability were assessed with graded response models (GRMs) and associations with these criterion variables were examined to assess the scores’ validity as a screening measure assessing RA. Patterns of missing data and associations with parent-report onset age and frequency of outbursts were also examined.ResultsConfirmatory models showed good fit for the hypothesized RAGA-16 factor structure. Total score reliability was also high (ωt = .98). RAGA-16 scores showed large significant positive associations with measures of aggression, impulsivity, and oppositionality (rs>.70, ps <.001). The association with hypomanic symptoms was significant but lower (r=.35, p<.01), and RAGA-16 scores were not significantly related to depression symptoms (p>.05). Though not intended as a diagnostic instrument, RAGA-16 scores statistically discriminated diagnoses associated with severe externalizing symptoms on par with how a validated depression measure classified depression diagnoses (both AUCs = .81). Boys had lower mean age of outburst onset was than girls. Endorsement of RA in more contexts (home, school, or elsewhere) was associated with high RAGA-16 scores. ConclusionsThese results support using the RAGA-16 as a clinical screening measure for RA behaviors in outpatient and inpatient contexts. It accurately detected DMDD diagnoses but did not over-generalize to all externalizing disorders. Overall, RAGA-16 scores showed a pattern of criterion associations consistent with the construct of interest along with high score reliability. The addition of inpatient and outpatient clinical score characteristics to the existing epidemiological norms makes the RAGA-16 a promising new tool in the RA assessment toolbox.