Psychometric Properties of the Emotional Outburst Inventory (EMO-I): Rating What Children Do When They Are Irritable

Objective: Existing measures of irritability rarely distinguish phasic and tonic forms, despite their different clinical implications. We developed the Emotional Outburst Inventory (EMO-I) as a brief screening tool for phasic irritability in youth in clinical settings. The EMO-I assesses outburst severity, frequency, and duration. This article reports on its psychometric properties. Methods: The sample included 2,552 youth (mean [SD] age = 12.1 [3.5] years) evaluated at a university outpatient clinic between February 2005 and June 2014. Parents of 1,772 youth (69.4%) endorsed some anger problem. We assessed convergent, construct, and incremental validity of the EMO-I using a variety of measures, including the Child Behavior Checklist (CBCL) dysregulation profile (DP) and the CBCL irritability subscale. We also examined associations with hospitalization, emergency department visits, atypical antipsychotic use, help-seeking for outbursts, and impairment. Results: The EMO-I severity had good internal consistency (Cronbach α = 0.83) and was significantly associated with other irritability constructs (median correlation, r= 0.66, all P< .01). Outburst severity was associated with impairment (β = .87, P< .01) and with hospitalization, emergency department referral, antipsychotic use, and help-seeking for outbursts (median odds ratio = 1.27, all P< .01). The EMO-I showed incremental validity over and above the CBCL-DP and CBCL irritability subscale (explaining an additional 2%–12% of variance) when examining associations with impairment, hospitalization, emergency department referral, antipsychotic use, and help-seeking for outbursts. Conclusions: The EMO-I showed good internal consistency and convergent, construct, and incremental validity. Outburst behavior severity had stronger associations with important clinical variables than did outburst frequency and duration.