Background: We examined the extent to which trait anger and psychopathic traits predicted post-discharge self-directed violence (SDV) and other-directed violence (ODV) among psychiatric patients.
Method: Participants were 851 psychiatric patients sampled from in-patient hospitals for the MacArthur Violence Risk Assessment Study (MVRAS). Participants were administered baseline interviews at the hospital and five follow-up interviews in the community at approximately 10-week intervals. Psychopathy and trait anger were assessed with the Psychopathy Checklist: Screening Version (PSC:SV) and the Novaco Anger Scale (NAS) respectively. SDV was assessed during follow-ups with participants and ODV was assessed during interviews with participants and collateral informants. Psychopathy facets and anger were entered in logistic regression models to predict membership in one of four groups indicating violence status during follow-up: (1) SDV, (2) ODV, (3) co-occurring violence (COV), and (4) no violence.
Results: Anger predicted membership in all three violence groups relative to a non-violent reference group. In unadjusted models, all psychopathy facets predicted ODV and COV during follow-up. In adjusted models, interpersonal and antisocial traits of psychopathy predicted membership in the ODV group whereas only antisocial traits predicted membership in the COV group.
Conclusions: Although our results provide evidence for a broad role for trait anger in predicting SDV and ODV among discharged psychiatric patients, they suggest that unique patterns of psychopathic traits differentially predict violence toward self and others. The measurement of anger and facets of psychopathy during discharge planning for psychiatric patients may provide clinicians with information regarding risk for specific types of violence.