Short-term risk assessment and imminent violence adjusted for repeated measurements: Differences among diagnostic subpopulations

This naturalistic prospective in-patient study investigates the association between the Broset Violence Checklist (BVC) and imminent violence, adjusted for repeated measurements. We particularly scrutinize if the prediction of imminent violence differentiates between diagnostic subpopulations. The study included all admitted patients in an acute psychiatric ward during one calendar year (N = 528). Generalized linear mixed model analyses were conducted. Adjusted for repeated measurements, the BVC was significantly associated with imminent violence. The association remained for four out of six diagnostic subsamples, according to the International Classification of Diseases (ICD 10).