Inpatient violence in forensic psychiatry: Does change in dynamic risk indicators of the IFTE help predict short term inpatient violence?

Abstract Inpatient violence is a form of recidivism in forensic psychiatric treatment and is stated as an adverse outcome of treatment and a predictor for recidivism after release from the institution. Dynamic Risk Indicators (DRI) are critical key indicators that can predict inpatient violence, but little is known about the effects of change in DRI during forensic psychiatric treatment on the prediction of inpatient violence. This study examines the effects of change in DRI on the prediction of short-term inpatient violence using the Instrument for Forensic Treatment Evaluation (IFTE). A group of 96 patients is followed from entering a high secure forensic hospital until their fifth measurement approximately three years later. The outcome measure is defined as any inpatient violence six months after measurement five. Repeated measures are used to study whether there was a difference in change in DRI between the group of patients who did or did not committed inpatient violence. Binary logistic regression is used to establish the extent to which changes in DRI add to the predictive power of the last measurement. At the group level, the extent of change in DRI did not discriminate between the two patient groups. A large part of the 96 patients already scored low on DRI when entering the hospital and did not (need to) change. At all five measurements violent patients had significant higher scores on DRI than nonviolent patients. Logistic regressions showed that the last measurement predicts inpatient violence sufficiently, the change in DRI during the first four measurements did not contribute to this prediction. The change in dynamic risk indicators does not help to predict short term inpatient violence. The last measurement is the most practical predictor for short term inpatient violence, but because of the dynamic nature of these indicators it is necessary to frequently monitor these indicators to detect imminent risks.