[Predicting aggression during the treatment of forensic psychiatric patients by means of the HCR-20].

BACKGROUND

A substantial number of forensic psychiatric patients also show aggressive behaviour while being admitted to a psychiatric hospital. Risk assessment can therefore be of importance not only for estimating the risk of recidivism after treatment, but can also be used to protect the hospital staff.

AIM

To find out to what extent scores on the Historical Clinical Risk Management-20 (HCR-20) can predict aggressive behaviour during inpatient treatment in a forensic psychiatric department.

METHOD

In total, 102 patients were included in our analysis. Of these, 43 patients had caused 174 aggressive incidents between January 2005 and August 2008. The incidents were recorded by staff members who used the Staff Observation Aggression Scale-Revised (SOAS).

RESULTS

On average, the group of patients involved in one or more aggressive incidents were found to have higher hcr-scores than patients who were not involved in aggressive incidents. The area under the curve (AUC)-value of the HCR-20 total score was 0.70 as far as the prediction of aggression was concerned. Logistic aggression analysis suggested that particularly the C-subscale items assessing impulsivity (item C4) and the patient's response to treatment (item C5) may be able to predict aggressive behaviour fairly accurately in a particular department.

CONCLUSION

The HCR-20 can predict to a certain extent which patients will engage in violent behaviour while receiving treatment in a forensic psychiatric department. These results correspond to those of similar earlier investigations which showed that the HCR-20 could predict that patients would engage in further violent or criminal behaviour after being discharged from hospital.