An investigation into the ability of psychological measures to predict violent behaviour for men detained in hospital under a section of the Mental Health Act(1983).

Recent public enquiries have highlighted examples of inadequate violence prediction and risk management by Mental Health Services in the UK. These inadequacies have had dire and sometimes fatal outcomes. A review of literature that spans two decades of research on the prediction of violent behaviour and risk management for mentally disordered people is presented. This study will explore whether a number of different measures drawn from the research literature can be incorporated into a routine clinical assessment within an inpatient setting to improve the accuracy of violence risk assessment. A sample of 40 male inpatients was drawn from a North London Medium Secure Unit. The HCR-20 was used to carry out a notes review. A sub-sample of 18 men agreed to complete self report measures which included the PCL-R, the NAS, the STAXI, the BIS-10 and the EIS. Violent behaviour was measured for the first month of admission for each man using the OAS. Scores for the HCR-20 significantly predicted inpatient violence during the first month of admission. Scores from self-report measures were examined for associations with inpatient violence. Associations found are discussed in the light of the small number of men consenting to self-report measures. Clinical implications of employing these measures as part of routine clinical practice are discussed in relation to levels of accuracy of violence risk assessment for psychiatric inpatients