Risk factors for violence among long-term psychiatric in-patients: a comparison between violent and non-violent patients.

OBJECTIVE The problem of the prediction of violence in psychiatric patients has led to a proliferation of research over the last decade. This study focuses on enduring patient related risk factors of violence, and investigates which long-term patients in Weskoppies Hospital (a specialist psychiatric hospital) are the most likely to commit violent acts. METHOD Nursing statistics on violent incidents and other security breaches were collected for 262 long-term in-patients over a six month period (April - September 2007). The 41 patients who committed violent acts were compared to the 221 non-violent patients in terms of demographic and clinical variables, using two-way tables and Chi-Square or Fisher's Exact Tests. RESULTS The prevalence of violence among the long-term patients was 16%. Fighting among patients was the most common form of violence (58%). The most significant risk factors of violence among the long-term patients are: A diagnosis of mental retardation; first hospital admission before the age of 40 years; total hospital stay >12 years; current accommodation in a closed ward; habitual verbal aggression; absence of disorganised behaviour; and being clinically evaluated as unsuitable for community placement. CONCLUSION The findings will help to identify those long-term patients most at risk of violence. The subgroup of patients with mental retardation is responsible for a isproportionately large number of violent acts in the hospital. The risk lies not so much in their psychiatric symptoms, but more in their cognitive ability, coping skills and inappropriate admission circumstances. Efforts should be directed - at a provincial level - towards their community placement.