UK researchers have tested the validity of a violence risk assessment tool, the Violence Risk Scale second edition (VRS-2) on a British sample of psychiatric patients, held in a medium security facility. The VRS-2 has recently been adopted by the UK Home Office as one of its standard measures of 'dangerous and severe personality disorder', yet to date it had not been extensively validated in the UK.

Researchers Mairead Dolan and Rachael Fullam completed the VRS-2 and the already well-validated Historical Risk Assessment Scale (HCR-20) using the admission notes of 136 patients held at the Edenfield Medium Secure Unit in Manchester. The patients were aged 35 years on average, and the majority had a diagnosis of schizophrenia. Most had been referred to the unit from prison, with 14 percent having previously committed murder.

The VRS-2 contains 6 'static factors' including age of first violent conviction and 20 so-called 'dynamic' factors (subject to change through intervention) such as 'criminal attitudes' and 'emotional control'. The HCR-20 contains 10 historical items such as 'age of first violent incident', 5 clinical items such as 'lack of insight' and 5 risk items, including 'lack of personal support'.

There was a strong correlation between scores on the VRS-2 and the HCR-20 showing the VRS-2 has strong 'construct validity. That is, it appears to be measuring what it is supposed to be measuring. The VRS-2 also showed reasonable 'inter-rater reliability '. That is, two people administering the tool to the same person's records tended to give similar scores.

Finally, looking at data from 80 of the patients, the researchers found that those who went on to be violent over the next 12 months tended to have scored higher on the VRS-2 and HCR-20 than did those patients who did not commit violence during that time. This demonstrated both scales have 'predictive validity'. This was especially the case for the dynamic items of the VRS-2 rather than the static items, and the clinical items from the HCR-20.

The researchers noted that 'Taken together our findings suggest that the more dynamic elements of these risk scales are the important factors in predicting institutional violence, at least in patients with major mental disorders.' Read more on this topic by following the links below.

Web Directions:

Dolan, M. & Fullam, R. (2007). The validity of the Violence Risk Scale second edition (VRS-2) in a British forensic inpatient sample. The Journal of Forensic Psychiatry and Psychology, 18, 381-393.

Further reading:

Violence prediction and risk analysis webpage from the Pacific Institute for the Study of Conflict and Aggression:

Information on violence risk assessment from the website of Canadian forensic psychiatrist Stephen Hucker:

'The challenge of preventing violence is not just an American problem'  This is an opinion piece from The Observer newspaper

Study shows clinicians were unable to predict which mentally ill patients would go on to commit violence in the next two years. This is from the British Psychological Society's Research Digest.