For more information about the Network
call 0121 415 6960 or email
Like-minded organisations in pursuit of restraint-free services

Developing a tool for violence risk assessment

Roger AlmvikFor his Keynote Speech at the 2015 Restraint Reduction Conference, Roger Almvik, Research Director at St Olav’s University Hospital/Norwegian University of Science and Technology will be discussing the Brøset Violence Checklist as an effective means of assessing the risk of violence. Here, he give us an insight into what delegates might expect from the presentation.



A structured risk assessment acts as an aide-mémoire, making sure that we collect all the relevant information… Best practice, therefore, is to use the data as the basis for a clinical team meeting. The teams do the work of evaluation and planning, once the structured instrument has ensured that the necessary information is to hand. The outcome is, of course, only as good as the clinical team. If staff lack training or experience, they end up by being overwhelmed by information they are unable to use.

(Maden, 2003)

Risk assessment in general and the risk of violence in particular has received more and more attention in mental health care during the last couple of decades. This is clearly illustrated by the level of media attention every time a mentally ill person is involved in serious crime. Despite both awareness of, and competence on, the topic, it remains a complex issue, influenced to some degree by whomever defines the risk.

The Brøset Violence Checklist (BVC) was developed on the basis of the empirical work of Linaker and Busch-Iversen (1995). They investigated episodes involving physical violence documented in the medical records of 92 patients admitted to a Norwegian maximum-security unit during the years 1988-1993. The criterion for inclusion was violence severe enough to require the use of physical restraint. In all, 48 episodes of violent behaviour by 32 patients were identified; 16 of these patients were involved in two episodes each. All the daily nursing reports during the five-year period were also examined at specified control dates for reports of symptoms and behaviours. Single behaviours that were significantly more frequent before violent incidents were identified and used to build a logistic regression equation. The occurrence or non-occurrence of violence was used as a dependent variable, while the behaviours constituted the independent variables.

The six most frequent behaviours that occurred in the screening period prior to an incident were confusion, irritability, boisterousness, physical threats, verbal threats, and attacking objects. Entering these six behaviours into a logistic regression equation indicated that all of them were predictive of violence and seemed to constitute frequent warning behaviours.

To assess or estimate risk is a very difficult task, but success benefits staff, patients and society in general. Staff feel, and actually are, safer; patients benefit from improved clinical environments; and the general anxiety in communities for people with a mental illness is reduced. There is no clear pathway to best practice in this area, but improvements in research quantity and quality bring us one step closer to an even better way of reducing violence in psychiatry. Also, ongoing debate in both professional and open public forums is needed to keep clinicians, researchers and the public informed about the pros and cons and updated on the latest development in the field of risk assessment.

Roger’s Keynote Speech is scheduled to take place on the afternoon of Friday 13 November.