Review finds insufficient evidence base for the view that violence, sexual, or criminal risk can be predicted

group silhouette

Normally, we at the Learning Disabilities Elf like to look at learning disability specific research, but we thought this review of the utility of risk assessment tools was interesting and relevant to people with learning disabilities who come into contact with psychiatric and criminal justice services.

Risk assessment research stresses the dynamic nature of predictors of offending, in line with notions of behaviour as the result of interactions between personal and environmental factors. Given the dynamism implied by these ideas, it may not be possible to reach a definitive judgment on risk that is stable over time.  Despite this uncertainty, there are a number of structured risk assessment tools used in clinical and criminal justice settings

The authors of this review looked at the use of currently available risk assessment tools to see if it was possible to come to a view on their predictive validity in relation to the risk of violence, sexual, and criminal behaviour.

They found risk assessments conducted on 73 samples covering 24,847 participants from 13 countries.

They found that risk assessment tools used to inform treatment and management decisions performed moderately well in identifying individuals at higher risk of violence and other forms of offending.

If the assessment tools were used as the sole determinant of sentencing and release or discharge decisions however,  then they were much more limited in their utility.

They point to a number of clinical implications of the findings, suggesting that there is not a sufficient evidence base for the view that violence, sexual, or criminal risk can be predicted in most cases and that the tools alone are not sufficient for the purposes of individual risk assessment. They suggest that they may be able to “roughly classify individuals at the group level, but not to safely determine criminal prognosis in an individual case.”

The authors conclude that the predictive accuracy of risk assessment tools varies depending on how they are used and that the

current level of evidence is not sufficiently strong for definitive decisions on sentencing, parole, and release or discharge to be made solely using these tools.

They call for further research into whether or not such tools improve clinical outcomes and reduce repeat offending

Use of risk assessment instruments to predict violence and antisocial behaviour in 73 samples involving 24,827 people: systematic review and meta-analysis Fazel S et al., in British Medical Journal, BMJ 2012;345:e4692

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+
Mark as read
Create a personal elf note about this blog
Profile photo of John Northfield

John Northfield

After qualifying as a social worker, John worked in community learning disability teams before getting involved in a number of long-stay hospital closure programmes, working to develop individual plans for people moving into their own homes. He worked for BILD, helping to develop the Quality Network and was editorial lead for the NHS electronic library learning disabilities specialist collection. This led him to found the Learning Disabilities Elf site with Andre Tomlin as a way of making the evidence accessible to practitioners in health and social care. Most recently he has worked as part of Mencap's national quality team and also been involved in a number of national website developments, including the General Medical Council's learning disabilities site.

More posts

Follow me here –