News

Providing up to date risk assessments of mental health service users is second nature to staff in this field. But few question the nature of the procedure despite often unjustified media criticism of mental health workers for their perceived failure to protect the public against violent clients.

For a picture of how risk assessments are carried out, we examined seven community mental health teams across England and Wales. There were substantial regional variations. Most used a tick box to rate their clients on a scale but the risk categories were not all the same. Despite the availability of formal risk assessment tools, staff were often directed to use tools created by their employers. Consequently, in multi-disciplinary teams various forms were in use as each specialism undertook its own assessment.

Worryingly, in forensic teams - usually those with the most risky clients - the most renowned tool is the Hare psychopathy checklist.¹ This has been criticised for its lack of attention to an individual's motives, which can be the most predictive clinical variable.² In addition, there are many variables associated with possible violence, such as being young, male and abusing alcohol or drugs. Clearly, some mental health users are in this category although they may also have psychotic symptoms such as hallucinations where voices order people to carry out violent acts. But, in reality, clients are likely to have only a few risk factors as a result of their illness. Others derive from social circumstances and personality.

Mental health inquiry findings emphasise that there should be no short cuts in risk assessment, which should include a patient's background, present mental state, social functioning and past behaviour. This takes significant time if carried out properly and has massive resource implications - it is not simply a matter of formalising good practice.³ But none of the teams had any extra resources, leaving risk assessments to be tagged on to the work already completed. They were likely to be rushed and inadequately made; those that were completed well were done so to the detriment of other interventions.

Also, in five of the teams there were wide variations in what happened or was supposed to happen once an assessment had been completed. Usually, the form took pride of place at the front of a client's file, but this was often all that would be done. Each team had a different approach, and within these there were different plans designed by each worker. This lack of standardisation is a concern in itself.

Another feature identified was a belief that an up-to-date assessment was a means of "covering your back", although failure to carry out any part of the risk plan could still be deemed negligent. This raises a question about whether staff would choose not to develop plans at all for fear of being unable to carry them out later. This was compounded by the fact that some staff felt that assessments and plans were out of date as soon as they were completed.

It is obvious that major problems exist in relation to staff attitudes to the whole process of risk assessment, with many feeling that the process is used mainly to keep tabs on clients who have a history of violence. Should we not be as concerned about the service we offer to all our clients, no matter how vulnerable they are?

Another point of concern is clients' human rights. Many teams did not discuss their risk assessment with the client. This is not only a major flaw in building a therapeutic relationship with the client, but also denied clients a right to know how their carers viewed their behaviour. Most risk assessments are formulated via a process of "Chinese whispers" and allowing the client to tell you their version of events may lead to more enlightened practices. 

Recommendations

- All risk assessments should be multi-disciplinary. 

- Regular reviews of all assessments.

- Assessments should be carried out in co-operation with client and carer.

- Implementation should not be the responsibility of one worker. 

- Management should ensure that all staff are aware of the risks posed by clients.

- Record-keeping of risk should not just consist of ticking boxes.

Lance Carver is an approved social worker and Mike Lehane is a community psychiatric nurse in a community mental health team in south Wales.

References

¹ R Hare, Manual for the Hare Psycopathy Checklist - Revised, Multi-health Sysytems, 1991

² G Andrews and R Jenkins, Managing Mental Disorders, Volume 2, World Health Organisations, 2000

³ R Grounds, " Risk management in clinical context" in J Crichton (ed) Psychiatric Patient Violence - Risk and Response, Duckworth, 1995