The meticulous report of John Dowsett on ‘Measurement of risk by a community forensic mental health team’ (Psychiatric Bulletin, January 2005, 29, 9-12) illustrates nicely the use of a standard risk assessment instrument, the HCR-20, in an inner city context. The high scores on the historical scale of the eight (out of 47) patients who re-offended in the 2.5 years following data collection very much reflect that adage of forensic psychiatrists, that previous violence is the core predictor of future violence.

However, although harbouring doubts about the limits of risk assessment (e.g. Szmukler, G., ‘Homicide inquiries: What sense do they make?’, Psychiatric Bulletin, January 2000, 24, 6-10) we consider that pragmatic reviews, such as this study, more importantly call into question the role of a forensic community team. For reasons of history, resource limitations and serendipity, in City and Hackney (an equivalent inner-city area) we have no such agency, restricted patients being routinely handed over to the community mental health teams. These do have an integrated forensic community psychiatric nurse, but the forensic/general psychiatry interface is of the simplest indoor/outdoor type. Dowsett’s report that there are a number of patients in his team who have remained ‘stable for some years’ (and therefore ‘could perhaps be handed back to generic services were it not for the fact that they committed a very serious offence’) certainly reflects part of our own experience with restricted individuals. They are often easier to manage than many ‘ non-forensic’ patients, because the nature of the restriction order and their history of institutionalisation generates therapeutic and social control.

Likewise, another group of Dowsett’s patients are noted to be perfectly manageable on ordinary acute wards, and again he considers that there would be advantages, in terms of quick admission, were they to be managed by a local generic service. His third group, namely those with what might be termed ‘historically established criminality’ also do not benefit from a ‘forensic’ team, since there is no specific psychological intervention known to have an impact. In which case, why have an expensive resource ‘looking after’ such individuals, with no evidence of benefit given that criminality, per se, is not a treatable disorder. The cynic might even suggest that maintenance of a stable mental state in such a group enhances their likelihood of offending.

If such findings reflect the case-load of forensic mental health teams elsewhere, and anecdotal reports very much suggest this is the case, then is there not an urgent need to rethink the notion of a separate forensic capacity? As Dowsett has pointed out, it is important for forensic services to ‘ demonstrate expertise in managing’ this criminal group, but use of the HCR-20 is not especially difficult and reintegration with generic community mental health teams would perhaps be a much better option. It would help break down the often difficult interface issues of parallel teams, would enhance resources for those who do benefit from psychiatric input and would put the issue of risk exactly where it belongs, at the heart of all routine clinical practice. In its current specialist location it sustains the de-skilling and as yet unproven notion that an expert elsewhere may be able to manage risk more effectively. It might even enable psychiatry to withdraw from its untenable and exposed position that has made us the whipping boys of the government’s public safety agenda.