Dr Maden’s editorial (Psychiatric Bulletin, April 2005, 29, 121-122) in response to the article by Higgins et al (Psychiatric Bulletin, April 2005, 29, 131-133) neglects a fundamental aspect of risk assessment and his recommendations therefore need to be treated with caution. Higgins et al refer to the different contexts in which risk assessment forms were being used, but Dr Maden does not appear to have taken this fully into account in advocating the routine use of the Historical/Clinical Risk-20 (HCR-20) in adult general psychiatry.

There is no doubt that the HCR-20 is a useful tool in forensic settings, where it is already widely used. However, there would be significant time and cost implications to introducing it routinely into general adult settings, and it could be argued that this would not be sensible or cost-effective. Although Dr Maden starts by advocating ‘a more systemic approach in marginal cases’ which it is hard to disagree with, his advocacy of the routine use of the ‘ideal’ HCR-20 does not appear to address whether this is a feasible option with non-marginal cases. The HCR-20 can be very effective in supporting teams in assessing and reviewing ongoing risks, but it is not clear that it is a practical solution to supporting, for example, the risk assessment and management decisions of a junior psychiatrist doing an assessment of a new patient in an accident and emergency department in the middle of the night.

Dr Maden notes that Higgins et al reported that many consultants did not attend the (presumably free) half-day training on violence risk assessment already on offer within their services. To suggest 3-day external fee-based training specifically for violence risk assessment seems a disproportionate response when other priorities in mental health (for example assessing the risk of self-harm) are not identified as requiring such expensive formalised training.

It is a pity that neither Higgins et al nor Dr Maden were able to expand on the possible utility of the CARDS project (Reference Watts, Bindman and SladeWatts et al, 2004) that the study of Higgins et al was part of. This appears to be a worthwhile collaborative attempt at developing a more standardised approach to risk assessment in general adult psychiatry. It is also free and potentially more easily integrated into routine clinical practice than the wholesale use of the HCR-20.