We read with interest the paper regarding the cost-effectiveness of intensive v. standard case management for severe psychotic illness (UK700 Group, 2000).

We feel very strongly that more comment should have been made on the topic of training (or lack of training) of the case managers involved. The findings of this large, well-designed trial are very similar to those of Muijen et al (Reference Muijen, Cooney and Strathdee1994), who found no difference in outcome between the use of community psychiatric nurses (CPNs) configured into case management teams v. CPNs working in a generic way. The main implication of this study was that merely reconfiguring services into different working arrangements provides no additional benefits to patients. However, it could be argued that benefits may accrue if training in research-based interventions is provided. Indeed, such training for CPNs and other health care professionals has been developed in the Thorn and similar programmes, which focus specifically on providing skills in assertive community treatment, family interventions, psychological interventions etc. One could argue that these groups of skills, which comprise what is loosely known as psychosocial interventions, are essential to effective case management.

In the UK700 study, we note that the case managers received a 2-day induction course in case management and an unspecified amount of instruction in outreach practice given by a team leader in the assertive community treatment service from Boulder, Colorado. It seems to us that such training input is insufficient to provide the skills necessary to deliver truly effective psychosocial care. (The Thorn programme comprises 36 days of direct training plus the equivalent of 50 days of further study and project work.) We are therefore not surprised that the case managers with smaller case-load sizes could not improve on the outcomes attained by those working with the more average size case-loads.

Surely studies of training per se are now needed, with random allocation of case managers to training in research-based interventions or to standard practice, and measurement of outcomes for both the trainees (in terms of skills acquisition and knowledge gain) and their patients (in terms of clinical outcomes).

We are at present spending enormous amounts of money on training throughout the National Health Service and yet the vast majority of this training remains completely unevaluated. Although randomised controlled trials of training interventions will be costly, the price of not knowing whether training makes a difference is much greater.