Sir: I read with interest the study by Harrison et al (Psychiatric Bulletin, August 2001, 25, 310-313) about which patients are suitable for a home treatment service. The authors are right to comment that little has been written about the type of patient suitable for this approach. However, this question may be premature, in that there is little agreement about what ‘the approach’ actually involves. Their own particular model is described as a hybrid between day hospital and home treatment, which is rather an unusual configuration for home treatment services.

The authors write as if there is a strong evidence-based rationale for the development of acute home treatment services. Despite their incorporation in the National Service Framework for Mental Health, there is not. Their referral to “key components of the model” is highly questionable. To borrow from Burns (Reference Burns2000), how do they know that any of these features are necessary or make a difference? The so-called ‘model’ has not been adequately defined. At present, the terminology of crisis intervention, or home treatment as it is otherwise known, is inadequate and confusing and prevents adequate conclusions being formed. What is needed is a well-defined model with tightly defined components followed by extensive testing of model fidelity. What the authors are reporting on here is the characteristics of those who were selected for their particular hybrid version of home treatment and day hospital, which limits the external validity of the research.