We thank Dr Slade for his interest in our paper and are grateful for the chance to clarify these crucial points. At the time this study was carried out D.K. was a general psychiatrist working full-time in the community. He did have a general cognitive therapy training and applied the techniques that he had learned for the treatment of depression and anxiety in an adapted form to his patients with schizophrenia (Reference Kingdon and TurkingtonKingdon & Turkington, 1991). We then designed this study to test the efficacy of these techniques as against a befriending control (Reference Kingdon, Turkington and CollisKingdon et al, 1989). The study was then carried out in the course of D.K.'s clinical work.

We accept that a shared formulation is a fundamental component of cognitive therapy with schizophrenia (Reference Fowler, Garety and KuipersFowler et al, 1995), which not only helps to direct the process of therapy but which can also help to predict the emergence of depression as a delusion recedes. The identification of key maladaptive core beliefs is part of the process of formulation and helps when there are blocks in progress and in the prevention of relapse. Psychiatrists do not normally identify such core beliefs in their case formulations and may not share all of the formulation with the patient. However, the formulations of psychiatrists do contain an aetiological component comprising such issues as genetic predisposition, birth traumas, early losses and personality. The precipitation of the psychosis in relation to any pertinent life events is included, as are maintaining factors such as isolation, poor adherence or high expressed emotion within the family (Reference Gelder, Gath and MayouGelder et al, 1983). Formulation is so central to psychiatric practice that it is a key component of the Royal College of Psychiatrists' membership examination. With such formulations psychiatrists can safely use cognitive-behavioural techniques as long as they remain aware of the risk of emerging depression or increased suicidal ideation linked to improved insight. Such phenomena are widely recognised in psychiatry when certain types of symptoms (e.g. grandiose or systematised delusions) respond to antipsychotic medication.

The purpose of our paper was to attempt to shift attitudes in psychiatry in order that we can be better engaged with our patients who have psychoses by working more directly with their symptoms rather than simply monitoring them and titrating antipsychotic medication. If general psychiatrists can make this shift in attitude, training workshops and supervision will be necessary. We expect that this shift could well facilitate the delivery of formal cognitive therapy by other community mental health teams trained in this approach and help with the implementation of other psychosocial interventions in this patient population.