McKenna (Reference McKenna2003) comments that Sensky et al (Reference Sensky, Turkington and Kingdon2000), in their trial of cognitive–behavioural therapy (CBT) v. befriending for the treatment of schizophrenia, found no advantage of CBT over befriending at the end of the 9-month intervention period. In his view, they were therefore not justified in making the claim that CBT is effective in treating negative as well as positive symptoms in schizophrenia. This assertion fails to recognise the different mechanisms by which CBT and drugs may benefit psychotic symptoms. While drugs are likely to produce a (relatively) immediate effect in altering neurotransmitter pathways, CBT (as is the case with other psychological therapies) is postulated to alter attachment-related memory (Reference GabbardGabbard, 2000) and develop an understanding of the illness. Cognitive–behavioural therapy utilises skills which, if successful, can be maintained by the patient long after therapy has ended. This would explain why Sensky et al (Reference Sensky, Turkington and Kingdon2000) witnessed a non-significant difference between the control and intervention groups at the end of the intervention period but a significant continued improvement in those receiving CBT (and not in those receiving befriending) at 9-month follow-up. It would not be expected that drugs would maintain a benefit 9 months after being stopped. Preliminary results of a 5-year follow-up of the cohort of patients in this study indicate that these gains in the CBT group have been maintained (D. Turkington, personal communication, 2001).

As a result of the distrust of psychological approaches, studies of CBT (e.g. Reference Kuipers, Garety and FowlerKuipers et al, 1997; Reference Sensky, Turkington and KingdonSensky et al, 2000) have invariably recruited patients whose symptoms are ‘resistant’ to medication. The fact that these studies have still shown significant improvement over either a control intervention or routine care is testament to the greater benefits that might be demonstrated if the patients enrolled in research were representative of those in clinical practice targeted for psychological intervention.

In any case, surely the question is not which is more effective, but how both pharmacological and psychological approaches could be combined for greatest effect.