There are many cognitive–behavioural therapies, with the precise form of the CBT shaped to the nature and context of the particular disorder. So, for example, CBT for panic disorder and CBT for bulimia nervosa (Reference Hawton, Salkovskis and KirkHawton et al, 1989), although sharing a basic orientation and broad therapeutic principles, are very different from one another. The form of CBT in which we were interested had as its principal focus the mother–infant relationship and aspects of infant management. The reason for this was quite clear. It is well established that many forms of treatment for post-partum depression, including counselling (Reference Holden, Sagovsky and CoxHolden et al, 1989), interpersonal psychotherapy (Reference O'Hara, Stuart and GormonO’Hara et al, 2000), ‘cognitive–behavioural counselling’ (Reference Appleby, Warner and WhittonAppleby et al, 1997) and fluoxetine (Reference Appleby, Warner and WhittonAppleby et al, 1997), have significant anti-depressant effects, but it has not been established that any of these interventions has an impact on the quality of the mother–infant relationship and child developmental progress, both known to be compromised in the context of post-partum depression. (The evidence for the efficacy of CBT in this context is, incidentally, less certain. Indeed, none of the three studies cited by Professor McGrath and colleagues in support of this form of treatment delivered an orthodox CBT; and one, in fact, was not a study of CBT at all, but of non-directive counselling.) We were interested in determining whether treatment that addressed the maternal role, as part of a wider supportive therapeutic relationship, would have wider benefits. The form of CBT we investigated was shaped by these concerns, and the discussion refers explicitly to this treatment and is, therefore, wholly apposite.

In several respects the findings of our trial were not what we had expected and were, to us, disappointing. However, the data were what they were, and it was our job to try to understand them. When the first trials comparing CBT with interpersonal psychotherapy for major depression were published in the 1980s, British clinical psychology reverberated with the chunterings of the CBT faithful whose instinctive reaction to the equivalence conclusion was to query the probity of the trial CBT therapists. With time, a more mature position was evolved. The findings of our study, along with the broad failure of the trials of preventive treatments for post-partum depression, would seem to us to be cause for pause and reflection, rather than instinctive defensiveness.