Placebo-Control Groups in the Study of Outpatient Depressives

To the Editor.—  The article by Murphy et al¹ on cognitive therapy and pharmacotherapy is timely and important. However, the group that was treated with tricyclic antidepressants (TCAs) may have contained a high proportion of placebo responders. This would compromise the usefulness of the TCA group as a relevant standard.Murphy et al contrasted four treatments (cognitive therapy, TCAs, cognitive therapy and TCAs, and cognitive therapy and placebo). There was no initial placebo washout period nor was there a parallel placebo group. The efficacies of the treatments were indistinguishable. Depending on the criteria, 44% to 56% of the TCA group were successfully treated.We suspect that a majority of these patients were placebo responders for the following reasons: Fig 2 indicates that the TCA-treated group had approximately a 12-point improvement in their Hamilton Rating Scale for Depression (HAM-D) scores during the first six weeks; two thirds of this improvement