Placebo response in depression

An unstated conclusion to Gavin Andrews' editorial (2001) is surely that placebo controlled trials are absolutely essential to our understanding of the true effects of antidepressants. Without placebo trials Andrews' main conclusion that the placebo effect is significant and worth potentiating would not be possible.

It is not sufficient to prove that new treatments are better than or equivalent to existing treatments because we do not know that the existing treatment is still better than ‘placebo’ treatments. Today's ‘placebo’ treatment may not be the same as that of 10 or 20 years ago when the original placebo trials were done. Further, there may be considerable differences between groups with the same diagnosis. This is all well demonstrated in the study of tricyclic antidepressants in children. Since it was thought unnecessary and unethical to do placebo trials in children and adolescents, new antidepressants were tested only against existing ones and found to be effective in 50 to 70 per cent of cases. Only after 20 or so years of such trials were placebo trials done and the ‘placebo’ treatment (probably the accompanying environmental, individual and family treatment) was found to be just as effective as the drug. In this time numerous children were treated unnecessarily with tricyclic antidepressants and several may have died from cardiac arrhythmia. This was not an ethical way to introduce new drugs.

Among additional reasons for placebo controlled trials are first, that far more people have to take part in a trial comparing a new treatment with an existing treatment because the difference in effect is much less than with placebo. Thus, more people will be exposed to a new treatment with unknown side-effects. Second, placebo controlled trials are the only way to get accurate knowledge of side-effects: essential information for clinicians.

Thus, the statement by Andrews that “the existence of proven treatments would normally render placebo trials unethical” is unwarranted. I believe it is unethical not to use placebo controlled trials even when there is a proven therapeutic method (since no method is perfect), so long as there can be no lasting harm from delaying treatment and the subjects fully understand the risks and voluntarily consent. I urge researchers and clinicians to press the World Medical Association to modify the latest version of the Declaration of Helsinki (World Medical Association, 2000), which contains this restriction on placebo controlled trials.