Ethics of placebo-controlled trials in depression
Some believe that introducing a placebo control into an antidepressant drug trial means that ill patients are being unnecessarily deprived of treatment. As a result, many clinicians and ethical committees are opposed to the use of such trial methodology. A contrary view i s that it is unethical nor to use a placebo control.
Depressive disorders are not clear-cut diseases of known aetiology. Any cohort of depressed patients, however carefully selected, is likely to be heterogeneous, and clinical and research findings in one population do not necessarily apply to another. Furthermore, one team of investigators may have different diagnostic criteria than those of other researchers in spite of the attempts that have been made to keep these to a minimum. Many depressed patients are unresponsive to antidepressants, and when the drugs appear to help, they may d o so as an adjunct to overall clinical management; rarely d o they seem to be the complete answer to a patient's problem.
Trials that show a statistically significant improvement in depression during treatment with a drug do not prove that the drug has antidepressant properties, because such an improvement can result from spontaneous remission or coincidental factors, including contact with a caring person and the placebo response. Whatever the reason for improvement, the drug can usurp the credit that should rightly be given to nonspecific psychotherapeutic aspects of patient care. Open trials therefore cannot prove that a drug necessarily has pharmacological antidepressant properties. Controlled trials in which a new drug is compared with a conventional tricyclic antidepressant may simply be demonstrating comparable non-specific responses, and the results of many older trials could be invalid because power statistics were not in widespread use at the time they were carried out.
Evidence for pharmacological effectiveness can only come from placebo-controlled trials. Most trials have