Selective serotonin reuptake inhibitors

Editor—Geddes and Cipriani fear that regulatory bodies overrate the importance of placebo controlled trials.¹ We disagree. The placebo controlled randomised trial is still the best test of efficacy available, although several improvements in conventional methodology are necessary to prevent selection biases.² However, selection bias owing to a placebo washout period has not yet been tackled.

Although no doctor uses a placebo washout in real life practice, most antidepressant trials use such a period before randomisation. To demonstrate their effect: say, 200 patients are available, and 20 placebo responders are excluded before randomisation. The remaining 180 patients are then randomised to treatment with antidepressant or placebo. However, as it is not known whether all placebo responders in the antidepressant arm would have recovered if they had been given antidepressants, the direction of this differential selection may favour antidepressants.

Other necessary methodological improvements are: more use of “active” placebos to allow better blinding as they contain substances that mimic the specific side effects of antidepressants³; better concealment of allocation;⁴ independent outcome assessment⁴; independent funding⁴; inclusion of relevant groups such as general practice patients and elderly people; and the use of the preferred more conservative intention to treat analyses.⁵

Most known biases inflate the effects of antidepressants. Intention to treat analyses would, for example, decrease the risk difference between antidepressant drugs and placebo from 28% to 19%.² A large, independently funded, good quality general practice trial of antidepressants versus placebo is much needed now, without placebo washouts.