According to Kirsch, antidepressant trials were also compromised by side-effect leakage: the drugs had side effects while the placebos did not, so patients knew when they were not taking a placebo. That's the equivalent of putting your finger on the scale. As far as I know, there has been no attempt to measure that distorting effect (Kirsch only mentions it), nor any clinical trials of antidepressants conducted on a level playing field, where the placebo had similar side effects to the drug. But I haven't searched, and if such studies exist, I'd welcome correction.
But keep in mind that ECT is only for crisis-level severe depression:
> ECT uses an electrical stimulus to the brain to induce seizures. It is prescribed for patients with crisis-level severe depression – who are catatonic (people who are so slowed down that they stop moving, talking and eating) or suicidal – or for patients with major depression who have not responded to medication.
Except in severe depression, antidepressants have no clinical advantage over placebos:
http://www.thedailybeast.com/newsweek/2010/01/28/the-depress...
According to Kirsch, antidepressant trials were also compromised by side-effect leakage: the drugs had side effects while the placebos did not, so patients knew when they were not taking a placebo. That's the equivalent of putting your finger on the scale. As far as I know, there has been no attempt to measure that distorting effect (Kirsch only mentions it), nor any clinical trials of antidepressants conducted on a level playing field, where the placebo had similar side effects to the drug. But I haven't searched, and if such studies exist, I'd welcome correction.