Then it is a good thing you have no say in our access to psychiatric pharmaceuticals. The only way known to distinguish one variety of depression from at least

Then it is a good thing you have no say in our access to psychiatric pharmaceuticals. The only way known to distinguish one variety of depression from at least a half-dozen others is to see which medication relieves it. So, any RCT of a depression treatment necessarily throws together people with a half-dozen or more independent conditions, some of which might be relieved by the treatment, others worsened. Repeat after me: RCT results depend absolutely on precise and correct diagnosis. If you cannot diagnose, you cannot design a valid RCT. Invalid RCTs are easy, though; people run those all the time. The RCTs that seem to prove psychiatric drugs are ineffectual are exactly those. We are in the condition of people getting arm and leg fractures, but we can't tell where. Splinting is seen to help; doctors who put splints on various places usually hit on a place where it manifestly helps. But our RCTs of splints placed on left shinbones seem to help only a precious few of our patients. Some people swear up and down, loudly and repeatedly, that they are dead certain this means splints are useless and a big scam.