@markhoro@awaisaftab@jack_goulder We can’t really say whether drug is moving everyone in Area C to Area A or some from C to B and others from B to A, so the total number affected could be anywhere between 11.5% improving by an average of 15.2 points and 23% improving by an average of 7.6 over placebo
@markhoro@awaisaftab@jack_goulder Area B is where the drug and placebo response distributions overlap. A and C are where they differ, accounting for about 11.5% of their respective distributions. The mean improvement in Area A is 16.9 points and in Area C it is 1.71.
@markhoro@awaisaftab@jack_goulder We also point out in the paper that the effect size does not change with study length. And even if the effect were short-term, that is different from your claim of no effect
@markhoro@awaisaftab@jack_goulder The studies go up to 16 weeks and the response patterns are very different from "a little bit high" or "numbing their emotions". The difference is in the proportion of patients that show considerable improvement: equivalent to an average of 16 points or so on the HAMD17
@markhoro@awaisaftab@jack_goulder What my paper shows is that the response patterns are very different from what we would reasonably expect to see from unblinding: a difference in the number of modes of response or different locations. The datan show neither
@markhoro@awaisaftab@jack_goulder Your unblinding argument in unfalsifiable. If 100% olf patients on drug went into complete remission and no one on placebo improved at all, you could claim the result was due to unblinding
@markhoro@awaisaftab@jack_goulder If 50% of people respond to drug and 20% respond to placebo, you can't say that the drug isn't effective because you can't say which of the 50%are experiencing placebo effect and which are experiencing placebo effect.
@markhoro@awaisaftab@jack_goulder You have responded and I have answered you. First is it unnecessary to show who those people are to demonstrate the effect.
@markhoro@awaisaftab@jack_goulder Your serotonin argument is based mostly on absence of evidence. Your principal basis for evidence of absence is your disingenuous interpretation of the drug-placebo difference
@markhoro@awaisaftab@jack_goulder Mark-
You conveniently ignore my paper, with which you are highly familiar, to repeat points that my paper has discredited. You continue to fallaciously treat the mean difference between drug and placebo as the modal response.
@mgoldingmd 2. There is not much reason to believe that patients who do not respond to one AD will respond to a different one. RCTs of patients who failed to respond to a given drug who are then blindly randomized to remain or the drug or switch to another do no better with the new drug