I think the way some trials are conducted is worrying. Of course there is the points you bring up; I had also heard (but can't find any link from a bit of a cu

I think the way some trials are conducted is worrying. Of course there is the points you bring up; I had also heard (but can't find any link from a bit of a cursory look around the net) that all subsequent trials of antidepressants show decreased treatment effect until you end up with meta-analyses that seem to show that they are no better than placebo, leading (jokingly) to Doctors, on the release of a new anti-depressant, saying 'Quick, let's use it before it stops working' Part of the problem I think is the model of the pharma industry (I have no suggestions on where to start to change this!), part is that students/researchers don't want to give up on publication and years of work with null results and so change the aims or methodology, or rig their study... Which is morally indefensible but understandable at the same time.. I think a lot of people involved with medicine are under no illusions as to the limitations of our current methods. However one of the beautiful things about the scientific method is that we are continually testing our preconceived ideas, and trying to tease out if the effects are real or some artefact of bias, statistics, population dynamics, etc. When I started Med School they tell us that 'Within 4 years, 50% of what you have learnt will be wrong and replaced. The difficulty is which 50%'. THAT is how quickly our knowledge is growing, and goes a long way to show that the people involved in the delivery of healthcare don't take this all for granted.