In the same spirit as the Placebo story, Ben Goldacre reports on a significant breakage in the way we run clinical trials. NB this doesn’t mean that the theoretical construct is broken, merely that we are human in its implementation and let other factors intervene (bigger labs and faster cars, as I used to paraphrase Feyerabend). This is important: we need to get all trials registered and monitored so that we can believe the outcomes.