That's not true at all. The way the data is analyzed could be done differently for example- there are some cases where previous clinical trials were brought bac

That's not true at all. The way the data is analyzed could be done differently for example- there are some cases where previous clinical trials were brought back up as possibly helping a subset of the individuals, but got thrown away because of p-values. When our disease labels are so bad this is the type of thing that needs to be especially carefully considered. But also I was talking about basic neuroscience. The pharma companies don't just magically start a clinical trial, as you said. They base the clinical trials on previous results from academia and their own basic research labs. Which mostly doesn't involve humans at all. Perhaps those labs are not putting forward the best candidates, or not doing as good a job as they could be at eliminating candidates.