> There are many problems with clinical trials. Read Ben Goldacre for examples Yes, and the statisticians in the FDA and the drug companies have a lot of exper

> There are many problems with clinical trials. Read Ben Goldacre for examples Yes, and the statisticians in the FDA and the drug companies have a lot of experience in this. I think they started looking at ways to end trials early back in the ADT (anti-HIV drug) days. It was partly due to popular pressure - people with AIDS didn't want to wait for the clinical trials. Lots of big angry protests. And the cat and mouse game between regulators and big pharma cooking the books is also pretty mature. There's problems with the system, but bad statistics is one of the smaller ones. The real problem is marginally effective drugs being used instead of better patent-expired ones, because the less effective (but not truly bad - they beat placebo) ones have bigger marketing budgets.