> I can prove it is dangerous with studies, but to prove it isn't dangerous seems like an impossible task. You don't see that you run the same damn studies but

> I can prove it is dangerous with studies, but to prove it isn't dangerous seems like an impossible task. You don't see that you run the same damn studies but get a negative effect? You can even do it solely off p-value (but instead should do a more complex calculation based on the harms discovered, etc.) If your "is it dangerous" study shows it is with p < 0.05 = dangerous, p > 0.10 = safe, 0.05 <= p <= 0.10 = more study needed.