[MauranKilom]

> Personal and eyewitness testimonies count for nothing now?

Can't see your original comment, but assuming this refers to side effects with sub-percent occurrences, then the answer is "no, random testimonies have no statistical significance without a base rate to compare to".

You could put hundreds of people into the same room whose testimonies show that eating bread is deadly - you just have to find enough celiacs. That doesn't make the baking industry a big conspiracy.