It probably means that you take the statistical evidence produced by massive double-blinded placebo-controlled randomized clinical trials as actual evidence

That is not really an evidence unless you yourself eliminate any biases or flaws in the trial methodology.

Even trained professionals fail to do that regularly...