Most people haven't evolved from blind trust in the shaman. The Placebo effect is a real phenomenon; someone believes something will cure their affliction, so it will. The real magic was in their confidence in recovery. This means there's little difference between a Roman pontiff sacrificing a chicken and muttering in Latin (reading entrails), and a man in a white lab coat sacrificing a rat and muttering in Latin. 80% of medical studies can't be replicated, and it gets worse in any long term study because conditions are seldom repeatable, and that's before fraud is even considered. Fraud itself is a massive problem, and peer review largely serves to cover it up. But, because of the magic of "science", which is largely restricted to short repeatable experiments, people transfer their trust into areas that are wildly inappropriate. "Magic man make computer work, maybe magic man cure illness too."