> You'd have to prove that increasing the risk of doctor infection would lead to a lower rate of patient infection.
The point isn't to infect doctors. The point is to make them (potentially) subject to the tools they're about to use as a way to ensure that they are sufficiently involved in designing and overseeing the systems that ultimately move the needle on patient outcomes.
When things are done sufficiently well these kinds of steps aren't necessary. We don't make the engineering team that designs a bridge walk/drive across it first because bridges have an exceedingly low failure rate and the failures that do happen aren't infant mortality (you designed it wrong) they're old (you maintained it wrong).
