> a physician can not limit himself to protocols. he has to know where and when to apply them.
I discussed this with mathematicians who analyze how doctors work. Main view is that (great) doctors do not follow protocols and general patterns, but they implicitly find in their memory a specific case that was similar to the current one in as many characteristics as possible and use lessons learned in that case. This is, of course, not working with a novel disease and a lot of disasters are happening while following generic protocols until doctors accumulate enough knowledge.