A recent blog posting by Dr Robert Centor, see here, introduced me to the term "naturalistic decision making" which seems to be one aspect of the broader area of the study of cognitive reasoning, or how we make decisions. In this regard one focus is on decisions in which there are high stakes, time pressure and complex situations as is common in emergent clinical situations.
Basically faced with that type of situation,such as the challenging chest pain case described in Dr. Centor's article, expert clinicians typically rapidly categorize the situation based on a pattern recognition ( as described by Kahneman as a System 1,fast and unconscious mental act) and then move on to use a deliberate,analytic System 2 approach involving,in part, a search for missing data and for discrepancies and then a simulation of what might occur next if the first plan based on the first impression were carried out.
The Nobel prize winning work of Kahneman and others working to elucidate how people think in have apparently fleshed out possible mechanisms of some of what Aristotle referred to as Phronesis.
Aristotle spoke of the virtues of the mind as
sophia (wisdom of first principles),
techne (technical knowledge)
nous (intuition) and
(practical wisdom or prudence).
medicine with its emphatic focus on techne and episteme may foster
the illusion that good clinical practice can be encapsulated in
guidelines and pathways and that quality lies in obedience to them but
it is the blending of those elements of science and technique with
practical wisdom that makes good practice. Treatment guidelines or algorithms come into play only after a diagnosis has been reached and in all but the most trivial cases a bit of phronesis might help.
In Kathryn Montgomery's 2006 book, How Doctors Think, we find the following quote which sounds a lot like what we learn from the field of Naturalistic decision making:
"Clinical judgment done well is the
intuitive and iterative negotiation of the patient's narrative of
illness...This focus come with experience"
And for that I don't believe there is an algorithm.