Skip to content Skip to navigation

When Doctors Quit

May 30, 2015
| Reprints

As I was thinking about this blog, a friend sent an opinion piece to me from The Washington Post entitled, “Why doctors quit,” by Charles Krauthammer. It was well written, succinct, and accurate. Like any writing, including this blog, it was necessarily incomplete. What intrigued me most was this sentence from the column:

“You may have zero sympathy for doctors, but think about the extraordinary loss to society — and maybe to you, one day — of driving away 40 years of irreplaceable clinical experience.”

It was the “irreplaceable clinical experience” that I found sufficiently incomplete as to warrant this post. The gist of it is that when you need a competent doctor, you cannot backfill that need with Google search, IBM Watson, a professional association’s clinical practice guideline, or the potentially relevant information sources that might have a greater economic interest in presenting the information we’re seeking in a biased way.  Clinical experience is a unique juggling act between genuine empathy, diagnosis, treatment, and communication. “Why doctors quit” alludes to this, albeit from a certain partisan posture. My concern is more focused on “When doctors quit” in terms of juggling that act, and specifically what irreplaceable value we’re talking about.

As quick background, I am huge fan of consumer convenience, so I am big fan of the advances being made by retail chains with telemedicine. There’s a great example and brief video discussion of these advances with Walgreen’s CMO, Harry Leider, M.D. here. No one viewing this thinks it is a complete substitute for a face-to-face, in-person interaction with a doctor when there are important benefits versus risks to arriving at the right decisions on diagnosis, treatment, or both.

What I want to share here is that there are four approaches people, including doctors, use to make decisions regarding what to do. These decisions are progressively sophisticated from 1) simple rules, 2) cumulative inference, 3) serial logic, and 4) parallel weighing of the resulting, often complex alternatives. Physician’s with irreplaceable clinical experience use all four approaches. It’s impossible to fully automate or otherwise replace the fourth approach, and silly to try. 

I think Krauthammer’s point is that for political and economic reasons, we are throwing the baby out with the bath water. And my point is that to make the right policy and economic decisions, we must recognize and value this very real, four-level consultative ability. Maybe the right word is simply “judgment.” That is what we risk losing when physicians quit.

What do you think?






Joe, another on point essay from you. I agree with four skills being used and the importance of the fourth. I will add a fifth, which is compassion. The physical presence of a physician adds to the healing process - the laying on of hands as well as the power of looking into your patient's eyes and communicating compassion. Physicians who are happy with their work seem to be working in systems where they are a team providing excellent care, and where they are valued by that system.

(Margaret Cary MD MBA MPH)

Maggi, I'm glad I spent a day pondering your comment before replying. I was in agreement initially. Now, however, I'd move compassion to number one and put the other more cerebral "skills" into the latter positions in terms of importance. It's probably more appropriate to cast all five skills as "talents", the distinction being that skills and experience are generally trainable, where as talents have more of an innate quality that can be improved with training, but generally not created. I'm using the word talent the way StrengthsFinder uses strengths.

It's critically important to have compassionate advocates to navigate the health system. Although the original post focused on the decision process talent, 'parallel weighing' which is uncommon in the general population, compassion is essential and a prerequisite.

Thanks for your comment.

Joe - clearly there is a growing problem and you've focused in on key areas that need attention. As the care delivery model continues to change, your points must be kept in mind otherwise the consequences become worse. Jim

Jim - Thanks for your comment. I know that you've led the successful development of several solutions in the computer-assisted coding space. These solutions were designed to help physicians by automating the tedious albeit important chart review necessary to compassionately and comprehensively decide what a patient needs, both clinically, and for administrative documentation. I'm currently helping do this to assess risk, in the sense of HCC risk adjustment. That is what led me to study how physicians actually make decisions. By extension, it has clarified for me what we are at risk of losing when we replace the mechanics of information processing through NLP, without considering the larger process. Thanks again, Jim.


[ a friend who wished to remain anonymous added this: ]

What is irreplaceable in a fine doctor (and too rare):

1) The ability to understand other minds, from those with intellectual disabilities to professionals--their goals, preferences, and capabilities

2) The ability to compare a) complex, fuzzy care needs with b) complex, fuzzy evidence of widely varying freedom from bias and c) patient goals, preferences, and capabilities and identify optimal (but often not ideal) diagnoses and action plans.

3) The ability to convey as much of 2) as feasible to the patient (and lay caregivers) and create a care plan together.

4) Monitor the care plan in light of changes in 2) and repeat 3, when needed.