The cognitive part of the practice of medicine spans between two extremes, from registering and recognizing the most minute nuances of human and biological expression to seeing the overarching big picture of complex constellations of details.
Like the arts of painting and photography, it requires us to see both the unique and the universal in the most ordinary manifestations of everyday human life. But instead of capturing with imaging tools what we see and perceive, we turn those impressions and observations into the understanding and interventions we call diagnosis and treatment.
The art of medicine involves both technical mastery of treatment and the carefully honed ability to register, analyze and evaluate a vast array of what we might call data. While there is still respect for masterful treatment, perhaps especially when it is of a technical or procedural nature, there is growing disdain and disrespect for what we used to call clinical judgement. In picking stocks, web design, marketing and many other human endeavors, experience seems to have retained or even expanded its value, but in medicine it is often downplayed or even ridiculed.
Non-physician healthcare thinkers have evangelically big hopes for “data”, entered into medical office computers by fallible, disillusioned (think “Meaningful Use”) and distracted humans and new generations of “connected” medical instruments, and analyzed by centralized computers at Medicare, research institutions or big insurance companies. The vision is that more data will unlock the hidden potential for economies of scale and unseen patterns of disease, and generate vastly improved accuracy and efficiency of diagnosis and treatment.
But more data doesn’t always lead to better insights. Borrowing from other arts, a well-written poem can sometimes convey to the reader as much as a novel. And the weight of each piece of “data” isn’t the same to an experienced physician as it is to a computer. Humans in the healing arts can do the work of a recording device, a lie detector, a microprocessor, a translator, a judge, a pastor and a teacher.
In real therapeutic encounters, the agenda is not always the stated one, the given history isn’t always accurate and the clinical exam isn’t always typical or even relevant; sometimes the physician gives more weight to the unspoken clues in a case, something a computer isn’t likely to do.
Even our definition of disease lends itself poorly to interpretation and intervention guided by a computer. Examples are obesity, diabetes and chronic back pain. The measurable parameters of these conditions, biometrics, average laboratory values or pain rating scales, tell little about what role the disease plays in the patient’s life. And, unlike routine cases of pneumonia or step throat, sometimes the disease defines the person across a whole lifetime, and takes on archetypal meaning. Just talking numbers isn’t likely to change the manifestations of such conditions. Only going to the depths of the subconscious can alter the trajectory in most such cases. It would be naive to think that computers can do anything for such patients. Only a human with considerable skill and wisdom can penetrate the layers surrounding the core of these conditions.
The art of medicine is making the connections on a personal, case-by-case level with the archetypes that most of us relate to on some level, but which almost never exist in the physical realm. But they exist in the inner lives of all of us, as heroes and villains, as our inner children, older selves, and better or worse incarnations of our own spirits.
Classic disease presentations are like archetypes; we look for them all the time, and we think we see glimpses of them, but we seldom see the true personification of them.
And, the most important archetypes of all in the realm of medicine, the Patient and the Healer, hover in the air above us in every clinic room, hospital ward and nursing home. Patients enter the therapeutic encounter with ancient perceptions of what healers can or should do for them, and providers have visions of how patients should behave; we fill these roles for each other in the slowly evolving ritual we call healthcare.
It is probably terribly inefficient, but modern life generally is; we are not machines, but an ancient species with stone age reflexes in a postmodern society.
You always write about the art and skill of physiicans with such grace and insight. The whole problem with AI medicine is that data has to be in some written form to be counted. The real art of medicine is in the expertise of recalling one incident of something 30 years ago. This expertise rarely makes it into a form which can be expressed as data. I wrote about this on Substack (https://ruraldocalan.substack.com/p/ai-common-sense-and-the-art-of-medicine) but I still feel as if I'm crying in the wilderness. :)
Yes!!! Right on Doc!