I had some recent surgery to repair an orthopedic issue. The surgeon is extremely good at what he does. He benefits from being very focused on a defined skill set and he does what he does repeatedly and often. Practice does make near perfect.
For the model of physician care which requires defined and discrete tasks whose execution requires little knowledge beyond the specific organ-limited pathology, this approach works fine. The orthopedist may repair my knee, the gastroenterologist may peer into their periscope, the dermatologist excise the skin cancer, or the cardiologist may place their stent. Assuming the patient survives the immediate post procedural period and the repair holds, any other problem the patient may have is someone else's concern.
This whole model is predicated on the assumption that problems can be treated discretely. Many problems can be approached this way and there are substantial incentives to stake out your claim as a physician in an arena where problems can be managed with such distinct beginnings and ends. However, one of the problems with this approach is a product of hyper-specialization; the loss of the ability to do much of anything else.
Who in this model sees the big picture? I believe in the ideal world, from the perspective of a patient, the most valued healer is someone who cares for them, not someone who can do something to them. Interventions may be a piece of the care package, but as those whose professional activities become more and more focused on a limited number of things they do and think about, they become less valuable in terms of integrating care. There may even be a point where their skill sets render them incapable of this. When a physician has done nothing but robotic prostatectomies, or ERCPs, or Mohs surgery, or cataract surgeries, or knee replacements for countless years, they likely are not able to keep sufficiently up to date to perform any function which requires them to have a broad medical knowledge base outside of their narrow focus of expertise.
There may be financial advantages to individual physicians who can claim to be incapable to having the understanding which would allow them to care for many facets of their patient's needs. From the standpoint of controlling one's time and being paid optimally for acquired skills, the ideal scenario is assume responsibility for only a very defined period of time and to be paid as handsomely as possible for a defined intervention. However, I think virtually anyone would be hard pressed to see this as a model which would be highly desired by patients under any circumstances.
In virtually all other realms of human interactions the person who is most highly valued is the one who is in charge. We value those who can coordinate human activity, identify where activity has created real value, and see where other activities represent only Brownian movement. In medicine, assuming long term broad responsibility is like holding the "old Maid" in cards. It means making difficult decisions, open ended commitments, dealing with lots of uncertainty, being responsive to all types of patient needs and getting paid badly. If the building trade were like medicine, the most poorly paid schmucks would be the general contractors.
We still have cadres of generalists and specialists who maintain a generalist knowledge base. These docs often find themselves fielding calls from patients on a broad variety of topics, even outside their realms of expertise. Patients are good at detecting physicians who have a modest knowledge base and exhibit the slightest proclivity toward problem solving. Why any physician would persist in this type of behavior in the current climate may be hard to fathom. To some it is a higher calling. To others perhaps it is reflective of some sort of personality disorder, trying to gain stature by demonstrating competence beyond their peers. Given the present incentive structure, it is not unreasonable to predict the behavior will become increasingly less common at a very time where it is needed more than ever.
The trend toward hyper-specialization is not going away. Perhaps it is accelerating. When do we reach a tipping point? If 10% of the physician workforce is completely incapable of understanding anything outside of their narrow focus of technical expertise, is that a problem? What about 50%? 80%? Should someone else fill that void?