I believe the appropriate use of ‘primary care’ was bastardized by the ‘diagnosis-centric’ orientation to care in medicine (especially facilitated by the payor side). A patient became, not a complex person with various problems, but a manila folder full of ICD-9 codes. You can’t divide up a person, all the parts and problems interact with each other, but you can divvy up the ICD-9 codes so that someone takes care of problem A, someone else problem B, etc., and no one really cares about the manila folder.

The diagnosis-centric approach has several pernicious effects on the care of patients (especially complex care). First, the importance of the underlying condition of the patient tends to be minimized, especially by the payment system. Even though a low-grade fever in a 6 year old who just returned to school is a very different problem than a low-grade fever in a cancer patient on chemotherapy, the lack of patient-centricity tends to push excessive care on the 6 year old (through guidelines, treatments for low probability events, etc.) and push inadequate care on the cancer patient (primarily through low payments and bureaucratic hurdles which ration physician time).

More importantly, though, interaction effects among health problems tend to be ignored. The diagnosis-centric system will pay Dr. #1 to take care of Problem A (let’s say $150 for example), and will pay Dr. #2 to take care of Problem B (another $150), but does not pay either one for the interaction effects between Problems A and B (of course, good docs pay attention, but the system has few incentives other than pride and honor). But if Dr. #3 wants to take care of both Problem A and B, she get paid less (probably $200 total instead of $300) and she is clearly responsible for the interaction effects too. Since the interaction effects can be as much, if not more, of a issue than each primary problem, she essentially gets paid a fraction more for three times the work. And since interaction effects escalate geometrically (three problems have 4 interactions, four problems have 11), the great majority of time need to be spent on them in any complex patient.

Occasionally a complex patient will enter the medical system in a true primary care situation. For example, an auto accident victim may be perfectly healthy and then suddenly have multiple medical problems related to trauma. Interestingly, the primary care in this situation is generally given by EMT, and then on delivery to a hospital, care quickly escalates from the emergency physician to a team approach with multiple physicians.

Most of the care given by internists, however, is not the initial entry into the medical system. I argue that chronic care of complex patients is not primary care because the patient has not left the medical system. The care needs are more analogous to the multiple trauma patient (although at a slower time scale) than the true primary care patient (such as the 6 year old with fever).

Payors love “generalist physicians” taking care of complex chronic patients because they pay less than if the patient saw different doctors for each problem. As an added benefit to patients the interaction effects among their problems get at least some attention. On the flip side, payors don’t want to pay much for true primary care, because in most cases lesser trained and less expensive personnel can treat a single new problem. And in the diagnosis-centric culture differences between a new problem in a healthy patient versus a new problem in a complex, chronic patient are underrecognized. So generalist physicians are leaving (or not entering) the field because their pay reflects the lesser value provided for true primary care, and when they provide a lot of complex care they capture none of the value added (payors capture the reduced payment, and patients capture the interaction-effect care).

I had a problem for which I’d seen specialist A, who’d made a diagnosis that fell within his field A. Then I’d seen specialist B for the same problem, and he made a diagnosis that fell within his field B. A and B both said that their respective diagnoses were non-contributory to one another — either A was correct, or B. So I went to see my internist, who said that in fact A and B diagnoses are related, and explained how. Then we talked about how this fit in with my broader medical history, which neither A nor B really could speak to. In other words, I’m very glad I have an intelligent, sensitive internist — I don’t feel like he has a set toolbox he’s eager to use, but instead seeks to find the right tool for me, as nuanced and particular as I might be.

Boy, things are really going to get interesting:

http://online.wsj.com/article/SB120710036831882059.html?mod=pj_main_hs_coll

(hope the link works)

Dr. Nurse, eh? Lots of great quotes in the article, but everyone should just read it for themselves.