An article in the March 24 NEJM called Specialization, Subspecialization, and Subsubspecialization in Internal Medicine might have some heads shaking: Isn’t there a shortage of primary care physicians? The sounding-board piece considers the recent decision of the American Board of Internal Medicine to issue certificates in two new fields: (1) hospice and palliative care and (2) advanced heart failure and plans in-the-works for official credentialing in other, relatively narrow fields like addiction and obesity. The essay caught my attention because I do think it’s true that we need more well-trained specialists, as much as we need capable general physicians. Ultimately both are essential for delivery of high-quality care, and both are essential for reducing health care errors and costs. Primary care physicians are essential, above all, so patients have a doctor who knows them over a long period of time, who cares about them and whom they trust. A well-educated, thoughtful family doctor or internist should be able to handle most conditions: prophylactic care including vaccinations, weight management, high blood pressure, diabetes, simple infections – like bacterial pneumonia or UTIs, gout and other routine sorts of problems. On the other hand, specialists can be lifesaving when highly-detailed expertise matters. There are limits to how much a general internist knows about chemotherapy, for example. Even within the field of medical oncology, a subspecialty of internal medicine, there are doctors who only see patients with particular kinds of cancers. When I had breast cancer I chose an oncologist whose practice consists almost entirely of patients with breast cancer and related diseases. If someone in my family has a lymphoma, I’d advise them to consult with someone who, for the most part, patients with lymphoma and similar disorders. Why? Because each of these cancers represent a group of disorders, really, and the treatment depends in part on the doctor’s knowledge of the specific subtypes and the relevant, current data for treatment of that specific tumor type. Treatment of lung cancer involves choosing among a different set of drugs than would be considered for brain or kidney cancer. I mention oncology, here, because I’m most familiar with this field. But the same is likely true for the subspecialty of Infectious Diseases: knowing about all the new HIV drugs, in pregnant women, children and adults, involves a different set of knowledge than knowing about parasites in the tropics, and that differs from knowing about viral and other, unusual infections in patients who’ve had kidney, heart or lung transplants. In each of these settings, expertise can reduce errors – because specialists are more likely, in the first place, to establish a correct diagnosis and, next, to prescribe the right therapy based on the best evidence available. You could argue that it’s impossible to provide these kinds of sub-sub-specialists to people in rural areas, or that it’s too expensive, but I don’t think either of these factors should be limiting. To a large extent, experts might work with primary care providers and communicate with patients via Telemedicine and Skype-like technologies. As for surgical subspecialties, it may be that patients would find it worthwhile to travel to a regional center where a specialized procedure is done routinely, as opposed to having an operation in a local hospital where the doctors perform a certain kind of surgery – say a laparoscopic splenectomy, for example — only a few times each year. There’s a tradeoff, as discussed in the NEJM piece, between increasing use of specialists and fragmentation of care. I think this concern is legitimate, based on my experiences practicing medicine and as a patient. But I do think we need specialists, and sub-specialists if we want doctors who can answer their patients’ questions, i.e. who really know what they’re doing. I was a bit surprised that the article mentions a survey of physicians in which the majority of respondents reported that “professional image” was the primary reason for seeking subspecialty credentials. While this may be true, I don’t think doctors’ motivation matters in this. From the public’s perspective, what’s important is that hand surgeons know how to do hand surgery, and that a heart specialist knows how to interpret an echocardiogram, and that the hospitals where they work not let them practice if they’re not appropriately credentialed. In cutting health care costs, or in trying to so, I don’t think it makes sense to reduce the number of physicians or short-cut their education by way of 3-year medical schools. Rather, we need well-trained primary care doctors and specialists, both. There’s no way around it.