The 2 position papers strongly affirm the importance of training good generalists. The underlying premise is that most internists should know how to provide front-line care for the major diseases in any specialty of internal medicine. The case for training good generalists is based on strong evidence. First, patients want good generalist physicians to take responsibility for their care (3). Second, many health care systems (for example, Kaiser Permanente, the Veterans Administration, Group Health Cooperative, the military health system, and the Palo Alto Medical Clinic) organize their practice around primary care physicians (4, 5). Third, many internal medicine subspecialists also need to function as generalists. While the position papers make a strong case for substantially more ambulatory learning time, the current system gives first priority to the care of fragile hospital patients. Shifting the balance toward more ambulatory care during training would be expensive, as shown by efforts to provide additional inpatient coverage in response to limits on residents' duty hours.

Education reforms address only 1 element among many that are responsible for a decline in interest in internal medicine as a career. Internal medicine needs payment reform and better organization of office-based care. While these problems may some day yield to skillful political advocacy, they are beyond internists' direct control. In contrast, internists control the curriculum of internal medicine residencies, so education reform is both a priceless opportunity and a test of our will.

The ACP and APDIM position papers describe the goal of reform, but they do not describe a strategy for reaching that goal. In this editorial, we focus on 2 key obstacles with which any strategy must reckon. First, internists do not control the resources needed to effect change in residency education. Second, internal medicine is ambivalent about what it stands for and what it should become.

The path to successful implementation of the ACP and APDIM reforms will require fundamental changes in the organization and funding of residency training. Control of graduate medical education means control of its financial support and accreditation. The sponsoring hospitals receive funds for residency stipends and teaching salaries largely from the Centers for Medicare & Medicaid Services (CMS). The hospital controls the flow of dollars to the training programs. The formulas for distributing money seem arcane at the federal level and mysterious at the local level. The department chairs delegate conduct of the residencies to program directors. Accreditation, unlike financing, is under the control of the profession. It is a key point of leverage because teaching hospitals want fully accredited training programs that will attract good trainees to provide care to patients. The Accreditation Council for Graduate Medical Education accredits individual residency programs through its 27 constituent residency review committees. It sets the standards for the content of training. The American Board of Medical Specialties, an organization of specialty groups, certifies physicians who have received specialty training. Hospital executives and accrediting institutions wield the power to effect change. The organizations that issued these position papers—the ACP and APDIM—have no direct control over any aspect of this complex process. To be sure, they are vitally involved in internal medicine residency programs (APDIM) and in the well-being of internists (ACP), and ACP plays a role in the accreditation process as one of the appointing organizations to the Internal Medicine Residency Review Committee. But they don't hold the power to effect change in graduate medical education.

Wouldn't it be nice to see organizations like the Association of Professors of Medicine (the chairs of medicine), the Council of Teaching Hospitals (the hospital directors), the Internal Medicine Residency Review Committee (accreditation of training programs), and the American Board of Internal Medicine (credentialing of internists) take a strong public stand on the content of internal medicine training? Without their active support, change is unlikely, because redesign means shifting more time, energy, and resources to ambulatory care training—at the expense of hospital care.

The second major shortcoming of the 2 position papers is that they presuppose a commitment to training excellent generalists. We think that internal medicine's actions bespeak ambivalence on this crucial point. Here we use a sports analogy: the golfer who can't decide how to aim a putt on a tricky green. For the APDIM and ACP redesigns to occur, internal medicine needs to "putt or get off the green." Getting off the green would mean giving up serious efforts to train internists to function as generalists in office practice, either as general internists or subspecialists. It would mean ceding the primary care field to others. Internal medicine would then become a confederation of specialists, each specialty with its own needs. Opting out of training generalists could have 1 positive consequence: Internal medicine would have fewer residency positions, which might make the specialty more competitive for U.S. medical students, who today fill fewer than 60% of the available spots (6).

Getting off the green could lead to difficulties. Balkanizing internal medicine is risky. The prosperous subspecialties may secede into independent departments, as happened with surgery. Perhaps because all internists now have a common training model and learn the same basic skills, they share allegiance to their parent discipline. It would be harder to maintain that cohesion, so necessary for effectiveness in the political arena, without the common experience of general internal medicine training, because the worlds of, say, rheumatology and cardiology are so very different. And fewer generalist internists would create a vacuum in primary care. Who would fill that vacuum is not clear, given what ACP has recently called "the impending collapse of primary care" (6) and the rapidly declining interest in family medicine residency training (7).

Putt or get off the green? Putting would mean making a serious commitment to training general internists for office-based practice—not just hospital-based practice. It would mean honoring the generalist trunk of internal medicine's tree by creating a good environment for future subspecialty internists to learn office-based practice skills outside their area of specialty expertise. Making this commitment would entail placing educational needs above service needs and working with the giant hospitals and graduate medical education funding sources to provide excellent service and education. Although achieving both goals would be expensive, we contend that leaders will find the money if they make them both a high priority, especially if CMS uses its financial support of graduate medical education to leverage change. The ACP and APDIM position papers outline what the new system would look like (1, 2). Another editorialist has recently described a related future scenario (8). But the ACP and APDIM reforms would require determined advocacy from the chairs of medicine, and we fear their attention is elsewhere—in building departmental research programs and the faculty practice plan, the main sources of prestige and income for departments of medicine.

Either of these options is preferable to the default position: trying to provide excellent ambulatory training while inpatient care needs drive internal medicine residency training. This model is not attracting enough of the outstanding students who will lead internal medicine clinically and academically. Moreover, the default position is leading to compromise of the teaching mission, continued waning of medical student interest in low-tech fields, and an increasing mismatch between learning during training and subsequent day-to-day practice. These outcomes would be a disservice to a tax-paying public that has a right to expect its generous support of graduate medical education to produce physicians that meet the public's needs.

The APDIM and ACP position papers contain an accurate diagnosis and present an inspired vision of what major redesign could accomplish. But they lack 2 crucial elements: endorsement by those empowered to make the necessary changes and, most important, a strategy for uniting internal medicine around a future course for the discipline. Without unity of purpose, we fear that internal medicine is headed down the default path of halfway measures.