It is widely recognized that the United States is going to experience a serious shortage of physicians in the coming years unless the number of physicians completing residency training and entering practice is greatly increased. Members of the academic medicine community have approached this issue by calling on Congress to eliminate the cap that currently limits the number of residency positions that Medicare will support. Simply eliminating the cap, however, will not ensure an adequate supply of physicians. In this commentary the author argues that decreasing the length of training required in core clinical specialties will be required to effectively address the workforce shortage by allowing more residents to be trained in core specialties without greatly increasing the number of training programs and the aggregate amount that Medicare currently spends on graduate medical education.
Dr. Whitcomb is Flinn Visiting Scholar, University of Arizona College of Medicine–Phoenix, Phoenix, Arizona.
Other disclosures: None.
Ethical approval: Not applicable.
Correspondence should be addressed to Dr. Whitcomb, 8393 N. Via Mia St., Scottsdale, AZ 85258; e-mail: firstname.lastname@example.org.
A general consensus exists within both the policy-making and academic medicine communities that the United States is on the verge of experiencing a major shortage of physicians. A number of reports indicate that certain regions of the country and certain specialties are already experiencing physician shortages. It is generally understood that the looming shortage can only be avoided by substantially increasing the number of residency positions available in the core specialty training programs (internal medicine, pediatrics, general surgery, etc.) in the U.S. graduate medical education (GME) system.1,2 However, there appears to be little understanding of the number of new positions that are needed and of the major challenges that will have to be overcome to make them available. To begin to answer these questions, I offer a brief description of the challenges the U.S. GME system must address in order to increase physician supply, and I propose a way to address the problem.
GME Expansion by the Numbers
To begin, it is important to recognize that if the number of positions in core specialty programs remains relatively constant, the number of physicians entering the workforce around 2020 will roughly equal the number leaving the workforce because of retirement, disability, or death. When that point of equilibrium occurs, the relative size of the workforce (i.e., the number of physicians per 100,000 persons) will actually begin to decline because the country’s population will continue to increase each year by approximately 3 million persons.3
Thus, if the relative size of the physician workforce at the point of equilibrium is approximately 250 physicians per 100,000 persons, an additional 7,500 physicians would have to enter the workforce each year after that simply to maintain the relative size of the workforce as the population increases.2 Accordingly, core specialty training programs would require 7,500 additional entry-level and subsequent training year positions to produce the required number of physicians. Since the average length of training required across the spectrum of core specialties is close to four years, approximately 30,000 new GME positions would be needed in core specialty programs before the end of the decade simply to maintain the relative size of the workforce that exists today. At issue, of course, is whether it is possible to accomplish this.
At present, the Medicare program contri butes to the funding of approximately 90,000 training positions at a cost to the federal government of approximately $9 billion per year (including both the direct and indirect medical education payments).4 Currently, there are limits on Medicare financing of residency programs, but if this funding cap were removed to allow the 30,000 new GME positions that need to be added to the system, the Medicare program would have to increase its current level of GME funding by approximately one-third, or roughly $3 billion. Given that the federal government is considering ways to decrease the growth in Medicare spending as part of the effort to address the country’s long-term financial problem,4 it seems highly unlikely that the government would allow Medicare expenditures for GME to increase to that degree. But even if Medicare GME funding were allowed to increase to whatever level might be needed, does the country’s GME system, as currently structured, have the capacity to add 30,000 new positions?
At present, the great majority of GME positions are sponsored by about 150 institutions—medical schools or major teaching hospitals. It is widely assumed that if the Medicare GME cap were to be removed, those 150 institutions would substantially increase the number of positions in the programs they sponsor. However, it is important to recognize that the number of new positions they could develop would still be limited, since the programs would have to adhere to existing accreditation requirements regarding the maximum number of residents that are allowed to be in each program.3 Available evidence suggests that most of the programs are currently close to their maximum size. And while it is not possible to accurately predict how minor teaching hospitals or hospitals that are not currently teaching hospitals might respond to the need for additional training positions in core specialties, it seems highly unlikely, given their history, that those institutions will develop a large number of new GME positions.
A Sensible Solution
Since funding constraints will likely remain in place even if the Medicare cap is removed, I believe that the physician supply will not increase to the level required unless the academic medicine community changes accreditation requirements to allow the length of training in core specialties to be shortened by at least one year. Shortening the length of training will not in and of itself guarantee that physician supply will increase to the level needed, but decreasing the length of training will make it possible to train many more residents in core specialties without increasing the aggregate amount that Medicare currently spends on GME. Equally important, decreasing the length of training would initially result in fewer residents in training within individual programs, providing major teaching hospitals with the capacity needed to increase over time the number of residents in each year of training.
I recognize that this suggestion will engender high levels of resistance from individuals within the academic medicine community who are deeply committed to a tradition that relates the quality of training to the length of training. Indeed, residency redesign initiatives conducted by several major specialties during the past decade had no effect on the length of training required, even though some members of the academic community have advocated for shorter periods of training.5,6 It is also worth noting that many in the community support the move to assess residents on the basis of their ability to demonstrate that they have met certain milestones defining the acquisition of knowledge and skills considered important for the practice of a specialty, even though this concept challenges the long-standing view that it is the length of training that determines readiness for practice.7
Finally, in reflecting on the need to decrease the length of training, it is important to understand that there is a historical precedent for decreasing the length of GME training. In the mid-1970s the yearlong internship experience that had been in place for decades was eliminated as a requirement for entering a residency program, thereby decreasing the total length of training required across the specialties without any apparent adverse effect on the quality of training provided within the GME system.8
It is important for the members of the academic medicine community to acknowledge that changes occurring in medical practice do not require residents to complete lengthy training programs before branching off into narrow fields of practice.2,9 Given the extraordinary challenge the country faces, I urge the academic medicine community to take the necessary steps to decrease the length of training in core specialties. If the academic community will not accept the responsibility to address the major physician workforce challenge now facing the country, I believe the federal government should take steps to ensure that residency training is redesigned in ways that will contribute to the effort to ensure that the country has an adequate supply of physicians to meet the health care needs of the American public. To do so would simply be good public policy!
1. Jolly P, Erikson C, Garrison G. U.S. graduate medical education and physician specialty choice. Acad Med. 2013;88:468–474
2. Ensuring an Effective Physician Workforce for America. Recommendations for Graduate Medical Education to Meet the Needs of the Public. 2011 New York, NY Josiah Macy Jr. Foundation
3. Cooper RA. It’s time to address the problem of physician shortages: Graduate medical education is the key. Ann Surg. 2007;246:527–534
4. Iglehart JK. Financing graduate medical education—mounting pressure for reform. N Engl J Med. 2012;366:1562–1563
5. Johns MM. The time has come to reform graduate medical education. JAMA. 2001;286:1075–1076
6. Whitcomb ME. Putting patients first: The need to reform graduate medical education. Acad Med. 2003;78:851–852
7. Nasca TJ, Philibert I, Brigham T, Flynn TC. The next GME accreditation system—rationale and benefits. N Engl J Med. 2012;366:1051–1056
8. Graduate Medical Education. Present and Prospective. A Call for Action. Report of the Macy Study Group. 1980 New York, NY Josiah Macy Jr. Foundation
9. Emanuel EJ, Fuchs VR. Shortening medical training by 30%. JAMA. 2012;307:1143–1144