Secondary Logo

Journal Logo


Is the Physician Shortage Real? Implications for the Recommendations of the Institute of Medicine Committee on the Governance and Financing of Graduate Medical Education

Salsberg, Edward S. MPA

Author Information
doi: 10.1097/ACM.0000000000000837
  • Free


On July 29, 2014, the Institute of Medicine (IOM) Committee on the Governance and Financing of Graduate Medical Education released its report calling for a major overhaul of the financing of graduate medical education (GME).1 The same day, several national organizations with an interest in GME issued press releases opposing the report.2–4 One of the key reasons cited for this opposition was that the IOM Committee recommendations would worsen physician shortages. However, this conclusion is based on two questionable assumptions: first, that the nation is facing a general physician shortage, and second, that the IOM Committee’s recommendations would make shortages worse. A strong case can be made that although there are some communities with shortages, a current and future general national shortage of physicians is unlikely. Although I have commented elsewhere on the report in general,5 in this paper I focus on the question of the likelihood of physician shortages and the implications of the recommendations of the IOM Committee for the supply and distribution of physicians.

The IOM Committee concluded that the current system of financing of GME, which relies heavily on Medicare funds distributed to teaching hospitals based on inpatient days and discharges of Medicare patients, is not an effective vehicle for physician workforce policy, is not equitable, has unintended negative consequences, and does not support health delivery reform. It recommended that the current overall funding level from Medicare (about $10 billion per year) be maintained and increased for inflation for at least the next decade but that the basic approach to distributing these funds be revised significantly. Their proposal would combine what are now two separate Medicare funding streams: one for direct costs of GME (DGME), which covers resident compensation, faculty, and other related costs, and a second stream known as indirect medical education (IME), which helps cover higher costs associated with teaching hospitals due to such factors as more intensive services and care for the uninsured. The IOM Committee recommended that the overall Medicare funding go into two new funds: an Operational Fund to continue support for existing Medicare-funded GME positions; and a Transformation Fund to encourage innovation and needed expansion of GME. Further, the IOM Committee recommended that the Operational Fund be distributed through a national per resident amount with geographic adjustments.

Although the IOM Committee did not recommend a general increase in funding to support an expansion of GME positions, they did recommend maintenance of the current funding level with annual inflationary increases in the overall Medicare GME funds. They also recommended that the Transformation Fund be used to support documented, needed expansion. The IOM Committee questioned the evidence for a projected national shortage and the need for new funding to support expansion. The concern with shortages was central to the statements of groups like the Association of American Medical Colleges (AAMC), the American Hospital Association, and the American Medical Association in their opposition to the report.

In assessing the IOM Committee recommendations and the validity of the concerns regarding shortages, there are two key questions to consider:

  1. Is a future national physician shortage likely?
  2. Would the IOM recommendations limit growth in GME and make shortages worse?

What Did the IOM Committee Say About Physician Shortages?

The report1 cites numerous studies to support the following conclusions about shortages:

Simply increasing the numbers of physicians is unlikely to resolve workforce shortages in the regions of the country where shortages are most acute and is also unlikely to ensure a sufficient number of providers in all specialties and care settings. [Section 2]

Concerns that the nation faces a looming physician shortage, particularly in primary care specialties, are common. The committee did not find credible evidence to support such claims. Too many projections of physician shortages build on questionable provider–patient ratios, fail to consider marked geographical differences in physician supply, and ignore recent evidence of the impacts of more effective organization, new technology, and deployment of health personnel other than physicians. [Section 5]

As the director of the workforce centers at the AAMC (2004–2010) and the federal Health Resources and Services Administration (2010–2013), I was responsible for the work on projections of supply and demand for physicians. I can strongly attest to the great challenges of projecting supply and demand especially in a period of health systems transformation.6 Although progress has been made in developing more sophisticated models and to improve the data inputs,7,8 there is always inherent uncertainty in projecting the future.

In addition, if one of the key goals of workforce policy is to ensure access to care, then national data and projections are of limited value because they tell us nothing about access at the local level, which is dependent on the distribution of physicians. Even if we had a million physicians in the United States, if they were all on the East and West Coasts or in urban areas, a significant share of the population would face severe access problems. The current maldistribution of physicians in the United States is well documented, and the federal process for identifying Health Professional Shortage Areas (HPSAs) has identified hundreds of communities with shortages.9 Increasing the national supply of physicians does not mean they will go to the areas and specialties with the greatest need. The adequacy of physician supply is best measured at a local or regional level because an excess in one area does not balance out a shortage in another area. Therefore, ensuring an adequate supply is best addressed by policies and programs targeted to the specific geographical areas of need rather than general increases in supply.

Is a Future National Physician Shortage Likely?

Articles appear regularly in the popular press citing concerns about a physician shortage, and most cite AAMC reports for national numbers of physicians. In March 2015, AAMC issued a new report projecting an overall shortfall in 2025 of between 41,100 and 90,400 physicians10; for primary care physicians, AAMC projects a shortage of between 12,500 and 31,100. Although these numbers represent significant decreases from the AAMC’s previous (2010) shortfall estimates of 130,600 physicians overall in 2025, and a primary care physician shortfall of 45,400,11 the new numbers still appear to overestimate the shortfall for reasons discussed below.

Having overseen the preparation of the original AAMC projections, I am very familiar with the basic methodology. The methodology assumes that the services that will be provided in 2025 and how people will use these services will be generally unchanged from the base year of the projections. Although the new AAMC projections update some key inputs, such as more recent population projections from the U.S. Census Bureau and a higher number of GME positions based on actual recent growth, the projections suffer from a number of significant shortcomings.

First, the new report, while acknowledging the importance of physician assistants (PAs) in increasing access to care, does not include PAs in the actual projections of future supply. There are now more than 100,000 active PAs12 in the United States, and the pipeline is growing rapidly: In 2014, 7,578 PAs passed the examination required for new entrants into the field compared with 4,337 in 2003—an increase of 75%—and the number of new programs is rising rapidly.13

Second, although the new AAMC projections did use a higher estimate for the number of nurse practitioners (NPs) in the future, the projection still undercounts the likely growth in this field. The AAMC projection is based on an estimate of 15,000 new NPs per year; however, according to the American Association of Colleges of Nursing, in 2014, 18,484 nurses completed NP programs, and the annual number of graduates is still increasing.13

Third, the projections do not include any reduction in demand for physicians as a result of incentives for systems redesign, efforts to reduce readmissions, or the use of new technologies. In fact, the projections assume increased demand for physicians due to the development of Accountable Care Organizations and integrated delivery systems.

Finally, the projections also rely on straight-line projections of supply and demand independently and fail to take into account the interaction of the two, including the ability of the delivery system to evolve in response to new demand.

The exclusion of over 100,000 PAs and the underestimation of NP growth by at least 35,000 over the next decade has an enormous impact on the bottom line. Although there is room to debate how much NPs and PAs substitute for physicians rather than supplement physician services, one committee established by the federal government, reviewing the available data on patient visits, concluded that, in primary care, an NP or a PA provides on average 75% of the number of services provided by a physician.14 Applying the 0.75 to an additional 135,000 PAs and NPs is equivalent to over 100,000 physicians.

In late 2013, the federal government published its own projections for primary care practitioners.15 These projections used a similar methodology to that used for the AAMC projections but more explicitly included the impact of the growing supply of NPs and PAs. As in the AAMC projections, the federal projections assumed a shortfall in the base year equal to the number of primary care physicians needed to eliminate the federal designation of primary care HPSAs—7,500 physicians. The federal projections estimated a shortfall of 6,400 primary care practitioners in 2020; this is less than the shortfall in the base year. The federal projections were limited to primary care practitioners, but the same methodological and data issues would be applicable to the projections for specialists.

Although there will continue to be specific geographic areas and specialties with shortages, just as there are today, there is evidence that shortages in the future are likely to be less than today’s shortages. Some of the key reasons for this conclusion are reviewed below.

Evolution of the health care delivery system

The delivery system is evolving rapidly; pressure and incentives to find ways to be more effective and efficient are likely to continue. The health care industry and policy makers are actively seeking ways to reform and improve the delivery system, including the development of new payment policies and incentives to reward improved performance. While the health care system was already moving in this direction, the Affordable Care Act (ACA) included new incentives and new strategies, such as accountable care organizations, patient-centered medical homes, value-based payment policies, readmission penalties, bundled payments, and significant funding for innovations.16 The net effect is that providers across the nation are trying new ways to deliver services more efficiently and effectively, including the use of new combinations of workers. The Federal Center for Medicare and Medicaid Innovations Health Care Innovation Awards program has provided billions of dollars for delivery system innovations to improve efficiency, effectiveness, and quality. Many projects include innovations in the use of workers such as nurses, technicians, care coordinators, patient navigators, community health workers, and aides and assistants. Although the results of these initiatives have not yet been reported, the funding has greatly stimulated new thinking on how to redesign services.

Demand for care versus demand for physicians

The aging population and increasing numbers of chronically ill will lead to an increase in demand for care, but the demand is for services, many of which can be provided in a variety of ways by a variety of practitioners/workers. Articles on potential physician shortages often state that demand for health services is rising because of population growth, aging, and advances in medicine that keep people living longer. But some of the care that may have historically been provided by physicians can also be provided by other clinicians; in some cases, nonphysician clinicians and other caregivers are able to provide some services even more effectively than a physician, especially if care is provided by interprofessional teams. This use of other health professionals frees up physicians to focus on patient needs appropriate to their unique knowledge and skills.

Many of the recent initiatives to control costs and improve outcomes have targeted those with the greatest expenditures. This is particularly important given that 15% of the population incur 75% of the medical expenditures.17 Many of the new initiatives are targeted to the chronically ill and include services that can be provided by nonphysicians.

The very significant potential impact of using a variety of health workers in an interprofessional team was documented by Altschuler et al in 2012.18 The authors calculated that a family physician working alone and providing all appropriate services to a typical population could serve 983 patients, but, if part of an effective interprofessional team, the physician could adequately serve 1,947 patients. This is more than a twofold difference. A strong case can be made that we are more likely to experience the second rather than the first scenario in the future.

Rapidly growing supply

The supply of health care practitioners is growing rapidly, and the evolution of the health system is being facilitated by this increasing supply of health care providers. While the supply of physicians is growing slowly, the supply of NPs, PAs, other clinicians, and assistive personnel is growing rapidly. The combined growth rate of all of these health workers has far exceeded the growth rate of the U.S. population.


The supply of physicians has been rising steadily for many years. While Congress essentially froze Medicare GME payments in 1997, over the past decade the number of positions for new entrants into GME has grown by about 1.7% per year.* The current cohort size of physicians entering medicine is approaching 30,000 each year.19 Today’s cohort size is well above the size of the cohort entering medicine 30 years ago (about 20,00020); this cohort has already or will shortly be exiting medicine. The bottom line is that even with the recent modest growth in the numbers of physicians entering GME each year, the total supply will be increasing steadily for years to come.

NPs/PAs/pharmacists/registered nurses.

In addition to the growth of the NP and PA pipeline cited above, the educational pipeline for many other nonphysician clinicians (e.g., pharmacists13 and registered nurses21) has grown significantly over the past decade. Even if the pipeline does not grow any further, this past growth will lead to increases in the overall supply in these professions for the next several decades.22

Other caregivers, like physical therapists, occupational therapists, care coordinators, medical assistants, and community health workers, have also been growing rapidly. In fact, despite the recent recession, health care employment has grown significantly over the past several years.23

Some people may worry that this increasing use of nonphysicians will reduce the quality of care. This is unlikely for a variety of reasons, including the increasing focus of the health system on safety and outcomes; the high professional standards used for educating and credentialing health professionals; and the increased adoption of team-based care which, if done correctly, allows for better use of the skills of each member of the team, including the physicians.

Technological advances

Technological advances like telehealth, eHealth, and electronic health records hold great potential to improve efficiency and effectiveness. Although many questions and challenges remain regarding the overall impact of technological advances, they clearly hold great potential, particularly as part of the overall effort to improve efficiency and effectiveness. For example, telehealth, including the use of e-mail and video visits, can reduce the need for visits to the clinician’s office, saving time and improving efficiency for both the patient and the physician.24

Although many would agree that the above factors will increase supply of practitioners and reduce demand, there is still a general perception and concern that the expansion of coverage under the ACA will increase demand significantly and outweigh these other factors. Several studies, including the recent AAMC projections, have found that the ACA is likely to increase demand between 2% and 3%.10,25,26 On the other hand, the work by Altschuler et al18 concludes that effective teams can more than double the number of patients served by a primary care physician. Other studies have had similar findings.27 It would seem that the use of teams with the increased supply of nonphysician practitioners and the general redesign of delivery has a far greater potential impact than the expansion of insurance under the ACA.

Would the IOM Report Recommendations Limit Growth in GME and Make Shortages Worse?

Although the IOM Committee concluded that there is little evidence that there will be a general shortage in the future, they did recognize that parts of the country currently have shortages in specific specialties and that maldistribution is a serious problem. For this reason, the IOM Committee recommended that the Transformation Fund be used to support expansion of GME in specialties and communities with documented need. The IOM Committee suggested that the Transformation Fund receive between 10% and 30% of the total Medicare GME funds per year over the decade depending on the year; this would result in between $1 billion and $3 billion per year for targeted increases and innovations.

The IOM report does not specify the number of positions that would be supported by the Transformation Fund, and there is no way of knowing now what percentage of the money would go to expansion compared with innovation. Nevertheless, the $1 billion to $3 billion per year Transformation Fund could support a significant increase in GME slots, and this support would be targeted on the basis of documented need.

Although elimination of IME would reduce funds to teaching hospitals, it is generally accepted that IME covers costs not directly related to training physicians, like care for the uninsured and high-cost tertiary services. Even excluding the Transformation Fund, the IOM proposal, by adding between 70% and 90% of current IME ($6.5 billion) per year (depending on the year) to the proposed Operational Fund (which would already include current DGME funds), would make roughly $45,500 to $58,500 additional dollars per year available for each of the approximately 100,000 residents and fellows whose positions might be covered by Medicare in the future. In other words, the majority of funds would be shifting from a general hospital subsidy to explicit support for GME.


Under this analysis, arguments that we face a general physician shortage do not stand up well. What the nation faces are shortages in specific areas and specialties. These would be most effectively and efficiently addressed by need-based strategies, including targeted GME expansion and programs like the National Health Service Corps. Although the location of graduate training is not a guarantee of practice in a region or state, according to recent data nearly half the physicians (47%) completing training in an Accreditation Council for Graduate Medical Education–accredited program stayed in the state where they completed their most recent GME.28 Making documentation of need a condition for receiving support from the Transformation Fund, as recommended by the IOM, would more effectively address the United States’ physician workforce requirements than would general increases in funding for GME. It may also be a more politically acceptable approach in tight budget times.

It has been suggested that the IOM Committee recommendation to eliminate the $6.5 billion IME funding stream would reduce the dollars available for training physicians. However, an actuarially appropriate level of indirect costs has been debated for some time; MedPAC29 has consistently concluded that the current level of funding cannot be justified as costs related to GME. Although the elimination of IME would undoubtedly present fiscal problems for many hospitals, a strong argument can be made that, if moved into the proposed Operational Fund as recommended by the IOM Committee, the dollars would be more directly tied to the education and training of physicians.


Any significant change in Medicare GME reimbursement policies, such as those recommended by the IOM Committee, requires federal legislation. The IOM GME report provides a very helpful framework for moving forward. Many details will need to be decided in moving from the framework to specific legislation and policies. This is a time for dialogue and discussion. Restructuring how we spend $15 billion per year (Medicare and other sources of GME funding) will involve many interests and challenges, but we owe it to the nation, including the physician and medical education communities, to find a better way to support the training of the future physician workforce.

Acknowledgments: The author wishes to thank Leo Quigley, George Washington University, for his assistance in editing this article.

* Based on an analysis by the authors F. Mullan and K. Weider of Accreditation Council for Graduate Medical Education and American Association of Colleges of Osteopathic Medicine GME data on entrants into pipeline positions between academic year (AY) 2004–2005 and AY 2013–2014.
Cited Here


1. Eden J, Berwick D, Wilensky GInstitute of Medicine, Board on Health Care Services, Committee on the Governance and Financing of Graduate Medical Education. Graduate Medical Education That Meets the Nation’s Health Needs. 2014 Washington, DC National Academies Press
2. Kirch DG. July 29, 2014 Washington, DC Association of American Medical Colleges Accessed June 4, 2015
3. Hoven AD. July 29, 2014 Chicago, Ill American Medical Association Accessed June 4, 2015
4. Fishman LE. July 29, 2014 Washington, DC American Hospital Association Accessed June 4, 2015
5. Salsberg E.. IOM graduate medical education report: Better aligning GME funding with health workforce needs. Health Aff Blog. July 31, 2014 Accessed June 4, 2015
6. Cunningham R, Scanlon WJ managers. Forum session: Projecting health care workforce needs in an era of practice and technological change. Health Policy Forum. April 19, 2013 Accessed June 4, 2015
7. Cecil G. Sheps Center for Health Services Research. . FutureDocs forecasting tool. University of North Carolina at Chapel Hill Accessed June 4, 2015
8. Dall TM, Gallo PD, Chakrabarti R, West T, Semilla AP, Storm MV.. An aging population and growing disease burden will require a large and specialized health care workforce by 2025. Health Aff (Millwood). 2013;32:2013–2020
9. Health Resources and Services Administration, U.S. Department of Health and Human Services. . Health professional shortage areas (HPSAs) Accessed June 4, 2015
10. Dall T, West T, Chakrabarti R, Iacobucci W. 2015 Washington, DC Association of American Medical Colleges/IHS Inc. Accessed June 1, 2015
11. Center for Workforce Studies. June 2010 Washington, DC Association of American Medical Colleges Accessed June 4, 2015
12. National Commission on Certification of Physician Assistants. . 2014 statistical profile of certified physician assistants. March 2015 Accessed June 4, 2015
13. Salsberg E.. The nurse practitioner, physician assistant, and pharmacist pipelines: Continued growth. Health Aff Blog. May 26, 2015 Accessed June 5, 2015
14. U.S. Department of Health and Human Services, Negotiated Rulemaking Committee on the Designation of Medically Underserved Populations and Health Professional Shortage Areas. October 31, 2011 Accessed June 4, 2015
15. U.S. Department of Health and Human Services, Health Resources and Services Administration, National Center for Health Workforce Analysis. 2013 Accessed June 4, 2015
16. Centers for Medicare and Medicaid Services. . Innovation models. Accessed June 4, 2015
17. Henry J. Kaiser Family Foundation. . Concentration of health care spending in the U.S. population, 2010. Accessed June 4, 2015
18. Altschuler J, Margolius D, Bodenheimer T, Grumbach K.. Estimating a reasonable patient panel size for primary care physicians with team-based task delegation. Ann Fam Med. 2012;10:396–400
19. Salsberg E.. The 2014 GME residency match results: Is there really a “GME squeeze”? Health Aff Blog. April 24, 2014 Accessed June 4, 2015
20. Salsberg ES, Forte GJ.. Trends in the physician workforce, 1980–2000. Health Aff (Millwood). 2002;21:165–173
21. Salsberg E.. Recent trends in the nursing pipeline: US educated BSNs continue to increase. Health Aff Blog. April 9, 2015 Accessed June 4, 2015
22. Salsberg E.. Sharp increases in the clinician pipeline: Opportunity and danger. Health Aff Blog. June 9, 2014 Accessed June 4, 2015
23. Altarum Institute. . Health sector employment surge intensifies. January 16, 2015 Accessed June 4, 2015
24. Iglehart JK.. Connected health: Emerging disruptive technologies. Health Aff (Millwood). 2014;33:190
25. Hofer AN, Abraham JM, Moscovice I.. Expansion of coverage under the Patient Protection and Affordable Care Act and primary care utilization. Milbank Q. 2011;89:69–89
26. Huang ES, Finegold K.. Seven million Americans live in areas where demand for primary care may exceed supply by more than 10 percent. Health Aff (Millwood). 2013;32:614–621
27. Green LV, Savin S, Lu Y.. Primary care physician shortages could be eliminated through use of teams, nonphysicians, and electronic communication. Health Aff (Millwood). 2013;32:11–19
28. Association of American Medical Colleges, Center for Workforce Studies. 2013 state physician workforce data book. Accessed June 4, 2015
29. Medicare Payment Advisory Commission (MedPAC). June 2010 Accessed June 4, 2015
© 2015 by the Association of American Medical Colleges