Study Identifies Approaches for Balancing the Mix of Generalists and Specialists in the Physician Workforce
During the early 1990s, as policy-makers began to express concern about the size of the physician workforce and its mix of generalist and specialist physicians, a number of national institutions and commissions recommended various strategies to reduce the growth of physician supply and alter the mix of generalists and specialists.
From 1992 to 1997, researchers at the University of Wisconsin-Madison Medical School conducted policy studies using different mathematical models designed to test the effects of these proposals.
Among the key findings reported in the more than 20 articles and papers produced are:
- Achieving a 50/50 mix of generalist and specialist physicians could take 50 years or longer.
- Nurse practitioners and physician assistants are already used extensively as substitutes for medical residents in teaching and other hospitals.
- HMOs use fewer physicians per 100,000 enrollees compared with the ratio of physicians to the total U.S. population.
- Large sustained cuts in the number of first-year residents being trained are needed in order to achieve targeted ratios both for the overall physician supply and for the ratio of generalists to specialists.
- Cutting residency slots currently occupied by international medical graduates alone would not sufficiently curb the growth in U.S. physician supply.
- Improving minority representation in the physician workforce would require huge increases in the numbers of blacks, Hispanics and Native Americans as first year medical residents.
The Robert Wood Johnson Foundation (RWJF) supported this project through two grants totaling $742,077.
Interest in balancing the supply and distribution of generalist and specialist physicians continued to grow during the early 1990s as policymakers watched the failure of what they had expected to be a market-driven rebalancing that began in the 1980s. The actual size of the physician workforce continued to grow during this period, and the distribution of providers reflected a continuing choice on the part of medical residents for careers as specialists over careers as generalists.
In response, a number of national organizations and commissions concerned with physician workforce issues advocated a variety of approaches to address actual and anticipated imbalances in supply and demand.
The Council on Graduate Medical Education (COGME) called for a reduction in first-year graduate medical education (GME) positions to 110 percent of US medical graduates and a target of a 50:50 generalist to specialist mix. Similar recommendations were voiced by the Physician Payment Review Commission and the Clinton Health Security Act, among others.
Against this backdrop, the Foundation awarded these grants for a series of policy studies designed to test the effects of these proposals and investigate a range of related physician workforce issues. At the time, the Foundation also supported three national programs the Generalist Physician Initiative, the Generalist Physician Faculty Scholars Program, and the Practice Sights program signed to improve the supply and distribution of primary care physicians; and one national program, Generalist Provider Research Initiative, to conduct additional research in this area.
To forecast the size of the physician workforce and its specialty composition under a variety of policy assumptions, the project created a mathematical model of the graduate medical education system.
Using this model and data from the Bureau of Health Professions, the Association of American Medical Colleges, and the American Medical Association investigators examined a range of policy options and results related to the size of the physician workforce and its mix of generalist and specialist physicians. They produced more than 20 papers, 10 of which have been published in peer-reviewed journals (see Communications and the Bibliography).
In the first phase of the project (ID# 019407), the investigators sought to determine the likely time necessary to achieve a 50:50 mix of generalist and specialist physicians, the extent to which nurse practitioners and physician assistants are already used as substitutes for medical residents, and the current mix of generalists and specialists used by five large (more than 100,000 enrollees) health maintenance organizations (HMOs) compared to the ratios of generalists and specialists for the US population.
Among the findings:
- Altering the balance of generalists and specialists will take a very long time. Assuming that 50 percent of all medical graduates entered generalist practice beginning in 1997, the proportion of all physicians who would be in generalist practice would not approach 50 percent until 2040. If generalist entry rates drop to the 20 percent level, as suggested by current medical student interest, the overall generalist practice percentage would fall to less than 25 percent by 2020 and would be 20 percent by 2040.
- Goals for generalist/specialist physician mix should be expressed in per-capita numbers (physicians per 100,000 population) instead of percentages because the total number of physicians is projected to continue growing until 2020. Therefore, any percentage will represent more physicians per capita in the future than it does now. For example, in 1990, if 50 percent of physicians were generalists, that would have translated into 79 physicians per 100,000 US population; for 2020, it translates into 130 per 100,000.
- Sharp cuts in specialist residency positions, but only modest increases in generalist positions, would be needed to achieve a 55 percent generalist physician workforce. Assuming the number of 1992 residents was held to 110 percent of US medical graduates, specialist residency positions would need to be cut by 40 percent and generalist positions would need to increase by only 3 percent to ensure that 55 percent of physicians in practice were generalists by 1999.
- Specialists are more likely to practice in urban areas, and in areas with higher per-capita income and a greater concentration of medical residents. Generalist physicians are more concentrated in rural areas and in areas with greater concentrations of elderly. Since generalists are more evenly distributed throughout the country than specialists, improving the supply of generalists should improve the distribution of physicians nationwide.
- Nurse practitioners (NPs) and physician assistants (PAs) already are used extensively as substitutes for first-year medical residents in teaching hospitals. 70 percent of teaching hospitals use PAs as substitutes, 54 percent use NPs, and 25 percent use both. Most PAs (almost half) are used in surgical departments; most NPs (about one third) are used in primary care departments. A survey of community hospitals revealed similar patterns of use.
- Staff and group model HMOs employ fewer physicians and a higher number of generalists per 100,000 enrollees compared to the United States as a whole. At the five HMOs studied, there were a mean of 127 physicians per 100,000 enrollees in 1992, of which 58 (46 percent) were generalists. In the United States, there were 180 MDs and DOs (minus residents and fellows) per 100,000 population; 71 (39 percent) were generalists.
During the second phase of the project (ID# 024109), investigators refined their mathematical model, analyzed the effects of reductions in the number of international medical graduates (IMGs) on US residency programs, examined the effects of changes in graduate medical education on internal medicine residency programs, and forecast the numbers of minority physicians that would be needed to match changes expected in the US population. Among their findings:
- The number of medical residents trained must decrease significantly to achieve recommended levels of physicians per 100,000 population. Assuming a target level of 192 physicians per 100,000 population by 2050, a figure considered "reasonable" under managed care, then the number of first-year residents in 2000 would have to be 28 percent lower than in 1995. Even if all slots occupied by IMGs were eliminated, the number of US medical graduates would still need to decrease.
- Regional growth in residency positions was strongly related to the growth in the number of IMGs. A reduction in the numbers of IMGs would have the greatest impact on cities with the greatest growth in residency positions from 1985 to 1995.
- Internal medicine residency program directors are open to reforms in GME. Program directors surveyed believed that the objectives of reform could be accommodated by internal medicine programs. Most supported institutional collaboration as a way to achieve reforms, but believed competition between larger programs in academic health centers and smaller programs in community hospitals could make collaboration difficult.
- Improving minority representation among physicians would require significant increases in under-represented minority GME residents and decreases of residents in other racial or ethnic groups. If the racial and ethnic mix of physicians is to match the mix present in the general population by 2060, the number of black first-year residents must increase 89 percent from 1995 to 2010, the number of Hispanic residents must increase 136 percent and the number of Native American residents must increase 220 percent. At the same time, the number of white first-year residents must decline 40 percent and the number of Asian residents must decrease 67 percent.
- Race and ethnicity appear to be important factors in how patients select a physician. Minority patients are five times more likely to report having a minority physician than are other patients. The phenomenon is most pronounced among Hispanic patients, who are 19 times more likely to have a Hispanic physician.
The project produced 10 published studies, which appeared in peer-reviewed journals including The Journal of the American Medical Association (JAMA), Health Affairs, Academic Medicine, The American Journal of Medicine, and the Journal of Community Health. An additional 11 papers have been submitted for publication. Project staff made eight presentations, six at annual meetings of the Association for Health Services Research between 1993 and 1997. See the Bibliography for a full list of publications and presentations.
AFTER THE GRANT
The project team is currently working on a Foundation-supported investigation of physician supply in 25 US cities (see Grant Results on ID# 030968). It has also pursued several other projects related to the research completed under these two grants:
- Affirmative action. The researchers modeled the effects of the termination of affirmative action programs in California and Texas on the future supply of minority physicians if such terminations were carried out nationally. The study found that minority admissions would drop by 75 to 80 percent, and, over time, the proportion of physicians from minority groups would diverge even further from minorities' current proportions in the general population if only GPA and MCAT scores were used to admit medical students.
- Physician immigration. An analysis of physician immigration rates begun under grant ID# 024109 is continuing. Survival analysis of Immigration and Naturalization Service data supports the findings of earlier surveys that 75 percent of foreign-born physicians entering the United States on visas remain to practice permanently in the country. These findings support the recommendations of COGME's Eleventh Report on international medical graduates.
- Residency program growth. An analysis of factors associated with the increase in size of residency programs in various regions of the United States and the effects that cuts related to GME reform would have in those regions, found that contrary to conventional wisdom, the growth of residency programs was not related to Medicare funding for GME, but instead to factors such as percent of population over age 65, growth rate of IMG residents, inpatient days per capita, presence of a medical school, and the percentage of hospital beds in the region that are in large hospitals.
GRANT DETAILS & CONTACT INFORMATION
Studies of the Mix of Generalists and Specialists in the Physician Workforce
University of Wisconsin-Madison Medical School
- Policy Studies on Generalist-Specialist Physician Mix
Amount: $ 368,518
Dates: September 1992 to October 1994
- Policy Studies on Health Workforce Issues
Amount: $ 373,559
Dates: November 1994 to May 1997
David A. Kindig, M.D., Ph.D.
(Current as of date of this report; as provided by grantee organization; not verified by RWJF; items not available from RWJF.)
Books and Reports
Rentmeester K and Kindig DA. "Physician Supply by Specialty in Managed Care Organizations." Report to The Robert Wood Johnson Foundation, August 1994.
Riportella-Muller R, Kindig DA, and Libby DL. "The National Experience with the Substitution of Physician Assistants and Nurse Practitioners for Medical Residents in Council of Teaching Hospitals' Member Hospitals." Report to The Robert Wood Johnson Foundation, January 1994.
Gray BM and Stoddard JJ. "Patient-Physician Pairing: Do Racial and Ethnic Congruity Influence Selection of a Regular Physician?" Journal of Community Health, 22(8): 247259, 1997.
Kindig DA, Cultice JM and Mullan F. "The Elusive Generalist Physician: Can We Reach a 50% Goal?" Journal of the American Medical Association, 270(9): 10691073, 1993. Abstract available online.
Kindig DA. "Counting Generalist Physicians." Journal of the American Medical Association, 271(19): 15051507, 1994.
Kindig DA and Libby DL. "How Will Graduate Medical Education Reform Affect Specialties and Geographic Areas?" Journal of the American Medical Association, 272(1): 3742, 1994. Abstract available online.
Kindig DA and Libby DL. "Domestic Production vs. International Immigration: Options for the US Physician Supply." The Journal of the American Medical Association, 276(12): 978982, 1996. Abstract available online.
Libby DL and Bagchi A. "US Physician Immigration Rates: 19861993." Unpublished manuscript.
Libby DL, Bagchi A, and Killian CD. "Generalist Physician Supply: A Test of Modeling Assumptions and a Revised Forecast." Unpublished manuscript.
Libby DL, Budetti PP, Fagan MJ, Weiss KB, Ramsay JS, and Kindig DA. "How Would GME Reform Affect the Size and Distribution of Residency Programs?" Unpublished manuscript.
Libby DL, Cooney TG and Rieselbach RE. "The Feasibility of Consortia for Internal Medicine Graduate Medical Education: Perspectives From a Survey of Residency Directors." Academic Medicine, 72(4): 301304, 1997. Abstract available online.
Libby DL and Kindig DA. "GME Downsizing: Scalpel or Scythe." Abstr Book Association Health Services Resources Meeting, 14: 215216, 1997.
Libby DL and Kindig DA. "Threats to Diversity in Medical Education and the Physician Workforce." Unpublished manuscript.
Libby DL and Kindig DA. "Can Medical Education Policy Redistribute Physicians Geographically?" Unpublished manuscript.
Libby DL, Kindig DA and Rentmeester K. "Network/IPA Model HMO Physician Staffing Study: Final Report." Unpublished manuscript.
Libby DL, Z. Zhou and Kindig DA. "Will Minority Physician Supply Meet U.S. Needs?" Health Affairs, 16(4): 205214, 1997. Abstract available online.
Rieselbach RE, Libby DL and Cooney TG. "The Potential Impact of Graduate Medical Education Policy Initiatives on Internal Medicine Residency Programs: Response from a Survey of Program Directors." American Journal of Medicine, 100(3): 251256, 1996.
Riportella-Muller R, Libby D and Kindig D. "The Substitution of Physician Assistants and Nurse Practitioners for Physician Residents in Teaching Hospitals." Health Affairs, 14(2):181191, 1995. Abstract available online.
Riportella-Muller R, Kindig DA and Libby DL. "Geographic Variation in the Supply of Generalist Physicians." Unpublished manuscript.
Stoddard JJ, Gray BM, Libby DL and Kindig DA. "Does the Number of Specialists Affect State Health Expenditures?" Unpublished manuscript.
To CW, Libby DL and Kindig DA. "Factors Affecting Geographic Distribution of Physician Assistants and Nurse Practitioners." Unpublished manuscript.
Presentations and Testimony
Gray BM and Stoddard JJ. "Patient Physician Pairing and Race: Do Minority Physicians Disproportionately Serve Minority Populations?" at the annual meeting of the Association for Health Services Research, Chicago, Ill., June 1995.
Kindig DA and Libby DL. "Domestic Production vs. International Immigration: Options for the US Physician Supply," at the annual meeting of the Association for Health Services Research, June 1996.
Libby DL, Riportella-Muller R, and Kindig DA. "Factors Associated with Primary Care Physician Location," at the annual meeting of the Association for Health Services Research, Washington, D.C., June 1993.
Libby DL and Kindig DA. "The Impact of Graduate Medical Education Reform on Specialties and States," at the annual meeting of the Association for Health Services Research, San Diego, Calif., June 1994.
Libby DL and Kindig DA. "GME Downsizing: Scalpel or Scythe?" at the annual meeting of the Association for Health Services Research, Chicago, Ill., June 1997.
Riportella-Muller R, Kindig DA and Libby DL. "State Variation in the Supply of Primary Care Physicians," at Third Primary Care Research Conference, Agency for Health Care Policy and Research, Atlanta, Ga., January 1993.
Stoddard JJ, Gray BM, Libby DL and Kindig DA. "Does the Number of Specialists Affect State Health Expenditures?" at International Conference on Health Economics, Vancouver, B.C., November 1996.
Zhou Z, Libby DL and Kindig DA. "Will Future Minority Physician Supply Meet National Need and Demand?" at the annual meeting of the Association for Health Services Research, Chicago, Ill., June 1995.
Report prepared by: Richard Camer
Reviewed by: Timothy F. Murray
Reviewed by: Marian Bass
Program Officer: Michael P. Beachler