The javascript used on this site for creative design effects is not supported by your browser. Please note that this will not affect access to the content on this web site.
Skip Navigation
Department of Health and Human Services
Council on Graduate Medical Education

Summary of Eleventh Report

International Medical Graduates, the Physician Workforce, and GME Payment Reform

March 1998

Executive Summary & Recommendations

It has long been the position of the Council on Graduate Medical Education (COGME) that the United States has too many physicians and these physicians are not appropriately distributed across medical specialties and geographic locations. In recent years the growth of managed care organizations has reduced the demand for physicians' services and magnified the size of the projected physician oversupply. To achieve a better balance between physician supply and demand and more appropriate distributions across specialties and locations, COGME has advocated three broad policy goals: (1) reduce the number of first year GME positions from 140 percent of the number of graduates of accredited schools of medicine and osteopathy to 110 percent; (2) encourage half the residents completing their training each year to enter primary care specialties; and (3) continue support for residency programs that provide care to substantial numbers of underserved people.

Table 1 - All Residents in Allopathic Programs

Academic Year


1988-89
1989-90
1990-91
1991-92
1992-93
1993-94
1994-95
1995-96

Foreign-Born
IMGs

7,227
8,726
10,949
12,881
15,621
18,558
21,199
22,565

U.S. Citizen
IMGS

4,329
4,595
5,067
5,258
5,272
5,162
4,481
4,198

Total IMGs


11,556
13,321
16,016
18,139
20,893
23,720
25,680
26,763

USMGs


71,235
73,675
75,762
77,016
77,716
78,562
78,074
77,849

Total


82,791
86,996
91,778
95,155
98,609
102,282
103,754
104,612

Source: Residency data from Association of American Medical Colleges

Table 2 - All Residents in Allopathic Programs

Academic Year


1988-89
1989-90
1990-91
1991-92
1992-93
1993-94
1994-95
1995-96

Temporary Visa IMGS*

2,173
2,299
3,615
5,041
6,787
9,325
11,068
11,545

All Other IMGS+

9,383
11,022
12,401
13,098
14,106
14,395
14,612
15,218

Total IMGs

11,556
13,321
16,016
18,139
20,893
23,720
25,680
26,763

USMGs


71,235
73,675
75,762
77,016
77,716
78,562
78,074
77,849

Total


82,791
86,996
91,778
95,155
98,609
102,282
103,754
104,612

Sources: Residency data from Association of American Medical Colleges; J-1, J-2, & H-1 B temporary data from various Journal of the American Medical Association medical education issues.

* This category includes IMGs on J-1, J-2, and H-1 B visas
+This category includes both permanent resident and U.S. citizen IMGs

COGME's 1997 Recommendations on the Physician Workforce - International Medical Graduates and GME Payment Reform

I. Reduce the Number of Graduate Medical Education (GME) Positions

A. Eliminate both Medicare direct graduate medical education (DME) and indirect medical education (IME) payments for new exchange visitor (J-1 Visa) residents beginning the year following implementation of this provision. Fund new exchange visitor residents from alternative sources, such as home country financing or foreign aid. Continue to make DME and IME payments for those exchange visitor residents who entered training prior to implementation.

COGME believes that Medicare GME payments should be available only to those residents expected to become part of the U.S. physician workforce. The original intent of the physician exchange visitor program was to strengthen international relations and further mutual understanding through educational and cultural exchange; the program was not intended to add physicians to the U.S. physician workforce. This recommendation would apply only to new exchange visitors; Medicare GME payments would continue to be made for exchange visitor residents already in the training pipeline.

COGME believes that GME funding for exchange visitor residents should come from either foreign sources or U.S. provided non-Medicare sources such as foreign aid or private sector sponsored assistance. Funding from sources such as the Agency for International Development (AID) or the private sector can provide a continuing U.S. commitment to preserving the exchange visitor program.

B. Base hospitals' resident counts for both Medicare DME and IME payment determinations on a 3-year rolling average beginning with the two years prior to implementation of this provision. This would provide a temporary financial cushion and incentive to reduce the number of residents reimbursed by Medicare.

  • On a hospital-specific basis, cap the total resident counts for both DME and IME payment determinations, and either

    - cap the non-primary care resident count;
    -or -
    - maintain or increase the primary care proportion of residents, at the level in the year prior to implementation of the cap, to limit further increases in the number of residents reimbursed by Medicare while protecting the number or proportion of primary care residents.

  • Cap the individual resident-to-bed ratio (IRB) ratio on a hospital-specific basis at the level prior to implementation of the cap, to prevent the hospital's IRB ratio and IME payments from increasing because of a decrease in the hospital's inpatient bed capacity.

Computing Medicare payments on a three-year rolling average of annual residents would reduce payments proportionately less than the number of residents. The rolling average provides hospitals with a limited incentive to reduce the number of residents and enable a smoother transition toward reducing the number. COGME believes that this "cushion" will provide resources for hospitals to restructure their organizations more efficiently, be less dependent upon residents for services, and include the appropriate use of other health professionals. By tagging the payment response to changes in resident numbers, this recommendation also provides an incentive not to increase the number of residents. Capping residency counts eligible for DME payment facilitates market forces by removing the hospital's incentive to increase the number of residents. Capping the IRB ratio removes any incentive hospitals might have to increase the number of residents or decrease the number of beds in order to raise the ratio. Additionally, capping non-primary care positions prevents any resident reductions from impacting disproportionately primary care residents, and allows flexibility to expand the training of primary care residents.

C. If recommendations I.A. and I.B. are not enacted, encourage additional demonstrations analogous to New York GME Demonstration Project, especially in states with high resident per-capita ratios.

A small number of states are especially dependent upon service delivery by residents. Hospitals in these states may need additional support to make a transition to an ambulatory-based service and training environment. The New York Medicare GME Demonstration Project provides:

  • incentives for hospitals to reduce the size of residency training programs in the state while providing transition funds to support the reorganization of service delivery and use of replacement personnel required, and

  • a somewhat smaller required reduction in resident counts if hospitals agree to promote primary care or to participate in a formal consortia with coordinated GME programs.

D. If, within three years of implementation of these recommendations, significant progress has not been made toward reducing the number of first-year residency positions to 110 % of 1993 U.S. medical school graduates, consider stronger policies aimed at reducing DME and IME payments that would result in a decrease in first-year residency positions to 110 % of 1993 U.S. graduates.

COGME believes that the use of financial incentives, such as those proposed, is the preferred method to adjust the market production of residents, particularly at a time of rapid change in the health care industry. However, there is a need to reduce both the number of first-year residents and the cost of GME. If the combination of Recommendation I.B. and the market does not accomplish the necessary reduction in residency positions and produce a better balance between physician supply and future requirements as advocated by COGME's "110%" recommendation, other measures may need to be considered to accomplish that goal.

II. Support for Communities and Shortage Areas Affected by Loss of Residents

Use a portion of the savings from these recommendations to support programs such as Community or Migrant Health Centers and the National Health Service Corps, where a substantial decline in residents creates continuing severe service shortages.

A substantial decline in the number of physician residents in communities that rely heavily on their services may produce severe service shortages. Allocating a portion of medicare savings to community-based service delivery programs, particularly "safety-net "programs, will permit hospitals to move to a more appropriate service delivery environment for both training and care.

III. Revise the Temporary Visa Programs

A. Phase out over a 4-year period the granting of J-1 waivers for purely service reasons as a move toward restoring the exchange visitor program to its original purpose. At the same time, the policy of considering waivers for uniquely qualified researchers in nationally and internationally significant research efforts should be continued.

The four-year phase out of the service-based waivers would allow development of domestic program strategies to provide long-term, permanent solutions to alleviate physician maldistribution. On the other hand, waivers for researchers will encourage medical and health related research of an advanced technological nature that would produce benefits to both the Nation and the international community.

B. Increase the J-1 visa return-home period from 2 to 5 years for exchange visitor physicians.

A five-year minimum return home period allows J-1 visa holders sufficient time to contribute the benefits of U.S. training to their home country physician workforce and permits reasonable time for reacculturation. This should reduce the probability that J-1 visa holders will return to the U.S.

C. Eliminate use of the H-IB visa program for physician residency training.

The elimination of H- IB visa program would stop the use of the H-IB visa to circumvent the J-1 visa "return home" requirement.

IV. Enhance Primary Care Residency Training

A. Provide Medicare DME payments to a wide variety of ambulatory teaching settings, including managed care plans.

B. Include time spent in ambulatory settings outside the hospital in the calculation of Medicare IME payments to hospitals.

C. Make Medicare IME payments to ambulatory settings outside the hospital when ambulatory cost estimates have been developed.

D. Identify and remove the DME and IME components of the Average Adjusted Per Capita Cost (AAPCC) from Medicare capitation rates and utilize these funds specifically for GME purposes.

Practitioner competency is dependent upon training in appropriate settings such as in community-based ambulatory sites. Physicians trained to provide primary care in ambulatory settings can provide comprehensive, continuing, longitudinal care to patients. The policy of providing direct and indirect GME payments only for hospital-based residents or DME payments to residents rotating in hospital based ambulatory clinics has restrained appropriate training for all physicians, generalists in particular, to provide such care. Medicare IME payments to ambulatory settings would provide a strong incentive to initiate such training.

The AAPCC payment system for Medicare risk HMO contractors presents a difficulty in financing HMO residency training. AAPCC payments include equivalent amounts of DME and IME for a relevant geographic area, but these GME dollars are not identified in the AAPCC and are paid regardless of whether the HMO engages in residency training. As a result, Medicare GME funds are spread among all HMO contractors, without being focused on those that actually have teaching programs or necessarily use teaching hospitals for services.

These amounts should be removed from the AAPCC and made available for GME in a wide variety of teaching settings, including teaching hospitals and managed care organizations and HMOs with teaching programs, to help rectify possible inequities to teaching hospitals and eliminate the current disincentives to HMOs who wish to establish or expand residency training activities but do not currently receive explicit reimbursement for their efforts.

E. Support Public Health Service Act Title VII education programs, which have ultimate underserved practice as a goal.

Efforts need to be strengthened to encourage the domestic production of competent generalists who will serve in these areas. Most Title VII physician education programs operate under a statutory funding preference for applicants who demonstrate success in placing graduates in underserved communities.

F. Encourage new generalist residency programs by permitting exceptions to Medicare GME payment caps (as proposed in recommendation I.B.) for new primary care residency programs in geographic areas with shortages of physicians, including residents.

If Medicare GME payment caps are enacted, there should still be opportunities for Medicare DME and IME payments to encourage primary care residency programs in areas where relatively few or no programs exist. Residents frequently remain near the hospital where they received their residency training. The competency of primary care practitioners is dependent upon training in the proper settings such as in ambulatory sites in community-based hospitals. Promotion of new primary care residency programs can provide this type of training to physicians. These primary care physicians can offset the adverse impact of residency reduction and changes in the health service delivery environment by providing continuing, longitudinal, comprehensive general care to Medicare beneficiaries and vulnerable populations.

G. For DME, reinstate the 1994-1995 freeze on non-primary care per-resident amounts for a two-year period, while continuing the Consumer Price Index for Urban Areas (CPI-U) updates for primary care per-resident amounts.

Such a freeze has historical precedent. Freezing payments to non-primary care training programs while continuing for two years the CPI adjustment for primary care programs creates a payment differential in favor of primary care programs, carried forward as the CPI adjustments are resumed for both primary and non-primary care training programs payments. COGME believes that this differential in payments will motivate hospitals to shift the specialty training more in favor of primary care residency training or at least not reduce primary care training. This differential should be examined periodically for effectiveness.

It is COGME's intention that this freeze not adversely affect the recruitment and retention of minority residents in any specialty. COGME is on record as considering under represented minority participation and advancement in medicine as particularly critical for the Nation.

V. The United States Role in International Medical Education

A. Recommend to the World Health Organization that other countries engage to a greater degree in physician workforce analysis and planning.

B. The U.S. government should cease to support undergraduate medical education of U.S. students in foreign countries through loans.

Currently, the Department of Education is in the process of reviewing the credentialing requirements for medical schools in other countries. The purpose of this review is to ascertain if credentialing requirements are similar to those of the Liaison Committee for Medical Education (LCME). Eligibility of foreign medical schools to participate in the Federal Family Education Loan program for U.S. citizen medical students would be accorded only to those schools with LCME-like credentialing requirements. U.S. citizens who receive medical training in such schools would improve their likelihood of becoming ECFMG certified and accepted into a residency training program. Satisfaction of LCME-like requirements would reduce the number of foreign medical schools eligible to provide U.S. government funded support for U.S. citizen students.