Mr. Chairman, thank you for inviting me here today to talk about the President's FY 1998
graduate medical education training proposals and the recently announced New York
Graduate Medical Education (GME) demonstration project. It is an opportunity I welcome,
since I have been involved in discussions of these issues for the last 20 years, as a charter
member of the New York State Council on Graduate Medical Education, as a. member of the
Prospective Payment Assessment Commission, and in many other roles. After all these years
of discussion, it is gratifying to believe that this may finally be the year in which some of
these reforms come to fruition.
Medicare payments are designed to adequately pay for care provided to Medicare
beneficiaries, including the direct and indirect costs of medical education. However, there is
broad consensus that there is an oversupply of physicians. In addition, with the increase in
enrollment in managed care and treatment in outpatient settings, many believe there is a
distributional problem, with too many specialists and not enough primary care physicians in
areas where they are needed. Despite the national consensus on oversupply, Medicare
currently provides hospitals with financial incentives to increase the number of residents they
The high-quality research and medical education that occurs in our academic health centers is
internationally recognized, and is something we should take pride in, especially since
Medicare has played an integral role in supporting medical education. It is important that
Medicare continue to play an important role. Over time, it has become clear to us that the
incentives provided through Medicare payment policy have overpowered market incentives
that might affect the number of residents. Some historical background on this will provide
context, after which I will describe our current legislative proposals which are designed to
more clearly align the incentives of Medicare payment policy with those of the marketplace.
Historically, Medicare paid for inpatient hospital services, including the training of interns
and residents, on a cost basis. However, when Medicare began paying for inpatient hospital
services through the Prospective Payment System (PPS) in 1983, we recognized that teaching
hospitals serve a special population and that the teaching process, itself, means that care in
teaching hospitals is more costly than in other hospitals. For this reason, Medicare makes an
adjustment to the payment rates for Medicare discharges provided in hospitals that train
residents. This adjustment, the indirect medical education (IME) payment, recognizes that the
higher costs of treating patients in teaching hospitals. Under current law, the IME payment is
based on the teaching hospital's ratio of interns and residents to beds. Because the hospital's
IME payment is based on this ratio, the hospital will receive a higher indirect teaching
adjustment for each Medicare discharge if it trains more residents. This system provides
incentives for hospitals to train more interns and residents. Medicare's IME payments
increased from $2.9 billion in FY 1990 to $5.1 billion in FY 1995, an increase of more than
The current formula for determining Medicare's indirect medical education payments provides
approximately a 7.7 percent increase for each 10 percent increase in a hospital's ratio of
interns and residents to beds. Studies by HCFA, the General Accounting Office, DHHS
Inspector General, and the Prospective Payment Assessment Commission have shown that
this formula compensates teaching hospitals in excess of the additional costs of treatment. If
enacted, the President's proposals would gradually bring Medicare's indirect medical
education payment adjustment down to 5.5 percent in 2002, closer to the actual indirect costs
of treating a Medicare patient in a teaching hospital.
When Congress created the Medicare PPS in 1982, it excluded direct hospital costs associated
with approved educational activities from payment through the PPS system. Since 1985,
Medicare has paid teaching hospitals for the direct costs of GME training on a per resident
amount. Direct GME costs include such items as salaries and fringe benefits for interns and
residents, supervisory teaching physician costs, and miscellaneous overhead associated with
operating a medical residency training program. Under this payment system, Medicare
determines a hospital's per resident costs in a base year and pays that amount for each intern
and resident working in the hospital, adjusted for Medicare inpatient utilization. Since 1987,
we have also been paying hospitals for intern and resident time spent in ambulatory training,
when the resident is involved in patient care and the hospital continues to pay the resident's
salary. Medicare payment for the direct costs of medical education is the second way in
which Medicare provides incentives for hospitals to train more interns and residents.
Medicare's direct GME payments have increased from $1.3 billion in FY 1990 to
approximately $2.0 billion in FY 1995, an increase of 50 percent.
Health resource experts have focused on the overall supply of physicians, the distribution of
physicians across the country, and the mix of physician specialists. Organizations such as the
Institute of Medicine, the Council on Graduate Medical Education, and the Pew Charitable
Trusts have raised concerns that the Medicare program is providing incentives to train too
many residents. Furthermore, given Medicare's policy of paying hospitals for training, these
and other organizations also have been concerned that Medicare provides incentives to train
too many specialists in inpatient hospital settings. Health experts have noted that medical
practice is changing; there is increased emphasis on primary care training, and the growth of
managed care means a greater demand for physicians who have strong skills in providing
primary care services. Just two weeks ago, the American Medical Association, the American
Association of Medical Colleges, and other national medical organizations held a press
conference, at which they issued a consensus statement confirming the oversupply of
physicians and the need for the Federal Government to realign its GME funding policies.
At the present time, Medicare does provide some financial incentives that promote training
of care physicians. First of all, Medicare's direct medical education payment reduces payment
for training beyond the time needed to train in one specialty. Second, Medicare pays hospitals
for direct medical education when a resident works in an ambulatory training site when the
hospital continues to pay the resident's salary. I should note, however, that under current law
we cannot count this resident time in the indirect medical education adjustment paid to
hospitals. And, since the indirect medical education adjustment is typically a much larger
share of the hospital's total medical education payments, Medicare's policy of not counting
resident time in an ambulatory setting for indirect medical education has been a significant
disincentive for hospitals to provide ambulatory care training.
Historically, when hospitals were paid charges or costs by all payers, the costs of medical
education were shared by all payers. However, as managed care companies and other insurers
negotiate with hospitals for lower payment rates, hospitals increasingly rely on Medicare to
help finance medical education. Over the next few years, we will continue to see major
changes in the configuration of the health services workforce. However, the need for large
sums of money to support the education of our health care providers will not disappear.
While it is the responsibility of government to support activities that are essential to the
welfare of its people, that does not preclude the need for other interested parties to do their
Senator Moynihan and others have introduced legislation to ensure that private payers make a
contribution to the cost of medical education as well as the public sector. There is a great deal
of interest in these proposals and we look forward to working on this issue with Senator
Moynihan and other Members in this session of Congress.
THE PRESIDENT's FY 98 BUDGET PROPOSALS
The President's FY 1998 Budget Proposals include a number of provisions affecting GME.
These Proposals are designed to more closely align the incentives created by Medicare
payment policy with market incentives, and to help achieve the following commonly
- to stop the rapid growth in the number of medical residency positions,
- to encourage more training in primary care,
- to encourage more training in non-hospitals settings for all residents,
- to provide more payment equity for teaching hospitals when they serve Medicare
managed care enrollees.
Our legislative proposals are not designed specifically to redistribute primary care and
specialty residency slots or to force programs to downsize. We do not think that it is
appropriate for Medicare to become involved in the micro-management of the details and
operations of individual residency programs; any attempt to do so would be fraught with
difficulties. Given the dependence of many teaching hospitals on Medicare GME funding,
any major or overly prescriptive changes in payment policy could have serious ramifications.
Let me now briefly describe our GME proposals in the President's FY98 Budget.
I. Cap on the Number of Residents:
The number of residents increased from 107,850 in FY 1990 to 129,434 in FY 1994, an
increase of 20 percent. To stop the rapid growth in the number of residents, we would limit
Medicare's payment to the number of residency slots at each teaching hospital in the 1996
residency year (ending June 30, 1997). We also would cap the number of specialist residents
at the number the program had in the same residency year. In this way, hospitals could
increase their number of primary care residents, as long as they reduced their numbers of
specialty residents, but not vice versa. We would also cap Medicare's payment for IME, so
that the hospital could not inflate the intern and resident to bed ratio to gain additional
payments by reducing its number of beds. This reform is supported by the American Medical
II. Use three-year rolling average of FTEs for GME and IME payments:
We believe that many hospitals continue to increase the numacilities will have to participate in
accredited teaching programs and bear the costs of the residency program (i.e., resident
salaries). We realize that other non-hospital settings, such as ambulatory surgical centers and
physician offices, are also important training sites. Our proposal also would give the
Secretary the authority to make GME payments to other non-hospital sites in the future and
would be an important step toward improved medical care access in rural areas.
IV. Allow hospitals to count residents in non-hospital settings for IME purposes:
To encourage more training in non-hospital care settings, and to eliminate the current
disincentive for hospitals to allow residents to rotate to these non-hospital settings, we
propose to allow hospitals to include such residents in the calculation of their IME payments.
Currently, hospitals cannot include residents in, non-hospital settings, other than in the
hospital's own outpatient department. To prevent additional increases in Medicare's IME
payments, this proposal would not allow the resident-to-bed ratio to exceed the ratio that
exists in the base year. However, the hospital's payment would not decrease if a resident were
transferred to a non-hospital setting, as long as the hospital continued to pay the resident's
salary. We believe this proposal will make hospitals more inclined to let residents rotate to
V. Reduce IME formula:
We also propose to decrease the IME formula to align Medicare's payments more closely with
the actual additional costs attributable to teaching. Our proposal would decrease the IME
payment so that instead of hospitals receiving a 7.7 percent increase in payments for every 10
percent increase in their resident to bed ratio, they would receive 7.4 percent in FY 1998, 7.1
percent in FY 1999, 6.8 percent in FY 2000, 6.6 percent in FY 2001 and 5.5 percent in FY
2002 and thereafter. HCFA, GAO and ProPAC analyses over the years have shown that an
adjustment in excess of 4 to 5 percent is not empirically justifiable. There was clear bipartisan
support for reducing the indirect medical education payments, and similar proposals appeared
in all of the budget reconciliation bills last year.
VI. Make GME/IME payments to hospitals for Medicare managed care enrollees:
Finally, while we do achieve savings in the IME formula and GME reforms, we propose to
extend our support to teaching hospitals by directly providing them medical education
payments of $10.7 billion over five years for their treatment of Medicare managed care
enrollees. Mr. Chairman, our proposal is very similar to the initiative you reported out last
year. As you well know, under current law, Medicare managed care contractors negotiate
payment rates with teaching hospitals. Many managed care contractors are not willing to pay
explicitly for the costs incurred in medical education. Medicare's payment rates to managed
care contractors are based on fee-for-service (FFS) spending in the county, and include
Medicare spending for medical education. Under our proposal, Medicare would pay teaching
hospitals directly for direct and indirect medical education. This payment would be based on
the number of Medicare managed care enrollees treated by the hospital. The mechanics of
this process would be simple: hospitals would provide bills for managed care enrollees, which
would enable us to calculate how much GME and IME the hospital would have received for
those cases under FFS. The hospital would receive a separate payment from HCFA for their
managed care enrollees. This would free teaching hospitals to negotiate more competitive
rates with managed care contractors, since their costs of teaching would be reimbursed
separately. We look forward to working with you on this and our other proposals.
GME REFORM PROJECTS
We realize the significant impact that changes to Medicare payment policy will have on
residency programs, and we appreciate the importance of residents in the delivery of health
care. For these reasons, HCFA is also supporting several projects that test medical education
reforms; demonstrations are very useful in testing the appropriateness of specific payment
methodologies in advance of full implementation.
The New York Medicare GME Demonstration
New York hospitals train a formidable 15 percent of the country's residents and receive 20
percent of Medicare's annual spending on graduate medical education activities, $1.4 billion
out of $7.0 billion in FY 1996. In view of New York's large number of residency slots,
HCFA recently announced a major Medicare GME demonstration project in New York which
serves a number of important national goals. It will lead to a reduction in the overall number
of residents, it will increase emphasis on primary care training and on training in ambulatory
sites, and it is designed to help teaching hospitals redirect their graduate medical education
programs so they can meet the demands of a more competitive health care market.
In New York, the number of residents increased 11.1 percent in the last four years. Although
many hospitals report that they have been considering downsizing their teaching programs in
response to competitive pressures, we have seen little actual evidence of movement. Both our
actuaries and the Congressional Budget Office assume continued modest growth in the
number of residents nationally over the next several years.
In New York's hospitals, teaching and service delivery missions have been highly
interdependent due largely to the State's regulatory system and to the need to serve a large
indigent population. In part because of favorable payments for residents under Medicare's
GME rules, New York hospitals have become highly reliant on residents to provide care. In
an increasingly competitive health care market, and given the widespread concern about
oversupply of physicians, such a heavy emphasis on residents is no longer appropriate.
This demonstration will help New York's hospitals transition to a more focused and
innovative teaching mission by providing transitional payments to assist participating
hospitals re-engineer the ways they provide care by reducing the number of residents by 20-25
The project is voluntary and the requirements for participation of hospitals are quite
demanding. HCFA set the overall thresholds for reductions -- while protecting primary care
training -- and required applicants to present strategic plans for both training and service
delivery. Participating hospitals and their medical leadership are responsible, within the
demonstration's parameters, for determining the implementation strategies that fit their unique
teaching missions and the service needs of their communities.
Specifically, 42 hospitals, either singularly, in joint ventures, or consortia, will participate in
Phase I of this demonstration. Each applicant selected an option to decrease the number of
residents they train by either 20 or 25 percent over the six-year life of the demonstration.
Hospitals that agree to increase the fraction of primary care residents they train by 20 percent,
or that are part of a formal consortium with coordinated GME programs, must reduce their
resident counts by 20 percent. All other hospitals must reduce their number of residents by 25
percent. Participants that fail to meet the agreed-upon targets face loss of all transition
payments. The number of resident slots reduced will be more than 2,000, from a base in New
York of over 14,000.
Medicare will make transition payments that will gradually decline to zero over the life of the
demonstration. In any year, these payments are a fraction of, and never more than, what
Medicare would have paid in the absence of the demonstration. We expect these transition
payments to total around $400 million over the demonstration period, but it is important to
realize that participating hospitals will in fact receive less in Medicare GME payments than
they would have in the absence of the demonstration. The transition payments are not "new"
money, over and above baseline spending, but are a portion of the reduced Medicare spending
resulting from the reductions in resident counts from the levels assumed in the baseline.
Overall, we estimate that under current law, Medicare will achieve net savings -- even after
making the transition payments - of at least $300 million, and perhaps as high as $650
million, over the six years of the demonstration. We are currently working to refine this
Over the longer run to the extent resident reductions are permanent and to the extent
participants create better training programs and more efficient and competitive service
delivery infrastructures, longer term savings to the Medicare program are also possible.
Later this year we expect to develop an additional aspect of the project designed to emphasize
care in ambulatory training sites. As medical care has moved to ambulatory settings, medical
training should do likewise. However, developing ambulatory training -- which requires both
mentors and space in clinics and doctors' offices -- is difficult, and Medicare payment policy
inhibits training outside the hospital.
A number of other organizations have expressed interest in the New York project. The
President's legislative proposal for basing Medicare GME payments on a three-year rolling
average of the count of residents would be very similar in effect to the way hospitals are
treated under the New York demonstration. This proposal could help other hospitals receive
transition payments as they reduce the number of residents they train.
Utah's GME Proposal
We have been working with the medical education community in Utah as they develop a
proposal for a consortium to coordinate all GME activity in the State. This consortium would
set training priorities and make resource allocations for GME in the State, based on work
force planning and goals. As planned, the consortium would receive funds for GME directly
from Medicare and Medicaid and possibly from private insurers as well. There is currently
legislation in front of the Utah State legislature to create this consortium. Once this
legislation has passed, we expect to work with the medical education community in Utah to
try to develop a proposal for a demonstration.
Michigan's Medicaid GME Proposal
Michigan is intending to revamp how its Medicaid program pays for GME, and we have been
in consultation with them about their proposed reform. The State is striving to achieve a
better alignment between the number and type of medical professionals trained and the
medical needs of its Medicaid population. Michigan would like to recapture GME funds that
have been included in its Medicaid managed care capitation payments and redirect those
payments to hospitals and managed care organizations that have entered into collaborative
partnerships between university medical and health professional schools.
The State would use three methods in its restructured GME payment system. One would
distribute funds based on historical cost of direct and indirect medical education. Another
would make payments based on the number primary care residents drawing salaries at each
hospital. The third would provide incentive payments to hospitals or managed care entities
for innovations in health professions education. This new design does not require
demonstration waivers and we have been working with the State to implement it in a way that
conforms with Medicaid payment rules.
There is broad consensus among the various players on the goals for reforming Medicare
payment for graduate medical education. We all agree that we need to let the market
appropriately determine how many and what types of residents are trained, and Medicare
should not be in the business of micro-managing our Nation's medical education programs.
The high-quality research and medical education that occurs in academic health centers is
internationally recognized, and we are rightly proud of their contributions. Medicare has
played an integral role in supporting medical education and will continue to do so. Clearly,
there is a need for reform, and I am heartened by the level of consensus that exists around the
various proposals for medical education reforms. I look forward to working with you and the
other members of this Committee to pass these reforms.