At the Annual Meeting of the Academic Orthopaedic Society in November 2001, Kenneth M. Ludmerer, author of Time to Heal: American Medical Education from the Turn of the Century to the Era of Managed Care
1 , stressed that education is the academic health center's reason for being
2 . (An academic health center refers to one of the 125 recognized institutions in the United States that has a medical school, a faculty practice plan, and an associated hospital. The term includes closely affiliated hospitals.) These centers are currently facing fierce competition from private markets, increasing pressure to self-support their varied and costly social missions, and shrinking federal support in the form of graduate medical education funding. As these institutions are being threatened financially, it is reasonable to be concerned about the quality of education that they provide.
Michael Simon, in an article in 2001, contended that the academic orthopaedic community must stress education over throughput, i.e., residents as students over residents as service providers
3 . However, teaching residents requires a great deal more effort and resources than does training residents. Thus, one key limitation to providing quality education is the amount of funding. Although it is difficult to quantify with precision, the direct cost of education alone, which includes resident stipends, teaching physician compensation, and associated overhead costs of residency programs, has been estimated to be $71,468 per resident in 1997
4 . Other, indirect costs, however, are thought to more than double this estimate.
Currently, the only explicit funding sources for resident education are Medicare and, in most states, Medicaid, which provide roughly $6 billion and $2 billion, respectively, each year for expenses related to graduate medical education
5 . Although these payments are deemed support for "medical education," their purpose is much broader in scope. These dollars, in fact, finance many of the additional social and academic missions that accrue to the nation's academic health centers.
Given their multiple uses, Medicare payments for graduate medical education are made directly to the academic health center administration, with only a variable portion making its way to the program directors at each institution. Typically, there is an understanding between the departments and the administration as to how much money is passed through for educational activities as well as how much is allocated to support faculty salaries for teaching time
6 . This system of payment for graduate medical education (directly to the academic health center) makes funding for education dependent on the importance that the hospital administration places on education as well as that institution's financial strength and fiscal strategy.
The current state of diminishing support for graduate medical education has important implications for residency education. A decrease in public support (chiefly through Medicare) and private support (mainly through managed care initiatives) for graduate medical education, use of funds for purposes other than educational programs, and the complicated and indirect allocation pathway for these funds are of most concern. Together, these factors continue to challenge the academic environment and force academic health centers to focus on fiscal solvency, rather than education, as their chief concern. In order for academic health centers to provide their unique and nationally important role of educating physicians, sufficient financial resources must be restored. This objective can only be achieved through policy change that aims to provide ample, stable, and dedicated educational funding.
Academic health centers provide a unique set of services compared with other health provider systems. These services include three academic pursuits: specialty care, research, and education. In addition, academic health centers have become a safety net for the poor, uninsured, and medically complex patients. As a result of these differences, academic hospitals have been shown to cost 30% to 50% more to operate than nonacademic hospitals
7,8 . The Commonwealth Fund, a private foundation that supports independent research on health and social issues, contended in a recent study
9 that a case at an academic health center costs, on the average, $10,655, whereas a case at a nonteaching hospital costs about half of that figure ($5034) (
Fig. 1 ). The same study suggested that academic mission-related activities at an academic health center account for 30% of the per case cost and case-mix differences account for 17% of the per case cost. (Mission-based activities include graduate medical education, standby capacity, and clinical research.) In another comparison, the cost per case at academic health centers was $9901, whereas the cost per case at nonteaching hospitals was $5412
10 .
Academic Health Centers Provide Special Services and Introduce New Modalities in Treatment
Academic health centers play a critical role in assuring access to specialty care and services that are not routinely available at community-based hospitals. Positron emission tomography scanning is an example of a costly service that is often provided only at academic health centers. Such institutions are also more likely to have highly specialized clinical care units. Despite accounting for only 2% of the country's hospitals, they contain 20% of the burn units and 34% of the level-I trauma centers
7 . Given appropriate financial support, academic centers are well suited to provide this specialty care. Pooling such care at these large institutions has been shown to improve outcomes and reduce the overall cost for many diagnoses
11-14 . Furthermore, the provision of such care is conducive to the teaching and research that occur within academic health centers.
Case-Mix Differences
Institutions that have resident-to-bed ratios of >0.25 (so-called major teaching hospitals) include only 6% of the hospitals in the United States but provide 41% of the uncompensated care
15 . In addition to generating less direct income for the hospital (which is discussed in more detail in the next section), the uninsured have been shown to present with later-stage disease, have more severe illness, and have higher overall mortality rates
16,17 .
Academic health centers also provide a disproportionate amount of care to both patients with AIDS (acquired immunodeficiency syndrome)
18 and transfer patients
7 . As a group, transfer patients are more severely ill and require more complex treatment
19,20 . Within academic health centers, the percentage of total hospital admissions represented by transfer patients increased from approximately 4% in 1988 to >8% in 1995, whereas the transfer rate was no more than 3% for nonteaching hospitals at any point during the same period
21 . While transferring complex patients to academic institutions may represent good clinical care, the net effect is that academic health centers (and the physicians who receive the transfers) lose money on many of these patients as reimbursement often does not compensate for the increased level of care. For example, the payment received for revision arthroplasties of both the hip
22 and the knee
23 has been shown to be insufficient in comparison with that received for primary replacements despite the fact that revisions require more physician time and skill, as well as more hospital resources.
In a July 2000 report, the Commonwealth Fund Task Force concluded that the greater severity of illness treated at academic health centers is not captured by the reimbursement coding systems
7 . Diagnosis-related group (DRG) reimbursement, which pays a set amount for each patient within a disease category regardless of severity, is one cause of inadequate compensation. Outlier payments are designed to ease the financial burden of an extremely costly case. However, they take effect only when cases are 2.5 times more costly than the median for that diagnosis-related group.
Research
In addition to specialty care, academic health centers are also well suited to conduct research as they combine bench-top studies with a broad patient population. These institutions currently host 28% of all biomedical research nationwide
7 . The federal government has been the biggest supporter of biomedical research, mainly through National Institutes of Health funding. Despite a budget totaling $23.6 billion in 2002 ($448.7 million of which went to the National Institute of Arthritis and Musculoskeletal and Skin Diseases), there are restrictions on the research costs that the National Institutes of Health money supports. Grant specifications usually require the research institutions and the researchers themselves to bear many of the project costs. For instance, laboratory space, administrative staff, salary support for senior physician-scientists, and costs for portions of a project that are already underway (full direct costs) are often excluded, or are insufficiently addressed, by National Institutes of Health grants
24 . Much of the funding to fill these gaps comes from the combined pool of graduate medical education funding and clinical revenue.
Support for many of the social and academic missions of academic health centers has traditionally come from cross-subsidies from patient care revenues
25,26 . Ludmerer estimated that 28% of the clinical income to academic hospitals goes to support the three "academic missions" outlined above
1 . In the days before managed care and prospective payment, academic health centers simply billed at a higher price for their services and directly received reimbursement for the amount that they charged. This practice allowed academic physicians ample support to teach, perform research, and provide both charity and complicated care to patients in need of these services. However, skyrocketing medical costs necessarily brought about an increasingly cost-conscious system. Containment came in the form of cutbacks in governmental support and an increase in competitive market forces in health care. As a result, operating margins for all hospital care underwent across-the-board reductions in the 1990s, but the effect was most dramatic at academic institutions because of their reliance on this operating margin to support their academic and social missions. To compensate for the decreased per case margins, academic administrators responded to the tighter health-care dollar by stressing "clinical productivity," which is achieved by performing a higher volume of revenue-generating cases.
Patient care revenue, now frequently assessed in the form of a "dean's tax" from faculty practice plans, has grown to represent a large source of income for academic health centers. Whereas <10% of medical school funding came from clinical services in 1960, this figure grew rapidly throughout the 1980s, approaching 50% by 1996
27 . The American Association of Medical Colleges estimated that the average academic health center assesses a 10% to 15% dean's tax to support their other institutional programs and departmental deficits
28 .
Academic Health Centers Provide a Disproportionate Amount of Care to the Poor and Uninsured
Before Medicare, treatment for the indigent was spread out more evenly among all providers, academic and nonacademic
1 . Since then, however, academic health centers have taken up a greater portion of the national burden of caring for those who do not generate income for the hospital. This trend has been especially true over the last decade when academic health centers have become widely viewed as safety nets for the poor and uninsured. From 1991 to 1996, the amount spent by academic health centers on uninsured cases increased by >40% as a percentage of gross patient revenues
29 . The American Association of Medical Colleges reported that the cost of caring for the uninsured borne by teaching hospitals was $4 billion in 1999
30 . Uncompensated care accounted for 7.2% and 6.8% of the total costs of academic health centers and other major teaching hospitals, respectively, in 2000
31 (
Fig. 2 ). In addition, a great deal of this burden has been shouldered by the academic physicians who provide the care.
Influence of Managed Care
As a result of the higher costs associated with their services, teaching hospitals are at a competitive disadvantage with respect to the pricing of patient care compared with nonteaching hospitals that do not engage in teaching, education, and specialty care. It is clearly not in the interest of companies whose success depends on profit rather than social good to pay for the additional costs that academic and social missions impose. Rather, a successful managed care group would aim to provide routine care at low cost for a largely healthy population. With rare exception, most managed care companies negotiate aggressively to pay rates that only cover the direct cost of care that their patients consume
32 . The net effect of managed care in medicine has been discount pricing and a rerouting of many managed care patients to lower-cost community hospitals. Correspondingly, private payer payment-to-cost ratios have decreased from 1.25 in 1989 to 1.15 and 1.05 for academic health centers and other major teaching hospitals, respectively, by 1998
4 .
Cumulatively, the effects of capped prospective payments (the diagnosis-related group system) and managed care have forced academic health centers to become more efficient in providing patient care. To that end, academic health centers have pushed to reduce their internal costs and increase throughput of patients. Some contend that the higher volumes at teaching hospitals enhance the resident experience. Others think that they have deteriorated the educational environment, replacing time spent learning diagnostic and clinical management tools with admitting, discharging, and surgical assistant duties for orthopaedic residents
1,3 . While measures to increase efficiency and throughput did relieve some of the financial strain on academic health centers for a period, the dam did not hold the tide of what was to come. Ultimately, falling operating margins negated any such gains that the increased volume had generated for the hospitals. Operating margins at academic health centers decreased in a linear fashion from 3% in 1995 to a remarkable negative value of -0.95% in 1999
9 (
Fig. 3 ). Data like these illustrate why ten teaching hospitals were sold over the past decade to for-profit entities; notably among them were those at Georgetown University, Tulane University, University of Minnesota, and George Washington University.
Changes to Resident Work Environment by the Residency Review Committees and the Accreditation Council for Graduate Medical Education
The recent effort on the part of the Accreditation Council for Graduate Medical Education (ACGME), the American Medical Association, and other organizations to change the work environment for residents complicates the issue of graduate medical education reform. Among the many points they have considered, the emphasis has been on ways to decrease resident work hours, increase dedicated educational activity, and enhance attending supervision. The ACGME, the controlling body of the Residency Review Committees, proposes limiting resident work hours to eighty, with ten hours off between shifts and a period of twenty-four hours off at least once a week averaged over a four-week span
33 . Federal legislation reflecting ACGME recommendations has yet to pass in Congress. Meanwhile, however, the ACGME has become more active in putting programs on probation or retracting accreditation from programs whose residents do not receive enough dedicated teaching time or rest.
These new requirements further strain the graduate medical education dollar for residency programs and teaching hospitals. In effect, they decrease the cost saving that resident work brings hospitals, while increasing the costs associated with resident education. New York, home to 15% of the residents in the United States, is currently the only state to have formally adopted these guidelines, making teaching hospitals subject to financial penalties if they do not comply. However, programs outside New York have already responded to the guidelines to ensure their continued accredited status from the ACGME and the Residency Review Committees.
Before addressing the key question, "Does graduate medical education currently warrant subsidy?," some historical perspective is required. Before 1965, the higher price of care at an academic health center was directly borne by the consumer. Recognizing the social good that academic health centers provide, the government formally began supporting them through graduate medical education funding with the advent of the Medicare program in 1965. Nonetheless, private payers still contributed to education by paying higher patient-care fees at academic health centers. Much of this money went to cross-subsidize the medical schools. Since the arrival of cost-competitive medicine and managed care, however, academic health centers have grown more dependent on Medicare's funding program for graduate medical education, which itself has undergone a number of changes in the methods and level of support provided over the years. Currently, total Medicare and Medicaid payments supply more than half of all incoming dollars to academic health center hospitals
24 .
Questions that pertinent policy reform needs to address in the near future include: Are the education, research, and specialty care missions of academic health centers worthy of support beyond that which the private markets will allow? If so, who will fund these activities? How much money is needed? What will be the role of state and federal government, managed care, private payers, hospitals, and trainees (residents) in bearing the costs? How will policy changes affect resident education?
A Fundamental Question: Is Graduate Medical Education Worthy of Public Funding?
Experts have long debated the applicability of a free market system to health care. Basic economic theory suggests that the government should use tax dollars to support public and social needs, such as the police force, roadways, and the armed forces. Private goods such as automobiles, computers, and clothing, on the other hand, should be left to the public to purchase individually. The appropriate amount of governmental support for a service should reflect the level of value that society places on that specific service. If graduate medical education represents a social good, then it deserves support.
From the beginning of the program, Medicare has supported graduate medical education on the basis of the recognition that the country needs highly trained physicians and that residents contribute to the quality of patient care of Medicare enrollees. However, the term "graduate medical education" funding is deceptive, as both residents and residency programs see little money directly allotted to them for the purpose of education. Rather, the funding is channeled to the academic health center itself, much of it going to support the other missions discussed above. As institutions suffered from cutbacks in funding, education did as well. The quasi-free market system over the past decade failed to provide for academic health centers in a way that would allow them to nurture a fair balance in their three-pronged mission.
Given the decreased support for graduate medical education, the increased administrative burden borne by residents, reports of rising failure rates on oral board examinations
3 , and the results of declining competency reported in surveys, we have cause to be concerned about the quality of resident education. In one study, 38% and 46% of graduating orthopaedic residents said they felt either "very unprepared" or "somewhat unprepared," respectively, to perform spinal procedures and cancer treatment
34 . However, only 1% of the residents reported feeling equally unprepared to perform total knee replacements. These findings highlight the need for increasing the diversity of cases and the breadth of experience that residents receive.
The institution of teaching represents inefficiency to a business-minded hospital. Initiatives to increase resident experience would require program restructuring and ultimately increased funding to compensate for the time that residents and attending physicians would spend putting education first and nonacademic work second. Perhaps this restructuring would necessitate the hiring of mid-level health-care professionals to fill the void, a practice that has been shown to increase overall expense
35 . Lecture and laboratory space, overhead to manage residency programs, and teaching tools such as CD-ROMs, books, materials, and equipment for mock surgical procedures are costs that are easily quantified. However, the costs associated with lost opportunities related to the time that attending physicians and residents spend on educational exercises, as well as the added time that they spend on delivering specialized care for complicated cases, are more difficult to calculate. Nevertheless, educating future physicians remains of vital importance to the country and, as such, represents an important social good. Given this premise, funding for education should not be subject to the same financial instability that has plagued academic hospitals over the past two decades. Furthermore, support for education should not be forced to compete directly with support for other social missions that fall to academic health centers.
Payment Methodology for Graduate Medical Education
Currently, patient-care revenue, faculty practice plans, grants and endowments, the Department of Veterans Affairs, the Department of Defense, and Medicaid (in forty-three states) all support the cost of educating residents. However, the largest source of funding for the explicit purpose of graduate medical education remains Medicare, which supplies 74% of these dollars according to the American Medical Association
36 .
Medicare reimbursement for graduate medical education is divided into direct (DGME) and indirect (IME) medical education payments. DGME accounts for the direct costs related to residents, which includes overhead expenses for graduate medical education, salaries and fringe benefits for residents, some compensation for teaching physician time, and the costs of the administrative staff for graduate medical education. IME is broader in scope. Indirect payments are intended to make up for the higher hospital operating costs associated with the presence of residency training programs. These include more complicated cases, additional tests ordered by residents as part of the learning process, uncompensated care, and reduced patient-care productivity by teaching staff members.
Direct Funding of Graduate Medical Education
From 1966 until 1985, Medicare payments were uncapitated. Payments to hospitals were based on whatever "Medicare-allowable" costs were reported and as a result varied widely from program to program by the early 1980s. Amis and Lantos reported that direct payments for graduate medical education ranged from as little as $7500 to $187,500 per resident depending on the program
37 . Since then, changes in the formula for funding of graduate medical education occurred to both limit these variations and alter the specialty and geographic mix of the workforce. Current DGME payment amounts are based on a complicated formula that defines each hospital-specific per resident amount. Per resident amounts were determined from teaching hospital audits of total expenditures for graduate medical education in a base year, fiscal year 1984 or 1985, divided by the number of residents. Each hospital has two separate per resident amounts paid to them on the basis of the specialty mix of their residents. Primary care programs, which include those in family medicine, general internal medicine, general pediatrics, preventive medicine, geriatric medicine, osteopathic general practice, and obstetrics and gynecology receive per resident amounts that are 5.6% higher than those received for nonprimary care programs, such as orthopaedics.
These per resident amounts still vary widely from program to program, but the goal of recent legislation (a provision of the Balanced Budget Refinement Act of 1999) is to bring each per resident amount across the country within a range of 70% to 140% of the national average per resident amount. To determine the DGME amount, the per resident amount is multiplied by the hospital's percentage of Medicare patients and is updated by an inflation factor. This formula provides the value received for most residents. However, Medicare pays only 50% of this calculated amount for residents who go beyond the minimum number of years for which an initial residency program is accredited, and it provides no more payment for residents in training beyond five years. This 50% DGME payment status is also given for fellows, residents who train in more than one field, and many of those who quit one specialty and start another. Exceptions include residents with a fellowship in geriatrics and those who train in combined primary-care programs, such as medicine and pediatrics; these residents receive 100% support throughout training. Recent legislation (a provision of the Balanced Budget Act of 1997) capped the number of residents for which programs can be funded and allows DGME funding to support resident experience in outpatient settings. The Department of Health and Human Services estimated that total DGME payments were $2.7 billion in 2000
9 .
Indirect Funding of Graduate Medical Education
IME payments, which accounted for a $5.1-billion Medicare outlay in 2000, represent a greater dollar amount of support for academic health centers than do DGME payments
9 . Similar to the payment methodology for nonteaching hospitals, Medicare classifies patient visits into diagnosis-related groups and pays a standard amount for every patient within the same disease category, adjusting slightly according to a geographic factor. For teaching hospitals, however, IME is a percent add-on to Medicare's diagnosis-related group payments for patient care. In 1983, this add-on was set to 11.59% for each 0.1 increment in a hospital's intern and resident-to-bed ratio. This rate was reduced to 7.7% in 1987 and remained at that value until the Balanced Budget Act of 1997, which called for further reduction of the IME add-on adjustment by 29% over four years to a rate of 5.5% by 2001. One driving force behind the cuts was the conclusion reached by a congressional advisory committee that the increased cost per case at academic health centers compared with that at nonteaching hospitals, solely accounting for higher direct patient-care costs, was on the order of 3.2%
38 . Believing that these cuts were too drastic, Congress passed the Balanced Budget Refinement Act of 1999 and the Beneficiary Improvement and Protection Act, maintaining the IME adjustment at 6.5% through 2002, when the rate was dropped to 5.5%. According to the American Association of Medical Colleges, the typical teaching hospital was expected to lose over $40 million between 1998 and 2002, as a result of the Balanced Budget Refinement Act of 1997 and the Balanced Budget Refinement Act of 1999
39 .
The original intention of payments for graduate medical education was to support the increased quality of care for Medicare patients as well as to cover the program's share of costs for educating physicians. Some experts have argued, however, that DGME and IME payments contribute more than the share that teaching intensity alone accounts for, although not enough to cover all of the traditional missions of an academic health center
4 . Third-party and other private payers, on the other hand, do not pay their share of graduate medical education costs.
Medicare policy on graduate medical education funding has been shown to have a substantial effect on residency programs. As the payment methodology depends on the number of residents, graduate medical education funding has been shown to be a determinant of the numbers of residents that hospitals employ
40 . In addition, changes in the formula have been used to shape the specialty workforce mix. The most pressing current issue, however, is the decrease in dollar amounts of graduate medical education payments, which has contributed to the change in focus of academic health center operations and has decreased the incentive for academic physicians to teach. Although increased attrition of academic orthopaedic surgeons, per se, does not appear to be a consequence
41 , cuts in academic physician salaries may make it difficult to recruit new talent
42,43 . The problem of recruitment is especially a concern in orthopaedics, given the higher expected salaries relative to other specialties.
Graduates of residency and fellowship programs often emerge with debt ranging from tens to hundreds of thousands of dollars. The American Medical Association estimates that 83% of medical school graduates in 2000 had debt averaging $93,000 per student
44 . Given this situation, financial considerations in career choice are to be expected. Recent surveys have confirmed what common sense suggests—that salary and personal financial considerations are among the issues considered to have the highest priority for graduates of residency training programs
45,46 .
However, salaries for academic physicians, particularly specialists, continue to be hit hard by cutbacks—first by a variety of individual measures that academic health centers took to remain fiscally solvent (such as increased dean's assessments on clinical revenues, higher overhead charges, and reductions in salary and benefits packages). Next, as academia tends to host highly specialized physicians, a greater proportion of academicians saw a decrease in relative compensation as a result of legislation enacted in the 1990s that continues to benefit rural and primary care physicians more than procedure-oriented and specialized practitioners. The change to a resource-based relative value scale for physician payment under Medicare Part B in 1992
47 and practice expense reform were examples of this movement. Despite the effort to compensate for financial pressures by spending more time treating patients and generating clinical revenue, academic surgeons and surgical specialists experienced no annualized compound growth in their salaries from 1993 through 1998, according to a study by Studer-Ellis et al.
47 . A likely result of this change in work environment is that academic physicians have less time to teach and to conduct research, receive less compensation, and have greater pressure to be clinically productive in order to maintain salary expectations. These factors decrease the attractiveness of academia to candidate surgeons. In addition, potentially outstanding mentors are leaving or turning down offers for academic positions. As a result, education suffers. Furthermore, several residency programs have expressed concern that community educators will not be willing to teach without compensation
4 .
The Balanced Budget Refinement Act of 1997 included a provision to create the Medicare Payment Advisory Commission (MedPAC). This committee of experts makes recommendations to Congress regarding Medicare payment policy. In their August 1999 report, the members concluded that Medicare should discontinue making separate payments to teaching hospitals for direct and indirect costs of education (DGME and IME). Rather, they advocated providing a "Teaching Hospital Adjustment (THA)" to teaching hospital diagnosis-related group payments, a system that would be similar in form to the current IME adjustment mechanism
48 . In a June 2000 follow-up report, they outlined three specific proposals to implement their recommendations—two of these recommendations decrease graduate medical education funding by about $1.5 billion, and one is budget neutral
38 . They also suggested that the teaching hospital adjustment be based upon a national average per resident amount, rather than hospital-specific amounts, thus having the effect of redistributing funds among teaching hospitals.
MedPAC's rationale for its recommendation was based on the economic theory that residents bear the cost of their own education and thus do not warrant specific funding. According to this theory, "it is clearly not in the interest of firms [host academic medical centers] to pay trainees [residents] more than the value of the services provided net of training costs."
40 Proponents of this theory would ask: Why would a company or hospital invest the time and money to educate an individual if that individual were just as likely to work for a competitor when the costs of his or her training finally paid off? The fundamental flaw in this theory is that it makes the assumption that residents are more akin to employees in training, rather than students.
Furthermore, MedPAC argued, why does education warrant subsidy when there is a surplus of applicants to these programs? The conclusion drawn from this rationale was that the most appropriate way to distribute graduate medical education funding would be as direct support to academic health centers as a teaching hospital adjustment. This represents what the Council on Graduate Medical Education calls a health-care model of graduate medical education reform, in which payments for clinical training are linked to patient-care activities
4 .
The Council on Graduate Medical Education proposed three other models of reform. They included a planning model in which graduate medical education funds are allocated on the basis of the decision of a committee made up of members from a variety of sectors who would assess and distribute the funding based on local workforce needs. The performance model would link payment to specific performance measures or objectives of education quality and workforce distributions. The education model, however, was the only proposal format that would provide stable and steady funding to individual programs, while treating graduate medical education funding as an educational cost, not a subsidy, to patient-care costs.
Response to MedPAC's Recommendation
MedPAC's approach to graduate medical education funding is in direct contrast to the academic mission of teaching hospitals. Rather, residents should be treated more like students. The best way to do this is to assure steady and stable funding sources for the explicit purpose of medical education. This approach requires that funding for graduate medical education be made separate from other worthy academic health center activities, such as specialty care, research, and treating the poor and uninsured. This also means excluding complicated and confounding methodologies in graduate medical education financing that aim to shape the medical workforce. We think that all who benefit from academic health centers and graduate medical education programs should financially support these important causes. Therefore, we present three specific recommendations.
1. Provide education funding directly to teaching programs.
Medicare's support for graduate medical education came into existence to account for residents' contributions to patient care and to support the quality of tomorrow's physician workforce. Because this funding is routed through the heavily burdened academic health center administration, however, it has been misused to support other academic programs. Theoretically, one could approximate the amount of resident education funding from the amount given by Medicare, less that used by academic health centers for these additional programs. Unfortunately, we are unaware of any data related to how much of the graduate medical education funding given to academic health centers is ultimately used for educational purposes.
In the September 2001 issue of The Journal of Bone and Joint Surgery, Michael Simon stated his opinion that "the quality of graduate orthopaedic education has continued to deteriorate" and called for graduate medical education funding to bypass the academic health center administration and be routed more directly toward educating our residents
3 . We agree with this position and believe that it would be best accomplished by the education model of graduate medical education reform. Under this proposal, DGME payments would be given directly to the entity most responsible (and accountable) for providing this activity. One such approach would be to route funding directly to the program director, who in turn would be responsible for its expenditure. The larger IME adjustment would continue to go to academic health centers to support other worthy social missions.
As resident education represents a social good, the federal government, through Medicare or another program, remains the appropriate agency for the delivery of graduate medical education funding. This approach is the only way to ensure that the interests of each constituent—whether it be the public, patient population, collective interest organizations, or budget committees—are represented. The role of the program directors in this model would be to allocate the funds most appropriately.
2. Eliminate the funding differential that supports primary care residents and obstetrics and gynecology residents at a DGME rate that is 5.6% higher than that for nonprimary care residents in specialties such as orthopaedics.
The 1990s effort to control health-care costs by producing more primary care physicians relative to specialists still exists in the form of differential funding for residency positions. In addition, federal reimbursement strategies have decreased academic and specialist compensation relative to that for nonacademic and primary care physicians. As a result, academic specialty departments, as well as their residency programs and faculty, have suffered the most from these changes.
Although the debate over the relative compensation between specialties is beyond the scope of this article, we believe strongly that graduate medical education payments should not be used as a tool to reshape the workforce. This approach holds education hostage to other agendas. Eliminating the discrepancy in funding between the specialties is a requisite for the recognition of residents as students, rather than as trainees. Furthermore, equalizing the subsidy can be done in a budget-neutral way.
3. Create an all-payer system for graduate medical education.
Proposals to implement an all-payer system to support graduate medical education gained much attention in the mid-1990s. At the time, Congress was concerned about health-care inflation and other data that predicted that the Medicare Part A trust fund would become insolvent by 2001
49 . The increases in health-care costs due to new treatment modalities and the aging of the population were largely responsible for these concerns. Meanwhile, many Medicare beneficiaries demanded better access to costly prescription drugs that the program could not afford. To save costs, the Balanced Budget Act of 1997 cut an estimated $116 billion
39 from the Medicare budget, a disproportionately large portion of which came out of graduate medical education and academic health center support.
Medicare cuts to graduate medical education, coupled with the arrival of managed care medicine, raised an important question in the health-care community. Why should Medicare and Medicaid be the only supporters of graduate medical education programs when clearly all of society benefits from the training of physicians and academic health center missions? Many believe that private industry and other payers should contribute to the pool of graduate medical education funding. However, initiatives to draw funding from a broader base, while widely supported by physician groups, the academic community, nonprofit medical think tanks, and congressionally appointed committees, continue to fail in Congress.
Annually, bills such as the All Payer Graduate Medical Education (GME) Act, proposed by Congressman Cardin (Maryland), and the Medical Education Trust Fund Act from Senators Reed (Rhode Island), Clinton (New York), and Shumer (New York) get little consideration in Congress. These bills propose a 1% to 1.5% premium on private health insurers for health-related services provided to its policyholders. A 1% premium is estimated to generate $4 billion
50 , an amount that could be used to support both education (DGME) and other academic health center missions (IME). The proposal is estimated to save the Medicare program $1.5 billion annually.
The medical community strongly supports the idea that both private industry and the public should bear their fair share of cost for graduate medical education. Proposals for the implementation of an all-payer system have varied. Some call for the creation of a trust fund to oversee its distribution. Another consideration is the role of support, if any, that general tax revenues should play. Regardless of the precise form taken, the advantage of an all-payer system is that it would create a more stable long-term funding source for education and more appropriately distribute the costs to those who benefit from it—everyone.
As Dr. Ludmerer alluded to in Time to Heal
1 , academic medicine is at a crossroads. The three-pronged mission of research, specialized care, and education is being tested to the limit as academic health centers are forced to compete with market mechanisms for revenue. However, rather than mirroring their competitors by adopting a high throughput, low-cost model to maximize financial standing, academic health centers should be encouraged to continue providing their unique set of services for which society has come to depend on them—treating the sickest and most needy patients, serving as fertile ground for cutting-edge research, and hosting the training sites for the country's future physicians and surgeons.
Resident education is at particular risk as the academic environment further deteriorates. This is largely due to the shrinking amount and overly-complex mechanism of graduate medical education funding, as well as an over-reliance on the Medicare and Medicaid programs. Policy changes should aim to support education in a way that views residents as students, not trainees, and recognizes the long-lasting societal benefit of the investment by providing stable, adequate, and broad-based support.
In addition to the American Medical Association, other specialty organizations, and congressmen who sponsor graduate medical education reform, the orthopaedic community has a new and unique forum to voice opinion regarding resident education—the recently established American Academy of Orthopaedic Surgeons Council on Academic Affairs and its Committee on Academic Advocacy, who are charged with representing the Academy regarding a broad array of academic concerns. It is important, however, that each of us understands how health care and physician education are financed so that we can effectively advocate for high-quality teaching programs.
Ludmerer KM. Time to heal: American medical education from the turn of the century to the era of managed care. New York: Oxford University; 1999.Â
Â
Ludmerer KM. The coming of the second revolution in medical education. Read at the Annual Meeting of the Academic Orthopaedic Society; 2001 Nov 9; Washington, DC.Â
Â
SimonMA. The education of future orthopaedists-déja vu. J Bone Joint Surg Am,2001;83: 1416-23. 831416Â
2001Â
[PubMed] Â
Council on Graduate Medical Education. Financing graduate medical education in a changing health care environment. COGME's 15th Report. Rockville, MD: Council on Graduate Medical Education; December 2000. www.cogme.gov.Â
Â
Committee on PediatricWorkforce. American Academy of Pediatrics: Financing graduate medical education to meet pediatric workforce needs. Pediatrics,2001;107: 785-9. 107785Â
2001Â
[PubMed][CrossRef] Â
BekesC. Faculty reimbursement for graduate medical education: who gives and who receives?. Crit Care Med,2000;28: 1688-9. 281688Â
2000Â
[PubMed][CrossRef] Â
Commonwealth Fund Task Force on Academic Health Centers. Health care at the cutting edge: the role of academic health centers in the provision of specialty care. Publication 390. New York: Commonwealth Fund; July 2000. www.cmwf.org.Â
Â
KassirerJP. Academic medical centers under siege. N Engl J Med,1994;331: 1370-1. 3311370Â
1994Â
[PubMed][CrossRef] Â
Commonwealth Fund Task Force on Academic Health Centers. Training tomorrow's doctors: the medical education mission of academic health centers. Publication 516. New York: Commonwealth Fund; April 2002. www.cmwf.org.Â
Â
MechanicR, Coleman K,Dobson A. Teaching hospital costs: implications for academic missions in a competitive market. JAMA,1998;280: 1015-9. 2801015Â
1998Â
[PubMed][CrossRef] Â
CebulRD, Snow RJ, Pine R, Hertzer NR,Norris DG. Indications, outcomes, and provider volumes for carotid endarterectomy. JAMA,1998;279: 1282-7. 2791282Â
1998Â
[PubMed][CrossRef] Â
KatzJN, Losina E, Barrett J, Phillips CB, Mahomed NN, Lew RA, Guadagnoli E, Harris WH, Poss R,Baron JA. Association between hospital and surgeon procedure volume and outcomes of total hip replacement in the United States medicare population. J Bone Joint Surg Am,2001;83: 1622-9. 831622Â
2001Â
[PubMed] Â
HannanEL, O'Donnell JF, Kilburn H Jr, Bernard HR,Yazici A. Investigation of the relationship between volume and mortality for surgical procedures performed in New York State hospitals. JAMA,1989;262: 503-10. 262503Â
1989Â
[PubMed][CrossRef] Â
GarciaFA, Miller HB, Huggins GR,Gordon TA. Effect of academic affiliation and obstetric volume on clinical outcome and cost of childbirth. Obstet Gynecol,2001;97: 567-76. 97567Â
2001Â
[PubMed][CrossRef] Â
Weinberger SE. Everything you wanted to know about teaching hospitals . . . but were afraid to ask! An AAMC publication. Washington, DC: Association of American Medical Colleges; 2002. www.aamc.org/members/shared/priselachospitals.ppt.Â
Â
Centers for Disease Control andPrevention. Self-assessed health status and selected behavioral risk factors among persons with and without health-care coverage-United States, 1994-1995. MMWR Morb Mortal Wkly Rep,1998;47: 176-80. 47176Â
1998Â
[PubMed] Â
FranksP, Clancy CM,Gold MR. Health insurance and mortality. Evidence from a national cohort. JAMA,1993;270: 737-41. 270737Â
1993Â
[PubMed][CrossRef] Â
Reuter JA. Patterns of specialty care: academic health centers and the patient care mission. Commonwealth Fund Task Force on Academic Health Centers. January 1999. www.cmwf.org.Â
Â
BernardAM, Hayward RA, Rosevear J, Chun H,McMahon LF. Comparing the hospitalizations of transfer and non-transfer patients in an academic medical center. Acad Med,1996;71: 262-6. 71262Â
1996Â
[PubMed][CrossRef] Â
JencksSF,Bobula JD. Does receiving referral and transfer patients make hospitals expensive?. Med Care,1988;26: 948-58. 26948Â
1988Â
[PubMed][CrossRef] Â
KerrHD,Byrd JC. Community hospital transfers to a VA Medical Center. JAMA,1989;262: 70-3. 26270Â
1989Â
[PubMed][CrossRef] Â
IorioR, Healy WL,Richards JA. Comparison of the hospital cost of primary and revision total hip arthroplasty after cost containment. Orthopedics,1999;22: 185-9. 22185Â
1999Â
[PubMed] Â
IorioR, Healy WL,Richards JA. Comparison of the hospital cost of primary and revision total knee arthroplasty after cost containment. Orthopedics,1999;22: 195-9. 22195Â
1999Â
[PubMed] Â
Commonwealth Fund Task Force on Academic Health Centers. Managing academic health centers: meeting the challenges of the new health care world. Commonwealth Fund. October 2000; Publication 408. www.cmwf.org.Â
Â
AndersonGF,Lave JR. Financing graduate medical education and biomedical research: whose responsibility?. Health Policy,1987;7: 3-12. 73Â
1987Â
[PubMed][CrossRef] Â
FreburgerJK,Hurley RE. Academic health centers and the changing health care market. Med Care Res Rev,1999;56: 277-306. 56277Â
1999Â
[PubMed][CrossRef] Â
KrakowerJY, Ganem JL,Jolly P. Review of US medical school finances, 1994-1995. JAMA,1996;276: 720-4. 276720Â
1996Â
[PubMed][CrossRef] Â
MacDonald AS. The governance and organization of faculty clinical practice for the new era. American Association of Medical Colleges/ECG Management Consulting Presentation: AAMC Executive Seminar for Deans. Washington, DC: American Association of Medical Colleges; 2000 Jan 31. www.ecgmc.com/expertise/pdf/0031984.pdf.Â
Â
Commonwealth Fund Task Force on Academic Health Centers. A shared responsibility: academic health centers and the provision of care to the poor and uninsured. Publication 443. New York: Commonwealth Fund; April 2001. www.cmwf.org.Â
Â
Association of American Medical Colleges. Protecting America's uninsured. Washington, DC: American Association of Medical Colleges; September 2001. www.aamc.org/uninsured.Â
Â
Dobson A, Koenig L, Sen N, Ho S, Gilani J, The Lewin Group. Financial performance of academic health center hospitals, 1994-2000. Publication 543. New York: Commonwealth Fund; September 2002. www.cmwf.org.Â
Â
GottliebS,Einhorn TA. Managed care: form, function, and evolution. J Bone Joint Surg Am,1997;79: 125-36. 79125Â
1997Â
[PubMed] Â
Stoll DA. Committee for Review of Program Requirements: Memorandum: Proposed resident duty hours language. Chicago, IL: Accreditation Council for Graduate Medical Education. 2002 Sept 18.Â
Â
BlumenthalD, Gokhale M, Campbell EG,Weissman JS. Preparedness for clinical practice: reports of graduating residents at academic health centers. JAMA,2001;286: 1027-34. 2861027Â
2001Â
[PubMed][CrossRef] Â
GreenBA,Johnson T. Replacing residents with midlevel practitioners: a New York City-area analysis. Health Aff (Millwood),1995;14: 192-8. 14192Â
1995Â
[PubMed][CrossRef] Â
Daugherty RM. Policy options for support of graduate medical education. American Medical Association Council on Medical Education Report 10 (A-99). Presented to the American Medical Association Reference Committee C; 1999.Â
Â
Amis ESJr,Lantos PR. Graduate medical education financing: a primer. Acad Radiol,1996;3: 507-11. 3507Â
1996Â
[PubMed][CrossRef] Â
Medicare Payment Advisory Commission. Report to the Congress: selected medicare issues. Washington, DC: Medicare Pay-ment Advisory Commission; 2000 June. www.medpac.gov.Â
Â
Association of American Medical Colleges. Issue Brief: America's teaching hospitals still hurt from the BBA. Washington, DC: American Association of Medical Colleges; 2000 Feb 2. www.aamc.org/advocacy/issues/medicare/ibba.htm.Â
Â
NewhouseJP,Wilensky GR. Paying for graduate medical education: the debate goes on. Health Aff (Millwood),2001;20: 136-47. 20136Â
2001Â
[CrossRef] Â
MealsRA, Bassewitz HL,Dorey FJ. Academic longevity and attrition of full-time orthopaedic faculty members. J Bone Joint Surg Am,2000;82: 1042-8. 821042Â
2000Â
[PubMed][CrossRef] Â
JacksonDW. The orthopaedic clinician-scientist. J Bone Joint Surg Am,2001;83: 131-5. 83131Â
2001Â
[PubMed] Â
MealsRA. Why orthopaedic surgeons leave full-time academic positions for private practice. J Bone Joint Surg Am,2001;83: 456-60. 83456Â
2001Â
[PubMed] Â
Allen R. Medical school financing, tuition, and student debt. American Medical Association Council on Medical Education Report 2 (I-00). Presented to the American Medical Association Reference Committee C; 2000.Â
Â
GabramSG, Hoenig J, Schroeder JW Jr, Mansour A,Gamelli R. What are the primary concerns of recently graduated surgeons and how do they differ from those of the residency training years?. Arch Surg,2001;136: 1109-14. 1361109Â
2001Â
[PubMed][CrossRef] Â
DohertyMJ, Schneider AT,Tirschwell DL. Will neurology residents with large student loan debts become academicians?. Neurology,2002;58: 495-7. 58495Â
2002Â
[PubMed] Â
Studer-Ellis E, Gold JS,Jones RF. Trends in US medical school faculty salaries, 1988-1989 to 1998-1999. JAMA,2000;284: 1130-5. 2841130Â
2000Â
[PubMed][CrossRef] Â
Medicare Payment Advisory Commission. Report to the Congress: rethinking Medicare's payment policies for graduate medical education and teaching hospitals. Washington, DC: Medicare Payment Advisory Commission; 1999 August. www.medpac.gov.Â
Â
GottliebS. The role of managed care in financing medical education: a survey of views. JAMA,1996;276: 757-8. 276757Â
1996Â
[PubMed][CrossRef] Â
Association of Program Directors in Internal Medicine. Medicare GME bills before Congress; 2001 Oct 1. www.apdim.med.edu.Â
Â