Teaching hospitals are generally subject to caps on the number of residents for which those hospitals are able to receive Medicare direct graduate medical education (DGME) and indirect medical education (IME) reimbursement. This issue’s GME Financing Essentials provides a discussion of the policy issues Congress sought to address in developing the two permanent unused resident redistribution programs that have been implemented since Medicare resident caps were established. These one-time resident redistribution programs were implemented separate from the “closed hospital program,” which provides permanent Medicare resident cap increases and was authorized under the Affordable Care Act (ACA). (See “CMS Announces Medicare Resident Slot Awards.”)
Congressional Intent in Implementing the Medicare Resident Caps
The Balanced Budget Act of 1997 (BBA) generally limited the number of residents for which teaching hospitals could be reimbursed to the number of residents reported in a prior year. The caps were set based on the number of allopathic and osteopathic residents training in the acute inpatient unit of the hospital and eligible non-hospital settings as of the hospital’s most recent Medicare cost-reporting period ending on or before December 31, 1996. In establishing the resident caps, Congress sought to achieve two principal goals—control the growth in physician supply and control the growth in Medicare spending.
In the mid-1990s, with the increasing prevalence of managed care as a preferred type of health insurance, there was growing concern expressed by policymakers and economists that the then-popular “gatekeeper model” being used by certain health insurance companies would drastically reduce the need for certain kinds of physician specialists. The gatekeeper model was built on the premise that requiring an enrollee of a health plan to first see his or her primary care physician before being able to access a specialist would reduce unnecessary utilization of specialty services and encourage more effective use of primary care services. In concept, the primary care physician would act as a “gatekeeper,” only referring patients to non-primary care specialists when such visits were truly needed and warranted. Certain physician workforce researchers (most notably, Jonathan P. Weiner, Dr. P.H., who estimated an oversupply of 165,000 physicians by 2000) analyzed the effect of staffing models of health insurance plans that used various forms of gatekeeper models and concluded that the nation was facing an impending oversupply of physicians. The Association of American Medical Colleges also issued statements expressing concern about an impending physician oversupply.
At the same time, the Federal government was facing a significant deficit and a weak economy. To address these concerns, there were calls from both political parties and the public for Congress and the Clinton Administration to significantly cut Federal spending to balance the budget. Congress sought ways to reduce some Medicare programs and put a cap on others, and since there was a significant body of research indicating a pending oversupply of physicians, Congress decided to limit the number of physicians that Medicare funding would spend to train each year.
After the Medicare resident cap was imposed, it became clear that the American public was largely rejecting the gatekeeper model and demanding more open access to specialty physicians to treat specific illnesses and conditions. As a result, workforce researchers concluded that there would be a need for additional physicians to be trained to meet the demands of the public and the aging of the baby boomer population. Specialty societies, the teaching hospital community, and medical educators began to call on Congress to lift the resident cap and provide Medicare support for the training of additional needed physicians.
The Section 422 Program
The Medicare Modernization Act (MMA), which President Bush signed into law on December 8, 2003, sought to address the concerns of the medical education community while not removing the Medicare resident cap. The MMA included a provision (Section 422) that mandated the one-time permanent redistribution of unused residency slots from teaching hospitals that were below their caps to teaching hospitals interested in receiving increases in their Medicare resident caps.* The rationale behind this approach was that Congress could address the concerns of the medical education community while still maintaining the same overall number of Medicare-fundable resident slots in the country.
Under the MMA, the Health Care Financing Administration (now the Centers for Medicare & Medicaid Services, or CMS) was directed to review each teaching hospital’s most recently submitted or audited Medicare cost report prior to 2002 and determine whether the hospital was above or below its resident cap. If the hospital was determined to be below its resident cap, CMS was directed to lower the cap by 75% of the difference between the cap and the hospital’s actual resident count. For example, if a hospital reported that it had a resident cap of 100 but was actually training only 80 residents in 2001, CMS would be required to lower the hospital’s cap on a permanent basis to 85 (since 100 - ((100 - 80) x 0.75) = 85).
If a hospital were participating in a Medicare GME–affiliated group as of July 1, 2003, whereby the hospital voluntarily adjusted its cap to accommodate cross-training with another hospital, the adjusted cap numbers of the hospitals in the group were taken into account. Similarly, if a hospital could demonstrate that it had expanded specific residency programs as of July 1, 2003, or had been approved for a new residency program, CMS could take that expansion into account in making hospital cap reductions.
The Medicare resident slots that would be generated through these reductions formed a pool that would be available to redistribute to teaching hospitals interested in receiving a resident cap increase. To ensure that the redistribution of slots would not be limited just to certain large teaching hospitals, the MMA directed that no applicant hospital could receive more than 25 resident slots. Even with this limitation, however, some policymakers were still concerned that rural hospitals that needed additional residency slots would be disadvantaged, so the MMA created a priority categorization among types of hospitals for the redistribution. The first priority for receiving the resident slots would be hospitals located in rural areas, followed by hospitals located in small urban areas, and then hospitals in a state applying for a specialty for which there were no other programs in the same state. Hospitals located in large urban areas, which is where most large academic medical centers and major teaching hospitals are located, would be last in line for receiving the slots. CMS was directed to develop an application process for distributing the available resident slots within each priority category, and adopted evaluation criteria to distinguish among the applicants. An applicant hospital was given a certain number of points based on whether it had high Medicare utilization, was located in a health professional shortage area, wished to start a geriatrics program, another primary care program, or an emergency medicine program, and other criteria.
To implement certain other policy preferences of some members of Congress, the redistributed slots would also be reimbursable at a level different from otherwise applicable payment rules. DGME slots awarded under the Section 422 program were reimbursed based on the locality-adjusted national average per-resident amount (rather than the hospital’s own specific per-resident amount). IME slots were reimbursed based on a teaching adjustment percentage of 2.7% as opposed to the usual 5.5%. By mandating these special payment rules, Congress essentially lowered the value of the redistributed residency slots.
Under the Section 422 program, in the aggregate, New York teaching hospitals had their resident caps reduced by 495 for DGME and 497 for IME, and experienced increases in their caps by 132 DGME slots and 70 IME slots (under the modified payment methodologies). These hospital cap changes were effective as of July 1, 2005.
The Section 5503 Program
The Affordable Care Act (ACA), which was signed into law by President Obama in March 2010, contained a one-time resident redistribution program that was modeled on the MMA Section 422 program. This new program (Section 5503 of the ACA) directed CMS to review each teaching hospital’s three Medicare cost reports most recently submitted prior to the ACA’s enactment and determine whether the hospital was at, above, or below its resident cap during each of the three years. Like the Section 422 program, a hospital that had entered into an affiliated group agreement and made an adjustment to its cap would have that adjustment taken into account in making this determination.
If the hospital was determined to be below its cap in all three years, CMS was mandated to lower the hospital’s applicable cap by 65% of the smallest difference between the hospital’s cap and the hospital’s actual resident count. For example, if a hospital had a cap of 100 but reported training only 80, 85, and 90 residents in each of the three years, CMS would use the year during which the hospital reported training 90 residents to determine an adjustment. CMS would be required to lower the cap on a permanent basis to 93.5 (since 100 - ((100 - 90) x 0.65) = 93.5).
As in the MMA program, the Medicare resident slots generated through these reductions formed a pool that would be available to redistribute to teaching hospitals interested in receiving a resident cap increase. Even though there were expected to be a smaller number of resident slots available as compared to the earlier MMA program, under the ACA applicant hospitals could receive up to 75 resident slots. Since, however, there were fewer slots available, the ACA reserved all of the redistributed slots just for defined areas of the country; 70% of the new positions would be provided to hospitals located in a state with a resident-to-population ratio in the lowest quartile; 30% of the new positions would be provided to hospitals that are a) located in one of the top 10 states with the greatest proportion of their population living in Federally designated health professional shortage areas, or b) located in rural areas. (Teaching hospitals located in New York, New Jersey, or Connecticut urban areas were not eligible for additional residency positions under this program.) CMS also adopted six evaluation criteria to use within the priority categories, with associated point scores, similar to what was used in the MMA program. Unlike in the MMA, the redistributed slots were reimbursable at a level using the otherwise applicable payment rules.
Under the Section 5503 program, New York teaching hospitals in the aggregate had their resident caps reduced by 205 for DGME and 124 for IME. These changes in the hospital caps were effective as of July 1, 2011.
*The unused resident slot review and redistribution process applied equally to a hospital's direct GME cap and indirect medical education cap. For simplicity of explanation, in this article, we will refer to a hospital's GME cap.