Comments to CMS on the proposed rule changes on Graduate Medical Education

Mark B. McClellan, MD
Administrator
Centers for Medicare & Medicaid Services
Department of Health and Human Services
Attention: CMS-1428-P
Room 443-G
Hubert Humphrey Bldg.
200 Independence Avenue, SW
Washington, DC 20201

Dear Dr. McClellan:

I am writing on behalf of the American Academy of Family Physicians, which represents more than 93,700 family physicians and medical students nationwide. We are responding to the Graduate Medical Education section of the May 18, 2004, proposed rule “Medicare Program; Proposed Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2005 Rates.” In analyzing and preparing a response to this proposed rule, the Academy has worked in close cooperation with the Organizations of Academic Family Medicine.

The Academy wishes to stress the importance of these regulations due to the impact they will have on family medicine training programs. Family medicine is critical to the presence of an adequate health care safety net, particularly in rural areas where family medicine represents the nexus between health care and the Medicare beneficiary. As such, the level of funding, and other means of support, for family medicine training programs is often what determines whether or not rural Medicare beneficiaries have access to health care services. GME support of primary care also has effects on public health systems as well as implications for homeland security and the effectiveness with which our nation and our communities deal with disaster, including the consequences of bioterrorism.

As you know, over the past two decades, there has been a growth in physician training being accomplished in non-hospital settings. Indeed, the Balanced Budget Act of 1997 (BBA 97) contained provisions intended to support physician training in rural areas, much of which is performed in the non-hospital site.

Written Agreements: The proposed rule calls for the elimination of written agreements currently required between a hospital and non-hospital site. In exchange for the elimination of this requirement, the Centers for Medicare and Medicaid Services (CMS) would require the hospital to make the payments it owes the non-hospital site by the end of the month following a month during which training occurred. The rule requiring the hospital to pay all or substantially all the costs associated with the residency remains unchanged. This rule in itself is not objectionable. But the provision proposing that “the actual cost of the time spent by teaching physicians in supervising residents in the non-hospital setting must be compensated by the hospital” the Academy views as seriously problematic.

The Academy urges CMS to recognize and make clear that many physicians perform clinical education, supervision and preceptorships on a volunteer basis. In fact survey data show that, in family medicine, as high as 85 percent of clinical faculty provide such services on a volunteer basis. Community-based teachers (CBT) have been used increasingly for ambulatory-based clinical education and residency training with demonstrated positive outcomes for both the student and the preceptor. (See Ferenchick, GS, Chamberlain, J, Alguire, P; Community-based Teaching: Defining the Added Value for Students and Preceptors, Amer J of Med, Vol 112(6), 512-517). In recognizing and clarifying that volunteer teaching is common, CMS should also inform its intermediaries that frequently no direct clinical faculty expense is incurred by the hospital at the non-hospital site and, therefore, no direct payment to the non-hospital site is necessary. In such instance, this should not be sufficient reason to deny time spent by resident in that non-hospital environment nor is it sufficient reason to subject the hospital to an audit and a demand for repayment.

In recent years, CMS has issued regulations citing the agency’s approval of the use of volunteer teachers in non-hospital sites. (Federal Register, July 31, 1998, pg. 40996) allowed for “the existence of volunteer teaching physicians” and (Federal Register, July 30, 1999, pgs. 41517-18) CMS reiterated its policy allowing such volunteer faculty.

Moreover, in a Program Memorandum (Transmittal No. A-98-44, December 1, 1998, Volunteer Teaching Physicians), CMS (then HCFA) once again acknowledged “situations where the non-hospital site has no supervisory costs and the physician is voluntarily participating in training.”

Moreover, eliminating the written agreement, then requiring payment within 30 days seems contradictory as it replaces one burdensome requirement with another. It should suffice for CMS to require evidence of payment made by the hospital for those direct expenses incurred by the non-hospital setting, during the same cost reporting period or within 30 days following the closure of the cost reporting period or upon reconciliation of the cost report for that period.

We appreciate the attempt to lighten the regulatory burden for hospitals complying with the regulations surrounding graduate medical education by removing this requirement. However, for the purposes of family medicine education, written agreements are already required by the Residency Review Committee (RRC) for Family Practice and are part of the accreditation process. Proof of accreditation, therefore, should be used and accepted by CMS as deemed status acknowledging the existence of written agreements. Therefore, time spent in these non-hospital sites is eligible for reimbursement.

Implementation of Moratorium: The proposed rule codifies the statutory exception for family practice. The provision stipulates that for calendar year 2004, there will be both a one year moratorium on CMS activity (denials) with regard to the use of volunteer faculty and that any agreement allowing the use of volunteer faculty during this year will be approved in subsequent audits (i.e., during the settlement of cost reports pertaining to 2004).

We are pleased that the agency clarified in the proposed rule its interpretation of the statute in this fashion. However, information from the field strongly suggests that CMS and intermediary activity is not consistent with the statute nor with the interpretation published in the proposed rule and accordingly we urge CMS to issue an explicit call for discontinuation of audit and denial activity with respect to time spent in non-hospital sites where the supervisory physician is volunteering his/her services.

Rolling average and (IRB cap):
The Academy does not believe that any new slots achieved by a program/hospital must necessarily be subjected to the rolling average calculation. The Congressional intent behind this section of the statute was to move positions that have gone unfilled to institutions where they are likely to be filled. When done, it should allow for the payment/reimbursement of residency positions that the institution was unable to have reimbursed under the caps set by the BBA and modified by the BBRA. Congress did stipulate how these new positions should be paid: DME would be paid at the national average per resident amount; IME would be at the equivalent of 2.7 percent. Consequently, we believe that the hospital/program should be reimbursed in that fashion for the time the new slots are filled by a resident.

Demonstrated likelihood eligibility criteria:
The proposed rule requires that certain criteria be met in order for an institution to be eligible to garner new slots above its current cap. In order to be eligible to be awarded new training slots, the hospital/program must meet one of the following four proposed eligibility criteria: (1) establishing a new residency program, (2) expanding an existing program, (3) current resident count exceeds cap, and (4) loss of accreditation is possible if hospital does not increase FTE residents.

Criterion 1: Establishing a new program
The proposed rule would require the hospital to demonstrate that each of the hospital's existing programs has a fill rate of at least 95 percent in 2001 through 2003, submit a cover page of employment contracts with current or future residents, or document that the specialty has a national fill rate of 95 percent.

While we concur with CMS's use of the national fill rate as an appropriate measure, we urge CMS to define "fill rate" as the number of residents training in a program or programs as of July 1 of each year. Such information is available on the ACGME Web Accreditation Data System.

Lastly, we would urge to make clear in the final rule that this rule-making does not replace or supercede the method established in the Balanced Budget Act (BBA 97) which allows rural hospitals to establish a new GME program.

Criterion 2: Expand an existing program
Again, we would urge to make clear in the final rule that this rule-making does not replace or supercede the method established in BBA 97 which allows rural hospitals to expand a rural program up to130 percent of their BBA-set cap.

Criterion 3: Resident count exceeds cap
To meet this criterion, the proposed rule would require a program/hospital to submit each of following three items: most recent cost report documenting resident caps and counts; 2004 match information showing intent of number of residents in program; and copies of recent accreditation letters (showing the number of residents for which the program is accredited).

The information essential to the determination of compliance with this criterion is the evidence that resident counts have exceeded the caps. In addition to the cost report, the Academy urges CMS to allow hospitals to submit Intern and Resident Information Survey (IRIS) data, contract cover pages and resident schedules that demonstrate that the actual resident FTE that could be counted for IME and DME purposes is greater than the cap allows. Moreover, rather than requiring 2004 match information, CMS should require the 2004 fill rate for reasons cited above.

Item three under this criterion should be deleted as this information is not useful in demonstrating that a resident count is above the institution’s cap.

Criterion 4: Residency program at risk of losing accreditation since numbers are too low
This criterion does not fit with the requirement to “demonstrate a likelihood” to fill and should be deleted.

Priority for Redistribution:
The Academy commends CMS for crafting the priority list to include items such as rural and underserved areas and minority institutions. As you know, there is substantial evidence demonstrating that family practice programs with as little as three months training in rural settings, results in the production of graduates in excess of fifty percent of the time who practice in rural areas. (See “The Case for the Development of Family Practice Rural Training Tracks,” Journal of the American Board of Family Practice [Sept-Oct 1998] and “Family Practice Residency Programs and the Graduation of Rural Family Physicians,” Family Medicine [April 1998]. These studies clearly demonstrate the direct relationship between time spent in rural training and production of graduates who serve in rural areas. Furthermore, evidence shows that rural training tracks, which have more limited rural time and that rotate more residents to rural areas have a much higher yield (in the range of 70 percent) of rural physician production.

It should also be noted that both of the above-cited studies show that Congressional intent for increasing rural access to physicians would be better served by fully funding, or giving additional support by increasing residency slots to, family practice programs with three or more months of training in rural sites.

Based on this information we urge CMS to reduce the current threshold requirement of 25 percent time in the current evaluation criterion three. This length of time (25 percent) in a family practice training program would be nine months. Since the data show that only three months training time in rural areas is necessary to produce effective outcomes (number of rural physicians, and since the family practice RRC also requires two years of continuity training with the same patient population, most programs cannot meet a 25 percent requirement. Accordingly, we urge CMS to reduce this threshold to no more than 12.5 percent.

A similar concern is registered with evaluation criterion seven which requires the hospital to be located in a health professional shortage area (HPSA) or Medicare physician scarcity county. In the context of the above data, the criterion is too narrow and Academy urges CMS to recognize and include time residents spend training in these areas, even when the hospital is not physically located in the HPSA or scarcity county.

Thank you for the opportunity to comment on the proposed rule and as always, the Academy looks forward to continued cooperation with CMS on these and other issues.

Sincerely,


James C. Martin, M.D., FAAFP
Board Chair