Over most of the past decade, CMS has constantly modified its Medicare policies governing payment for direct graduate medical education (DGME) and indirect medical education (IME). Last year was no exception. Unlike in prior years, however, where the focus has largely been on Medicare affiliation agreements or non-hospital training agreements, CMS’s focus in 2006 was on didactic time. CMS has announced a new policy, which it has labeled a “clarification,” that significantly could affect hospitals’ ability to claim reimbursement for resident time spent in didactic activities. CMS’s new policy on didactic time—that is, time spent by residents in journal duties, seminars, classroom lectures, and other scholarly pursuits—appears both illogical and difficult to implement, as commenters have pointed out. Nevertheless, CMS remains firm in its conviction that this new approach is appropriate. As a consequence, hospitals can expect that their Medicare contractors will make additional payment cuts in IME and DGME reimbursement for future years and, because this is a policy “clarification,” for past years.

Background
General Rules
Medicare reimburses hospitals for both the direct and indirect costs of their medical education programs. The DGME payment is made to compensate hospitals for their direct costs—resident stipends, teaching physician salaries, and the like—and is based on a hospital-specific per-resident amount multiplied by the number of full-time equivalent (FTE) residents training at the hospital times the hospital’s Medicare patient volume (the “patient load”). 42 U.S.C. § 1395ww(h). The IME adjustment, similarly, is based largely on the number of FTEs training at the hospital. Unlike the separate DGME payment, however, the IME adjustment, authorized by 42 U.S.C. § 1395ww(d)(5)(B), is an adjustment or “add-on” to the Medicare diagnosis related groups (DRG) payments made to each hospital. Congress’s stated rationale for enacting the IME provision was to address several factors that contribute to higher costs incurred by teaching hospitals as a result of the hospitals’ teaching programs. These factors include teaching hospitals’ treating patients with a greater severity of illness; the hospitals’ offering more specialized services; and the hospitals’ incurring additional costs associated with training residents, such as the additional tests ordered and extra staffing demands. See H.R. Rep. No. 98-25 at 140–41 (1983), reprinted in 1983 U.S.C.C.A.N. 219, 359-60. See also 71 Fed. Reg. 18,654, 18,655–56 (Apr. 12, 2006).

Both DGME and IME payments are available for training that takes place in the hospital and in nonhospital locations. For training that takes place in the hospital, Medicare has typically allowed FTEs time to be counted, for DGME, as long as the resident activities take place in the “hospital complex,” irrespective of the form of those activities. 42 C.F.R. § 413.78(a). For IME, CMS requires that the residents’ activities take place in a part of the hospital that is subject to acute care PPS or in a hospital outpatient department. 42 C.F.R. § 412.105(f)(1)(ii). The agency has also stated, since 1988, that intermediaries are not to count an FTE as part of the IME adjustment if “[t]he individual is engaged exclusively in research.” Provider Reimbursement Manual (PRM) § 2405.3.F.2. In 2001, CMS clarified this point as part of the inpatient prospective payment system (IPPS) update, stating as follows: “Resident time spent ‘exclusively’ in research means that the research is not associated with the treatment or diagnosis of a particular patient of the hospital. Therefore, although the research component may be part of an approved program, the time that residents devote specifically to performing research that is not related to delivering patient care … may not be counted for IME payment purposes.” 66 Fed. Reg. 39,828, 39,896 (Aug. 1, 2001). See also 42 C.F.R. § 412.105(f)(1)(iii)(B) (2005); 66 Fed. Reg. at 39,933–34 (research that is not associated with the treatment or diagnosis of a particular patient is not countable).

As noted earlier, hospitals may count for both DGME and IME not only in-hospital time but also time spent by residents training in nonhospital locations, subject to certain conditions. For DGME, the hospital must incur all or substantially all of the costs of the training and the care must be spent in “activities relating to patient care.” 42 U.S.C. § 1395ww(h)(4)(E). Similarly, for IME, the hospital must incur all or substantially all of the cost of the training, and the time must be spent in “patient care activities.” 42 U.S.C. § 1395ww(d)(5)(B)(iv).

Related to Patient Care
Although the statutory nonhospital training site provisions contain a related-to-patient-care requirement, there is no parallel statutory requirement contained in either the DGME or IME inpatient provisions. Nevertheless, CMS has taken the position that, for purposes of the IME adjustment, there must be a nexus to patient care for both inpatients and outpatients and that this requirement is nothing new. In response to comments questioning its position in the context of research time and the IME adjustment, CMS asserted in the August 1, 2001 rule that its change in regulations “addresses our longstanding interpretation of existing regulations and reflects longstanding general Medicare reimbursement principles” under which, “as reflected in § 413.9, costs incurred by a hospital generally must be related to patient care in order to be reimbursed by Medicare.” 66 Fed. Reg. at 39,897.

Although CMS has said this, in applying the related-to-patient-care principle in the past, CMS has limited its discussion to research time and at no time has suggested a principle of broader application. Certainly, CMS has not suggested that didactic time lacks a sufficient nexus to patient care to be allowed as part of the IME calculation. Indeed, CMS’s prior statements suggested just the contrary. In a letter sent from CMS in 1999, the agency expressly stated that, in the context of nonhospital settings, CMS interpreted the phrase patient care activities quite expansively, to include any patient-care-oriented activities that are part of the residency program, which it said includes both resident participation in “1) the direct delivery of patient care, such as clinical rounds, discussions, and conferences, and 2) scholarly activities, such as educational seminars, classroom lectures, research conferences, patient care related research as part of the residency program, and presentations of papers and research results to fellow residents, medical students, and faculty.” In its inpatient PPS rule for FY 2007, however, CMS has performed an about-face and has taken the position that time spent in didactic activities is for the most part, unallowable, a position that could lead to the rejection of significant training time spent by residents as part of their residency programs.

CMS’s Policy—A “Clarification” That Will Surprise Many
On April 25, 2006, CMS issued proposed changes to the hospital inpatient prospective payment system for fiscal year 2007 rates. 71 Fed. Reg. 23,996 (Apr. 25, 2006). Included among these proposed changes was a policy “clarification” addressing time spent by residents in didactic activities and how that time is to be treated in the calculation of DGME and IME payments. 71 Fed. Reg. at 24,114–15. In its clarification, CMS took the position that time spent by residents in didactic activities in nonhospital locations cannot be included as countable time for purposes of either DGME or IME payments and that, similarly, time spent in didactic activities in the hospital setting must be excluded from the IME count. Only time spent in DGME activities conducted in the hospital was proposed to be exempt from this limitation.

CMS’s so-called clarification prompted many comments critical of the proposal, but, in its final rule, CMS continued to hew to its position. CMS concluded that didactic time spent in nonhospital locations is not allowable for purposes of either DGME or IME calculations; that didactic time spent in hospital locations is not allowable as part of the IME count, but is allowable for purposes of the DGME count; and that this newly articulated policy is nothing more than a recapitulation of past, long-standing policy. 71 Fed. Reg. 47,870, 48,072–99 (Aug. 18, 2006).

CMS’s rationale for its nonhospital site position is that the statute, at 42 U.S.C. § 1395ww(h)(4)(E), restricts the payment for training time in such sites to that time “spent in activities relating to patient care” under an approved medical residency program, for purposes of DGME, and similarly restricts payment for such time, for purposes of IME, to time spent by the intern or a resident in “patient care activities” under an approved medical residency program, 42 U.S.C. § 1395ww(d)(5)(B)(iv). CMS then makes a leap of logic and, while acknowledging that it has never explicitly defined in regulations the term patient care activities, reasons that the term’s plain meaning refers to the care and treatment of particular patients, or to services for which a physician or other practitioner may bill. 71 Fed. Reg. at 48,080–81. Based on this reasoning, CMS concludes that “[t]ime spent by residents in other activities in the nonhospital site that do not involve the care and treatment of particular patients, such as didactic or ‘scholarly’ activities, is not allowable for direct GME and IME payment purposes.” 71 Fed. Reg. at 48,081.

In taking its “didactic-time-is-not-related-to-patient-care” position, CMS explicitly rejected the earlier letter that it wrote related to didactic activities in nonhospital locations, labeling that position an “erroneous response to a question” that “inaccurately stated our interpretation of the phrase ‘patient care activities’.” 71 Fed. Reg. at 48,081–82. Similarly, CMS was unpersuaded by arguments made by commenters that the agency’s application of the term patient care activities is contrary to CMS’s historical interpretation of that term, under which CMS had always allowed conferences and seminars for hospital employees and similar activities as costs related to patient care. 71 Fed. Reg. at 48,086. CMS concluded that IME is a payment “specifically for patient care costs” and that past regulations and subregulatory guidance concerning “cost related to patient care,” cited by commenters, does not determine what constitutes patient care for purposes of medical education. 71 Fed. Reg. at 48,087.

In its August 18, 2006 discussion, CMS also explained the reason for its position allowing in-hospital didactic time for DGME, but not allowing it for IME. CMS explained that, for DGME, “residents in an approved program working in all areas of the hospital complex may be counted,” citing 42 C.F.R. § 413.78(a). Thus, as long as the resident is in the hospital complex—the hospital and its hospital-based providers and subproviders—the distinction between patient care activities and nonpatient care activities is not relevant to the DGME FTE count determinations, according to CMS. 71 Fed. Reg. at 48,081. For IME, however, CMS said that the rules are different, and “consistent with the regulations at § 413.9 [the reasonable cost regulations governing activities being related to patient care], only time spent in patient care activities in the hospital may be counted.” Again, for IME, CMS asserted that “the requirement for residents to spend time in patient care activities is fundamental to including the FTE resident time in the count. . . .” This policy, CMS said, is “rooted in the creation and the purpose of the IME adjustment,” which is to reimburse teaching hospitals for their higher costs of patient care. 71 Fed. Reg. at 48,081–82.

In adopting its didactic time position, CMS rejected comments that didactic activities are an integral component of residents’ patient care activities, stating that it was not convinced that didactic time is “frequently integrated with patient care activities” and thus should be allowed. 71 Fed. Reg. at 48,085. CMS similarly dismissed arguments that higher costs are incurred by teaching hospitals and that removing didactic time from the IME count is, thus, contrary to the statute’s purpose of reimbursing those higher costs. CMS countered that, while engaged in didactic activities, the residents are not “participating in or contributing to more intensive or inefficient patient care.” 71 Fed. Reg. at 48,088. Moreover, since direct DGME payments are made to cover the “explicit educational costs of training residents,” CMS maintained, Congress did not intend for “the IME adjustment to duplicate those educational payments.” Id.

CMS’s Documentation Policy
CMS explained generally how it envisions time spent in didactic activities should be treated for purposes of the FTE count. Under CMS’s policy, if a resident, for example, were to work 80 hours in the course of a week and spend two hours of that work week in didactic activities, one-fortieth of the resident’s time could not be counted by the hospital for IME (or nonhospital GME). See generally, 71 Fed. Reg. 48,086. At first blush, this would appear to mean that both the hospital and the resident have to engage in detailed recordkeeping so as to “capture” all didactic time. That may not be the case, however.

The single bit of good news in CMS’s policy regards CMS’s recordkeeping expectations. In its preamble discussion, CMS recognized that maintaining documentation for scores, and often hundreds, of residents is no easy task, particularly if training activities are reported in short increments of less than a day. Accordingly, CMS said that it would impose, for cost reporting periods beginning on or after October 1, 2006, a “one workday” requirement. 71 Fed. Reg. 48,091–92. Under this policy, if a hospital’s rotation schedule reflects time elements of a day or more, didactic time will be disallowed only if a resident’s workday consists “entirely of scheduled didactic activities and no scheduled patient care activities,” in which case “that workday must not be recorded as ‘patient care.’” 71 Fed. Reg. at 48,091. Stated another way, as long as there are some patient care activities that occur in the course of that workday—that is, a 24-hour workday—the schedule need not reflect time has been spent in other than patient care activities (i.e., in didactic activity). Conversely, if a full workday is spent in didactic activity, or if, in fact, the hospital maintains rotation schedules with increments of less than a full day that reflect didactic activities, that time must be subtracted from the count.

It is here that problems could arise. CMS’s latest policy encourages teaching hospitals to maintain rotation schedules using time increments of no less than a full day. If the teaching hospital uses smaller increments—such as time records with blocks of one or four hours—and if those increments show any didactic activity, that didactic activity time must be subtracted from the overall count. On the other hand, if the teaching hospital maintains rotation schedules using the time element of a single workday and if, within that workday, time is spent both in didactic and patient care activities but the time on the daily rotation shows patient care activities, the teaching hospital would be permitted to claim the full day as a patient care day, according to CMS.

CMS’s documentation policy is fine insofar as it goes. The problem is that CMS has also made it clear that, if the intermediary learns that residents have engaged in didactic activities within a given day, even if this is not reflected on the rotation schedule, the fiscal intermediary may rely on this fact to deduct time from that otherwise claimed by the hospital. Thus, if a hospital has on hand evidence of the hospital’s residents having spent time in didactic activities and if the intermediary learns of this evidence, FTE time claimed by the hospital could be disallowed. This could place the hospital in a bit of a bind. If, for accreditation or other reasons, the hospital needs to maintain accurate records of how many hours its residents spent in didactic activities, those records could conceivably be demanded by the fiscal intermediary and used to disallow time related to those activities, even if the rotation schedules do not reflect didactic activities. Moreover, it is unclear how the OIG or the DOJ might later view a claim of a full FTE, which claim is supported by the “full workday” rotation schedule, when the hospital knows that the resident, in fact, spent a portion of that day in didactic activities. Could someone later suggest that this was a false claim? This is a troubling ambiguity in CMS’s policy.

Conclusion
There are unanswered questions flowing from CMS’s latest didactic activity position. What is almost certain, however, is that this policy will be challenged. Although CMS has gone to great lengths to justify its policy, the many comments that it received are all sound, and CMS’s attempt to issue its policy as a “clarification” is quite questionable. Although CMS’s policies are given great deference by the courts, this is one policy, I suggest, that may well fall when challenged.