|
Chairwoman Johnson, Representative Stark, members of the Subcommittee, thank you for the opportunity to share some thinking about reasons to change Medicare payment policies and a new framework for making appropriate changes. The Rural Policy Research Institute is committed to completing analysis that helps develop policies that address the needs of people in the rural places where they live. This reflects a belief that there are differences in place that ought to be recognized in public policies, including Medicare, with policy choices directed to the places where people live. I will conclude this testimony describing a policy choice for Medicare payment policy that does just that. Before reaching that conclusion I will use the current conceptualization of policy to make three points: 1) health care providers in small, remote rural communities continue to struggle to survive financially; 2) the existing policy framework forces the use of incremental changes to payment formulas that were not designed to address the fundamental problem of delivering services where operating revenue will not sustain services; and 3) the approach of payment policies has led to accumulating technical changes that overwhelm the systems used in payment. THE CURRENT FINANCIAL STRUGGLE OF RURAL PROVIDERS While not a perfect indicator of the financial condition of providers in the small, remote communities that are the focus of my testimony, the plight of small rural hospitals is a good proxy indicator. In its March, 2001 report, the Medicare Payment Advisory Commission (MedPAC) presented data showing that rural hospitals experienced negative margins on their total Medicare business, -2.1% in 1998 and -2.9% in 1999 (Figure 5-8). The picture is especially bleak for rural hospitals with fewer than 50 beds that are not included in any special payment category, 54.5% of those institutions experienced negative margins in 1999 (Table B-5), and their aggregated overall Medicare margin was -5.6% (Table B-10). Losses generated in serving Medicare beneficiaries might be offset by other sources, including local taxes and foundations and/or payment from other insurers, but not in all of hospitals in this group. In the aggregate, the total margin reported by MedPAC for those hospitals in 1999 was 2.1%, much less than the overall 4.9% reported for all hospitals (Table B-15). In its December 14, 2000 report of hospital margins, MedPAC presented data reflecting improvement in hospital margins, but even so, 34% of hospitals, mostly rural, had negative total Medicare margins. The National Advisory Committee on Rural Health, on which I serve, recently learned of the plight of small hospitals in some of the most remote territory I have seen, in Northern California. The 72 rural hospitals in that state averaged a -2.2% patient-operating margin in 1999, and 74% of them lost money on their operations in 1999. We visited one rural hospital that recently converted to Critical Access status, after being very near bankruptcy. Its most recent fiscal year showed a slight loss on operations, but thanks to an aggressive plan to expand services available in the community, it is on a path to recovery, albeit with continued very narrow margins. Two images were burned in my mind in that visit. First, that hospital is only barely surviving financially, because even though it is doing everything a good consultant would recommend, it suffers the natural disadvantage of being located in a remote mountainous area. Second, as difficult as the struggle is for that hospital, it is planning to affiliate or merge with a neighboring (28 miles away in an area that can receive as much as 20 feet of snow in the winter) hospital that is literally bankrupt. This is a tale of difficulty about which aggregate data can only give us hints. For me, this demonstrated the value of combining case study data with aggregate data in our research, giving you a better sense of the range of circumstances that exist in rural America. The example of Northern California is repeated elsewhere in rural areas – providers that are in precarious financial condition but finding ways to continue providing services. The access to services they provide today is assured only if they are able to continue to balance operating losses with other sources of revenue, and doing everything possible to reduce those losses. CURRENT POLICY FRAMEWORK Current policies present imperfect choices to address the needs of rural communities. Medicare payment policies are designed, as summarized by MedPAC, to create a system of efficient payment for services delivered to beneficiaries, not to recognize needs caused by environments in which providers cannot survive under financial models assuming constant gains in efficiency. Yet those policies are the tools available to address the needs of a unique group of providers serving the small, remote rural communities. Therefore, proposals such as setting a floor payment in the wage index, or equalizing the base payment in the prospective payment system, appeal to the providers and their associations who can calculate an improvement in revenue that may "stop the bleeding." Other suggestions, such as adjusting payment systems to increase per case payment in situations where the number of cases are few (low-volume adjustment) are intended to bring the operating margin closer to, if not in balance. But because there are multiple payment streams to multiple providers in the same communities, the incremental approach would need to apply to all of those payment systems. This is the nature of the segmentation inherent in decisions that track with the type of service and category of provider. It should not be surprising that another approach to this problem gains favor; that of creating a separate, cost-based payment system for vulnerable providers. This would seem to be the most certain means of meeting the needs of providers in target communities, since it would neutralize the effect of Medicare payment on the financial condition of those providers. However, two problems remain. First, even cost-based payment may not be sufficient if "allowable" costs are not inclusive and if there is no possibility to build up reserves needed to maintain quality services. Second, the policy is still linked to providers, which presents complications in trying to be sure the definition of provider matches with communities that need those particular providers (the concept of "essential provider"). Despite these problems, this approach would appear to be succeeding, for the most part, in the designation of Critical Access Hospitals, 91% of which are located in counties designated as Health Profession Shortage or Medically Underserved Areas, 65% are the only hospital in their county, and nearly 83% are located in counties with higher than state averages of persons aged 65 or over. UNINTENDED CONSEQUENCES OF MULTIPLE, SEGMENTED TECHNICAL SOLUTIONS The legacy of numerous incremental attempts to address the special financial circumstances of rural providers, combined with repeated efforts to refine and establish prospective payment systems, has overwhelmed the capacity of administrative systems. The numbers of providers in different categories is some indication of the complexity being created: as of 1997 there were 1804 independent rural health clinics (RHCs) and 1525 provider-based RHCs; the most recent PPS update identified 165 Rural Referral Centers, 667 Sole Community Hospitals, and 328 Medicare Dependent Hospitals. None of these classifications are necessarily inappropriate, but the mere existence of separate provider types necessitates at least different applications of payment systems and regulations, if not completely separate regulations. The inherent strain of multiple policies is most apparent when new policies are enacted. For example, not all provisions of the Balanced Budget Act of 1997 have been implemented through publication and application of final rules. Issues have arisen in the implementation of new systems to certify and reimburse Critical Access Hospitals, resulting in legislative amendment in 1999 and 2000. Implementation of prospective payment for outpatient services has been slow, affecting timely completion of cost reports and therefore delaying fiscal year end reconciliation, and forcing new policies based on cost report data to start late or use old data. Frequent and multiple "fixes" to payment systems can easily outpace the ability to use appropriate data, both for developing systems and for measuring their impacts. The delays in cost report data described earlier means wage data (including occupational mix) used in prospective payment formulas are reliable only as of some years prior to the payment. Recent increases in wage scales and mix of employees (due to changes in medical practice and/or new regulations) will not be reflected in the payment. Delays in implementing new systems might be only bumps in the road were it not for the effects on the small, financially vulnerable providers that are the focus of this testimony. A seven month delay in reconciliation based on the year-end cost report could result in a serious, even crippling, cash flow problem for a provider expecting a positive result. Problems in implementing legislation that initially does not include the specific authority intended (for example, cost-based payment for lab services provided by CAHs) can be resolved, and retroactively (as was the case in CAH lab payment). But, again, that means a delay in receiving expected revenue. The resolution to this pattern of ever-increasing complexity is not "administrative simplification," since that only makes each regulation simpler and does not slow down the increase in volume every time a payment system is altered. A completely different approach would offer a better solution. PAYMENT BASED ON COMMUNITY With access to locally-based services as an objective of public policy, a new framework would identify those communities within which any provider would be unable to sustain services based on operating revenue. The RUPRI Center is currently developing a means of designating "vulnerable communities" as an alternative to designating specific types of health care providers. Our aspiration is to use information that can be obtained from Census data and other public sources in a formula that accounts for low expected patient-based revenue. We have begun by mapping areas based on population aggregation, to at least 1500 persons in a "community" (town and surrounding census tracts). Within each community, in addition to total population, variables expected to be related to operating revenues are: household income, percent of population enrolled in Medicaid, percent of population enrolled in Medicare, percent of the population at or below federal poverty level in income; percent of workforce unemployed, and population per square mile. Population per square mile is one measure of isolation, but physical terrain will need to be operationalized and either incorporated into a formula or used as second-level criteria. By building the definition of community using census tracts rather than county boundaries, this approach can deal with the problems of having small isolated communities within very large counties that include areas that are well-served. This is a problem in western counties, like the ones the National Advisory Committee on Rural Health toured last week. We are currently testing the early phases of our modeling with data regarding communities in Nebraska, and so far our thinking passes the "face validity test" of identifying communities we know have difficulty recruiting and retaining health care practitioners and sustaining institutional providers. Once communities are identified as "vulnerable" policies devised at the federal, state, and local levels of government can be used to maintain the economic viability of providers based in those communities. For the federal government the policies can be a combination of payment from Medicare and availability of grant funding for special needs such as capital investment or conversion of information systems to develop new systems of quality improvement. From state governments the policies can include special payment from Medicaid and direct financial assistance. Local policies can include access to a tax base (which may be broader than the single community), direct assistance, and/or help in organizing community-based foundations. Hospital financial data indicate, to at least some degree, communities who are self-identified as "vulnerable." Small rural hospitals (under 50 beds) with substantial income from government subsidies and non-patient revenue are financing care through these means. Of the types of hospitals as classified by MedPAC, the category with the highest percent gain from "other government payers and subsidies" are the small rural hospitals. For those hospitals, a loss of non-patient revenues would have resulted in net losses in 1999 (Table B-15). Of course, those are aggregate numbers, and the reason to develop a community-based policy is target resources effectively within clusters of hospitals, or any other health care providers. If this or some other means of identifying communities is found to be effective, non-payment policies, including special grant programs and flexibility in meeting conditions of participation, can be targeted to specific places. The same is true for grant programs, including technical assistance and promoting innovative means of meeting local needs, such as developing networks. I am suggesting a framework for considering policies targeted to issues of access to care in rural areas. The focus of this framework is place, not provider. Short of finding acceptable means of imposing this framework, policies targeting providers should be based on underlying assumptions about the communities they serve. HOME Updated Tuesday, November 04, 2008 09:58:05 -0600
|