Statement of Kathleen Dalton, Ph.D., Research Faculty Member, 
Cecil G. Sheps Center for Health Services Research,
University of North Carolina at Chapel Hill

Testimony Before the Subcommittee on Health
of the House Committee on Ways and Means

Hearing on Rural Health Care in Medicare

June 12, 2001

Madame Chairman and members of the committee, thank you for inviting me to speak to you today. My name is Kathleen Dalton and I am a research faculty member at the University of North Carolina, where my field is health care finance. In recent years I have participated in several funded research projects that investigate rural hospital cost and reimbursement issues. I have been asked to describe some of our findings to you, focusing specifically on the role of Medicare payment policies in addressing special financial problems faced by rural health care institutions.

Rural providers and constituent groups are alarmed at the deteriorating Medicare and overall financial margins of their health care providers over the past three years. Several advocacy groups are examining specific components of Medicare's prospective payment formulas to identify ways in which to obtain some regulatory or legislative relief. To help in assessing some of the specific issues with regard to Medicare's payments, my remarks will cover the following topics: 1) The hospital area wage index; 2) the Medicare disproportionate share adjustment to hospital prospective payment rates; 3) the gap between rural and urban Medicare payment ratios; and 4) special payments for categories of rural hospitals, particularly the expansion of cost-based reimbursement alternatives.

Hospital Area Wage Index. The wage index is used to adjust PPS rates for regional differences in the cost of labor. Many rural providers and advocates believe that the wage index penalizes rural hospitals because it overstates average wage differences and is based on poorly defined labor markets. Our studies have led us to conclude that the index does not overstate regional differences in average labor costs. In our research, in fact, we find little evidence that the wage index contributes to low Medicare margins. The rural labor markets as currently defined are too large and appear to incorporate multiple market areas. Yet we found that this deficiency has the effect of protecting hospitals in the very smallest and most rural communities because they are grouped with higher-wage areas. In earlier years there was some evidence that hospitals in the larger rural communities were penalized by the broad rural market definitions, but by the mid-1990's many of these hospitals were allowed to be reclassified into neighboring urban labor markets, without bringing down the index values of the markets from which they transferred.

Any set of labor market boundaries will create individual cases where there are arbitrary differences between facilities on either side of a boundary. The existing exceptions process, however, allows for geographic reclassification to regroup cases to neighboring markets if they can meet certain criteria. Reclassification cannot fix all boundary problems, but we believe we have demonstrated that, for rural hospitals as a group, the exceptions process has substantially improved the equity of the wage index. Last year's legislative and regulatory changes to the reclassification rules will improve the index further.

The hospital wage index is now used to adjust prospective payment rates for home health and skill nursing services in addition to hospitals, but without the benefit of the geographic reclassification rules. The reclassification provisions were critically important to our findings, and we conclude that the wage index adjustments to these non-hospital rates will not be equitable unless similar reclassification provisions are available. If reclassification continues to be decided at the individual institution level this process could add considerable administrative burden to the new payment systems. If this is unacceptable, it should be possible to design a provision where the skilled nursing and home health facilities in a given county are paid under the same reclassified index values of that county's hospitals, wherever a majority of its hospitals have already proven that the reclassified labor market is more appropriate.

Disproportionate Share Adjustments (DSH). The stated purpose of Medicare's DSH adjustment has changed over the years since it was first adopted. We now appear to have a consensus that the DSH policy objective is to permit the Medicare program to share in the cost of indigent care, similar to the way in which other payers share when uncompensated care costs are factored into hospital price structures. Under this approach, additional DSH payments are not intended to reflect the higher costs of delivering care to Medicare patients, but rather, the distributed costs of delivering care to other, uninsured patients. DSH is thus acknowledged as a policy-driven, rather than a cost-driven, adjustment to the national standard rates under PPS. This underlying premise has been the justification for MedPAC's DSH-related recommendations over the past several years. MedPAC has recommended parity in the DSH formulas across all hospitals, and that HCFA should develop an improved measure to use a proxy for each hospital's indigent care burden. The Benefits Improvement and Protection Act of 2000 gave rural hospitals parity in eligibility for DSH adjustments, but not in the formulas that govern the size of the proportional adjustment.

If Congress accepts the premise that DSH adjustments should underwrite Medicare's share of indigent care, then several payment policy changes flow logically from this position that go beyond MedPAC's recommendations for formula parity.

Medicare payment ratios. Medicare's inpatient payments historically have been more generous to urban hospitals than to rural hospitals. This is evident in the long-standing differences in their respective average payment-to-cost ratios. When we analyze payment ratios from recent years we find that nearly all of the differential can be attributed to two of the special policy adjustments that Congress has created-- the disproportionate share and the indirect medical education (IME) add-ons to the per-case payment rates. (1)

In fiscal 1998, for example (the last year for which complete hospital data were available from HCFA, and the first year in which the effects of the Balanced budget Act of 1997 were felt) Medicare's inpatient PPS payments averaged 15.1% above costs among urban hospitals, but 6.9% above costs among rural hospitals. Among non-teaching hospitals the gap is much smaller (10.4% compared to 6.8%). If we remove the DSH payments from the calculation, the payment ratios are quite similar for both groups (close to 5%, though actually slightly higher for rural); if we look only at rural and urban hospitals that qualify for neither adjustment, we find that their average payment-to-cost ratios are also about 5%. Thus, the gap in urban and rural PPS margins appears to be more a reflection of federal policy than an indication that rural hospitals, as a group, are unable to compete under the discipline of the prospective payment system.

Graph of Excess of Medicare Inpatient Payments as a percent of Costs, average for rural and urban hospitals, federal year 1998

Recent changes in the DSH policy will make more rural hospitals eligible for these payments, which will reduce the difference in margins between rural and urban non-teaching facilities. If MedPAC's other recommendations regarding the DSH formulas are implemented, the gap will narrow even further. We could not expect the gap to be eliminated, however, because the policy objectives being met by these two payments address needs that are unequally distributed between rural and urban communities. Teaching hospitals are predominantly located in urban areas, and urban hospitals have proportionally higher indigent care loads, so even under identical DSH formulas they would tend to have higher DSH adjustment rates.

Even without the effects of policy adjustments, similarities in average payment-to-cost ratios across groups still mask real differences in their distributions. The proportion of hospitals receiving Medicare PPS payments that are below their cost of delivering services is always greater for rural than for urban facilities. A subgroup of very small rural facilities has always struggled to cover costs under the PPS rate structure. Many of these have adapted by becoming cost-based Critical Access Hospitals (CAH). If the facilities that are struggling are important to protect in order to maintain rural beneficiaries' access to care, then the task of lawmakers is to identify whether and how to address the problems of these sub-groups with further payment policy adjustments.

Special rural payment exceptions and the expansion of cost-based alternatives. The various Medicare prospective payment systems already define several categories of rural health care providers that are eligible for special payments or exemptions from rate limits. These categories target different groups of facilities with varying levels of success. Some provide only small supplemental payments to groups that are in need of substantially more relief, while others provider substantial payments to facilities that qualify on the basis of a geographic characteristic, yet may not demonstrate any financial need. Much attention recently has focused on Critical Access Hospital (CAH) status, which allows very small, isolated hospitals to withdraw from prospective payment systems altogether for inpatient, outpatient and swing bed services, and receive cost-based reimbursement. Proposals to expand cost-based alternatives to other rural hospitals have been circulating among rural advocacy groups; at the same time, MedPAC is proposing a new low-volume adjustment to the inpatient PPS rates, in order to provide an alternative to cost-based payments for the smallest providers.

The proliferation of special exceptions and protected status makes it difficult for providers, analysts and lawmakers alike, to evaluate the adequacy of Medicare payments in rural settings. Our analyses of the economic and related health-delivery problems of rural communities have revealed great geographic variation. Severe isolation and sparse populations predominate in the west, while low-volume but only moderately isolated providers predominate in the mid-west. In the southeast and Appalachian regions we find multiple proximate hospitals, with low average census figures and very low average occupancy, but they serve communities that have very low incomes, poor employment and high levels of indigent care.

When we examine the Medicare margins as well as the overall finances of rural hospitals over periods of several years, we are struck by two characteristics in particular. First, even among those located in the most disadvantaged of rural markets, some facilities manage to earn good margins on Medicare and/or other patient services. Thus any Medicare policy offering automatic payment adjustments based on geographic criteria without a demonstration of financial need is potentially misdirected. Second, the margins for any one institution are unstable from year to year, particularly in the smaller providers. Hospitals that opt for cost-based alternatives to Medicare PPS are guaranteeing that they will not have to deliver Medicare services at a loss during bad years, but they are also giving up the possibility of ever earning a surplus during better times. In an uncertain environment where many providers are able to earn a surplus on some PPS services in some years, the expansion of cost-based reimbursement should be a last resort alternative.

We think that a well-designed DSH policy can alleviate many of the problems of rural providers located in very poor areas, without over-committing Part A trust fund support. Many of these communities may continue to need additional local and state-level public support, or even a different level of federal support, but those needs should be considered over and above Medicare's share of the burden.

Apart from the DSH adjustment, the underlying justification for special exemptions, payment add-ons or other subsidies should relate back to the Medicare beneficiaries and the need to provide adequate access. There is a certain class of rural provider that should be protected, if there is evidence that the loss of these providers could jeopardize beneficiaries' access to care. That class may be difficult to identify, but we think that the criteria should relate to characteristics of the community (or beneficiaries) rather than to the financial condition of the individual institution. The class should certainly include very isolated providers, but it may also include providers that are not isolated, possibly are underutilized, but that serve as an important focal point of a poor community's health infrastructure (for example, where the loss of the hospital would jeopardize the town's ability to retain physicians, pharmacists or emergency medical personnel). Even if the class is broadly defined in terms of the number of qualifying providers, that group is not likely to represent a very large share of total Medicare payments. Whether Congress chooses a broad or narrow definition of the group, the question remains, what is a better way for Medicare to support it that would not overwhelm the system with multiple options, yet would still target only those institutions in need??

Many of the special payment provisions now in place could be rationalized by simply making a single commitment to this class of protected health care providers: that the Medicare program will never pay less than cost for the care of its beneficiaries, in any year. Paying everyone under the PPS rules but setting a cost-based floor (while excluding the DSH payments from the payment-to-cost ratio calculations) could provide consistent help to marginal hospitals, periodic help to providers suffering from temporary shifts in demand, and would avoid unnecessary subsidies to providers that are able to operate successfully within the discipline of the various PPS systems. The incremental Medicare payments from such a policy are likely to be very low (based on our historical estimates from the inpatient component), and the administrative burdens would probably be less than those required to support the current system.

In summary, my points are as follows:

1. The IME adjustment is theoretically cost-based, but the formula intentionally computes an adjustment that is greater than the expected cost differences in teaching hospitals. The portion of the IME add-on rate that is attributable to this intentional increase (we estimate it at 30%- 40% in the FY 1998 data) is viewed as a policy adjustment.