Post-Acute Care Issues for Medicare: Interviews with Provider and Consumer Groups, and Researchers and Policy Analysts

05/01/2000

U.S. Department of Health and Human Services

Post-Acute Care Issues for Medicare: Interviews with Provider and Consumer Groups, and Researchers and Policy Analysts

Korbin Liu, Urban InstituteJennie Harvell, U.S. Department of Health and Human ServicesBarbara Gage, The MEDSTAT Group

May 2000

PDF Version: http://aspe.hhs.gov/daltcp/reports/2000/pasissue.pdf (49 PDF pages)


This report was prepared under contract #100-97-0010 between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and The Urban Institute. For additional information about this subject, you can visit the DALTCP home page at http://aspe.hhs.gov/_/office_specific/daltcp.cfm or contact the ASPE Project Officer, Jennie Harvell, at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, S.W., Washington, D.C. 20201. Her e-mail address is: Jennie.Harvell@hhs.gov.

The views expressed in this report are solely those of the authors and should not be attributed to The Urban Institute, the Department of Health and Human Services or The MEDSTAT Group, Inc.


TABLE OF CONTENTS

ACKNOWLEDGMENTS
EXECUTIVE SUMMARY
I. INTRODUCTION
II. THE BBA AND SKILLED NURSING FACILITIES (SNFS)
A. Base Payment Levels for SNFs
B. Impact on Non-Therapy Ancillaries of the RUG-III Casemix System
C. Effects on Therapy Services
D. Impact of SNF PPS and Consolidated Billing on SNF Services
E. Impact on Therapy Services of Part B Cost Caps
F. Additional SNF Issues
III. THE BBA AND HOME HEALTH AGENCIES
IV. REHABILITATION AND LONG-TERM CARE HOSPITALS
A. Respondent Concerns about the Rehabilitation Hospital Provisions
B. Respondents Concerns about the Long-Term Care Hospitals Provisions
V. CHANGES IN THE POST-ACUTE CARE SYSTEM AS A WHOLE AND THEIR IMPLICATIONS FOR THE CONTINUUM OF CARE
VI. RESEARCHERS AND POLICY ANALYSTS PERSPECTIVE
A. SNF-Related Issues
B. HHA-Related Issues
C. Rehabilitation Facility and Long-Term Care Hospital-Related Issues
D. Post-Acute Care System-Related Issues
VII. FUTURE RESEARCH
VIII. END NOTES
NOTES
APPENDICES
APPENDIX A:  
APPENDIX B:  

ACKNOWLEDGMENTS

We are indebted to the many knowledgeable individuals, and their organizations, for sharing their views on policy-related issues regarding Medicare's post-acute care services. The information presented in this report was derived almost exclusively from our interviews with those individuals. Clearly, we obtained much more information than is reported here. We attempted to synthesize and highlight the most salient issues based on analysis and interpretation of information derived in the interviews.

EXECUTIVE SUMMARY

A. Introduction

Until the implementation of Medicare's hospital prospective payment system (PPS), post-acute care provided by skilled nursing facilities (SNFs), home health agencies (HHAs), rehabilitation facilities and long-term care hospitals accounted for only a small part of Medicare spending. After implementation, post-acute spending began to grow at a remarkably rapid rate. Concerned about this trend, policy makers enacted multiple provisions in the 1997 Balanced Budget Act (BBA) aimed at curbing the spending growth. Beyond the spending trend, the increased level of post-acute care activity created other issues that warranted further Medicare policy consideration. For example, as the patient populations served by different provider types appeared to blur with increasing Medicare spending for post-acute care, the problem that Medicare was paying different amounts for similar patients also intensified.

Widespread interest in Medicare post-acute care issues, particularly after the BBA passed, prompted the Office of Disability, Aging and Long-Term Care Policy in the Office of the Assistant Secretary for Planning and Evaluation (ASPE) within the Department of Health and Human Services to sponsor an "early indication" qualitative study of the potential effects of the BBA provisions and other related issues. To obtain timely information on such issues, this study collected information through interviews with individuals knowledgeable about Medicare post-acute care issues. They included representatives of post-acute care providers (including discharge planners) and their organizations, quality of care experts, consumer groups, researchers, and health policy analysts.

It is important to note that while we interviewed a large number of individuals affiliated with many different organizations, the respondents were clearly not representative of all parties interested in or involved with Medicare post-acute care services. Hence, the issues and concerns discussed in this report should be viewed as those of a purposive sample of individuals. This report records most of the issues that were mentioned to us, but we did not select nor attempt to weigh the relative importance of different issues in terms of their importance to providers, consumers, or the Medicare program.

B. Provider and Consumer Perspectives

Skilled Nursing Facility Issues

According to the provider and consumer respondents, the SNF-related issue that warrants the most urgent policy consideration is ensuring that patients in need of intense rehabilitation services or who are medically complex have access to SNF services they need. Directly accounting for non-therapy ancillary services (possibly through outlier payment policies) and refining BBA policies that restrict payment for therapy services were suggested. In addition, respondents noted that refinements to the SNF PPS classification system may be needed simply to reflect changes in types and amounts of resources used in SNFs (e.g., new prescription medications) since the casemix classification system was originally designed.

Respondents indicated that refinements to the SNF PPS are needed because access for some types of SNF patients has reportedly been constrained by the PPS. Nursing homes now have very strong disincentives to admit patients with high non-therapy ancillary costs, because such costs were not explicitly accounted for in the design of the payment system. In addition, the inherent incentives of a price-based PPS have induced some facilities to reduce therapy costs by providing less of it or by shifting such functions to staff other than certified therapists. In general, the provider and consumer respondents questioned whether the casemix adjustments in the SNF PPS adequately distribute nursing and therapy costs across patient groups. Several respondents noted that many providers have not yet been able to calculate the effects of the new payment system on their facility, and that the uncertainty tends to lead to conservative cost cutting practices, raising questions about patient access and quality of care.

The need for cost cutting associated directly with the SNF PPS is also compounded by other BBA provisions, including a requirement that all Medicare costs for nursing home patients be billed only by SNFs. When implemented, this "consolidated billing" requirement means that some services and supplies (e.g., durable medical equipment) that used to be provided by external vendors and billed by them directly to Medicare Part B must be accounted for under the SNF PPS payment rate. The SNF, in turn, pays the external provider or supplier a fee that is negotiated between them. Consolidated billing further raises the cost consciousness of the SNFs while increasing their administrative burden.

The rehabilitation therapy industry, by most accounts, has been strongly affected by several BBA provisions, including Medicare Part B caps on physical therapy and speech and language pathology, and occupational therapy expenses, and the cost savings incentives in the price-based SNF PPS. Layoffs or reductions in compensation have been widespread among all types of therapists.

Home Health Agency Issues

Provider and consumer respondents had a smaller set of concerns about the BBA home health provisions than they did about those for SNFs. A major concern centered on their perception that the limits on allowable costs per beneficiary, instituted by the HHA interim payment system (IPS) as one "stop gap measure" before the HHA PPS is implemented, have "taken much more money out of the system" than Congress intended. They thought that payments are less than agencies need to care for many types of patients with long episodes, such as wound care patients, heart patients, and diabetics.

Another issue cited by the respondents was that some agencies are closing because they often have no capital and no reserves to absorb higher cost patients. As HHAs close, the concern is that higher cost beneficiaries will have more trouble accessing home health care. In addition, beneficiaries that are accepted by agencies might not receive the amount of services (e.g., therapy visits) that they need. Respondents further suggested that this reduced access may lead to increased Medicaid nursing facility use.

A third issue noted by the respondents was an uneven impact of the IPS across HHAs. Some felt that agencies most likely to close because of BBA changes were freestanding, for-profit agencies in dense markets. Hospital-based HHAs (as was also suggested for hospital-based SNFs) may have some long-term protection relative to free-standing facilities, because their link to the hospital setting may give them the opportunity to select patients first, referring high-cost patients to freestanding agencies.

The respondents suggested that the retroactive payment limits on per-beneficiary costs be eliminated as one way to improve the short-term health of the most vulnerable agencies. With respect to the future, some respondents were looking forward to a casemix-adjusted PPS as a viable option for HHAs, particularly if the BBA-required 15 percent payment cut currently planned for the PPS is eliminated.

Rehabilitation Facility and Long-Term Care Hospital Issues

The BBA attempted to address the inequities between new and old providers and reduce the incentives to reduce costs in years following the provider's base year with various policy changes including modifications to reduce variations in payment caps and limits. The BBA also required the Secretary of Health and Human Services to begin on October 1, 2000 a three year phase-in of a casemix and wage adjusted Medicare PPS for rehabilitation hospitals and units. For long-term care hospitals, the BBA required the Secretary to develop a legislative proposal for establishing a casemix-adjusted PPS be submitted to Congress by October 1, 1999.

For both rehabilitation facilities and long-term care hospitals, respondents concerns focused on the reduced "bonus/relief payments" (already implemented). Respondents said that all facilities are feeling the financial pains of these reductions.

With respect to prospective payment for rehabilitation facilities, the main concerns addressed technical aspects of the system's design. Respondents universally favored a per-case rather than a per-diem PPS. The fundamental point they made was that a per-diem system would provide an incentive for facilities to increase lengths of stay, at the same time it limited the resources they could use on a given day--a combination that tends to defeat the goal of rehabilitation (to restore function and discharge home). In contrast, with a per-case unit of payment, facilities would have greater clinical flexibility to provide needed care during a stay and improve rehabilitation outcomes. Most respondents preferred a PPS based on the Functional Independence Measure/Function Related Group (FIM/FRG) classification system previously developed and tested in rehabilitation facilities.

The major options suggested by respondents are now being addressed by HCFA, which announced in July 1999 that the rehabilitation PPS would be a per-case, FRG-based system. HCFA also is studying ways to refine the resident assessment instrument used for SNFs in order to better explain the characteristics and needs of rehabilitation facility patients. It is also addressing measurement issues basic to monitoring quality of care in rehabilitation facilities.

With respect to a PPS for long-term care hospitals, respondents concerns were less immediate, since long-term care hospitals will continue to be paid on a cost basis beyond the time when other post-acute care settings have switched to prospective payment, and work on the design particulars of such a system is still in the formative stages. Respondent concerns focused on their preference for an episode-based payment system for long-term care hospital providers based on the DRG system used in short-stay acute hospital payments.

Post-Acute Care System Issues

The post-acute care system, in the opinion of some respondents, had achieved a continuum of care in which patients were shifted from one setting to another according to what was most appropriate for their conditions. In general, more and more complex patients were being shifted from acute care hospitals "downstream" to post-acute settings. They thought that the BBA payment changes have arrested and are possibly reversing this trend. While many respondents suggested this was a negative outcome, others thought that it may be more medically appropriate to serve certain complex patients in medically intensive settings (i.e., acute care hospitals rather than post-acute care placement).

Various observers thought that the previous system, prior to the BBA, tended to blur the distinction between services provided by the different providers and the patients they serve. Because of the growth in SNFs under the old system, the concern used to be that their patients were overlapping increasingly with the patients served in rehabilitation facilities and long-term care hospitals. As the post-BBA system is evolving, the concern about blurring boundaries is being directed more at overlaps between the patient groups served by rehabilitation facilities and long-term care hospitals.

Home Health and SNF Perspectives. Industry representatives argued that the payment changes were effectively removing the option of home health care for some high cost patients, such as those with medically complex conditions or who require extensive rehabilitation. One consequence is the increased likelihood that high-need patients will remain in institutional settings. At the same time, given the SNF PPS as implemented, nursing homes reportedly are also reluctant to admit extensive rehabilitation and medically complex patients. Such patients, according to this viewpoint, are backing up in hospitals due to lack of placement opportunities in post-acute settings.

Rehabilitation Facilities and Long-Term Care Hospitals. In contrast to the home and nursing home settings, admissions directly from hospitals to rehabilitation facilities are increasing, according to reports. Rehabilitation facilities reportedly now have waiting lists and little competition from SNFs and HHAs for the more complex patients. But they are having increasing difficulty finding home health services following discharge from rehabilitation facilities. Long-term care hospitals reportedly are also seeing increased admissions directly from acute care hospitals and longer lengths of stay. As with the rehabilitation facilities, long-term care hospitals reportedly have waiting lists of medically complex patients that SNFs and HHAs no longer serve.

Quality of Care. The big issue raised with respect to quality of care was the perceived exodus of and reductions in competent specialized staff (such as clinical nurse specialists and therapists) from SNFs and HHAs. The concern was that this would reduce the quality of care available to patients. Interestingly, this view of quality of care trends was not universally shared by our respondents. Some noted that it was too soon to assess the situation, while others thought that quality of care might actually be improved if some of the potential SNF or HHA patients received more care from hospitals.

C. Researchers and Policy Analysts Perspectives

The health care researchers and policy analysts discussed their perspectives on some of the same BBA-related issues raised by the provider and consumer respondents, and offered views on more general issues related to Medicare post-acute care services and the need for future research.

Skilled Nursing Facility Issues

The most important SNF-related issue noted by many of the analysts was a need to make refinements to the casemix payment system for SNFs. Consistent with the providers and consumer representatives' view that the system does not adequately account for the costs of some patients, the analysts saw several areas in which the casemix system could be adjusted to better reflect the range of patients' needs and costs, including non-therapy ancillary costs.

Some analysts thought that, beyond having to address the high non-therapy ancillary costs of some patients, the SNF casemix system needed to be refined to better recognize the high skilled nursing service costs of Medicare SNF patients, while reducing some of the incentives to provide rehabilitation therapy services. Because the rehabilitation therapy categories in the SNF casemix classification system generally have higher payments than the medical and clinical categories, incentives are strong for SNFs to classify patients into one of the rehabilitation therapy categories. The analysts noted the possibility of an increase in "non-traditional" therapy patients (e.g., those with congestive heart failure, chronic obstructive pulmonary disease) classified as therapy patients as an example of the casemix system's incentives.

The analysts also noted that the SNF PPS is likely to affect facilities differentially depending on a host of factors, including historical costs, payment mix of patients, and size. Consequently, access to SNF services by subgroups of patients, even those patients with high non-therapy costs, will also vary. In general, while the analysts expect potential access problems to SNF care for some patient groups (e.g., medically complex), the extent of the problem is unclear. They also were not convinced that increased hospital lengths of stay were evidence of access problems.

In discussing ways to refine the SNF PPS, the analysts suggested that a combination of options may be appropriate, after it is determined how many people are "underpriced" by the payment system. Adjustments should be made without creating perverse incentives, and particular types of services requiring payment adjustments could be addressed more or less readily depending on the resulting potential for "gaming" the payment system. The researchers thought that, in addition to refining the casemix classification system, some type of outlier payments might be warranted (e.g., prosthetics, but not prescription medications).

Home Health Agency Issues

The analysts expressed concern for medically needy or rehabilitation intense home health patients who require either extensive durations of, or high intensity, services. Both situations would equate to high numbers of home health visits and relatively high per-person costs. It seemed plausible that agencies' costs would exceed their per beneficiary payment limit, and that such patients could have, therefore, higher risks of hospitalizations or emergency room use.

Although the casemix-adjusted HHA PPS will be implemented soon, it is not clear how well that system will account for the medically needy home health patients. The analysts generally felt that, regardless of the details of the HHA PPS, it is possible that more research will be required to refine the way that Medicare pays for care provided by HHAs. The analysts expressed the opinion that future research on casemix classification systems should focus on the needs of patients, rather than on the services they received.

Several analysts also noted the fundamental need for a policy clarification (and consensus) of the goals of the benefit. Medicare's home health benefit appears to cover two general populations, individuals with chronic conditions requiring skilled nursing supervision, and those needing restorative care for acute medical conditions. Whether Medicare's home health benefit should cover both groups should be explicitly addressed.

Rehabilitation Facility Issues

The BBA provisions affecting rehabilitation facilities were regarded by the analysts as important improvements in the way that such Medicare providers are paid. The prior payment methodology was regarded as "broken," because it unduly favored new hospitals relative to older ones. Although the PPS for rehabilitation facilities is still under development, the analysts agreed generally that a per-case payment unit, rather than a per diem one, better suits the nature of the product of rehabilitation facilities--a bundle of intensive, time-limited services leading to functional improvement. Moreover, measurement tools applicable to rehabilitation facilities already exist to quantify outcomes of patient care in those facilities (e.g., changes in functional status between admission and discharge).

Post-Acute Care System Issues

Discussion with the analysts about the post-acute care system focused on the extent to which the payment methodologies for the various post-acute care providers could realistically be integrated, and factors that determine utilization patterns of post-acute care providers.

The importance of payment integration in future Medicare payment policy reform efforts depends largely on the extent to which there is overlap of similar patients served by the post-acute care providers. Some analysts thought that, beyond having similar DRGs in the prior hospital stay, there is considerable variation in the needs of patients treated by different providers. Although it might be possible to develop an integrated payment system for some patients (e.g., those with hip fractures), the heterogeneity of the post-acute care population may limit the generalization of that option.

The pattern of post-acute care provider use is determined, in large part, by the supply of particular provider types in a given area. Because of the wide geographic variation in supply of provider types, utilization patterns, even for patients with similar needs, also vary widely by geography. The relatively small number of rehabilitation facilities and long-term care hospitals particularly highlight the potential geographic variability in patterns of use of post-acute care providers.

The supply of provider types is determined, in part, by incentives provided by the Medicare payment systems. The analysts noted, for example, that more SNFs are apparently being certified as long-term care hospitals, since the latter are still paid on a facility-specific, cost-related basis. To deal with the revenue maximizing opportunities in the Medicare post-acute care system, the analysts agreed that it is necessary to first learn about the relationship between payment and outcomes in post-acute care settings. This need increases the importance of developing better measures of quality of care.

Research Issues

In discussing important areas for future research on Medicare post-acute care services, the researchers and policy analysts focused on concerns about the impact of BBA provisions on services for subgroups of the post-acute care population, and ways to measure the benefits of Medicare-funded post-acute care services.

The analysts expressed concern about the circumstances of "underpriced" patients, regardless of post-acute care setting. Are they, for example, experiencing problems with access, and is there a problem with the quality of care provided to those patients who are admitted? The analysts thought it important to identify the number of people who might be adversely affected by the various provisions of the BBA and to estimate the personal and program effects of the provisions. Because the BBA provisions were only recently implemented, data had not been available to systematically measure their effects. In the coming year, such empirical research can be conducted with various types of administrative data collected on Medicare beneficiaries and their use of services.

The important longer range research issues described by the analysts focus on deriving a better understanding of the outcomes of post-acute care. One major dilemma is that, without such information, Medicare cannot readily judge what it is purchasing when it pays for post-acute care. As a consequence, it has been difficult to establish an appropriate price for that "product." In addition, it is important to be able to compare patients treated by different post-acute care providers. This information is important for understanding whether an integrated post-acute care system is needed. Such information is also needed to accurately compare outcomes across providers. A common data set of patient characteristics applied across post-acute care providers (as currently being developed by HCFA) will be an essential first step.

Finally, the researchers and analysts noted that in creating casemix adjustments for payment systems, regardless of information on outcomes, research should focus more on patient or condition-specific needs, in contrast to services used. Toward that goal, the analysts indicated that further research is needed to develop normative standards for what patients need given their medical and functional profiles.

D. Endnotes

While this study was in progress, Congress enacted the 1999 Balanced Budget Refinement Act (BBRA), which effectively increased Medicare payments for post-acute care providers. Some of the recommendations noted above were adopted, while other BBA provisions were temporarily suspended until more research could be conducted on the issues.

Despite the changes created by the 1997 BBA and the 1999 BBRA, many important post-acute care policy issues remain unresolved. One is the goal of the Medicare home health benefit. Ambivalence about whether individuals requiring extended personal assistance with skilled supervision (who also tend to have high per-person costs) should be covered has resulted in the use of reimbursement policy strategies, such as the per-beneficiary limit in the HHA IPS, to contain Medicare costs of those recipients. Reimbursement policies, however, do not directly address the underlying issues which are about eligibility and coverage.

Second, provider and consumer representatives, and the researchers and policy analysts, expressed concern about access problems faced by individuals with particularly high service needs, such as SNF patients with high non-therapy ancillary costs, rehabilitation patients who exceed the Part B therapy caps, and home health patients whose costs exceed the per beneficiary limits. In addition, one of the health policy analysts raised the notion that some beneficiaries might have needs (and costs) that Medicare's system of post-acute care services (and payments) cannot accommodate. It will be important to learn about the prevalence of such individuals and the extent to which they have service access problems.

Finally, it is important to recall that the 1997 BBA provisions have only partially been implemented (and modified by the 1999 BBRA) and, in the near future, Medicare's post-acute care environment will continue to be dynamic. Many issues identified by many respondents refer to BBA policies that are in effect or are being designed. Many other issues, extant prior to the BBA, also warrant further research and policy consideration.

I. INTRODUCTION

Until the implementation of Medicare's acute care hospital prospective payment system (PPS) in 1984, follow-up care after hospital stays, often referred to as post-acute care, accounted for only a small part of Medicare spending. After the hospital PPS was implemented, Medicare spending for post-acute care providers, such as skilled nursing facilities (SNFs), home health agencies (HHAs), rehabilitation facilities, and long-term care hospitals, began to grow rapidly. For example, between 1990 and 1996, SNF and HHA expenditure growth averaged approximately 25 percent per-year. The change in the spending trend for Medicare post-acute care services emerged largely because the fixed price payments under the hospital PPS provided strong incentives for hospitals to discharge patients as quickly as possible to other settings for continuing care.

In addition to the expenditure growth, the rapidly changing environment of post-acute care services led to other policy concerns, such as an increasing likelihood that patients with similar needs were being served by different types of post-acute care providers. Because Medicare payment methodology differed by provider types, the amount of Medicare payments for such similar patients would have varied simply by where they received their care. Although numerous post-acute care issues warranted consideration by policy makers, they, not surprisingly, focused their concern primarily on the rapidly increasing expenditure growth. They acted on that concern by passing multiple provisions in the 1997 Balanced Budget Act (BBA) aimed at curbing the growth.

Prominent among them are mandates for prospective payment systems for SNFs (phased-in beginning in 1998), HHAs (to be implemented in 2000 with an interim payment system effective in 1998), and rehabilitation facilities (to be phased in during 2000-2002). Also included are spending caps on the amount of Medicare Part B rehabilitation therapy payments a beneficiary can receive in a calendar year (implemented in 1998). For rehabilitation facilities and long-term care hospitals, the BBA contained provisions that reduced wide variations in the relative payments between hospitals of each type. Such variations were due partly to differences in how baseline costs were established, which varied, in turn, by when hospitals became certified by the Medicare program. The BBA has an additional provision requiring the Department of Health and Human Services (HHS) to recommend a PPS for long-term care hospitals (a report to Congress on this was required by October 1999).

The BBA provisions that have been implemented to date appear to have been effective in achieving that objective. Some observers have argued, however, that the BBA provisions have also added to the list of ongoing policy issues regarding Medicare post-acute care services that require further attention. Because the BBA provisions had important implications for the payment methodology--and the amount and distribution of Medicare payments--for almost all of the post-acute care providers, widespread concerns about the effects of the provisions surfaced promptly after their passage.

Given the broad public interest in BBA post-acute care issues, the Office of Disability, Aging and Long-Term Care Policy in the Office of the Assistant Secretary for Planning and Evaluation (ASPE) within the Department of Health and Human Services initiated an "early indication" qualitative study of the potential effects of the BBA provisions along with extant issues related to Medicare post-acute care services. This report presents the views of two sets of respondents knowledgeable about such issues.1 The first set consisted of post-acute care service providers (including discharge planners) and their organizations, quality of care experts, and consumer groups. The second set consisted of researchers and health policy analysts. Interviews were conducted with the respondents in 1998 and 1999.2

Methodology and Context. To obtain timely information on such issues, this study collected information through interviews with individuals knowledgeable about Medicare post-acute care issues. They included representatives of post-acute care providers (including discharge planners) and their organizations, quality of care experts, consumer groups, researchers, and health policy analysts.

Although we interviewed a large number of individuals affiliated with many different organizations (Appendix A), the respondents were clearly not representative of all parties interested in or involved with Medicare post-acute care services. Hence, the issues and concerns discussed in this report should be viewed as those of a purposive sample of individuals. This report records most of the issues that were mentioned to us, but we did not select nor attempt to weigh the relative importance of different issues in terms of their importance to providers, consumers, or the Medicare program. For example, a particular concern could have been expressed by many respondents, while another concern was voiced by only one respondent, it was not possible for us to determine the relative importance of the issues for Medicare spending, beneficiary access, or quality of care.

The information gathering methodology used in this study was one way by which we could obtain early indications of the potential effects of the BBA provisions. Although the actual impact of those policies will not be clearly understood until empirical studies can be conducted, findings from our interviews provide an initial set of questions that might be important to investigate with sufficient elapsed time and data availability.

There was considerable variation in the process of arranging for and conducting the interviews, largely because of our objectives to obtain information quickly and minimize respondent burden and costs. For example, while some of the interviews involved meetings with one organization at a time, others involved a gathering of multiple organizations. Some of the meetings were conducted only in person, while others were conducted only through telephone conference calls. At times, we met in person with some representatives of the organizations, while other representatives were "patched in" by telephone. The discussions usually lasted about an hour and a half to two hours.

Prior to each interview, we sent the respondents a list of issues that we thought were important for consideration in ongoing discussions about Medicare post-acute care services (Appendix B). We informed the respondents that the list was made available to help guide discussion, but that the issues in the list were not exclusive to our study aims nor was it intended that they address every issue on the list. We encouraged the respondents to raise other issues they deemed important, and we asked them to focus on the issues that were most important to them or about which they were most knowledgeable.

The interviews were conducted between March and November, 1999. During this data collection period, a number of the BBA mandated provisions, including the SNF PPS, the HHA interim payment system (IPS), and the Part B therapy service caps, were in effect. The HHA and rehabilitation prospective payment systems were still being developed when we conducted our interviews for this study. Many of the concerns expressed by our respondents, therefore, focused on the SNF PPS, the HHA IPS and the therapy caps. Not surprising, concerns regarding rehabilitation facility, long-term care hospital, and HHA prospective payment systems focused on design features.

Because our interviews were conducted after some of the BBA provisions were in effect, some of the concerns expressed by the respondents were already being addressed by HCFA or being examined by other agencies of the government (e.g., GAO, DHHS/OIG). HCFA, for example, was in the process of conducting a study to refine the SNF PPS with a particular focus on adjusting payments for non-therapy ancillary services. HCFA also was in the process of conducting a broad-based study to monitor the impact of many of the BBA provisions. In addition, the data collection period of this study took place prior to the promulgation of the 1999 Balanced Budget Refinement Act, in which Congress responded, in fact, to some of the recommendations voiced in some of this study's interviews.

Finally, it is important to recognize that this study was conducted in the middle of a very dynamic process in which major and wide ranging reform of payment systems for Medicare post-acute care service providers are being implemented. Responses to many of the BBA provisions are still being developed, while refinements or moratoria have been imposed on some recently implemented policies. Findings from this study are most relevant, therefore, to a slice of the dynamic policy reform process that is in progress. The research was intended, however, to synthesize a vast number of issues important to providers, beneficiaries and the Medicare program. Similar studies would be useful to help track old and identify new issues rising from the policy changes.

Focus. The provider and consumer respondents focused almost exclusively on problems they perceived were occurring in post-acute care service provision and access to care in the wake of Medicare post-acute care provisions in the 1997 BBA. The researchers and health policy analysts, on the other hand, stressed that it is too early to gauge the impacts of the BBA provisions. Their particular focus was on the research evidence needed to assess the impacts of the BBA provisions on post-acute care provider services and whether and how those impacts ultimately affect quality of care as reflected in patient outcomes. The perceptions of the provider and consumer group respondents may be particularly valuable in helping researchers formulate the research questions that should be given priority. Individuals in this group also discussed policy-related issues, beyond those directly related to the BBA, that they thought had important long-term implications for Medicare's post-acute care services.

Report organization. The next four sections of the report synthesize the concerns of the providers and consumer groups with respect to skilled nursing facilities (Section II), home health agencies (Section III), rehabilitation and long-term care hospitals (Section IV), plus a discussion of changes in the post-acute care system as a whole and their implications for the post-acute continuum of care (Section V). The last two sections of this report synthesize the responses of the researchers and health policy analysts, divided into perspectives on current issues (Section VI) and research that needs to be undertaken (Section VII).

II. THE BBA AND SKILLED NURSING FACILITIES (SNFs)

Major features of the BBA's PPS for SNFs include (1) national payment rates adjusted for differences in patient needs based on a casemix classification system called RUG-III, (2) a four-year transition during which most SNFs receive payments based on a combination of the casemix-adjusted national rate and, in declining proportions, their own historical costs, and (3) a single SNF payment that includes routine operating, ancillary, and capital costs, which had previously been paid under separate cost-based methodologies. In addition, the new system requires that almost all Medicare services provided to patients during their nursing home stays be billed exclusively by the SNFs themselves (called consolidated billing), in place of the previous system under which other providers (e.g., therapists and ambulances) could bill Medicare directly for the services they provided to nursing facility patients. This consolidated billing provision was intended to facilitate Medicare's ability to account for all the services and costs of nursing facility patients. Because of the demands on HCFA's computer systems to be Y2K-compliant, however, initial implementation was limited to Medicare Part A SNF patients.

Here are the major concerns expressed by our consumer and provider respondents with regard to the BBA provisions.

A. Base Payment Levels for SNFs

Respondent concerns about the SNF PPS amounts focused on both the overall base payment levels and the potential inequities among SNF types in the transition payment period.

Payment Levels. The SNF PPS payments starting in 1998 were derived from baseline SNF costs in 1995 trended forward on the basis of the SNF market basket. Designing a payment system based on staff-time measurements effectively locks in the resources at the levels derived during the design period. Similarly, using spending levels in a base year to trend forward in later years locks in costs from the base year. Respondents noted that these lock-ins are a problem because there have been changes in service delivery since the SNF PPS was designed. These include the use in SNFs of very expensive drugs that were not previously used in this setting, and the heavy resource needs of patients (such as those needing respiratory therapy) who were not specifically accounted for in the RUG-III system. Hospital representatives also remarked that hospital-based patients, who were under-represented in the development of the RUG-III system, may have more complex problems than patients in freestanding SNFs, so their costs might not be adequately accounted for in the RUG-III casemix system.

Industry respondents also re-iterated concern that the market basket index, which was derived from an analysis of hospital input prices, underestimates the actual increase in the input costs of SNFs. They also felt the provision for the years 2000 and 2001, under which the payment levels will only be adjusted by the market basket minus one percentage point--would discriminate against SNFs whose patient loads have become more resource intensive since the 1995 base period.

Transition Period. Respondent concerns also included the potentially disparate fortunes of hospital-based and freestanding SNFs during the payment transition period. The reason for this is that hospital-based SNFs had higher average historical costs than freestanding facilities. Since the post-BBA payment level is based on a blend of hospital-based and freestanding costs, hospital-based facilities face a greater loss than if their payment was based only on their own costs. As the blend moves increasingly away from the historical cost component, this disparity can be expected to increase. Freestanding SNFs, in contrast, benefit increasingly as the national rate replaces average historical costs in the payment formula. The respondents feared that hospital-based SNFs may tend to leave the Medicare program. Despite their relative advantage, freestanding SNFs would also have incentives under the PPS to move toward providing more services to less costly, more traditional long-term care patients.

Policy Options Suggested by Respondents. Options suggested for increasing the payment rate included rebasing to bring in more dollars, dropping the deduction of one percentage point from the market basket index, and using alternative market basket indices (for example, basing the market basket on the Producer Price Index kept by the Bureau of Labor Statistics). For the transition period, it was suggested that SNFs be able to choose whether to continue under the transition method specified in the BBA or to transfer immediately to the full national rate.3

B. Impact on Non-Therapy Ancillaries of the RUG-III Casemix System

Prior to the BBA provisions, ancillary services provided to SNF patients were reimbursed by Medicare on a cost-related basis. Under the SNF PPS, ancillary services are included in the all inclusive, casemix adjusted price. Because the RUG-III classification system, which is the basis of the SNF PPS, sorts patients into payment groups by averaging only nursing resources and physical, occupational, and speech therapy staff time needed for different patient conditions, it does not include any calibration of non-therapy ancillary service and supply needs (e.g., respiratory care, prescription drugs, prosthetic devises) of different patient groups. RUG-III does not, therefore, make specific provisions for individuals with extraordinary high non-therapy ancillary costs.

The concerns expressed by respondents centered on the potential impacts on Medicare SNF patient access and bed supply.

Patient Access. Respondents were concerned that SNFs would become less willing to admit Medicare patients with particularly high non-therapy ancillary resource needs, because the RUG-III classification system distributes non-therapy ancillary payments more evenly among different patient groups than use patterns actually require. Patients identified by respondents as requiring more non-therapy ancillary resources than are accounted for under the current system include those receiving respiratory therapy (not included in RUG-III measurement of therapy hours), hyperbaric oxygen treatments, intravenous medications, orthotics and prosthetics (e.g., wheelchairs of nonstandard size), wound care (e.g., decubitus ulcers), and total parenteral nutrition. Non-therapy ancillary costs are also high for persons with end-stage renal disease, with AIDS, and more generally, with many co-morbidities.

We were told that many SNFs have already chosen to reduce admissions of patients needing respiratory therapy and, even when such patients are admitted, facilities are increasingly using lower-paid staff to administer the therapies. Industry and consumer concerns include, for example, that registered nurses (and aides under their supervision) may be performing respiratory therapy when neither have had specialized training.

Prior to the PPS, when both therapy and other ancillary services were paid on a cost-related basis, SNFs had strong incentives to provide such services. In the process, SNFs that wanted to specialize in providing intensive ancillary services became "subacute care" providers, which reportedly permitted faster hospital discharge and encouraged competition between SNFs and rehabilitation facilities (and possibly long-term care hospitals). The new PPS system has reduced incentives for SNFs to treat resource intensive patients (which may have high-end therapy needs and nursing needs, as well as high non-therapy ancillary needs) and, in the opinion of respondents, may even have increased the demand for admission to the types of post-acute care facilities that continue to be cost-reimbursed (i.e., rehabilitation facilities, long-term care hospitals).

SNF Bed Supply. Because of the new incentives not to take high-cost patients, respondents fear that the number of Medicare SNF beds is shrinking and say that some, particularly hospital-based, SNFs have closed down altogether. In addition, smaller facilities, which tend to be independent operators, may be more likely to leave the Medicare program because they do not have sufficient volume to balance cases requiring large amounts of non-therapy ancillary services with lower cost patients. A similar situation may apply to hospital swing beds when they are eventually subject to the SNF PPS, because hospitals with swing beds, by definition, do not provide SNF care to many patients in a year.

When asked why hospitals would forgo the additional SNF payments for Medicare patients during an episode of care, we were told that many hospitals were simply more comfortable absorbing such costs in the inpatient setting. We were also told that, where possible, hospitals were providing incentives for freestanding SNFs to accept patients needing specialized care, by such measures as contributing to these SNFs the services of respiratory therapists.

Policy Options Suggested by Respondents. HCFA has recognized the problem of SNF PPS not directly accounting for non-therapy ancillary costs in its methodology and is sponsoring additional research on the possibility of integrating an index for those costs within the existing system or expanding the number of casemix categories. However, providers feel it is urgent to address the problem before these studies can be concluded. Two alternatives were mentioned to address payment for patient groups with resource needs inadequately paid for under RUG-III. The first is an outlier policy, modeled on Medicare's PPS for hospitals, under which a portion of any patient costs exceeding specified levels would be reimbursed as outliers. The second alternative suggested is to modify RUG-III to include additional payments for special conditions (e.g., diabetes) or special service needs (e.g., at least 30 minutes of respiratory therapy a day).4

C. Effects on Therapy Services

The initial classification of Medicare SNF patients into RUG-III casemix categories is based on an assessment by the 5th day after admission. To be able to classify patients needing the highest levels of care in the rehabilitation categories (which have the highest payment levels), it is generally necessary, according to respondents, for patients to receive extensive amounts of therapy on each of the first five days of care. Providing therapy in this manner creates potential quality of care problems. As discussed below, respondent concerns centered on possibly perverse admission behavior due to the assessment "windows," potential under-treatment due to the payment rates, and the likelihood that SNFs would reduce therapy staff.

Therapy Window. One of the main respondent concerns relates to the requirement that initial assessments be completed by the 5th day of admission. In order to classify patients needing the highest levels of care in the rehabilitation categories with the highest payment levels, extensive therapy must be provided on each of the first five days of care. The problem perceived here is that SNFs are reportedly striving to admit such patients only on Mondays (because therapy is typically not given during weekends). This is reportedly resulting in acute hospital backups and increased lengths of hospital stay. Three grace days are allowed to mitigate this danger (thus, extending the initial assessment period to the 8th day of admission), but respondents say that SNFs are reluctant to use these days because fiscal intermediaries (FIs) are reportedly using them as a red flag to trigger greater monitoring or an audit. The second assessment has a similar problem if the assessment period spans a weekend. Patient care concerns include both too little therapy for patients who could not be classified correctly because of admission date and patients who receive therapy too early (i.e., before their medical conditions are stable) in order to ensure they get into the high-end classification.

Incentives to Administer Too Little Therapy. A general concern was that any system that caps payments contains incentives to reduce service levels or restrict access for higher cost patients. The particular concern was that during the transition period, any SNF whose historical costs were above the national rate, which is casemix adjusted, will have an incentive to under-serve patients whose resource needs are expected to be above the facility's historic cost base. The payment only becomes fully casemix adjusted in the final year of implementation. Even when the payment rate is fully casemix adjusted, however, incentives will remain to reduce costs by limiting services or denying access for certain patients. SNFs are reportedly either actively avoiding particular cases on which they could lose money or focusing on breaking even across their total patient caseload.

Reductions in Therapy Staff and Payments. Because of the array of cost controlling features affecting therapy provision, respondents say that SNFs have been reducing amount and types of therapy services and shifting the function, as noted, to lower paid staff. We were told, for example, that some SNFs are having speech and language pathologists train occupational therapists (again lower paid) to perform barium swallow examinations. Anecdotal evidence also suggests that SNFs have been releasing staff therapists and negotiating much reduced payments for specialized therapy services when provided under contract by rehabilitation agencies. Agencies then pass these payment reductions on to therapists in the forms of job loss, conversion from salary to wage payments, and benefit loss. Providers indicated that such changes could be harbingers of poor quality of care.

Policy Options Suggested by Respondents. The same options as before were suggested (1) to better calibrate the RUG-III payment rates to therapy and nursing resource needs and (2) provide SNFs the option of immediately moving to the national rate. Clarifying the use of grace days was also suggested.5

D. Impact of SNF PPS and Consolidated Billing on SNF Services

The BBA required that, with few exceptions, payments for all Medicare costs incurred for nursing home patients were the responsibility of the SNF, regardless of whether the service was provided by the SNF itself or other providers. Prior to the BBA, it was not unusual for independent rehabilitation agencies and durable medical equipment (DME) suppliers, for example, to provide services to SNF patients and bill them directly to Medicare Part B. The consolidated billing provision, as implemented, requires SNFs to be responsible for all costs incurred for patients during Part A-covered stays. In addition to assuming this new administrative function, because of PPS, SNFs are now responsible for all "care plan services" for Medicare covered patients except those for which SNFs are not considered appropriate settings (such as MRIs, CT scans, cardiac catheterizations, ambulatory surgery with the use of an operating room, and emergency room care).

Respondent concerns focused on the implications of transferring these costs from Part B to Part A and uncertainty about which services are to be included in consolidated billing.

Transferring Costs from Part B to Part A. Some services and supplies, such as transportation by ambulance, specific orthotic and prosthetic devices, were formerly provided and billed to Medicare directly under Part B by outside organizations. During Part A covered stays, the payments for such services now are to be made by SNFs, because the estimated costs of such services were included in the derivation of the PPS payment rates. Hence, for Part A-covered patients, SNFs now must directly provide or negotiate with outside organizations to provide all such services. In addition, because the PPS rate is an all inclusive rate paid on behalf of Part A-covered patients, outside vendors can no longer collect copayments (allowed under Part B, for therapy services, for example) for individual supplies and services during Part A stays.

Uncertainty about What is Included. Respondents view the care plan requirement as too inclusive. Discussions between provider organizations and HCFA resulted in clarification of services covered under the SNF PPS.

Options Suggested by Respondents. With respect to the uncertainty about which services were subject to consolidated billing, the provider organizations felt strongly that additional procedures be added to the list of procedures excluded from consolidated billing on the grounds that they are more appropriately provided in a hospital outpatient setting (chemotherapy, barium swallow, stress tests, hyperbaric oxygen treatments, Doppler studies, nuclear medicine, and gastrointestinal procedures).6

E. Impact on Therapy Services of Part B Cost Caps

The BBA established two $1,500 cost caps on Part B spending for therapy services per-beneficiary per-calendar year. One is for occupational therapy, the other for physical therapy and speech and language pathology services. The only setting excluded from these cost caps is hospital outpatient departments. Prior to the BBA requirements, Medicare only capped (at $900) services provided by physical therapists in independent practice.

Respondent concerns focused on perceived adverse effects on certain patients, differentially harsh effects on nursing facility patients, and the risk that states now choosing to cover therapy under Medicaid may stop doing so.

Adverse Effects on Particular Patient Groups. Respondents singled out chronically ill and medically complex patients at risk of one or more acute events as being adversely affected by the caps. These include patients with Parkinson's disease, aphasia, stroke, or hip fractures. Therapy costs for Parkinson's patients, for example, reportedly reach the caps in four weeks, even though the condition reportedly warrants continuing therapy services. Stroke or hip fracture cases reach the caps in about two months. One consequence for those affected may be that they remain in (or are discharged from) SNFs without receiving needed therapy services, leaving them at increased risk of being readmitted to hospitals. We were also told that some patients forgo prescribed therapy in order to save money under the cap, for services possibly needed later in the year. Others reportedly have to choose "between walking and talking," since the two types of therapy come under a single cap.

Unique Penalties for Nursing Facility Patients. Although the cost caps were intended to be applied on a per-beneficiary basis, they were actually applied (because of Y2K constraints) only to the costs of individual providers. Thus, most beneficiaries could be treated by multiple providers with payment to each individual provider limited to the capped amount. However, this was not the case for nursing home patients. Under the therapy cap provisions, SNFs are required to submit all claims for therapy services provided to their patients to Medicare Part B. Outside therapists (e.g., contracted or independent therapists) can no longer bill Medicare for these services. Because SNFs are required to be the "billing agent" it was possible to track the amount of therapy expenditures made on behalf of their patients. As a result, the institutionalized population was the only group for whom the caps were applied on a per-beneficiary per-year basis. While nursing homes could send their patients to hospital outpatient departments for unlimited therapy services, they had disincentives to incur associated transportation costs.

Possible Shrinking of Medicaid Help. In certain states, but by no means all, Medicaid picks up some of the costs of additional therapy services to nursing home patients after Medicare's coverage ends. Respondents expressed concern that the Medicare caps would raise Medicaid therapy costs and cause some states that now provide such reimbursement to drop it.

Options Suggested by Respondents. One suggestion was an outlier policy to deal with those patients whose conditions dictated therapy over the Part B cost caps. It was also suggested that the amount of allowable additional services should be decided by therapists, not by FIs whose nurses conduct medical review to certify services. A third option was to apply the cost caps to an episode of illness rather than to a calendar year. This option would at least eliminate the need for beneficiaries to "save" therapy resources in case of greater need later in the year. A fourth suggestion was to increase the capped amount or eliminate the caps for nursing home residents.7

F. Additional SNF Issues

In addition to the five most heavily emphasized issues, respondents raised two other issues they perceived to be important. These relate to FI service coverage definitions and administrative burden.

SNF Service Coverage as Viewed by FIs. Casemix systems are intended to reduce subjectiveness in coverage determinations. Some respondents indicated, however, that FIs continue to use old guidelines to make coverage determinations. One result is that even patients classified into one of the top 26 (out of 44) RUG-III resource categories, who would otherwise receive "deemed" coverage, may still have to meet "medically necessary" criteria for coverage. The example given was a pneumonia patient with therapy needs for a collateral condition. Such a patient may be classified in a rehabilitation category, but may not meet coverage criteria because rehabilitation for pneumonia patients may not be considered medically necessary. Hence, while that patient qualified according to RUG-III, FIs might not certify the patient because she did not meet the medical necessity criteria.

Respondents also felt that FIs were not sufficiently forthcoming about how their decisions were made, and that more explicit coverage guidelines from HCFA were needed. In addition, respondents noted that the 1998 elimination of the need for observation, assessment, and education from the list of qualifying skilled services rendered some patients who would previously have qualified for Medicare-covered SNF care no longer able to qualify.8

Administrative Burden. Respondents felt that the BBA requirements have added to the administrative burden of SNFs in several ways. First, the requirement that MDS-based assessments, to classify patients in the RUG-III system, be completed more frequently for each patient has forced some SNFs to hire additional administrative personnel as well as use more nursing staff time for this purpose.9 The high fixed costs associated with MDS assessments are particularly onerous for SNFs with small Medicare volume. Second, consolidated billing requires SNFs to keep extra records in order to bill for services that used to be billed directly by outside vendors. Third, the fixed PPS payment gives SNFs a strong incentive to spend time negotiating price reductions with outside vendors of specialized services.

Options Suggested by Respondents. One suggestion was that HCFA restore the provisions that qualified beneficiaries in need of overall management and evaluation of the care plan, observation and assessment of the patient's changing condition, and patient education for SNF care. As noted, HCFA restored these qualifying services in the final SNF PPS rule. Another was that FIs be retrained. A third was that an add-on be provided to account for the new costs associated with completing the additional assessments now required. A final suggestion was that payments be adjusted to respond to changes in technology and service delivery.

III. THE BBA AND HOME HEALTH AGENCIES

Major BBA provisions for HHAs included reducing payment levels, moving some home health care coverage out of Medicare Part A into the Part B Supplemental Insurance Program, and implementing a casemix-adjusted PPS by 1999 (later delayed to 2000). The BBA also instituted an interim payment system (IPS) for HHAs, effective in 1998, to operate until the HHA PPS is implemented. Concurrent with the implementation of the new payment policies, HHAs were required to report periodic information on the health and functional status of their patients under the OASIS system.

Before the BBA, HHAs were reimbursed their costs up to a limit of 112 percent of the national mean cost per-visit for freestanding providers, with an add-on for rural agencies. Agency payment at the end of the year equaled actual costs or the agency's aggregate cost limit, whichever was lower. The IPS reduces the national cost limits to 105 percent of the median cost per-visit and limits agency payment at the end of the year to the lower of actual costs, the agency's aggregate cost limit, or a new average per-patient expenditure limit. This new limit--a new reimbursement concept for home health agencies--is 98 percent of a 75/25 percent blend of an agency's and its region's 1994 average per-patient expenditure. In addition, agencies are required to return to the federal government any overpayments for services delivered between October 1, 1997 and August 1998 (when the final IPS regulations were issued).

Providers had a smaller set of concerns about the BBA home health provisions than they did about those for SNFs. The concerns they did express focused on the payment level and its impacts on patient access and quality of care, and provider equity.

Payment Level. Respondent concerns here were centered on their perception that the limits on allowable costs per-beneficiary have "taken much more money out of the system" than Congress intended, and that payments are less than agencies need to care for many types of patients with long episodes, such as wound care patients, heart patients, and diabetics. The perceived problems with the reduced payments were compounded by the additional administrative costs incurred to meet the OASIS reporting requirements.

Patient Access. According to respondents, some HHAs are closing because they often have no capital and no reserves to absorb higher cost patients. And some may even be driven to bankruptcy, they argue, by the payback requirement. As HHAs close, the concern is that beneficiaries will have more trouble accessing home health care. In addition, beneficiaries that are accepted by agencies might not receive the amount of services (e.g., therapy visits) that they need. Respondents further suggested that this reduced access may lead to increased Medicaid nursing home use.

Uneven Provider Impacts. Respondents told us that the HHAs most likely to close because of BBA changes were freestanding, for-profit agencies in dense markets outside the Western region of the country. Western agencies will have some protection because managed care is strong there, and managed care capitation has already led to relatively lower per-capita spending levels. Hospital-based HHAs may have some long-term protection relative to freestanding facilities, because their link to the hospital setting may give them the opportunity to select patients first, referring high-cost patients to freestanding agencies. Agencies in areas with less competition will be protected to some extent by their market power.

Options Suggested by Respondents. Eliminating the retroactive payment limits on per-beneficiary costs was suggested as an obvious way to improve the short-term health of the most vulnerable agencies. With respect to the future, some respondents were looking forward to a casemix-adjusted PPS as a viable option for HHAs, with the proviso that the 15 percent discount on existing payment levels currently planned for the PPS be eliminated. This, they felt, would bring PPS payments closer to actual costs.10 The 15 percent discount, notwithstanding, provider representatives were clearly in favor of moving from the IPS to a PPS.

IV. REHABILITATION AND LONG-TERM CARE HOSPITALS

Rehabilitation and long-term care hospitals have historically been paid on a cost-related basis, subject to limits per-discharge (that is, per-case). The Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) set inpatient, facility-specific limits (TEFRA limits) equal to the product of a facility's base year costs per-discharge, updated to the current year, and its total discharges. Facilities with operating costs below their payment limit ceiling received costs plus a "bonus payment" as an incentive to reduce costs. The incentive payment was equal to the lower of 50 percent of the difference between costs and the target, or 5 percent of the target. Facilities with costs above the ceiling were paid the ceiling plus a "relief payment" equal to half the difference between their excess costs and the ceiling (up to 110 percent of the ceiling). Facilities could also request payment exceptions.

Establishing TEFRA limits using each facility's costs in a base year created inequities between newer and older facilities. Newer facilities, relative to older providers, have more recent base years in which they were able to inflate base year costs, establish higher TEFRA limits, and in subsequent years reduce their costs while retaining the higher limit. The BBA attempted to address the inequities between new and old providers and reduce the incentives to reduce costs in years following the provider's base year. First, the BBA changed incentives for new rehabilitation facility and long-term care hospital providers by eliminating the exemption from cost limits and limiting these providers' TEFRA limits to the lesser of costs or 110 percent of the national median target amount for that class of providers. Second, it allowed providers whose base years are before 1990 to rebase using their average costs from their three recent cost reporting periods. Third, the BBA limits facility target amounts to the 75th percentile of the 1996 target amounts, adjusted for inflation. In addition, the BBA specifies a formula for increasing target amounts based on each facility's costs relative to its target amount (the more a facility exceeds its limits the greater the increase). The BBA also reduced the amount of bonus and relief payments.

The BBA also required the Secretary of Health and Human Services to begin October 1, 2000 a three year phase-in of a casemix and wage adjusted Medicare PPS for rehabilitation hospitals and units.11 The BBA provided the Secretary significant discretion in designing this PPS including flexibility in: (1) specifying the groups of patients upon which the classification system would be based; (2) assigning weights to each group that reflects the relative facility resources used by the group; (3) establishing outlier policies if needed; (4) determining the unit of payment (e.g., per-day or per-episode); and (5) determining a prospective payment rate for each group.

For long-term care hospitals, the BBA required the Secretary to develop a legislative proposal for establishing a casemix adjusted PPS for long-term care hospitals. This proposal was to be submitted to Congress by October 1, 1999.

A. Respondent Concerns about the Rehabilitation Hospital Provisions

With respect to the current environment, respondents concerns focused on the reduced incentive/relief payments (already implemented). With respect to prospective payment to be implemented starting in 2000, the main concerns addressed the unit of payment, design of the classification system, and the assessment instrument that will be the prospective payment system's foundation.

Incentive/Relief Payment Reductions. Respondents said that all facilities are feeling the financial pains of these reductions. Respondents expressed concern that reducing incentive payments could result in some facilities increasing their costs while not exceeding their TEFRA limits. According to respondents, if all facilities raised their costs to their respective ceilings (e.g., by raising lengths of stay), 1999 Medicare costs from increased relief payments could exceed the amounts estimated to be saved through these payment reductions. Because of the financial hardship caused by these payment reductions, respondents oppose any delay in implementing PPS for rehabilitation hospitals.

Unit of Payment. Respondents universally favored a per-case rather than a per-day PPS. The fundamental point they made was that a per-diem system would provide an incentive for facilities to increase lengths of stay, at the same time it limited the resources they could use on a given day, a combination that tends to defeat the goal of rehabilitation (to restore function and discharge home). In contrast, money made available by a per-case unit of payment would provide facilities greater financial and clinical flexibility to provide needed care during a stay and improve rehabilitation outcomes.

Classification System. Respondents were fearful of applying an extension of the RUG-III, per-diem-based classification system to rehabilitation populations. They were concerned that the resource use of patients would be inadequately measured in the study being conducted by HCFA because they considered the study's sample to be too small. They also expressed concern that a RUGs-like, per-diem system reduces the incentive to provide intensive therapy when appropriate. Most respondents preferred a PPS based on the Functional Independence Measure/Function Related Group (FIM/FRG) classification system previously developed and tested in rehabilitation facilities.

Assessment Instrument. Respondents were equally concerned about use of the MDS-PAC, which they feel is, at its core, an instrument designed to measure quality of care in nursing homes, whereas hospitals, including rehabilitation facilities, have other mechanisms (e.g., existing JCAHO and CARF standards) for measuring quality. In their opinion, the MDS-PAC has an administrative burden that far outweighs its clinical utility. In addition, use of MDS-PAC could eliminate the value of historically collected FIM data and make future analysis of trends in performance (e.g., length of stay, outcomes) in rehabilitation facilities difficult.

Options Suggested by Respondents. Some of the major options suggested by respondents are now being addressed by HCFA. As already noted, respondents were strongly in favor of an episode-based system. HCFA has accepted this recommendation, having announced in July 1999 that the rehabilitation PPS would be an episode-based, FRG payment system.12 The other major concern of respondents was quality measurement. Respondents suggested that review of quality of care be based on clinical outcome measures (e.g., change in functional status between admission and discharge) and possibly use a telephone follow-up component (as required by CARF). It was also suggested that any patient assessment instrument be restricted to the relatively few elements actually necessary to assess quality and make appropriate payment. HCFA is addressing ways to refine the MDS instrument to better explain the characteristics and needs of rehabilitation facility patients; the refined instrument has been named MDS-PAC (minimum data set, post-acute care). It is also addressing measurement issues basic to monitoring quality of care in rehabilitation facilities.

B. Respondents Concerns about the Long-Term Care Hospitals Provisions

Long-term care hospitals are still operating under a cost-based system with the same payment limits as described above for rehabilitation hospitals. They are also subject to the same reductions in incentive/relief payments now in effect for rehabilitation facilities. Respondents raised the same concerns regarding the incentive payments as for rehabilitation hospitals, which will not be repeated here.13

With respect to a PPS for long-term care hospitals, respondents concerns were less immediate, since long-term care hospitals will continue to be paid on a cost basis well beyond the time when other post-acute care settings have switched to prospective payment and work on the design particulars of such a system is still in the formative stages. Respondent concerns focused on their preference for an episode-based payment system for long-term care hospital providers based on the DRG system used in short stay acute hospital payments. Similar to issues regarding rehabilitation facilities, they expressed concern about the adequacy and applicability of MDS-PAC for long-term care hospitals. One respondent told us that the medication section of the MDS, for example, addressed less than 10 percent of the medications typically used in the long-term care hospital setting. Additional problems included an inadequate procedures and treatment sections and no reflection of physician interventions.14

V. CHANGES IN THE POST-ACUTE CARE SYSTEM AS A WHOLE AND THEIR IMPLICATIONS FOR THE CONTINUUM OF CARE

With respect to the post-acute care system as a whole, respondents expressed concerns that the BBA provisions appeared to be changing the relationships among providers and the relative demand for and supply of services of different types of providers--creating, as a consequence, disruptions in the continuum of care.

Prior to the 1997 BBA, the post-acute care system, in the opinion of some, had achieved a continuum of care in which patients were shifted from one setting to another according to what was most appropriate for their conditions. BBA payment changes were perceived by at least some of our respondents as changing discharge patterns in a way that made this shifting more a matter of reimbursement incentives and less a matter of what was most appropriate for the patient. Whereas the previous trend in health care had been to shift more and more complex patients out of acute care hospitals "downstream" into post-acute settings, they felt the BBA payment changes have arrested and are possibly reversing this trend. While many respondents suggested this was a negative outcome, others thought that it may be more medically appropriate to serve certain complex patients in medically intensive settings (i.e., acute care hospitals rather than post-acute care placements).

It must also be noted, various observers thought that the previous system, prior to the BBA, tended to blur the distinction between services provided by the different providers and the patients they serve. Because of the already-noted growth in SNFs under the old system, the concern used to be that their patients were overlapping increasingly with the patients served in rehabilitation facilities and long-term care hospitals. As the new system is evolving, the concern about blurring boundaries is being directed more at potential overlaps between the patient groups served by rehabilitation facilities versus long-term care hospitals. The actual extent of overlap--whether appropriate or otherwise--is not known.

The Home Health Perspective. Industry representatives argued that the payment changes were effectively removing the option of home health care for some patients. High-cost cases will have difficulty getting access to home care because the IPS contains financial disincentives for agencies to take on high cost cases. The ultimate concern is that access to home health care will be reduced because beneficiary expectations will be lowered. One consequence is the increased likelihood that high-need patients will remain in institutional settings.

Nursing Homes. Some respondents felt that the new payment system for SNFs is having the effect of eliminating "subacute" care providers, as nursing homes are declining to admit intensive rehabilitation and medically complex patients. This problem is aggravated because hospitals are no longer finding it financially advantageous to maintain hospital-based SNFs and HHAs, to which they used to discharge patients after relatively short hospital stays. Such patients, according to this viewpoint, are backing up in hospitals due to lack of placement opportunities in subacute settings. In some cases, hospitals may be sending patients home without any nursing home or home health care, which is reportedly linked to increasing emergency room use and unplanned hospital readmissions. Respondents felt that this may be especially true in rural areas.

Rehabilitation Hospitals. In contrast to the home and nursing home settings, admissions directly from hospitals to rehabilitation units/facilities are increasing, according to reports. Rehabilitation facilities reportedly now have waiting lists and little competition from SNFs and HHAs for the more complex patients. But they are having increasing difficulty finding home health services following discharge from rehabilitation facilities.

At the same time that facility demand is increasing, the new system is making it more difficult for certain non-qualifying patient groups to gain admission. This is because Medicare conditions of participation for rehabilitation units/hospitals require that 75 percent of patients admitted be classified in one or more of 10 qualifying conditions. Non-qualifying conditions include cancers, pulmonary diseases, and cardiac diseases or events. Providers report that the percentages of patients with non-qualifying conditions in their caseloads is increasing and, in some cases, approaching the 25 percent limit. Some of this increase, they say, is due to technological advances in available therapies. But some of it is also attributed by respondents to access difficulties elsewhere on the care continuum. Respondents urged a review and updating of qualifying conditions.

Long-Term Care Hospitals. Long-term care hospitals are also reportedly seeing increased admissions directly from acute care facilities and longer lengths of stay. We were also told that more long-term care hospital admissions are coming directly from managed care organizations. Respondents further suggested that, as a result of the BBA, rehabilitation and long-term care hospitals are increasingly serving overlapping patient groups. As with the rehabilitation facilities, long-term care hospitals have waiting lists of medically complex patients that SNFs and HHAs no longer serve. Long-term care hospitals are also challenged by the difficulties they report experiencing when it is time to discharge patients.

The overall concern expressed was that the current system is marked by lack of understanding about what constitutes appropriate care and the settings capable of providing it. This observation led directly to expressed concerns about quality of care generally in the post-acute care system.

Quality of Care. The big issue raised with respect to quality of care was the perceived exodus of and reductions in competent specialized staff (such as clinical nurse specialists and therapists) from SNFs and HHAs. This was reported to impact the amount of time available for interdisciplinary team meetings and care planning activities. Further, respondents reported that therapists were being increasingly used on an as needed basis, rather than being generally available in facilities, and that institutions were relying increasingly on nurse aides to conduct rehabilitation activities. The concern was, of course, that this would reduce the quality of care available to patients. Concern was also expressed that this reduction in the availability of therapists would also affect the care provided to long-term nursing home residents.

Interestingly, this view of quality of care trends was not universally shared by our respondents. Some noted that it was too soon to assess whether or not these changes are impacting the quality of care delivered to Medicare beneficiaries in post-acute care settings. Others actually felt that the trend toward retaining some patients in acute care hospitals or placing some patients in rehabilitation hospitals rather than "subacute" SNFs could improve quality of care (and rehabilitation outcomes) for those patients.

Options Suggested by Respondents. The most drastic suggestion was that a unified payments system across post-acute care providers be created, based on patient needs. A particular form suggested was that such a system be financed by a life-time cap based on the costs of serving persons with disabilities (not the general population) and that it include: (a) an effective care management system, (b) flexibility to manage services across the continuum and episodes of care, and (c) quality standards. It was also remarked that such a unified payment system could be successful only if it is based on an understanding of outcomes.

For a system that continues to vary payment by provider type, the following more modest suggestions (some of which have already been noted in the facility-specific sections) were made:

  • Enhancing payment for non-therapy ancillaries in SNFs
  • Refining the RUG-III classification system to better account for resource use and changes over time.
  • Increasing the percentage of patients with non-qualifying conditions permitted in rehabilitation facilities.
  • Increasing understanding of the optimal outcomes of care that could be achieved across post-acute care settings.

VI. RESEARCHERS AND POLICY ANALYSTS PERSPECTIVE

The health care researchers and policy analysts with whom we talked were familiar generally with the Medicare post-acute care policy issues raised by the provider and consumer groups. In our discussions, the analysts shared their perspective on some of those issues, and offered insights on more general issues related to access, quality, and costs of Medicare post-acute care services.

A. SNF-Related Issues

The most important SNF-related issue identified by many of the analysts was a need to make refinements to the casemix payment system for SNFs. Consistent with the providers and consumer representatives' view that the system did not adequately account for the costs of some patients, the analysts saw several areas in which the casemix system could be adjusted to better reflect the range of patients' needs and costs.

Non-therapy ancillary costs of some patients were identified as an area where refinements in the SNF casemix payment system was needed. Because the classification of patients in the RUG-III system is based on their relative nursing and rehabilitation therapy resource requirements, the classification system does not necessarily reflect the relative non-therapy ancillary costs (e.g., for drugs, prosthetic devices) of SNF patients. It was pointed out that HCFA also recognized this need and was in the process of developing some refinements.

Some analysts thought that, beyond having to address the high non-therapy ancillary costs of some patients, the SNF casemix system needed to be re-tuned to better recognize the high skilled nursing service costs of Medicare SNF patients, while reducing some of the incentives to provide rehabilitation therapy services. In effect, this opinion suggests that adjustments in the SNF casemix system be considered that would redistribute Medicare SNF payments toward skilled nursing costs and away from rehabilitation therapy costs. It was noted that this situation arose, in part, because the development of the RUG-III system was based on a sample of nursing homes that was oriented toward traditional, long-stay patients, and that the sample of Medicare patients, particularly those with intensive skilled nursing and other clinical needs, was under-represented.

Because the rehabilitation therapy categories in the RUG-III system generally have higher payments than the medical and clinical categories, incentives are strong for SNFs to classify patients into one of the rehabilitation therapy categories. The analysts noted the possibility of an increase in "non-traditional" therapy patients (e.g., those with congestive heart failure, chronic obstructive pulmonary disease) classified as therapy patients as an example of the casemix system's incentives. This was noted as a potential problem given how the RUG-III system classifies some patients; notably a patient could be grouped by therapy categories according to use or receipt of services, rather than by need according to some clinical standard. Moreover, whereas the SNF PPS controls the price per-day for each RUG-III category, aggregate payments associated with increased therapy use may still increase as proportionally more patients are classified into one of the RUG-III rehabilitation groups.

The analysts also noted that the SNF PPS is likely to affect facilities differentially depending on a host of factors, including historical costs, payment mix of patients, and size. Consequently, access to SNF services by subgroups of patients, including those patients with high non-therapy costs, will also vary. In general, while the analysts expect potential access problems to SNF care for some patient groups (e.g., medically complex), the extent of the problem is unclear. They thought that, in general, SNF patients requiring traditional rehabilitation therapy would have minimal access problems during Medicare Part A covered stays, but patients, such as those requiring respiratory therapy (not accounted for by the RUG-III system) or those needing extensive non-therapy ancillary services, including those with a combination of intensive medical and therapy needs, might encounter difficulties.

Researchers were not convinced that increased hospital lengths of stay were evidence of access problems. Access to Part B therapy services was viewed as less of a problem, given that only about 13 percent of Part B therapy patients in the total population has been estimated to have costs exceeding the $1,500 caps. On the other hand, estimates of beneficiaries residing in nursing facilities suggest that 23 to 33 percent may exceed the cost caps.

In discussing ways to refine the SNF PPS, the analysts suggested that a combination of options may be appropriate, after it is determined how many people are "underpriced" by the payment system. Adjustments should be made without creating perverse incentives, and particular types of services requiring payment adjustments could be addressed more or less readily depending on the resulting potential for "gaming" the payment system. The researchers thought that, in addition to refining RUG-III, payments for outliers could be warranted in some cases. For example, paying for prosthetic devices on a pass through basis seemed reasonable because the likelihood of gaming this option is remote. On the other hand, paying for prescription medicines on a cost-related basis may require more consideration because of incentives to use higher cost prescriptions drugs when equally effective, lower cost, alternatives exist.

B. HHA-Related Issues

The analysts expressed concern for medically needy or rehabilitation intense home health patients who require either extensive durations of, or high intensity, services. Both situations would equate to high numbers of home health visits and relatively high per-person costs. It seemed plausible that agencies' costs would exceed their per-beneficiary payment limit, and that such patients could have, therefore, higher risks of hospitalizations or emergency room use.

Although the casemix-adjusted HHA prospective payment system will be implemented soon, it is not clear how well that system will account for the medically needy home health patients. Also of importance, in light of the concern about long duration patients, is how payment episodes will be defined (e.g., number of days) and how sequential episodes will be paid. The analysts generally felt that, regardless of the details of the HHA PPS, it is likely that more research will be required to refine the way that Medicare pays for care provided by HHAs. More research may be required, for example, to refine the casemix classification system for HHAs. The analysts expressed the opinion that future research on casemix classification systems should focus on the needs of patients, rather than on the services they received.

Beyond the importance of refining the way that Medicare home health is reimbursed, several analysts noted the fundamental need for a policy clarification (and consensus) of the goals of the benefit. Medicare's home health benefit appears to cover two general populations, individuals with chronic conditions requiring skilled nursing supervision, and those needing restorative care for acute medical conditions. Whether Medicare's home health benefit should cover both groups should be explicitly addressed. Establishing normative standards for care for eligible groups would then be a next step that would facilitate refinements in how home health services are reimbursed. To the extent that interim solutions may be required to manage the home health benefit (and other Medicare services) for chronic care conditions, establishing a "case management" function was favored by some of the analysts over the use of "macro payment policies."

C. Rehabilitation Facility and Long-Term Care Hospital-Related Issues

The BBA provisions affecting rehabilitation facilities were regarded as important improvements in the way that such Medicare providers are paid. The TEFRA payment methodology was regarded as "broken," because it unduly favored new hospitals relative to older ones. Some of the BBA provisions, in effect, reduced the effect of facility-specific baseline costs on payment amounts, and established rates that were more reflective of the whole group of facilities. Under TEFRA, rehabilitation facilities and long-term care hospitals had incentives to establish very high baseline costs to receive subsequently very high payment levels. This policy was particularly susceptible to "gaming" by start-up providers, in contrast to older facilities with baseline levels that had been established by their historical costs. In addition, under TEFRA, the payment methodology did not contain a means to adjust for cost differences of patients (i.e., casemix).

Although the PPS for rehabilitation facilities is still under development, the analysts agreed generally that an episode-based payment unit, rather than a per-diem one, better suits the nature of the product of rehabilitation facilities--a bundle of intensive, time-limited services leading to functional improvement. Moreover, measurement tools applicable to rehabilitation facilities already exist to quantify outcomes of patient care in those facilities (e.g., changes in functional status between admission and discharge).

D. Post-Acute Care System-Related Issues

Discussion with the analysts about the post-acute care system focused on two issues. The first was the extent to which the payment methodologies for the various post-acute care providers could realistically be integrated. The second was the factors that determined utilization patterns of post-acute care providers.

The importance of payment integration depends largely on the extent to which there is overlap of similar patients served by the post-acute care providers. Some thought that, beyond having similar DRGs in the prior hospital stay, there is considerable variation in the needs of patients treated by different providers. Although it might be possible to develop an integrated payment system for some patients (e.g., those with hip fractures), the heterogeneity of the post-acute care population precludes the generalization of that option. In groups of individuals with the same prior hospital DRG, for example, very sick people do not go to rehabilitation facilities and those without adequate informal care in the community cannot be sent home with home health care. In addition, level of functional disability and the strength of informal caregiver networks are factors that differentiate the type of post-acute provider that is used by persons with similar hospital DRGs.

The pattern of post-acute care provider use is determined, in large part, by the supply of particular provider types in a given area. Because of the wide geographic variation in supply of provider types, utilization patterns, even for patients with similar needs, also vary widely by geography. The relatively small number of rehabilitation facilities and long-term care hospitals particularly highlight the potential geographic variability in patterns of use of post-acute care providers.

The supply of provider types is determined, in part, by incentives provided by the Medicare payment systems. Analysts noted that SNFs are increasingly being certified as long-term care hospitals, since the latter are still paid on a facility-specific, cost-related basis. Another observation pertained to increasing rates of transfer of rehabilitation facility patients to SNFs, which is likely to occur when patients require extended care. It is also more common when patients receive services from vertically integrated health care systems. The analysts agreed on the need to revisit the revenue maximizing opportunities in the fragmented Medicare post-acute care system that were not addressed by recent legislation.

VII. FUTURE RESEARCH

In the discussion about important areas for future research on Medicare post-acute care services, the research and policy analysts identified both short and long-term issues. The former reflected concerns about the impact of BBA provisions on the provision of services to subgroups of the post-acute care population, while the latter focused on deriving better ways to measure the benefits of Medicare-funded post-acute care services.

As noted in prior sections, the SNF PPS does not appear to account adequately for some subgroups of (high cost) patients. Most readily identifiable are individuals with high non-therapy ancillary costs. The analysts expressed concern about the circumstances of such "underpriced" patients. Are they, for example, experiencing problems with access, and is there a problem with the quality of care provided to those patients who are admitted? Beyond the concern about the patients, the analysts questioned the potential cost consequences for the Medicare program if hospital readmission or emergency room use increased as a result of SNF access and quality of care problems. Similar issues were raised about HHA patients with extensive care needs (e.g., exceeding the per-beneficiary cost limits of the IPS) and nursing home patients whose costs exceeded the Medicare Part B cost caps.

The analysts thought it important to identify the number of people who were affected by the various provisions of the BBA and to estimate the individual and program impacts of those provisions. Because such BBA provisions were only recently implemented, data had not been available to systematically measure their effects. In the coming year, such empirical research can be conducted with various types of administrative data collected on Medicare beneficiaries and their use of services. For example, the analysts recommended studying the long-term medically complex or rehabilitation intense patients, given the disincentives for several types of post-acute care providers (e.g., SNFs, HHAs) to serve them. They also suggested patient backlog in hospitals should be evaluated.

The important longer range research issues described by the analysts focused on deriving a better understanding of the outcomes of post-acute care. The major dilemma is that Medicare does not currently have information to know exactly what it is purchasing when it pays for post-acute care. As a consequence, it has been difficult to establish an appropriate price for that "product." Currently, Medicare's post-acute care payment policies address services used (e.g., visits, discharges). There is a paucity of information, however, on the outcomes that result from the use of any quantity of services. Outcome data was viewed by respondents as essential information for defining the product of post-acute care. Moreover, to address the issue of appropriate payment for individuals in need of post-acute care, regardless of provider setting, it is necessary to derive common information across provider types so that comparisons of patient needs, services received and outcomes can be made to derive prices that are patient-centered rather than provider-centered. HCFA is in the process of deriving a patient assessment instrument that is adaptable to the multiple providers of Medicare post-acute care services.

Some of the prior research on patients of rehabilitation facilities was mentioned as "a good model" of the type of information system that can be useful for decision making about the benefits of particular post-acute care services. Because patient specific conditions are measured at both admission and discharge, for example, it is possible to determine the extent to which improvements in health and functioning were made during the stay in a rehabilitation facility. The inter-stay information on outcomes within a specific provider type also helps to elucidate the purpose and value of each provider's services in episodes of care. Measures of care provision can also be included in this framework in order to make comparisons between process and outcomes, given the needs of specific types of patients.

In addition, there is a need to understand the characteristics of patients treated by different post-acute care providers. This information is important in understanding whether an integrated post-acute care system is needed. Such information is also needed to accurately compare outcomes across providers. A common data set of patient characteristics across post-acute care providers will improve our understanding of the similarities and differences among patients across providers. Clarification of patient subgroups is important in the long-term for analyses to determine what is needed to achieve various outcomes, but it is also important in the short run to define the casemix adjustments for the prospective payment systems for different post-acute providers, as prescribed by the BBA. Even though the latter purpose does not necessarily contribute to the longer range goal of measuring the benefits of post-acute care services, it does help refine the payment methodology under current circumstances. In addition to the casemix studies required by the BBA, the analysts thought that research should be conducted to casemix adjust the amount of therapy services provided to beneficiaries under Medicare Part B.15

The analysts noted that in creating casemix adjustments for payment systems, regardless of information on outcomes, research should focus more on patient or condition-specific needs, in contrast to services used. The rehabilitation therapy categories in the SNF casemix classification system, for example, groups patients according to intensity of therapy services received rather than by patient specific conditions that are indicative of amounts of therapy needed. Toward the development of the latter approach, the analysts indicated that further research is needed to develop normative standards for what patients need given their medical and functional profiles.

VIII. END NOTES

While this study was in progress, Congress and the Administration enacted the 1999 BBRA, which contained numerous provisions that affected 1997 BBA policies directed at post-acute care providers. In general, the 1999 BBRA provisions increased Medicare payments for such providers, reflecting the sentiment that the 1997 BBA policies had a greater impact on spending than Congress had intended. We noted many of these refinements in the preceding discussions.

Despite the achievements of the 1997 BBA, and those of the 1999 BBRA, many important post-acute care policy issues remain unresolved. One issue that has not been directly addressed is the goal of the Medicare home health benefit. Existing eligibility and coverage rules enable two general groups of patients to receive Medicare home health services. The first are patients requiring short-term restorative or rehabilitation care after an acute medical episode, and the second are those with chronic conditions who receive extended personal care services (given their need for skilled supervision). Although the first group is readily seen as a Medicare-type patient, there is some ambivalence about the second group which has been viewed as a long-term care population. Use of services by the latter group contributed largely to the increasing Medicare HHA spending rate prior to the 1997 BBA. The per-beneficiary cost limit in the IPS appeared to be a mechanism to curb such spending but, as our respondents noted, there is concern that diminished access to needed services may be creating severe health care problems for some beneficiaries and possibly higher costs for the Medicare program in general. Moreover, it seems that the IPS was simply a reimbursement policy approach for addressing what is essentially an eligibility and coverage policy issue.

As noted by the researchers and policy analysts, there is still a paucity of information about the "product" of post-acute care services. Although elaborate data systems are gradually being installed for many post-acute care providers, considerably more research needs to be conducted before information is available on the relationship between types and amounts of services provided and patient outcomes. This effort is further complicated by the need to identify these relationships across post-acute care provider settings. Until the information becomes available, questions will remain about both the quality of, and the appropriate price to pay for, post-acute care. In the interim, the BBA provisions that moved most post-acute care providers from a cost-based to a price-based payment system provide an opportunity to derive insight, through market behavior of providers, on whether Medicare is paying appropriately for post-acute care services.

Related to the preceding issue, the payment systems being developed and implemented for Medicare post-acute care services are provider specific, and major questions remain about whether the "system" of post-acute care services is being used efficiently. Observed utilization of multiple post-acute care providers, in different combinations, cannot readily be interpreted as reflecting necessary and efficient care versus revenue maximizing behavior. Although the utilization patterns are expected to change as a result of the BBA and BBRA-mandated payment policy changes, the question about appropriate and efficient use of the system of Medicare post-acute care services remain.

It is important to highlight a clear theme that emerged from our discussions with the provider and consumer groups, and the researchers and policy analysts. This is concern about access to needed services by the high cost "outlier" patients in any of the post-acute care settings, such as individuals with very high non-therapy ancillary costs in SNFs, with rehabilitation therapy needs exceeding Part B costs, or with home health care needs exceeding per-beneficiary cost limits. Particularly interesting, one of the health policy analysts raised the notion that some beneficiaries might have needs (and costs) that Medicare's "system" of post-acute care services and payments cannot accommodate. It seems important to learn more about the prevalence of such high cost patients and the extent of their access problems to needed services, either from individual provider types or the system of post-acute care services.

Finally, it is important to recall that the 1997 BBA provisions addressing Medicare post-acute care services have only partially been implemented (and modified by the 1999 BBRA). Implementation of remaining mandates, and refinements to newly established prospective payment systems, promise to maintain a dynamic environment for all post-acute care providers, as well as for other Medicare services, for the foreseeable future. This report addressed the views of knowledgeable individuals on the initial phase of the BBA-mandated program changes. Because the major goal of the BBA provisions was to derive Medicare savings, many other post-acute care issues extant prior to the BBA continue to be ones requiring further attention. Issues identified by individuals with whom we spoke added many additional area for future research and policy consideration.

NOTES

  1. This report is the second in a series, supported by the Office of the Assistant Secretary for Planning and Evaluation (ASPE) in the Department of Health and Human Services, designed to bring together available information on Medicare's post-acute care benefits. The first report reviewed available research on utilization and expenditure trends, Medicare policies governing post-acute care providers, the distribution and changes in the supply of such providers over time, and determinants of service use. Views in expressed in this report were derived from personal interviews with individuals associated with numerous organizations, listed in the Appendix.

  2. Since those conversations, the 1999 Balanced Budget Refinement Act (BBRA) was passed which addressed a number of the issues raised. We indicate, in footnotes, specific changes made by the 1999 BBRA when the issues are addressed in the text.

  3. The 1999 BBRA contained a provision that enables a SNF to choose to go to a full federal rate payment either immediately or over the multi-year transition period originally prescribed by the 1997 BBA. In addition, the BBRA increased the federal share of the SNF PPS by 4 percent for fiscal years 2001 and 2002.

  4. A provision in the 1999 BBRA temporarily increased the payment rates in SNFs by 20 percent for 15 RUG-III categories (i.e., for patients classified in extensive services, special care, clinically complex, high rehabilitation and medium rehabilitation). The goal of this provisions was to increase payments for some high cost patients. In addition, the BBRA provides that the SNF PPS payment for certain SNFs be split 50/50 between the federal and facility specific rates if the SNF had historically served a high percentage of immuno-compromised patients (e.g., patients with AIDS).

  5. In the SNF PPS final rule, HCFA clarified the conditions under which use of the additional 3 day grace period would be appropriate for completing assessments.

  6. Services that were not subject to the consolidated billing responsibilities of SNFs were clarified by HCFA in the SNF PPS interim final rule and subsequently through HCFA Program Memorandum #A-98-37. The memorandum expanded the categories of services exempt from consolidated billing. The 1999 BBRA further expanded the list of services exempt from consolidated billing.

  7. The 1999 BBRA contained a provision that suspended the Part B therapy caps for years 2000 and 2001. At the same time, the Secretary of HHS is to report to Congress on an alternative payment methodology based on patients' diagnosis, function, prior service use and other clinical conditions. HHS is also required to conduct medical review of these services in SNFs.

  8. This issue was resolved in the final SNF PPS regulations.

  9. MDS refers to minimum data set, which is an assessment instrument that has been developed over time to describe clinical characteristics of nursing home patients, as well as indicators of quality of care and treatment needs.

  10. The 1999 BBRA contained provisions that delayed the 15 percent reduction in payment for HHAs for at least one year, and required the Secretary of HHS to report to Congress on adequacy of payment levels. It also provided that the per-beneficiary limits be increased and that incremental payments be made to offset the costs of reporting OASIS data.

  11. In May 2000 HCFA announced the implementation of the rehabilitation PPS would be delayed until April 2001.

  12. The 1999 BBRA required the Secretary of HHS to implement a per-discharge PPS based on function related groups.

  13. The 1999 BBRA increased bonus payments for long-term care hospitals. The increased payments are available for cost reporting periods beginning on or after October 1, 2000 and before September 30, 2002 (i.e., the start of the BBA required PPS for long-term care hospitals).

  14. The 1999 BBRA required the Secretary of HHS to develop by October 1, 2001 a DRG-based PPS for long-term care hospitals and implement this system by October 1, 2002.

  15. The 1999 BBRA required the Secretary of HHS to report to Congress on a casemix adjusted payment system for Part B therapy services.

APPENDIX A

The views about current issues regarding Medicare post-acute care services discussed in this report are based on our personal interviews with numerous individuals who are representatives, affiliates, faculty members, or staff of the following organizations, as well as individual providers represented by some of the organizations:

American Academy of Physical Medicine and Rehabilitation Alzheimer's Association of America American Association for Respiratory Care American Association for Retired Persons American Association of Continuity of Care American Association of Homes and Services for the Aging American Association of Physical Medicine and Rehabilitation American Bar Association Commission on Legal Problems of the Elderly American Health Care Association American Hospital Association American Medical Rehabilitation Provider Association American Occupational Therapy Association American Physical Therapy Association American Speech, Language, and Hearing Association Association of Health Facility Survey Agencies

Brandeis University

Career Nurse Assistants Program Catholic Hospital Association Center for Beneficiary Rights Center for Medicare Advocacy, Inc. Commission on Accreditation of Rehabilitation Facilities Community Health Accreditation Program Consumer Coalition for Quality Health Care

Federation of American Health Systems

Health Strategy Associates Home Health Services and Staffing Association

Joint Commission on Accreditation of Healthcare Organizations

Long Term Acute Care Hospital Association of America

Medicare Payment Advisory Commission Medicare Rights Center

National Association for Home Care National Association of Long Term Hospitals National Citizens Coalition for Nursing Home Reform National Senior Citizens Law Center

Paraprofessional Institute

RAND

U.S. General Accounting Office U. S. Health Care Financing Administration University of Colorado

APPENDIX B

Policy Research Issues on Medicare Post-Acute Care (PAC)

  1. Were the increasing expenditure rates evidence that Medicare's PAC benefit was broken? Is the current spending rate about right?

  2. To what extent are "similar" patients served by different types of PAC providers?

  3. To what extent are inconsistencies in payment methodology among PAC providers a problem? Does sequential use of different providers reflect patient needs or payment incentives?

  4. What is the goal of Medicare's home health program?

  5. What are possible access, quality and cost consequences of the BBA provisions?
    • Inadequate non-therapy ancillary payments and Part B caps created access problems For patients who are clinically complex or require extensive rehabilitation.
    • RUG-III does not adequately explain costs (nursing, therapy, non-therapy ancillary, drugs) of certain patients.
    • Hospital admissions, re-admissions and LOS, and ER use are increasing.
    • Administrative burden (from consolidated billing, MDS+ assessment) increased.
    • Access to HHA care will be problematic for high cost or long-stay patients.
    • Quality of care at risk because of fewer staff, shift of functions to lower/different skilled staff, etc.

  6. How will the supply of post-acute care providers change and how will these changes affect Medicare beneficiaries?

  7. What are important eligibility and coverage issues that need to be addressed?

  8. Is integration of services on the basis of a "patient-centered" payment system a solution for Medicare's post-acute care benefit?
    • Any chance for an integrated payment system? What is it going to look like?
    • Do we have the knowledge to make "optimal placement" decisions?
    • A bundled DRG payment? Who's in charge?
    • How will the per-diem-based and per-episode-based systems be reconciled?
    • How extensively can measurement of patient outcomes be incorporated into an integrated payment system?

  9. What PAC issues need priority research?


SYNTHESIS AND ANALYSIS OF MEDICARE POST-ACUTE CARE BENEFITS AND ALTERNATIVES REPORTS AVAILABLE

Medicare's Post-Acute Care Benefits: Background, Trends, and Issues to be Faced
Executive Summary   http://aspe.hhs.gov/daltcp/reports/1999/mpacbes.htm
HTML   http://aspe.hhs.gov/daltcp/reports/1999/mpacb.htm
PDF   http://aspe.hhs.gov/daltcp/reports/1999/mpacb.pdf
Post-Acute Care Issues for Medicare: Interviews with Provider and Consumer Groups, and Researchers and Policy Analysis
Executive Summary   http://aspe.hhs.gov/daltcp/reports/2000/pasisses.htm
HTML   http://aspe.hhs.gov/daltcp/reports/2000/pasissue.htm
PDF   http://aspe.hhs.gov/daltcp/reports/2000/pasissue.pdf

To obtain a printed copy of this report, send the full report title and your mailing information to:

U.S. Department of Health and Human ServicesOffice of Disability, Aging and Long-Term Care PolicyRoom 424E, H.H. Humphrey Building200 Independence Avenue, S.W.Washington, D.C. 20201FAX:  202-401-7733Email:  webmaster.DALTCP@hhs.gov


RETURN TO:

Office of Disability, Aging and Long-Term Care Policy (DALTCP) Home [http://aspe.hhs.gov/_/office_specific/daltcp.cfm]Assistant Secretary for Planning and Evaluation (ASPE) Home [http://aspe.hhs.gov]U.S. Department of Health and Human Services Home [http://www.hhs.gov]

Preview
Download

"pasissue.pdf" (pdf, 1.45Mb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®