8

Reimbursing to Improve Quality of Care

Over the past ten years, quality assurance initiatives for long-term care have focused on regulatory programs, such as strengthening the survey and certification process for providers. In contrast, relatively few efforts have highlighted the role that reimbursement can play in promoting or inhibiting the quality of long-term care. Contributing to the lack of emphasis on reimbursement is the paucity of conclusive data on the topic, a situation that has changed little since the 1986 Institute of Medicine (IOM) report on nursing homes. Some studies have linked poor quality of care in nursing homes to low Medicaid payment rates, but others have posited that quality-of-care deficiencies should be attributed to factors such as excess demand (Nyman, 1993). Although relatively little is known about the effect of reimbursement on quality of care in nursing homes, virtually nothing is known about its impact on other settings or on home and community-based services.

Two recent developments have directed new attention to the relationship between reimbursement and the quality of long-term care. First, the federal Balanced Budget Act of 1997 repealed federal standards for reimbursing nursing home care under the Medicaid program (the Boren amendment), giving states virtually unlimited freedom in setting nursing home payment rates. For Medicaid home and community-based waiver

Much of the information in this chapter draws from the background paper commissioned from J.M. Wiener and D.G. Stevenson for use by the committee.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 235
Improving the Quality of Long-Term Care 8 Reimbursing to Improve Quality of Care Over the past ten years, quality assurance initiatives for long-term care have focused on regulatory programs, such as strengthening the survey and certification process for providers. In contrast, relatively few efforts have highlighted the role that reimbursement can play in promoting or inhibiting the quality of long-term care. Contributing to the lack of emphasis on reimbursement is the paucity of conclusive data on the topic, a situation that has changed little since the 1986 Institute of Medicine (IOM) report on nursing homes. Some studies have linked poor quality of care in nursing homes to low Medicaid payment rates, but others have posited that quality-of-care deficiencies should be attributed to factors such as excess demand (Nyman, 1993). Although relatively little is known about the effect of reimbursement on quality of care in nursing homes, virtually nothing is known about its impact on other settings or on home and community-based services. Two recent developments have directed new attention to the relationship between reimbursement and the quality of long-term care. First, the federal Balanced Budget Act of 1997 repealed federal standards for reimbursing nursing home care under the Medicaid program (the Boren amendment), giving states virtually unlimited freedom in setting nursing home payment rates. For Medicaid home and community-based waiver Much of the information in this chapter draws from the background paper commissioned from J.M. Wiener and D.G. Stevenson for use by the committee.

OCR for page 235
Improving the Quality of Long-Term Care service states have always had complete freedom in determining reimbursement levels. The nursing home industry has warned that Medicaid reimbursement levels already are too low and that further reductions would adversely affect the quality of care. Second, the Balanced Budget Act of 1997 also dramatically altered Medicare reimbursement methods for nursing homes and home health agencies and combined these changes with large budget savings. In some cases the changes have been major. As states gain new freedom to set Medicaid nursing home reimbursement levels and the federal government reduces Medicare payments, it becomes increasingly important to understand whether and how these changes might affect access and quality in long-term care. As reported below, changes in payment policies are creating great turmoil in the long-term care sector. The withdrawal of substantial resources from long-term care providers is troubling, especially because many of the recommendations in this report require more, not less, funding. REIMBURSEMENT AND QUALITY Research on reimbursement and its potential impact on the quality of care generally focuses on two broad areas of concern. First, what is the relationship between the costs of long-term care and the quality of care? This policy question is important because like most other areas of Medicaid policy, nursing home reimbursement levels and methods vary dramatically by state. For example, average Medicaid nursing home reimbursement rates for 1998 varied from a low of $62.58 per day in Nebraska to a high of $329.62 per day in Alaska. Second, does the method of payment (e.g., flat rate, prospective payment, use or type of casemix adjustment), independent of its level, affect the quality of care? This is potentially very important since government policy makers have considerable control over these policy levers. Level of Cost or Payment and Quality Although measuring cost and payment levels is comparatively straight-forward, measuring the quality of care is not, and the way quality is assessed can significantly affect the results of studies that examine the relationship between the two. All of the studies examined here focused on nursing homes. Most studies have analyzed the relationship between cost or payment and quality by using some form of input (e.g., staffing levels) or process indicator as the measure of quality. For example, using 1995 –1996 Online Survey and Certification Assessment Reporting (OSCAR) System data, Harrington and colleagues (1998b,c) found a small but positive relation-

OCR for page 235
Improving the Quality of Long-Term Care ship between the amount of Medicaid reimbursement and nurse staffing levels (except for nursing assistants) and reported fewer certification deficiencies in facilities with higher staffing levels. Questions can be raised about the appropriateness of using structural measures of quality because increased inputs imply, almost by definition, higher costs. That is, input and cost are likely to be positively related regardless of whether there is a relationship between cost and “real” quality. Several studies have found that higher reimbursement is associated with higher staffing, but they failed to find a significant relationship to other measures of quality (Nyman, 1988b; Zinn, 1994; Cohen and Spector, 1996). For example, Zinn (1994) found that higher reimbursement was associated with higher registered nurse staffing but, surprisingly, also with more use of restraints and a greater proportion of residents who were not toileted. Only a few limited studies have examined the relationship between facility costs and quality using outcomes-based quality measures. Using 1983 data from Iowa, Nyman (1988b) found that costs were not significantly greater in nursing homes with higher quality as measured by various outcomes (including wearing clean clothing, being fully dressed, and having clean hair). At the same time, these outcome measures of quality were found to be associated with nursing time per patient. Similarly, using 1987 National Medical Expenditure Survey data, Cohen and Spector (1996) did not find a statistically significant relationship between reimbursement level and outcomes-based quality measures (including mortality, change in functional status, and presence of decubitus ulcers), while at the same time they did find a positive relationship between reimbursement level and staffing intensity. It appears that many aspects of quality care (e.g., staff attitude or administrative philosophy) do not require large expenditures and are not significantly related to facility costs (Ullman, 1987; Nyman, 1988b). Thus, improved quality might not necessarily imply higher costs, and poor quality might not simply be a result of inadequate resources. The results of the Nyman (1988b) and Cohen and Spector (1996) studies illustrate the complexity of the relationship among costs, inputs, and outcomes and the dilemma for states in trying to establish payment rates that are adequate to produce quality care. Both found a relationship between cost or reimbursement level and staffing intensity, and both analyses found that professional staffing had a positive and significant relationship to quality of care in terms of outcomes. However, the effects of higher cost or reimbursement levels on staffing and of staffing on outcomes were not large enough for cost or reimbursement to have a statistically significant impact on quality as measured by outcomes. Although there does not appear to be a simple relationship between

OCR for page 235
Improving the Quality of Long-Term Care cost and quality, logic suggests that there is some minimal level of reimbursement below which it will be either difficult or impossible for nursing homes to provide an adequate level of care. Moreover, current quality-of-care problems in long-term care should make policy makers alert to the possible negative impact of reducing the resources available to long-term care providers. Recommendation 8.1: The committee recommends that, before making decisions to reduce reimbursements, state officials carefully assess the impact on access to services and on quality of care of any proposed reductions in Medicaid reimbursements for nursing home, home health and other home and community-based services. Recommendation 8.2: The committee recommends that the Department of Health and Human Services fund and support research to better understand the effects of payment policies on accessibility and quality of long-term care services, including the following: the effects of low reimbursement rates or changes in Medicare and Medicaid reimbursement policies on providers of nursing home, home health, or other long-term care services; the effects of current payment systems, such as prospective payment for nursing facilities and interim payment systems for home health agencies, on the accessibility and quality of services; and whether states with low Medicaid reimbursement rates (adjusted for geographic variation in prices and other state-specific requirements) have lower quality of nursing home care. Method of Reimbursement and Quality Setting Medicaid reimbursement rates for long-term care is one way in which states control expenditures and shape the long-term care market. To achieve savings, states focus not only on the overall level of reimbursement but also on the payment methodology used to reimburse long-term care providers. These policies differ most fundamentally on two levels: (1) whether they base payment on facility-specific costs or on a set of flat rates (set independently of an individual facility's costs) and (2) whether rates are set retrospectively or prospectively.

OCR for page 235
Improving the Quality of Long-Term Care Facility-specific rates (set either prospectively or retrospectively) are based on an individual facility's costs, usually up to some ceiling. Under this type of payment, higher-cost facilities receive higher payments than lower-cost facilities. Under flat rates, nursing homes are paid a rate that is not based on the individual facility's costs. Typically, flat rates are based on the cost experience of all facilities in an area (sometimes adjusted for facility or patient characteristics such as the casemix of a nursing home's residents). Under retrospective cost-based payment, nursing homes receive a facility-specific interim payment rate based on costs for some base year, with adjustment for inflation. If the actual costs (usually up to some ceiling) are different from the interim rate, either the state pays the facility or the facility pays the state the difference. This methodology encourages facilities to spend more (perhaps improving quality) because they can be reimbursed for their expenses, although ceilings on allowable costs and lags in altering rates affect on the incentive for facilities to increase spending. Almost all states use prospective payment systems to pay nursing homes. Under prospective payment, providers receive a rate set in advance for a bundle of services, without adjustment for actual costs. Like a capitated payment for managed care, providers are at financial risk if facility costs exceed payments; alternately, providers can keep the surplus as profits should payments exceed their costs. To the extent that facilities make money by curtailing services, quality may be adversely affected. In theory, this should be more of a problem for flat-rate systems because there is no relationship between an individual facility's costs and its reimbursement. In contrast, facility-specific prospective payment systems periodically recalculate a facility' s base costs. Thus, if a facility dramatically reduced its expenses, its future rate would also be reduced, limiting the extent to which it is in the nursing home's interest to reduce costs. Empirical studies tend to support these theoretical expectations. Using cost-reporting data from eight states from 1978 to 1980, Holahan and Cohen (1987) found strong evidence that the cost containment incentives in state reimbursement systems appeared to have a real impact on cost increases: prospective and flat-rate systems generally reduced cost growth more than retrospective payment. Aggressive cost control strategies were also found to have a constraining effect on spending for direct care services (and therefore might adversely affect patient care). Patient care-related costs were constrained more than non-patient care-related costs in reimbursement systems with stronger cost-controlling incentives. At least in the short term, cost containment measures did not negatively affect access for Medicaid recipients. Using 1981 Medicare and Medicaid Automated Certification Survey

OCR for page 235
Improving the Quality of Long-Term Care files and Medicare cost reports, Cohen and Dubay (1990) found that as cost containment incentives became stronger (e.g., the use of flat-rate payments), nursing homes responded by decreasing the severity of their casemix (e.g., limiting access for heavy care residents) and decreasing staffing levels. Nursing homes in states that paid flat rates had fewer nurses per bed than similar homes in cost-based reimbursement states. In addition, access for Medicaid recipients was worse in states with flat-rate reimbursement and better in states with prospective reimbursement. Finally, prospectively paid nursing facilities did not have costs that differed significantly from retrospectively paid facilities. However, facilities that were paid flat rates under Medicaid had significantly lower costs. Prospective payment also can affect patient care services differently, depending on the bundle of services included in the unit of payment. For example, prospective payment rates may or may not include ancillary services such as prescription drugs and therapy services. In a study of five states, Moore and White (1998) compared New York—which includes prescription drugs in its prospective payments for nursing homes —to four states that reimburse prescription drugs on a fee-for-service basis (i.e., separate from the prospective rate). Their study found that prescription drug utilization for the treatment of selected medical conditions was significantly lower in New York than in the comparison states, suggesting that nursing homes responded to financial incentives of fixed payment. Casemix Reimbursement One variation of flat-rate reimbursement that many states have adopted is casemix reimbursement. An undesirable incentive of pure flatrate payments (and, to a lesser extent, facility-specific methodologies) is for nursing homes to avoid costly residents who are severely disabled, since reimbursement is no greater for residents with heavy care needs than for those with light care needs. Casemix reimbursement systems are designed to mitigate the effects of these perverse incentives by matching the payment level to an individual's care needs. Under casemix reimbursement, nursing homes receive higher reimbursement when individuals require more services. The major theoretical strength of casemix reimbursement is to make nursing homes indifferent to the relative care needs of the individuals they admit. Despite these advantages, casemix reimbursement also creates disincentives for nursing homes to rehabilitate residents (since they are paid more for more disabled residents) or to provide services that diminish their profits (since profits represent the difference between reimbursement and expenditures). These systems also create incentives for nursing homes to misreport resident conditions or services received. In their analy-

OCR for page 235
Improving the Quality of Long-Term Care sis of 1987 data from six states, Butler and Schlenker (1989) found some evidence that these problems did in fact occur and concluded that casemix reimbursement systems must include explicit ways to measure and ensure quality of care. In their review of the casemix reimbursement literature, Weissert and Musliner (1992) concluded that casemix payment by itself generally did little to improve quality of care. Higher casemix payments were not necessarily used to increase nursing home staffing levels or expenditures (Davis et al., 1998). In contrast, research is more positive on the use of casemix systems to improve access to care, although improving access for residents who require heavy care can create access difficulties for those who require light care (and arguably should be served outside of nursing homes). Prospective Payment for Medicare Post-Acute Care Services The Balanced Budget Act of 1997 (BBA 97) mandated the establishment of casemix-adjusted prospective payment systems for various Medicare post-acute care services, including nursing facility and home health services. These changes were expected to result in substantial reductions in Medicare expenditures for home health agencies and nursing facilities compared to what expenditures would have been without the changes. Reflecting concerns about the potential impact of these changes on people who use nursing home and home health services, the Medicare Payment Advisory Commission (MedPAC) recommended that DHHS establish systems to monitor quality of care as prospective payment is implemented for nursing homes and home health agencies (MedPAC, 1999). The BBA 97 also moved nursing facility services into a prospective payment system (PPS), transitioning over three cost-reporting periods starting in July 1998. The new payment system is supposed to pay appropriately for the level of care needed and to control costs. Nursing facilities receive a casemix-adjusted, per diem payment based on a blend of national and facility-specific payment amounts. The payment bundles nursing, therapy, and capital payments into a single per diem amount. This is expected to constrain the growth in therapy or ancillary service use, where much of the recent growth in expenditures has occurred. Many observers argue that the methodology does not adequately account for the costs of nontherapy ancillaries such as prescription drugs (GAO, 1999e). A major concern is that the methods and data for casemix-adjusted payments are inadequate to design incentives that discourage providers from skimping on care or from avoiding consumers who need greater amounts of care (MedPAC, 1998). MedPAC recommendations for refining the nursing facility prospective payment system include (1) improving the system's ability to predict resources associated with patient need for

OCR for page 235
Improving the Quality of Long-Term Care nontherapy ancillary services (e.g., respiratory therapy, drugs); (2) updating payment weights to reflect such factors as changing technology and care patterns; and (3) developing methods to review the accuracy of facility assessments and classification of residents for payment purposes (MedPAC, 1999). The committee did not investigate payment issues at this level of specificity, but these recommendations appear reasonable in light of the committee's concerns. In addition, the BBA 97 requires that nursing facilities reimburse therapists directly for services delivered in the facility, regardless of whether the service was provided by in-house staff or external organizations. Independent therapists will no longer be able to bill Medicare directly. As part of the transition, the cost limits for fiscal year 1998 will be adjusted to incorporate the cost limits frozen in 1994 and 1995. Like all prospective payment systems, a risk is that facilities may reduce what they spend on care in order to increase profits. Although the BBA 97 also requires the implementation of a prospective payment system for home health, it gives the Health Care Financing Administration (HCFA) great discretion in its design. A home health prospective payment system was originally scheduled to be implemented in October 2000, but has been postponed by subsequent legislation. Reflecting congressional belief that aggregate payments were too high, reimbursement rates are to be adjusted downward by 15 percent when the prospective payment system is implemented. In the meantime, an interim payment system based on modified cost limits is in place for home health care. Agencies are paid the lowest of their costs, 106 percent of the median cost for freestanding agencies, or average per-beneficiary expenditures. This third ceiling, average perbeneficiary expenditures, which is 75 percent based on facility-specific experience and 25 percent based on regional average costs, makes it difficult for agencies to change their service mix to provide either more expensive nursing services or more home health aide visits per patient. The interim payment system for home health care is very controversial. Some researchers have warned that the system could restrict access to care, especially for high-use consumers (Komisar and Feder, 1998; Lewin Group, 1998). In addition, as originally specified by BBA 97, the interim payment system could be particularly difficult for several types of providers including home health agencies in rural areas, small agencies that have a large number of high-use residents, agencies that serve a more disabled population than in 1994 (the base year for payment calculations), and agencies that have formed as a result of recent mergers or acquisitions. Implementation of the provisions of BBA 97 created great turbulence in the nursing home and home health industries. Nursing homes have complained that the budget cuts are too large, that consolidated billing is

OCR for page 235
Improving the Quality of Long-Term Care administratively burdensome, and that nontherapy ancillaries are not adequately recognized in the payment rate. Stock prices for publicly traded nursing home chains have plummeted dramatically, and seven national nursing home chains have filed for Chapter 11 bankruptcy protection, accounting for nearly 10 percent of the nation's nursing home beds (Childs, 2000; Editor, 2000; Vickery, 2000). While the industry blames changes in Medicare reimbursement policy for these problems, the General Accounting Office (GAO, 1999e) largely attributes them to poor business decisions. The impact of the reimbursement changes on home health agencies has been even more dramatic, with Medicare expenditures falling drastically. Medicare home health expenditures declined 45 percent between 1997 and 1999 (CBO, 2000). In addition, the number of Medicare-certified home health agencies dropped from 10,444 in 1997 to 7,747 in 1999, a figure that is still far above the number of agencies in the early 1990s (NAHC, 2000). A GAO study of early implementation of the interim payment system found that it had not caused significant access problems for beneficiaries (GAO, 1998c). In part, this reflects the fact that the number of agencies nearly doubled between 1989 and 1997. The GAO did note the possibility of access problems for high-cost beneficiaries, but characterized them as relatively few and possibly related to other characteristics of the Medicare program. The 1999 MedPAC report noted concerns about the interim payment system but did not recommend further changes. The report noted the difficulties of designing a prospective payment system for home health care that “appropriately classified patients who require both short- and longer-term home health services” (MedPAC, 1999, p. 92). Responding to the uproar caused by the Balanced Budget Act, in 1999 Congress passed and President Clinton signed the Medicare, Medicaid, and the State Children's Health Insurance Program (SCHIP) Balanced Budget Refinement Act of 1999. This legislation provides modest financial relief to a variety of providers, including nursing facilities and home health agencies. For nursing facilities, the legislation temporarily would increase the federal per diem by 20 percent for 15 Resource Utilization Groups (RUGs) in the categories of “extensive services, ” “special care,” “clinically complex,” “high rehabilitation,” and “medium rehabilitation.” According to the BBA, increased payments are made starting April 1, 2000, or with the implementation of a refined RUG system. In fiscal years 2001 and 2002, the federal per diem rate will increase by 4 percent for each year. In addition, starting in April 2000, separate payments are made for certain ancillary services. For home health agencies, the legislation will add $1.3 billion in home health reimbursement over five years by delaying the 15 percent reduction in payments until after the first year of the prospective payment

OCR for page 235
Improving the Quality of Long-Term Care system. The law also slightly increases per-beneficiary cost for some agencies, removes the consolidated billing requirement for durable medical equipment, provides a $10 payment for completing a patient assessment, and revises the surety bond requirements. Given the repercussions of these reimbursement changes for the long-term care industry, it is critical that the HCFA and Congress have accurate, up-to-date information. Reimbursement Incentives to Improve Quality of Care Some researchers have proposed that reimbursement be linked directly to quality of care (Zinn, 1994). Pointing out that reimbursement and quality assurance are typically defined by independent systems with separate objectives, Shaughnessy and Kurowski (1982) detailed several areas of research that have to be addressed before this linkage is possible, including the development of better process and outcomes measures, identification of quality norms, and development of incentives to change provider behavior. Although their article was written 18 years ago, most of the same limitations in the research base remain today. There have been some experiments with outcomes-based incentives. A demonstration in San Diego in the early 1980s tested the effectiveness of monetary incentives in improving the health of nursing home residents and reducing Medicaid expenditures. These incentives for improved outcomes were found to have beneficial effects on quality, access, and number of hospital transfers (Norton, 1992). Other initiatives in Illinois, Connecticut, and Michigan have not been evaluated. LIMITED NURSING HOME BED SUPPLY AND QUALITY OF CARE Several studies done in the 1970s and early 1980s found a strong relationship between poor quality and a high percentage of Medicaid residents in nursing homes (Anderson et al., 1968; Kosberg and Tobin, 1972; Gottesman, 1974; Fottler et al., 1981; Weissert and Scanlon, 1985). These results are often interpreted as evidence that Medicaid nursing home reimbursement rates were too low to provide good quality care. If this were the case, then the quality-of-care problem could be alleviated simply by raising Medicaid reimbursement rates. An alternative explanation is that the relationship between homes that are heavily dependent on Medicaid and low quality is attributable to an insufficient supply of nursing home beds available for Medicaid residents (i.e., excess demand), which means that facilities need not compete by providing high-quality care (Nyman, 1993, 1988a,b,c). Excess demand

OCR for page 235
Improving the Quality of Long-Term Care exists when not enough beds are available for consumers demanding care at a given market price, a condition that may be optimal for both the state and the nursing homes if not for the residents. Nursing homes typically charge two different rates—one for Medicaid residents and a higher rate for private-pay residents (Minnesota and North Dakota require that nursing homes charge the same rate to private and Medicaid recipients). Since private-pay rates are always higher than Medicaid reimbursement levels, profit-maximizing nursing homes would rather admit a private-paying individual than an individual supported by Medicaid. In an analysis of data from 43 states for 1969–1973, Scanlon (1980) found that excess demand resulted in a segmented nursing home market, with private-paying residents obtaining all desired care and public-paying residents filling any remaining beds. If excess demand exists for nursing home beds, it is excess Medicaid demand. Studies of Wisconsin and the 10 national long-term care channeling demonstration sites in the early 1980s also found evidence of excess demand (Nyman, 1985, 1988a,c). Under excess demand, nursing homes can attract as many Medicaid residents as they want, regardless of quality. Although Medicaid residents prefer a higher quality of care, they cannot exercise these preferences under excess demand and must instead choose among the limited number of available beds. Hence, high-quality care is necessary only to attract more private-pay nursing home residents. In a market with a surplus of nursing home beds, however, nursing homes cannot act on their preference to admit private-paying residents over Medicaid-subsidized residents—they will accept either type in order to fill their beds. In this environment, both private- and Medicaid-paying residents would be able to exercise their preference for high-quality care. Consequently, nursing homes should increase quality in an attempt to attract residents. Using 1978–1979 and 1983 data from Wisconsin, Nyman (1985, 1988a,c) found evidence to support his theory that excess demand was responsible for lower-quality nursing home care. He found that the low-quality–high-Medicaid relationship was stronger under conditions of excess demand. In addition, Nyman found that nursing homes in counties with a surplus bed supply spent more on patient care per empty bed than counties with a tight bed supply. Nyman posited that this evidence cast doubt on the assumption that the relationship between high Medicaid and low quality was simply from lower Medicaid payments. Based on 1987 nursing home survey and certification data, Zinn (1994) reached similar conclusions. In addition to quality, the presence of excess demand in the nursing home market creates concern about access to nursing home care, especially for Medicaid residents. Private demand should not be affected by Medicaid demand or bed supply since private-pay residents always have

OCR for page 235
Improving the Quality of Long-Term Care admission preference. Using 1982–1984 data from the National Long-Term Care Survey and Area Resource, Ettner (1993) found that Medicaid residents had poorer access to nursing homes on average and that these differences seemed to exist mainly in areas in which bed supply was low or private competition for beds was greater. In addition, some nursing homes (e.g., Vencor, Inc.) have begun to focus on the private-pay and Medicare markets to the exclusion of the Medicaid market (Moss and Adams, 1998). Few facilities, however, can afford to exclude Medicaid residents completely since Medicaid pays for the large majority of nursing home residents. There is some evidence that excess demand has declined in more recent years. In an evaluation of 1988 data from Wisconsin, Minnesota, and Oregon, Nyman (1993) found no evidence of excess demand (for Wisconsin, this finding contrasted with his analysis of 1983 data). In addition, national nursing home occupancy rates declined from 92 percent in 1985 to 88 percent in 1995 (NCHS, 1997). The number of nursing home beds also decreased substantially in a similar time period (Harrington et al., 1998a). However, there remains an extremely strong relationship between nursing home bed supply and use, leading most states to fear that a greater number of nursing home beds will result in higher Medicaid expenditures (Wiener et al., 1999). As a result, most states control the supply of nursing homes through certificate-of-need programs or moratoriums on new construction or certification for Medicaid. State efforts to shift the balance of the long-term care system from institutional-based to home and community-based care by expanding home care services and using case management and preadmission screening efforts to encourage placement in settings other than nursing homes may also reduce excess demand. In all but a few states, however, home and community-based services are only a small proportion of Medicaid long-term care expenditures for the elderly. In addition, Medicare home health expenditures have skyrocketed since 1989 and the program has become more long-term care oriented. As the availability of home and community-based services increases, nursing homes will have to compete more actively with other types of long-term care providers (e.g., assisted living facilities and home health agencies) as well as with other nursing homes. CONCLUSION The impact of changes in reimbursement on the quality of long-term care is difficult to assess. Almost all of the research literature on the relationship between financing and quality is limited to nursing homes, is based on very old data, and does not reflect the regulatory changes

OCR for page 235
Improving the Quality of Long-Term Care required by the Omnibus Budget Reconciliation Act of 1987. Most studies on this topic were published in the mid- to late 1980s and relied on data from the late 1970s and early 1980s. Moreover, several studies focused on data from one or a few states, making it hard to generalize to the nation as a whole. In addition, the measures of quality in most of these studies were quite rudimentary, especially in terms of outcomes. Again, even less is known about reimbursement for nonmedical home care services and its effect on quality of care. The additional research recommended here is important to help policy makers make better-informed decisions about payment for long-term care. In the meantime, because the quality of long-term care is already problematic, states and the federal government should be cautious in their quest for Medicare and Medicaid savings. Because many of the recommendations proposed in this report will likely mean additional costs for providers, (e.g., for additional staff), the withdrawal of substantial resources from long-term care providers is a matter of concern. Although research on nursing homes suggests that higher rates are not necessarily used to improve resident care and that many elements of quality care do not require spending more money, direct resident care expenditures are particularly vulnerable to rate reduction initiatives. Lowering Medicaid nursing home reimbursement rates may be especially problematic in states with high levels of excess demand, since nursing homes in these states do not have to compete for consumers on the basis of quality. Reducing excess demand would lower the quality risks of reducing nursing home rates, but probably would result in higher Medicaid expenditures overall.