Why Medicare Matters to People Who Need Long-Term Care

Introduction

Medicare was enacted to provide health insurance to the elderly (and later, the disabled) population. Not only was it not intended to pay for long-term care (LTC); its statute explicitly excluded coverage for custodial care—the assistance with basic activities of daily living (ADLs) (such as bathing, dressing, and eating) that constitute LTC. Although the Federal-State Medicaid program, unlike Medicare, does finance LTC, its protection does not prevent financial catastrophe resulting from LTC needs. Rather, it supports service only after people have become impoverished. Given the limitations of public programs and of private insurance, today, as in 1966, people face the prospect of financial catastrophe when they need extensive LTC services.

Although Medicare was not designed as, and has not become, an LTC program, it matters enormously to people who need such care. First and foremost, Medicare's functionally impaired beneficiaries depend on Medicare to finance the substantial medical care they require. Second, beneficiaries are affected by Medicare policies regarding its postacute benefits, home health, and skilled nursing facility (SNF) care. Although SNF care remains overwhelmingly related to acute rather than LTC, Medicare's home health benefit is of growing importance to a segment of the LTC population.

This population also matters to Medicare. The 13 percent of beneficiaries with substantial LTC needs accounts for 32 percent of Medicare's expenditures. Growth in expenditures for home health and SNF care is contributing disproportionately to rising Medicare costs. Policymakers seeking to control Medicare costs, in general and for these benefits, must pay careful attention to balancing the importance of slowing spending growth with the importance of meeting the needs of beneficiaries, including beneficiaries who need LTC.

Which Beneficiaries Need LTC?

The 37.6 million elderly people and people with disabilities covered by Medicare are generally healthy and do not need extensive health care. About 72 percent of persons over 65 years of age report excellent or good health (Rice, 1996). Additionally, more than one-half of beneficiaries required Medicare reimbursement of $500 or less in 1993, with more than 18 percent reporting no Medicare expenditures (Rice, 1996).

However, a significant subset of beneficiaries has functional limitations that necessitate LTC. In 1993, about 9.3 million, or 25 percent, of Medicare beneficiaries needed assistance in one or more ADLs or were in an institution (Figure 1).¹ Almost one-half of these (4.8 million) had substantial LTC needs—that is, they were in nursing homes or had three or more ADL limitations and lived in the community. For this subset of Medicare beneficiaries, medical and LTC services are essential to leading healthy and safe lives.

Figure 1. Medicare Beneficiaries With Long-Term Care Needs: 1993.

Since Medicare's inception, its population has changed in ways that increase the likelihood that beneficiaries will need LTC. First, a larger proportion of Medicare beneficiaries are under 65 years of age with disabilities. The share of all beneficiaries who have a disability or end stage renal disease (ESRD) grew from 8 percent in 1974 to 13 percent in 1996 and is expected to grow to 16 percent by 2015 (Figure 2) (Health Care Financing Administration, 1996). Second, Medicare's older population is increasing. The proportion of Medicare beneficiaries who are over 85 years of age rose from just over 8 percent in 1978 (Health Care Financing Administration, 1995) to 11 percent in 1994 (Gornick et al., 1996). People over 85 years of age have average Medicare spending per enrollee that is about twice that of beneficiaries between ages 65 and 69 (Rice, 1996). They are also more likely to become nursing home residents: In 1990, 1.4 percent of persons 65-74 years of age resided in nursing homes, compared with 18.6 percent of those 85-89 years and 33 percent of those 90-94 years (Gornick et al., 1996).

Figure 2. Growth in the Number and Percent of Medicare Beneficiaries With Disabilities or End Stage Renal Disease (ESRD): Selected Years 1966-2015.

These patterns are highlighted by the disproportionately high representation of the people with disabilities and older beneficiaries in Medicare's LTC population (Figure 3). People under 65 years of age with disabilities or with ESRD account for 15 percent of beneficiaries with substantial LTC needs, compared with 11 percent of all beneficiaries. The oldest Medicare beneficiaries—those over 85 years of age—account for 35 percent of this LTC population, more than three times their proportion in the general Medicare population (11 percent). Additionally, a greater proportion of beneficiaries with substantial LTC needs are women: 67 percent among the LTC population versus 57 percent for the total Medicare population (Figure 4).

Figure 3. Percent Distribution of Medicare Beneficiaries, by Age: 1993.

Figure 4. Percent Distribution of Medicare Beneficiaries, by Gender: 1993.

Because Medicare explicitly excludes LTC, it is Medicaid that beneficiaries count on if they need such care and cannot afford it. Medicaid is the primary payer for LTC in the United States, covering nearly one-half of all nursing home expenditures (Levit et al., 1996) and financing care for about two-thirds of all nursing home residents (Harrington, Thollaug, and Summers, 1995). Medicaid plays a critical role in assisting Medicare beneficiaries who need substantial LTC. Although only about 16 percent of the full Medicare population also have Medicaid coverage, 45 percent of beneficiaries with substantial LTC needs receive Medicaid (Figure 5).

Figure 5. Percent Distribution of Medicare Beneficiaries, by Medicaid Status: 1993.

Why Does Medicare matter To the LTC Population?

People with LTC needs have disproportionately high medical costs. Alongside their disabling conditions, people with chronic disabilities or ADL limitations are more likely to have acute illnesses that involve expensive treatment (Iezzoni et al., 1994). Beneficiaries with substantial LTC needs constitute only 13 percent of beneficiaries, but they account for about 32 percent of Medicare spending (Figure 6). Medicare spends, on average, $8,960 per person with substantial LTC needs, compared with an average of $2,840 per beneficiary without these needs (Figure 7).

Figure 6. Number of Medicare Beneficiaries and Expenditures for Beneficiaries With and Without Substantial Long-Term Care Needs: 1993.

Figure 7. Medicare Expenditures per Beneficiary for Beneficiaries With and Without Substantial Long-Term Care Needs: 1993.

Nearly 80 percent of the $8,960 results from hospital and physician services. Expenditures on these services for people with substantial LTC needs ($7,070 per beneficiary) are more than double the levels ($2,675 per beneficiary) for beneficiaries without such needs. In addition, Medicare's postacute services—home health and SNF care—are far more significant to the high-need population. About 6 percent of average Medicare expenditures for those without significant LTC needs are for postacute care, while more than 20 percent of the average spending for beneficiaries with significant needs is for home health and SNF services (Figure 7). However, this difference between groups in spending on postacute services ($1,730) accounts for less than 30 percent of the overall spending differential ($6,120).

Despite Medicare's valuable insurance protection, beneficiaries in general, and beneficiaries who need LTC in particular, incur substantial costs that Medicare does not cover. Even without taking LTC spending into account, out-of-pocket spending for health insurance, medical services, prescription drugs, and medical supplies absorbed 18 percent of seniors' after-tax income, more than three times higher than the comparable percentage for people under 65 years of age (Health Care Financing Administration, 1996). Low-income seniors spend an even larger proportion (24 percent)—six times the share for seniors in the top income quintile (Health Care Financing Administration, 1996). Despite Medicaid's support, Medicare beneficiaries with substantial care needs, including LTC, face substantial financial risk.

Medicare's Home Health Benefit

Medicare's home health benefit was established to facilitate hospital discharge. Perceived as an alternative to hospital stays, it was not intended to cover long-term personal care. Furthermore, until the late 1980s, it was administered explicitly to prevent extended service. Legal action in the late 1980s, however, significantly altered the program's capacity to limit coverage. The result has been a significant expansion in Medicare-financed home health care. Although much of that care remains tied to acute illness, both the character of the benefit and the nature of its recipients have made Medicare's home health care a significant element in LTC financing.

Although the Medicare statute explicitly prohibits Medicare coverage of custodial care, the program's acute care coverage includes financing for care at home for beneficiaries who are homebound, under a physician's care, and in need of skilled nursing care on a part-time or intermittent basis, or in need of physical, speech, or (continuing) occupational therapy. People who qualify as needing skilled care may also receive aide services, subject to certain limitations. In contrast to most other services, beneficiaries pay no cost-sharing for home health services.

From early in Medicare's history (Callender and LaVor, 1975), administrators and the Congress have struggled with implementing home health coverage, seeking to balance the desire to minimize care in institutions with a concern about program costs. Too narrow a benefit would minimize use of home health as an alternative to hospital care. Too broad a benefit would extend program coverage beyond acute to long-term custodial care, exceeding Medicare's boundaries.

Even before 1970, questions arose about how that balance was being struck. Concerns about inappropriate coverage led program administrators to instruct intermediaries (insurance plans responsible for administering claims) to more carefully distinguish uncovered from covered care. Agencies liable for the costs of denied claims responded. Many agencies dropped out of the program, and claims and expenditures dropped significantly (Callender and LaVor, 1975). After that initial restriction, however, Congress enacted legislation to make the narrower home health benefit more accessible (Moon, 1993). Initial cost-sharing requirements for a portion of the benefit were eliminated in 1972; requirements for prior hospitalization and limits on visits were eliminated in 1980. Provider-participation requirements were also modified, bringing considerable numbers of proprietary agencies into the home health business. Not surprisingly, these changes brought increased program expenditures (more than 40 percent per year between 1980 and 1983) as more beneficiaries received more service, primarily limited to short-term care (Health Care Financing Administration, 1995).

In the 1980s, concern about cost increases again led to administrative restrictions on coverage (Moon, 1993; Bishop and Skwara, 1993). Restrictions appear to have been particularly aggressive in response to the increased demand for home health care that followed the implementation of prospective payment for hospitals. The key to these restrictions—and the key to keeping the home health benefit short-term— has been interpretation and enforcement of coverage rules, both by HCFA and by its intermediaries. Because home health agencies were financially liable for uncovered claims, the availability of services tended to closely reflect the coverage rules. In the course of the 1980s, these rules were challenged in the Congress and the courts for the vagueness of HCFA guidelines, the inconsistency of interpretation across areas, and the specific interpretation of eligibility criteria. Particularly at issue was what it meant to be “home-bound” or in “part-time” or “intermittent” need of “skilled care.” Alongside disputes about what these terms meant came charges that arbitrary benefit limits and claims denials were being used to limit expenditures. The U.S. General Accounting Office (1996) reported that denial rates increased from 3.4 percent in 1985 to 7.9 percent in 1987.

These challenges brought changes in HCFA's coverage policy (Moon, 1993; Bishop and Skwara, 1993; U.S. General Accounting Office, 1996). As of July 1, 1989, HCFA both broadened and clarified its interpretation of skilled care and the terms on which beneficiaries could receive it. Skilled care was explicitly extended beyond specialized services to include judgmental services such as skilled observation, patient assessment and management, and evaluation of patients' care plans. The meanings of “part-time,” “intermittent,” and “homebound” were clarified to facilitate, rather than limit, provision of care at home. As before, people who qualified as satisfying these conditions became eligible not only for skilled services, but also for other home health services, including support services provided by home health aides.

Following the promulgation of these regulations, the proportion of beneficiaries receiving home health care has consistently increased, as has the number of visits per person served. From 1989 to 1994, the number of persons served per thousand enrollees increased from 50 to 87, and the number of visits per person served more than doubled (27 to 65) (Figure 8) (Prospective Payment Assessment Commission, 1996). Although payments per visit have grown very slowly over the period, increases in the volume of service produced a fivefold increase in Medicare spending on home health care, from $2.6 billion in 1989 to $13 billion in 1994, at an average annual growth rate of more than 35 percent per year (Prospective Payment Assessment Commission, 1996). Between 1986 and 1994, the number of home health agencies increased by 25 percent (Lewin-VHI, Inc., 1995).

Figure 8. Average Number of Medicare Home Health Visits per Person Served: 1983-94.

Increasingly important in the expansion of services—and significant with respect to LTC—has been the increase in the number of persons receiving extensive home health visits, particularly visits for aide, rather than professional, services (Bishop and Skwara, 1993). Between 1988 and 1991, the proportion of home health users with more than 100 visits more than doubled from 26 percent to 53 percent. Aide visits accounted for about 60 percent of the increase in total visits per user and grew from one-third to 44 percent of visits from 1986 to 1991. These changes provide strong indications that Medicare's home health benefit moved significantly toward LTC.

In the 1990s, Medicare's home health benefit nevertheless remains to a considerable extent the short-term, postacute benefit the initial legislation intended. In 1994, about one-half of all home health users received fewer than 30 visits (Mauser, 1996). In 1993, about one-half of all users received no aide visits (Health Care Financing Administration, 1995). For users with fewer than 100 visits, the average annual Medicare reimbursement was $1,750 per user (Mauser, 1996). At the same time, most Medicare beneficiaries needing substantial LTC and living in the community did not receive home health care. In 1992, more than three-quarters of beneficiaries who needed assistance in three or more ADLs did not receive home health care (Mauser and Miller, 1994).

Despite the continued emphasis of the benefit on short-term care, a small proportion of Medicare users appears to need LTC and to get a significant amount of personal care from the program. In 1994, about 10 percent of users received more than 200 visits (Mauser, 1996). The one-fifth of users with very long episodes of care (166 days or more) received three-fifths of the program's visits. About one-half of the visits for these high users were aide visits (Prospective Payment Assessment Commission, 1996). In 1992, just over 40 percent of home health users with more than 100 visits needed assistance in three or more ADLs. Interestingly, about 40 percent of high users (more than 150 visits) were also eligible for Medicaid—the program that explicitly covers LTC in the home (Mauser and Miller, 1994).

Although high users represent a small proportion of beneficiaries using Medicare home health, they account for a sizable proportion of dollars spent on the benefit. On average, high users (more than 200 visits per year) cost Medicare $17,420 per user for their home health care in 1994 (Mauser, 1996). The 10 percent of users with more than 200 visits accounted for 42 percent of total home health spending (Figure 9).

Figure 9. Medicare Home Health Spending, by Number of Visits: 1994.

It is difficult to definitively characterize the role of Medicare's home health benefit in LTC. Treatment of acute as well as chronic conditions may require substantial visits. Furthermore, variation in the volume of home health visits across regions and between proprietary and non-profit agencies has raised questions about the role of provider efforts to generate revenues, rather than beneficiary needs, as a significant contributor to increasing service (U.S. General Accounting Office, 1996). Nevertheless, it seems undeniable that the home health benefit's increased importance to a portion of the population needing LTC is significantly intertwined with increases in program spending.

Medicare's SNF Benefit

Like Medicare's home health benefit, its benefit for nursing home care has grown and changed in recent years. In contrast to home health, however, that change does not appear to be attributable to a shift toward LTC. On the contrary, it seems to be related to increased reliance on nursing home care as an alternative to hospitalization.

Medicare covers SNF care for patients who have been in the hospital for at least 3 days, are admitted within 30 days of that stay, and require skilled nursing or rehabilitative therapy on a daily basis. Coverage is limited to 100 days, and copayments (set equal to one-eighth the hospital deductible and amounting to $92 in 1996) apply after the 20th day. This benefit, like the original home health benefit, was included in Medicare legislation to facilitate hospital discharge.

Experience with the SNF benefit paralleled experience with home health. Before 1970, an administrative response to substantial and unanticipated claims experience dampened both participation and coverage (Feder, 1977; Moon et al, 1995). In contrast to home health care, however, Congress did not take other action to promote SNF use. It was Medicaid, not Medicare, that became the primary supporter of LTC in nursing homes during the following decade.

As with home health care, enforcement of coverage criteria plays a critical role in determining Medicare coverage of nursing home care. For the SNF benefit, the primary issue has been the definition of “skilled care.” Until the late 1980s, coverage guidelines narrowly defined what constituted skilled care, placing heavy emphasis on the provision of specific treatments and on patients' conditions—instability, high probability of complications, or (for therapy services) the presence of “rehabilitation potential” (Smits et al., 1982). Interpretation of these guidelines varied considerably across regions and intermediaries. Coverage determinations were made only after care was delivered and claims filed, putting nursing homes at financial risk for submitting uncovered claims. Finally, Medicare payment rates reflected nursing homes' average cost per patient. To the extent that these patients required greater-than-average staff time and care, homes were reluctant or unequipped to admit them. Given restricted coverage and financial risk, Medicare played a limited role in nursing home coverage. Homes' willingness to participate and to accept Medicare patients depended heavily on whether the bulk of their patients, Medicaid or private, required similarly intensive care (Feder and Scanlon. 1982).

In April 1988, Medicare altered its guidelines for covering SNF care, broadening the terms under which care would be considered “skilled” and providing clarifications that reduced the uncertainty and variability of coverage. Three months later, the Medicare Catastrophic Coverage Act became law, further expanding the scope of Medicare's coverage by eliminating the requirement for prior hospitalization, lengthening maximum coverage from 100 to 150 days, and rearranging and reducing cost-sharing requirements. The number of persons receiving Medicare SNF benefits increased by almost 50 percent, and the number of Medicare-covered SNF days nearly tripled between 1988 and 1989 (Moon, 1993).

Although Congress repealed the Medicare Catastrophic Coverage Act only a year after its passage, Medicare coverage for SNF care has not declined. Medicare expenditures for SNF care have increased from $3.5 billion in 1989 to $8.3 billion in 1994, at an annual growth rate of nearly 20 percent (Figure 10) (Prospective Payment Assessment Commission, 1996). However, in contrast to patterns for home health care, greater spending has little to do with long stays reflecting a shift to LTC. Although the number of days per person served averaged 39.9 days in 1994, up from 27.8 days in 1988, the number of days per person has changed little since 1990, and the average Medicare SNF stay is only about 15 percent of the average length of stay (263 days) for a Medicaid nursing home resident (American Health Care Association, 1995). Rather than longer stays, the primary factors driving continued expenditure increases are increased numbers of users and, more notably, increased expenditures per person using the benefit (Prospective Payment Assessment Commission, 1995).

Figure 10. Medicare Skilled Nursing Facility Payments.

This increased use and cost of Medicare's SNF benefit appears to have much to do with systemwide changes in service delivery associated with managed care and with Medicare-specific payment policies (Lewin-VHI, Inc., 1995). Nursing homes, along with other providers, have become attractive sources of relatively intensive, acute-related care for managed care plans seeking lower cost alternatives to hospital stays. Legislation enacted in 1987 required Medicaid nursing homes to meet more demanding staffing and other quality standards, increasing homes' readiness to provide more intensive care. Nursing homes providing subacute care to private patients are more equipped to serve similarly ill Medicare patients. At the same time, Medicare's prospective payment policies have encouraged early discharges from hospitals. Equally important, Medicare SNF payment policies have accommodated increasingly intensive service provision, through exceptions processes and the absence of limits on specialized (ancillary) services.

The increased demand for and profitability of service to Medicare patients needing intensive service have affected both providers' willingness to offer SNF services to Medicare patients and the kinds of patients they are serving. From 1986 to 1994, the number of Medicare-certified SNFs increased substantially; the number of freestanding SNFs grew 29 percent and the number of hospital-based SNFs almost tripled (Prospective Payment Assessment Commission, 1996). Between 1988 and 1993, ancillary charges for SNF admissions grew from $772 million to $4.9 billion (Lewin-VHI, Inc., 1995).

These patterns suggest that increased Medicare SNF spending has more to do with a change in the locus of acute care services than with the provision of long-term nursing home care. Although some recipients of the SNF benefit may become long-term nursing home residents, Medicare covers primarily short-term, acute-related rather than long-term custodial care. Medicare's share of nursing home revenues has grown as its spending has risen (from 1.4 percent in 1985 to 8.2 percent in 1994), but it is Medicaid that still finances the bulk (47.4 percent) of nursing home care (Levit et al., 1996).

Medicare and the Future of LTC

Pressure to control growth in the Federal budget and to bring Medicare spending more closely in line with its revenues is eliciting proposals for fundamental changes to the Medicare program. The need to control spending, both in general and on Medicare's postacute services in particular, is understandable. However, the importance of cost containment should not obscure the importance of Medicare protections to people who need LTC. This population's need for intensive and expensive service puts them at considerable risk in any aggressive effort to control Medicare costs.

Efforts under consideration include restructuring Medicare to enhance competition among private health plans for enrollment of beneficiaries. Competition that is already occurring reveals a significant problem: the tendency to segregate the sick from the healthy. The more aggressive the competition, the more likely there is to be discrimination against people, such as the severely impaired, who incur high Medicare costs. The result may be segregation of sicker people in higher cost, lower quality plans or inadequate provision of necessary services, regardless of the plans these people are in.

Restrictions on provider payment may also reduce access or quality for the LTC population. Of particular concern in this regard are potential changes in policy toward home health and SNF benefits. Growth in expenditures on these services has appropriately called attention to inefficiencies in Medicare's payment policies. Balanced-budget proposals have included initiatives that would replace cost-based reimbursement with prospective payments that would limit payments per day, per visit, per episode of care, or per beneficiary. Variation in proposals reflects, in part, considerable uncertainty as to how to establish limits that slow spending growth while reflecting differences in patient need and not unduly restricting appropriate care. Unless carefully designed, proposals to limit payment may penalize providers serving patients who need more intensive home health or SNF care, reducing service for those patients most in need.

Finally, changes in Medicare policy do not occur in a vacuum. For the LTC population, the interaction of Medicare with Medicaid policy is critical to securing adequate care. Medicaid as well as Medicare restructuring has become a major policy issue. If Medicare limits its longer term home health care, Medicaid may not pick up the slack. If Medicare policies increase the demand for and profitability of short-term nursing home care, long-term Medicaid patients may face difficulties finding nursing home beds. In general, if significant restrictions on Medicaid funding accompany changes in Medicare, people who count on both programs could find both their health care and LTC in considerable jeopardy.

Over the last 30 years, Medicare has provided health insurance for, and more recently, has contributed some LTC protection to, people who need LTC. In the next 30 years, the number of people who need these protections will grow substantially. As we pursue policy changes to prepare for the 21st century, we must make certain our action strengthens, rather than undermines, the fundamental support Medicare (along with Medicaid) now provides.