Integrating Acute and Long-Term Care for the Elderly

Integrating Acute and Long-Term Care for the Elderly By Joseph Alper and Rosemary Gibson

One of the principles that help explain the logic-or lack of logic-in the health care system is this: financing drives the organization of services. This principle is especially important for frail or physically disabled people who depend largely on government to finance their health care. Two distinct financing-and, therefore, delivery-systems that do not coordinate with each other have developed: the first system, Medicare, the federal government program for people over 65 and severely disabled individuals, pays for acute medical care. Except for stays immediately following hospitalization, it does not pay for nursing home care; nor does it cover supportive services, such as help in cleaning and cooking. The second system, Medicaid, a federal and state program for poor people, does pay for nursing home care and, in some states, supportive services. The two systems are cumbersome and difficult to navigate, particularly for poor, elderly people who are eligible for both Medicare and Medicaid.

A challenge for those trying to improve the care of poor older people is making the financing and delivery of services more rational and understandable. The challenge has been picked up by The Robert Wood Johnson Foundation, whose interest flows naturally from one of its three goals: to improve care and support for people with chronic health conditions. The Foundation's grant making has focused on improving acute, chronic, and supportive care for people with chronic conditions. However, improvements can happen only when financial incentives for providers are changed and when new models of organizing care emerge.

This chapter by Joseph Alper and Rosemary Gibson chronicles the Foundation's efforts to develop and support programs that integrate the financing and delivery of long-term care services for poor elderly and disabled people. The Foundation's grantees have devised strategies-many of them based on managed care-from which a number of creative models have emerged. But many older people, when given a choice, have rejected managed care approaches and the loss of independence they imply. The goal of developing coordinated services accessible to the majority of those with disabling conditions remains unmet.

Joseph Alper is a free-lance writer specializing in health care. He contributed a chapter to last year's Anthology on assisted living for elderly people. Rosemary Gibson, a senior program officer at The Robert Wood Johnson Foundation, has played a key role in the Foundation's chronic care grant making. She took the lead in designing a national program to improve services for the developmentally disabled, and has developed a set of national programs to improve the quality of care for terminally ill people.

At the start of the twenty-first century, many Americans face difficulties in trying to get the best care from a fragmented health care system that is often hard to navigate. This is particularly true for people with multiple chronic conditions that require them to see different physicians and other health care professionals in geographically dispersed settings. The challenge is especially burdensome for the nation's elderly, since they bear a disproportionate number of chronic conditions.

But as big a problem as this is today, it is poised to become far worse. With the aging of the American population, the need for medical care in all its forms-acute and long-term-is set to rise dramatically. With that rise, the health care system is bound to become even more difficult to navigate, particularly for the frail elderly.

Why is the health care system in the United States so difficult to navigate? One reason is that for the most part, the nation's acute and long-term care systems function separately, with little integration between them. Acute care is provided by physicians in hospitals and doctors' offices; its chief aim is to diagnose and treat medical conditions and to perform periodic activities to maintain health and prevent illness. For the nation's elderly, most acute care involves managing episodic flare-ups of problems related to long-term chronic conditions such as heart disease and diabetes.

Long-term care is assistance given over a sustained period to people who have long-term difficulties in functioning because of a chronic condition or disability. Long-term care can be community-based- in the home or in an adult day services program, for instance-or it can be in an institutional setting such as a nursing home or an assisted-living facility. Provided mainly by nurses and nurses' aides, it consists of two elements: management of chronic conditions and supportive (or custodial) services such as assistance with bathing, dressing, and cooking. Ideally, acute and long-term care systems should work together to ensure that a patient's entire health and daily living needs are met. Nursing home residents are likely to have more severe chronic illnesses than other older persons. This is all the more reason they need care to manage their existing chronic conditions and to prevent new ones. Yet many nursing home residents do not get the kind of care that promotes good health and looks after their chronic conditions. As a result, they are often sent to hospital emergency rooms for treatment of an acute illness that might otherwise have been avoided by better management of an underlying chronic condition.

An important reason that acute and long-term care are not effectively integrated is that these services often are paid for by different government programs, each of which has its own purposes and rules. Medicare provides health insurance coverage for people 65 and older and for those with long-term disabilities, irrespective of income. Administered by the federal government, it covers acute medical services, especially physician and hospital care. Nursing home care is available only for a limited duration for patients who need skilled care after having been hospitalized-recovering from a hip fracture, for example. Medicare does not cover nursing home care for elderly people whose primary needs are custodial-those, for example, who require help with daily activities such as bathing and meal preparation. Similarly, although Medicare's home health coverage was expanded in 1990, it is still relatively limited. In short, Medicare is not an insurance program for those who need long-term care.

In contrast, Medicaid is a state-administered program, funded jointly by the federal government and the states, that provides health insurance coverage for the poor. Medicaid pays for physician's services and hospital care and, in addition, covers the cost of nursing home care, including custodial care, for low-income seniors. In many states, Medicaid funds some community-based care.

Although integration of acute and long-term care services is important for all seniors, there is a subset-poor elderly people-who could benefit from making both the delivery and the financing of acute and long-term care more rational. They are the so-called "dual eligibles" who qualify for Medicare because they are over age 65 and for Medicaid because of their low income. Dual-eligible seniors often have complex chronic care needs. They often need long-term care-both to manage their chronic conditions and to provide supportive services-and have frequent contact with the acute care system.

In 1997, the number of dual-eligible seniors was estimated to be 6.7 million-they constituted 17 percent of all Medicare beneficiaries but absorbed 28 percent of Medicare expenditures.¹ Dual-eligible beneficiaries use more Medicare services across the board-hospital, skilled nursing facility, and home health care. Of these dual eligibles, 33 percent use emergency rooms in a given year, compared with 18 percent of Medicare-only beneficiaries. On the Medicaid side, dual eligibles represented 19 percent of all beneficiaries in 1997 but accounted for 35 percent of all spending.2

The Health Care Financing Administration, or HFCA, which oversees both the Medicare and Medicaid programs, has allowed states to develop projects to explore new ways of integrating Medicare's acute benefits with Medicaid's long-term care benefits to provide a more seamless system of care. HCFA does not provide startup funds for the projects, but does enable them to bill Medicare and Medicaid for the services they provide.

What would an integrated system of care mean for frail elderly people? Integration means that multiple systems act and feel as a seamless one.³ For those residing in nursing homes, integration means having good skin care to prevent bedsores and good primary care to prevent pneumonia from developing among the most frail group of elderly patients. For seniors living in the community, integrated care might mean a physician arranging for a home visit from a social worker to identify and address risk factors that could cause a fall and a broken hip.

Since the early 1980s, The Robert Wood Johnson Foundation has nurtured projects aimed at integrating acute and long-term care. One of the first of those, the Program of All-inclusive Care for the Elderly, or PACE, is an attempt to develop models of a seamless service delivery system for very frail seniors.

A fully integrated model of acute and long-term care was developed by On Lok Senior Health Services in San Francisco in the late 1970s. By 1983, On Lok's model had become a federal demonstration project, and this status allowed it to receive financing from Medicare and Medicaid that was placed in a single pool. To build on the On Lok model, in 1988, The Robert Wood Johnson Foundation provided $5.8 million for the Program of All-Inclusive Care for the Elderly. The four-year program supported the replication of the On-Lok model of integrated care for poor seniors at six sites (Boston, Massachusetts; the Bronx, New York; Columbia, South Carolina; El Paso, Texas; Milwaukee, Wisconsin; and Portland, Oregon).

The PACE model is characterized by a focus on the significantly impaired frail elderly; a team-managed approach to care that integrates a comprehensive package of acute and long-term health services in both inpatient and outpatient settings; and capitated, or fixed-rate per patient, financing from both Medicare and Medicaid. Like On Lok, PACE provides a viable alternative to nursing home care for very frail seniors.

To enroll in PACE, a person must be 55 or older, need nursing home care based on the state's standards, and reside at home in the program's defined geographic area. An interdisciplinary team based at the local PACE Center, which also provides adult day care, orchestrates all care. Though enrollees receive routine care at the PACE Center or at home, specialty and inpatient care are readily available whenever needed.

"What we've done with PACE is to completely integrate services and financing for both acute and long-term care, and by doing so we've created a fully integrated managed care system for the frail elderly," says Jennie Chin Hansen, On Lok's executive director and a PACE board member. "Our overall goal, which I think PACE achieves, is to maximize each enrollee's autonomy and continued community residence and, at the same time, provide quality care at lower cost to Medicare and Medicaid."

Hansen explains that under terms negotiated with HCFA, a PACE Center receives monthly capitation payments from Medicare and Medicaid, as well as payments from individuals who are financially ineligible for Medicaid and who pay that portion of the monthly fee themselves. Medicare bases its payment on a methodology, developed to reimburse Medicare health maintenance organizations, or HMOs, that accounts for regional disparities in the cost of medical care and that is then adjusted to reflect the frailty of PACE enrollees compared to the average Medicare beneficiary. In 1998, the monthly Medicare capitation payment to PACE programs ranged from $877 to $1,775. PACE Medicaid payments are based on state-specific estimates of the fee-for-service costs for comparable long-term-care populations. In 1998, Medicaid monthly capitation rates ranged from $1,750 to $4,301.

"One key to the PACE model is that all payments are pooled by the PACE site and used for any program-eligible senior without regard to traditional restrictions, giving the site greater flexibility to deliver whatever services are needed," Hansen says. The allocation of those services, she explains, is determined through continual assessments of each patient's needs by the interdisciplinary team of primary care physician, nurse, social worker, rehabilitation therapist, home health workers, and others.

In the Balanced Budget Act of 1997, Congress authorized PACE to graduate from a federal demonstration program to a permanently approved Medicare provider, joining the ranks of the thousands of other accepted providers that receive regular payments from Medicare. In addition, the Balanced Budget Act made PACE a state option under the Medicaid regulations-each state can now add local PACE program to its list of permanent regular providers. The Act also increased the authorized number of PACE sites nationwide from 15 in 1997 to 40 in 1998, with an additional 20 added to this cap each year thereafter, enabling a systematic expansion of the model. Finally, the Act also allowed for up to 10 sites to be run by for-profit entities. This step of making PACE a permanent Medicare and Medicaid provider was based on these considerations:

In preparation for the growth spurred by the 1997 Balanced Budget Act, the National PACE Association, with funding from The Robert Wood Johnson Foundation, completed the development of standards and an accreditation review process in 1997. The National PACE Association is now conducting a pilot study, in collaboration with an accrediting body, to test the standards and make necessary modifications in preparation for formal accreditation. The accreditation process is a way to ensure that new PACE programs will remain faithful to the original PACE model.

As of February, 2000, about 6,000 frail seniors were enrolled at twenty-five PACE sites around the country, reflecting the growth of the program. Enrollment in PACE is modest in relation to the number of frail elderly who could benefit from integrated acute and long-term care. But the significance of PACE on health care for the elderly is larger than its numerical impact might imply. In the 1980s, before it became a government-approved demonstration site, PACE developed as a neighborhood program and not as a large-scale health plan. At the time, before the involvement of The Robert Wood Johnson Foundation, start-up capital was scarce, yet the program was still able to build itself into a working model. The principles of managed care and managing risk were being tried and tested, and there was little, if any, experience in the country in managing all the health care and supportive service needs of the frail elderly under a capitation payment. PACE tackled each of these problems.

The challenge that remains is to take the integration of acute and long-term care to scale. Because of its reliance on the PACE Center, an adult day services facility of sorts, the program may not work well in sparsely populated rural areas. And PACE, as it currently exists, is designed to treat groups of less than 200 adults. More and larger efforts are needed.

Minnesota Senior Health Options is an attempt to expand significantly the integration of services for frail elders who are eligible for both Medicare and Medicaid. With the help of a Foundation grant, Minnesota became the first state to receive approval from HCFA for a statewide program to integrate acute and long-term care with funding from both Medicare and Medicaid. A program that has the potential to serve all dual eligibles in a state makes it possible to reach many more frail seniors, in contrast to demonstration projects at a single or only a few sites.

In 1997, Minnesota Senior Health Options, a five-year demonstration project, began enrolling its first participants in the seven-county metropolitan Minneapolis-St. Paul area. The central feature of the program is its reliance on Minnesota's extensive network of managed care providers and the fact that HMOs have a long history in the state and are widely accepted as providers of care by state residents. "That's one advantage we have here in Minnesota," says Pamela Parker, the program's director. "Minnesotans accept the role of managed care and have had good experiences with it over a long period of time."

Minnesota has contracted with three of the Minneapolis-St. Paul metropolitan area's five largest health plans to provide all Medicare and Medicaid services, including home and community-based services and 180 days of nursing home care. If a frail senior needs more than 180 days of care in a nursing home, the nursing home is reimbursed on a fee-for-service basis. Since 77 percent of enrollees reside in a nursing home, nursing home care is an important part of Minnesota Senior Health Options.

The program's home and community-based services include case management, companion services, caregiver training, extended home health aide, extended personal care assistance, adult foster care, adult day care, assisted living, residential services, homemaker services, home delivered meals, respite care, home modification, and supplies and equipment. Each senior has a "care coordinator" to assist with care planning and service access.

As of January 2000, enrollment in the program stood at 3,420, or about 25 percent of the dual-eligible population in the seven-county metropolitan area. Thirty-eight percent of enrollees are over age 85. In a survey conducted at that time, half of the Options program enrollees noticed a marked improvement in services when they enrolled, and 70 percent said their transition to the new program was smooth. "To me, it was like a gift from God," said Mae, a sunny woman in her mid-60s who participates in Senior Options. Another enrollee, Marge, said, "Now I see a doctor every six months, and if I don't call to make an appointment, someone calls me. I think that's good." For Sarah, who couldn't get any services before enrolling in the Minnesota program, the program has been a lifesaver: "They take very good care of me now. My coordinator makes sure I get the care I need when I need it," And for Ralph, "It has given me peace of mind. I think that's what it boils down to. I didn't know what would happen to me. This was a gift."

"Overall, we feel that we've been pretty successful so far, and it's been pretty smooth running on a day-to-day basis," Parker says. "The place where we are disappointed is in how far we've been able to take the program. We've pushed the marketplace forward toward being more focused on the elderly, but reimbursement issues and fear of risk under a capitated payment system have kept two of the state's big plans from joining the program. We're trying to work with HCFA to solve that, but for now it's a sticking point."

Building on the experience of Minnesota Senior Health Options, in 1996 The Robert Wood Johnson Foundation established the Medicare-Medicaid Integration Program, an $8 million grant initiative. Under this program, seven states and one regional organization have received grants up to $300,000 to test ways to integrate acute and long-term care. To date, the Foundation has awarded grants to Colorado, Florida, New York, Oregon, Texas, Virginia, Washington, and the New England Consortium (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont). Each grantee is trying a different approach to integrating acute and long-term care, depending on the target population-urban or rural, for example. And though many of the states have found it difficult to develop a program and are only in the early stages of doing so, states such as Texas have been more successful in enrolling seniors. "It's been a trying experience for many states, and each has had to find its way, but we are seeing results now and westarting to learn some important lessons," say Mark Meiners, the national program director at the University of Maryland Center on Aging. The Texas STAR+PLUS project and the Colorado Integrated Care and Financing Project are two examples of efforts funded under the Medicare-Medicaid Integration Program.

The Texas project, STAR+PLUS has big ambitions. The entire Medi- caid population that needs acute and long-term care in Harris County, which includes Houston, is the target group for this pilot. In January, 1998, participation in STAR+PLUS became mandatory for some 55,000 Harris County residents. Half of these 55,000 residents are over age 65 and also eligible for Medicare; the others are under age 65 and eligible for Medicaid only. STAR+PLUS is different from the other projects, since it integrates acute and long-term care only under Medicaid. It doesn't address the integration of Medicare financing or delivery systems.

STAR+PLUS contracts with three HMOs, which are responsible for providing acute and long-term care, including nursing home care and expanded home and community-based services. "We felt that this gives the HMOs an incentive to provide innovative, cost-effective care from the outset in order to prevent or delay the need for more costly institutionalization," says Pamela Coleman, director of STAR+PLUS. HMOs are required to conduct a health assessment for each new enrollee within 30 days of enrollment. For the 1,500 nursing home residents who are enrolled, the nursing home staff provides the care in the nursing homes and the HMOs make monthly visits to check up on them.

In addition, the state requires the HMOs to provide each enrollee with a care coordinator who is responsible for ensuring that the patient receives integrated acute and long-term care. Care coordinators make home visits and identify unmet needs, some of which are not the usual long-term care needs but are nonetheless important. For example, some care coordinators authorize pest control in unhealthy home environments and others install smoke alarms.

The main challenge has been the less-than-enthusiastic response from long-term care providers. "This has been a wakeup call for them," Coleman says. With a low nursing home occupancy rate of 81 percent, the nursing homes seem worried that the community services being offered under the program might reduce the number of people requiring institutional care. The rates paid to the nursing homes are the same for now, but the HMOs that pay the nursing homes might change the methodology for determining the rates in the future.

The picture for STAR+PLUS in Texas over the long run is uncertain. State legislators placed a moratorium on any expansion of Medicaid managed care programs in the state. STAR+PLUS has gotten caught up in this moratorium.

Colorado's Integrated Care and Financing Project was developed as a pilot project to integrate Medicare and Medicaid financing for dual-eligible residents of Mesa County, a largely rural county in western Colorado. Rocky Mountain HMO, which has a long history of providing managed acute health care in the county's largest city, Grand Junction, was to work with Mesa County Options for Long-Term Care, part of the Department of Human Services, to provide integrated acute and long-term care for the target population.

As the first step in this demonstration, the Colorado Department of Health Care Policy and Financing applied for approval from HCFA in September, 1995, to integrate Medicare and Medicaid financing. Two years later, HCFA granted approval to the state to go ahead, but a stumbling block emerged-HCFA and the state were unable to agree on the amount of the Medicare capitation payment that HCFA should pay to the HMO. After a series of negotiations, Colorado decided to have the Medicare contract with the HMO stay as is and providers would be paid on a fee-for-service service basis. The experience illustrates some of the difficulty in agreeing on how much health plans should be paid to care for the dually eligible population.

While the state was working intensely with HCFA to get the Medicare financing ironed out, Rocky Mountain HMO withdrew from the program because of an unrelated issue resulting from internal operating changes within the HMO. Colorado will now attempt to develop a demonstration program in Denver under the Medicare-Medicaid Integration Program.

The Wisconsin Partnership Program, funded under a different Foundation-funded national program-Building Health Systems for People With Chronic Illness-represents yet another approach to integrating acute and long-term care. Like the PACE model, the Wisconsin Partnership uses an interdisciplinary team of a physician, a nurse, a nurse practitioner, daily living assistants, and a social worker or an independent living coordinator. The teams provide acute and long-term health care for seniors eligible for both Medicare and Medicaid in four demonstration communities around the state. In contrast to PACE, where enrollees are required to use the primary care physician from the PACE team, enrollees in the Wisconsin Partnership Program can choose a primary care physician from a network of participating physicians.

One of the sites, the Community Health Partnership of Eau Claire in east central Wisconsin is particularly interesting because it serves a mostly rural population of 180,000, spread across three counties encompassing 2,500 square miles; the only major city in the area, Eau Claire, has 12,000 residents. In contrast, Milwaukee (site of another of the state's demonstrations) contains just over 960,000 people within its 241-square-mile area. "We knew it would be a challenge to serve such a sparsely populated area," says Steve Landkamer, project manger of the Wisconsin Partnership Program. "But we also knew that with the large number of dual-eligible seniors living in rural Wisconsin we were going to have to face this problem eventually. So why not do it now, before we get steamrollered by the coming wave of baby boomers reaching retirement age?"

The job of setting up the Eau Claire site fell to Karen Hodgson, chief executive officer of Community Health Partnership. "We cheated a little in getting started by focusing our initial efforts on the area within a 25-mile radius of Eau Claire," Hodgson says. "But that way, we were able to get up to speed without stretching ourselves too thin."

The 187 seniors who were enrolled as of February, 2000, have "very high levels of health care needs, the frailest of our citizens," according to Hodgson. When someone enrolls in the program, the team identifies any immediate health care needs and addresses them. The medical director at Community Health Partners, Mark Deyo-Svendsen, notes, "About 25 percent of members meet diagnostic criteria for depression." This means that there is a lot of upfront work to deal with health care needs that have been neglected, often for a long time.

After that initial assessment, the team's focus turns to prevention, an approach that appears to be working, both for the seniors enrolled in the program and for Community Health Partnership's bottom line. The project staff notes that over the three years that the program has been operating, participants are spending fewer days in the hospital and the nursing home compared to the time before they joined the program. Meanwhile, visits to the primary care physician have increased, reflecting the emphasis on prevention. "At first, we were spending heavily on physician visits to get our seniors onto the prevention track, but that spending has paid off and we now expect to be running a surplus on our budget," Deyo-Svendsen says.

A key feature of the Eau Claire project is the emphasis on in-home prevention, with social workers making home visits at least once a month, and sometimes weekly. While in the home, the social worker can spot a senior who is struggling with alcoholism, which might be hindering management of the patient's diabetes. The social worker can arrange for community resources to help the elderly person cope with the addiction.

The daily living assistants who are members of the interdisciplinary team are, in essence, the first-line care providers. Some are certified as nursing assistants who provide personal care such as dressing, feeding, and toileting. Some daily living assistants help around the house with cleaning and grocery shopping. One of the biggest challenges in the program, is "finding people who are willing to work for $7 an hour and show up for work every day," according to Hodgson. "We really count on them, and most of them do this work because they love the people they are caring for."

Team members work hard to instill a sense of responsibility so that the seniors themselves ask for and receive the appropriate amount of care. "We let our members know that although they have access to more services than they were used to, this isn't a blank check to give everyone everything that they wanted," Hodgson says. This tough message evidently sits well with members-only four people have left the program in more than three years. "Sure, people grumble sometimes, but the bottom line is that they are healthier and they are getting far more care than they used to get under the old fragmented system."

Rationalizing fragmented and unresponsive health care delivery systems is a worthy but difficult endeavor. The number of people benefiting from integrated acute and long-term care is modest in relation to the millions of elderly people who have to fend for themselves in the fragmented health care systems that are the norm in this country. The PACE model has shown that it is possible to create a seamless system of care for a very vulnerable population of frail elders. PACE programs work well because they are small, and because dedicated professionals work together to create their own system of care to meet the needs of patients. In fact, the team approach is likely to be a key factor in making integration of acute and long-term care work. It can lead to better communication among multiple providers caring for the same person and to a more consistent strategy for caring for the patient.

In contrast to the team approach is an approach that simply tries to coordinate the care offered by the long-term and acute systems. While the team approach involves restructuring the system of delivering services, the coordination approach merely patches some of the holes in it. Nevertheless, since health care delivery systems are unlikely to undergo a major reconfiguration any time soon, coordination may be the most viable strategy to improve care on a large scale within the existing fragmented system. Care coordination will still have to juggle the opposing cultures of acute care-which exists to cure, treat, and sometimes prevent illness-and long-term care-which exists to make life better for people when there is no cure.

The Medicare-Medicaid Integration Program has stressed the importance of integrated financing between Medicare and Medicaid as key to making delivery systems work better on the ground. But making the financing of acute and long-term care more rational has proven to be challenging. It is, after all, an ambitious undertaking to get the nation's two major health care insurance programs, Medicare and Medicaid, to provide seamless financing. Yet integrated financing is not a panacea, and certainly it will not make much of a difference if integrated delivery systems are not in place. The original On Lok model of delivering services, which was replicated by PACE, grew from the ground up in San Francisco, based on what frail elders and their families needed. Then the integrated financing from Medicare and Medicaid was put in place to support the model. And while it is often the case that reimbursement procedures shape health care delivery systems, On Lok and PACE have shown us that it is possible for the delivery system to shape how care should be reimbursed.

Some state efforts to integrate acute and long-term care seemed to give priority to melding financing streams. While this is important, it's not clear that the states had health care delivery models ready that would integrate acute and long-term care. Wisconsin appears to be a step ahead; at the same time that it was revamping financing, its demonstration sites were creating new team approaches to care for frail elders. Other states relied on HMOs, nursing homes, home care providers, and others to render care in a more coordinated way. The changes needed in each of the delivery systems, and across those systems, to integrate care are not insignificant, and it's not evident that major changes in the delivery system were, in fact, made. Perhaps the attention given to the financing preempted attention to the development of models of delivering care. Developing integrated models of care needs to be a high priority and not simply follow the financing of services.

As health care providers and states attempt to develop models, component parts of integrated acute and long-term care can be distinguished: flexible benefit packages, a broad range of providers, mechanisms for effective care integration, and financing. They can be considered the building blocks, and some lessons can be drawn from the experience to date.⁸ These lessons include:

States and providers are taking some steps toward integrating acute and long-term care. In theory, at least, integration does offer the opportunity for improvement in health outcomes. But whether such outcomes will actually result from integration is a question worthy of research.

Large-scale efforts to integrate Medicare and Medicaid financing are unlikely in the foreseeable future. Instead of integration of the whole range of acute and long-term care for a large segment of the frail elderly, it is more likely that bits and pieces will be integrated. Housing, supportive services, and some medical care are being integrated in assisted living facilities. Advances in technology are making diagnostic testing-and better primary care-available while patients remain at home. These are small yet important steps that can be enormously helpful to patients and their families. While they will not reduce the fragmentation across the continuum of care, they are, at least, steps in the right direction.