Ambulatory and community-based services.

The shift in the site of service delivery from inpatient and institutional to ambulatory and community settings has been prompted by concerns over cost and the prospect for improving the quality of life. In response to these concerns, Medicare has implemented several demonstrations that emphasize ambulatory and community-based services. In this issue, articles are presented on four demonstrations, which focus on the extent to which coordinated care models reduce health care costs, and the cost effectiveness and beneficiary outcomes of disease-specific programs. Two articles are included on home health. One examines home health care in relation to the other Medicare post-acute benefits, and the other focuses on the use of home health care in the treatment of end stage renal disease (ESRD). Finally, two articles report on Section 1915c Medicaid home and community-based waiver programs.

Introduction

Ambulatory and community-based services increasingly have become important in the delivery of health care, as the reliance on inpatient services and institutionalization has diminished. This shift has been prompted by concerns over cost and the prospect for improving the quality of life, which may result from care delivered in less restrictive environments. In response to these concerns, Medicare has implemented several demonstrations that emphasize ambulatory and community-based services. Similarly, through the home and community-based care waiver provision of the Social Security Act, Medicaid has fostered access to a wide range of community-based services for low-income populations with severe disabilities who might otherwise be institutionalized.

In this issue, articles are presented that examine various issues related to ambulatory and community-based care—ranging from integrated and coordinated care models, which emphasize ambulatory rather than inpatient care, to a mainstay of community-based care—home health care. Included are program evaluation results of two coordinated care demonstrations: the Social Health Maintenance Organization (S/HMO) and the Consolidated Omnibus Reconciliation Act of 1985 (COBRA) Demonstration of preventive health services in a Medicare HMO. These evaluations focus on the extent to which coordinated care models reduce health care costs. Also presented are evaluations of the Alzheimer's disease and Medicare high cost case management demonstrations. In these articles, the cost effectiveness and beneficiary outcomes of disease-specific programs are compared with care provided through usual fee for service (FFS). Two articles involve research on home health care. One examines home health care in relation to the other Medicare post-acute benefits, and the other focuses on the use of home health care in the treatment of ESRD. Finally, two articles report on Section 1915c Medicaid home and community-based waiver programs. One provides the background, trends, and issues of these programs, and the other focuses on the patterns of home health care use by acquired immunodeficiency syndrome (AIDS) patients in New Jersey.

While these articles touch on a wide variety of ambulatory and community-based care issues from the perspective of different programs, some common themes can be found. Many of the issues discussed in these articles relate to effectiveness of case or care management strategies, more typically practiced in the context of capitated managed care, in reducing inpatient care and/or for allowing substitution of community-based treatment. Also examined are the broader issues of cost effectiveness relative to more traditional ways of providing care—either in inpatient settings, through traditional FFS, or in a standard HMO. Integrating care across the range of providers, and the role of financial incentives, are discussed by a number of these authors.

Demonstrations

Through its demonstration programs, the Medicare program tests alternative ways to provide care to its beneficiaries-including programs that emphasize coordination of care and preventative services as ways to reduce the need for inpatient care and institutionalization. The S/HMO concept, which became operational in 1985, demonstrates an integrated acute/long-term delivery approach. The first generation sites (S/HMO I) provided traditional Medicare acute care benefits, supplemental acute care benefits, such as prescription drug, and limited long-term care services (skilled nurse facility [SNF] services and homemaker, personal health aide, medical transportation, adult day health care, respite care, and case management). Dowd, Hillson, VonSternberg, and Fischer compare the 1990 clinical expenditures of a S/HMO I site with a Tax Equity and Fiscal Responsibility Act (TEFRA)-risk Medicare HMO operated by the same parent HMO. This research found the S/HMO I site clinical expenditures were 18 to 19 percent higher than the TEFRA HMO enrollees (excluding expenditures on services covered only by the S/HMO) after controlling for self-reported functional status, general health status, and chronic conditions. The authors suggest that the higher costs and inadequate capitated revenue led to the decision for this S/HMO I site to withdraw from this Medicare demonstration in 1994.

In another demonstration project, Medicare tested managed care innovations as a possible way to benefit the health and well-being of average-risk beneficiaries. Under the COBRA demonstration, a preventive health services package was added to an existing Medicare HMO plan. Although preventive care is considered a hallmark of managed care, the purpose of the demonstration was to test a somewhat more purposeful and coordinated set of services than are customarily associated with managed care, including health risk assessment, counseling, and classes (e.g., exercise). There also was an attempt to integrate the package with the beneficiaries' regular physician and nurse care within the HMO. In their evaluation of the project, Patrick, Grembowski, Durham, Beresford, Diehr, Ehreth, Hecht, Picciano, and Beery considered whether elderly beneficiaries responded to the preventive approach by adopting lifestyle changes and experienced better health outcomes and lower utilization and costs.

The results indicated short-term benefits in health behavior changes and self-ratings of health and well-being, but little of the impact persisted by the 4th year of observation. No significant cost and utilization differences between the treatment and control groups were found. A higher mortality risk in the treatment group 75 years of age or over was observed. This could have been due to chance, some unmeasured effect of the program, or possibly from an increase in advance directives, attributable to add-on counseling services that covered advance directives. In light of the mortality differences and the general health decline of persons 75 years of age or over, the authors concluded that a uniform package of preventive services may not be efficacious for older beneficiaries. To design a preventive package for Medicare, more research is needed to target specific preventive services appropriately for various subgroups of beneficiaries.

Other Medicare demonstrations also attempted to adapt techniques from traditional managed care, though these projects focused specifically on high cost patients. The Medicare Alzheimer's disease demonstration, which was initially mandated under OBRA 1986, was designed to: (1) test the cost effectiveness and impact on health status and functioning by providing an expanded benefit package of home and community-based services to the Alzheimer patients, and (2) to determine if these services reduced the need for institutionalization. Traditional Medicare benefits were expanded to include adult day care, companion services, family counseling, education of caregivers, and homemaker/personal care services. Two case management models were used to coordinate and authorize care obtained through the expanded benefit package. Beneficiaries were liable for copayments up to a maximum amount per month, at which the benefit level was capped. In one model, case management techniques were less intense (a case manager-to-client ratio of 1:100), but the monthly benefit cap was lower. In the second model, case management techniques were more intense (a casemanager-to-client ratio of 1:30), but the benefit cap was higher. This dichotomy would help test the influence of copayments on benefit use, as well as the intensity of case management.

Newcomer, Miller, Clay, and Fox evaluate the Medicare Alzheimer's disease demonstration for HCFA. Their primary finding is that the added cost of case management did not offset the savings generated from lower clinical care expenses. This is somewhat surprising, since managed care techniques have been assumed to reduce total health care costs. The authors discuss potential reasons for this finding, as well as policy implications and research designs for future demonstrations. It is postulated that the effectiveness of case management may be a longer-term effect, which may increase clinical costs in the short run. One design limitation was that high-risk cases were not jointly identified in an integrated approach by providers and case managers, which may have reduced the effectiveness of the targeting function.

In a related article, Fox, Maslow, and Zhang examine the notion of applying medically oriented functional impairment criteria to people with Alzheimer's disease to determine eligibility to long-term care. Using data from the Alzheimer's disease demonstration, the authors conclude that the five basic activities of daily living (ADLs) scores understate behavioral and cognitive status. Consequently, additional information, such as behavioral and mental status tests, should be considered in determining the long-term care (LTC) eligibility for demented patients.

A second demonstration that focused on high-cost patients was the Medicare high cost case management demonstration, which was launched under an OBRA 1990 mandate. The demonstration was targeted to FFS beneficiaries with catastrophic illnesses (e.g., congestive heart failure, chronic obstructive pulmonary disease, and high medical costs). It was hoped that case management could serve as an investment in prevention; that is, through reducing relapses and sequelae of serious illness, the high utilization associated with catastrophic illness would be reduced. Case manager activities include patient education in self-care and prevention, increased monitoring, and co-ordination of community services such as medical social work. In three different experiments with three different sponsorships, approaches varied in terms of training case managers (usually nurses), structuring their activities, making use of patient education, and services coordination.

Schore, Brown, and Cheh evaluate the Medicare high cost case management demonstration and find no evidence that case management produced cost savings. According to the authors, an important difference between successful case management experiences reported in the literature and the demonstration outcomes concerned the role of the physician. The FFS physician and case manager were not teamed, and the physicians usually continued their normal activities without coordinating with the case manager. The authors also conclude that the case managers were not adequately prepared in the clinical aspects of their patients' conditions and in the generic techniques required to manage patients in the community. Finally, the demonstration designs incorporate no financial incentives to achieve the hypothesized outcomes of case management.

Home Health Care

Post-acute care (PAC) services are the fastest growing component of the Medicare program. Home health and covered SNF services increased from 6 percent of program expenditures in 1989 to 15 percent in 1996. The number of Medicare beneficiaries using home health nearly doubled to 3.7 million while the volume of services substantially increased as more visits per user were delivered. Part of this growth is likely related to Medicare policies that prompted a shift from expensive (i.e., inpatient hospital) to less expensive care settings.

Recent policy developments portend changes in the current relationships among acute-care inpatient and PAC treatment settings. Medicare PAC services are comprised of SNF care, rehabilitation facility care, and home health care. It is generally recognized that the home health benefit not only provides PAC, but also serves as a source of long-term community based care for some users (Feder and Lambrew, 1996). The Balanced Budget Act of 1997 (BBA) mandated major payment reforms affecting all three settings, including implementation of prospective payment systems (PPSs).

Gage focuses particularly on the impact of the BBA on access to PAC services. In her article, she analyzes the variation in PAC modalities for selected diagnosis-related groups. Building upon the work of others, Gage builds a statistical model using 1995 patient level data to predict the propensity to use PAC modalities for selected medical conditions controlling for age, sex, Medicaid eligibility, race, geographic region, etc. Using these results, Gage postulates that, depending upon the incentives built in the PPSs, substitution among PAC modalities may occur and the growth in PAC may be restrained.

In another article looking at access to home health care, Kauf and Shih find that the use of home health services by ESRD patients is more likely for patients who are elderly (over 85 years of age), female, white, switched dialysis modalities, and having diabetes. If the BBA mandated PPS results in a reduction in the number of home health care providers, access issues may evolve for this high user group. Because of this potential, the authors recommend monitoring the use of home health care services for ESRD patients so that the necessary information will be available to policymakers.

Home and Community Based Waivers

Over the past 25 years, Medicaid program changes have given States flexibility to design optional home and community-based benefits. These changes have been termed “the primary engine” for a movement to community-based living that places high value on independence and autonomy, in contrast to the social solutions of a generation ago (Master and Taniguchi, 1996). Miller, Ramsland, and Harrington trace the recent growth in Section 1915c Medicaid waiver programs for home and community-based care. The programs are targeted at Medicaid sub-populations who might otherwise be institutionalized and persons with disabilities. Reflecting the larger social backdrop, as well as budget pressures, some innovations use managed care approaches as well as non-traditional care settings.

The States have established more than 200 waiver programs through 1997 targeted at six specific populations, such as Medicaid beneficiaries with AIDS and persons with serious mental illness. The variation in the number and types of waiver programs, the State-specific proportion of LTC Medicaid expenditures allocated to the waivers, and the growth in the waiver programs' share of LTC spending since 1990 are documented in this study. The service composition of waiver programs, the spending by service category, and the services offered and expenditures associated with various eligibility categories also exhibit great variation. These results highlight the geographic inequities in access to comprehensive community services. With certain conditions, such as the severely mentally ill, for whom only four States have targeted programs, the variations in State decisionmaking accentuate inequities in access.

The authors suggest that research is needed to better understand factors in State decisionmaking that are ultimately responsible for the patchwork of resources offered to individuals who would otherwise depend on institutional care. The trend and composition data also highlight opportunities for the next generation of research into waiver program effectiveness. Better research may now be possible through improved characterization and measurement of the intervention specific to each waiver program. Along with isolating the intervention, the authors call for research based on a clearer definition of effectiveness goals sought by patients, their advocates, and policymakers.

Sambamoorthi, Collins, Crystal, and Walkup examine the impact of one specific waiver program—New Jersey's program of home and community-based care for AIDS patients. Medicaid is the single largest financing source for AIDS patient care, and inpatient care accounts for the bulk of AIDS-related medical care expenditures. There is some evidence in the literature that home care for persons with AIDS (PWAs) can prevent inpatient stays, which suggests that home care provision can save Medicaid money while offering improved quality of life for AIDS patients. Thus, there is strong policy interest in containing costs and enhancing well-being by substituting home care for inpatient care.

The authors find that in New Jersey, only about one-third of PWAs on Medicaid used home care—either from nurses or para-professionals such as home health aides and personal care assistants. However, persons enrolled in the State's specialized waiver program for PWAs were more than three times as likely to use home health services than persons not enrolled, and average monthly expenditures on these services among users were nearly twice as high in the waiver program. Controlled analysis of utilization and cost suggested that institutional care was avoided under the waiver program, although overall expenditures were not reduced. Another impact of the program is that home care utilization differences by such characteristics as race, AIDS-risk group, and geographic location that normally arise in New Jersey's Medicaid program were mitigated among waiver participants. While this suggests that, once enrolled in the waiver program, some types of patients may experience fewer access inequities than might otherwise be expected, other data for one important risk group—intravenous drug users—suggests relatively low home care expenditures per user, regardless of participation in the waiver program. Thus the need continues for the study of access barriers of different AIDS subgroups.

Conclusion

The articles in this issue elucidate several issues related to ambulatory and community-based care. Utilizing case management for special populations was not found cost effective in several HCFA demonstrations. While PAC services continue to grow, the impact of BBA provisions on the availability and substitution of future PAC services is questioned. Although home and community-based waivers have proliferated, research is recommended on factors that are needed to make decisions at the State level. These findings and recommendations add to our knowledge of ambulatory and community care delivery. With the changing demographic patterns over the next twenty years, health care policymakers should ask how best to deliver this care and how it will interact with other modalities.