No one measure adequately captures LTC system performance in terms of the breadth of the population covered, choice of services, and breadth and intensity of services provided. For this reason, we used a combination of measures to capture variation in system performance across states, including:
- Percentage of Medicaid LTC expenditures allocated to HCBS;
- Percentage of LTC users receiving HCBS;
- Ratio of per-recipient spending on HCBS to spending on institutional care; and
- Percentage of people potentially eligible for Medicaid LTC who used Medicaid HCBS.
The first two measures are commonly agreed-upon indicators of the degree to which states have balanced their LTC systems toward HCBS use and spending relative to institutional care. However, available data sources (e.g., state federal financial participation claims on the CMS "64" forms) have provided data only on spending. These data are available for all states through 2009. Our use of 2006 MAX data makes it possible to obtain (albeit for 37 rather than all 51 states, including the District of Columbia) not only spending data but unduplicated counts of service users and to develop additional measures that require data on expenditures for services, numbers of service users, and age and other characteristics of users of various types of LTC services.
The third measure derived from the MAX files captures the extent to which state spending on HCBS per user is similar to state spending on institutional care. Higher ratios of per-user spending on HCBS relative to per-user spending on institutional care are not necessarily indicative of "better" performance; rather, this measure provides additional insight into how a state is allocating financial resources between HCBS and institutional care. Medicaid law pertaining to HCBS waivers requires such services to be "cost effective" relative to institutional care and defines "cost effectiveness" as spending per HCBS waiver participant that on average does not exceed per-user spending on institutional care that waiver participants would otherwise require. Medicaid regulations further specify that states must spend up to the average amount spent per-capita on institutional care if the state's individualized needs assessment process determines that an HCBS waiver eligible individual requires that level of covered services. Medicaid law and regulations pertaining to other HCBS benefits neither require states to cap individual expenditures relative to institutional costs nor require states to cover the costs of all services assessed as necessary for a given individual (in other words, states may set lower coverage limits so long as these limits apply equally to all beneficiaries).
Some advocates of "re-balancing" argue that public program participants residing in the community who qualify for institutional coverage should be entitled to receive HCBS costing at least as much as what Medicaid would spend on their institutional care -- if a professional assessment indicates that they need care costing that much. Others point out, however, that Medicaid institutional care reimbursements cover room and board costs that account for at least one third of the total and that Medicaid law prohibits HCBS benefits to cover any room and board costs. By this standard, average per-capita HCBS spending per-capita ought not to exceed two thirds of average institutional care spending. At the same time, the Medicaid spending per-user statistics do not reflect the full cost of LTC services -- especially institutional care -- because beneficiaries with Social Security pension/disability benefits or other personal income insofar as Medicaid beneficiaries are required to contribute all but a small personal needs allowance to pay for care and Medicaid. Post-eligibility financial contributions toward the cost of nursing home care are substantially greater than for HCBS or for care in ICFs/IID. Thus, rather than making judgments of state performance based on this indicator, we simply examined whether higher ratios of HCBS to institutional spending per LTC user correlate with the agreed-upon indicators of "re-balancing" toward greater reliance on HCBS.
Many individuals who need human assistance with personal care (basic activities of daily living [ADLs] such as bathing, dressing, transferring from bed to chair, moving from room to room, toileting, and eating), and who have a level of need for such assistance similar to that of nursing home residents, nevertheless continue to reside in the community and to rely exclusively on unpaid help from family, friends, and neighbors. In many cases, such persons are not financially eligible for Medicaid and cannot afford to pay out-of-pocket for personal care services. In principle, low-income persons with personal care needs would be expected to seek Medicaid-covered HCBS rather than rely exclusively on informal help which could impose considerable burden on family members. However, a number of factors may prevent low-income individuals with personal care needs from accessing Medicaid-covered HCBS. These factors include being income-qualified but having assets in excess of the Medicaid allowable level, limits that the state has set on the numbers of qualifying Medicaid beneficiaries who can be served under HCBS waivers, and inadequate supply of HCBS providers to meet demand. To measure the extent to which state residents who potentially qualify for HCBS based on their level of income and need for assistance with personal care who are actually receiving Medicaid HCBS, we report the ratio of HCBS users to the numbers of low-income state residents in need of human assistance with personal care tasks as reported in the ACS. This fourth measure provides policy context for other performance indicators and is assessed in our subgroup analyses.
We examined two additional measures and included them in the Appendix D summary tables:
Share of total Medicaid expenditures for LTC users spent on enrollees using HCBS.7
Percentage of nursing home and ICFS/IID residents who used HCBS prior to their spell of institutional care (Ballou et al. 2011).
The purpose of the first measure -- the share of total Medicaid expenditures on LTC users spent on HCBS users -- is to adjust the LTC spending share measure for any differences between services captured in our definition of HCBS and institutional care. For example, prescription drugs may be included in nursing home payments, whereas our definition of HCBS excludes such services. Although this measure differs from the LTC spending share measure, the general analysis results were relatively consistent across states and subgroups and so are not presented here. Finally, we include in Appendix D tables (those for the aged and people with ID/DD only), measures developed by Ballou et al. (2013) to capture how often institutional residents had used HCBS prior to entering a nursing home or ICFS/IID. These measures indicate HCBS penetration as part of the continuum of care leading to traditional institutionalization.