Important Dimensions of Long-Term Care System Performance. HCBS spending as a percentage of LTC spending is the most commonly used measure of LTC system transformation. In this study, we examined differences across states and subgroups between this traditionally computed expenditure-based measure and several additional systems performance measures that may be relevant to a discussion of the relative balance between HCBS and ILTC: a utilization-based measure (percentage of LTC users who used HCBS), a relative per-user expenditure ratio (per-user HCBS spending to per-user institutional care spending), and a measure capturing whether Medicaid HCBS are reaching individuals that may need them (percentage of potential Medicaid LTC eligibles who used HCBS). Examination of differences across states on these measures illustrates that alternative indicators of LTC system performance provide different insights and an interpretive context for cross-state comparisons of the percent of total Medicaid LTC spending on HCBS.
There is high overlap between the highest scoring states on two "balance" measures: the percent LTC spending on HCBS overall and percent LTC users receiving HCBS overall (eight states are in the top ten on both these measures). However, there is considerably less overlap between the top scoring states on these two indictors and those in the top ten with respect to ratio of HCBS to ILTC spending per user overall. Only four of the ten states with the highest ratios of per-user HCBS to ILTC spending are in the top ten with respect to the percent of LTC spending on HCBS and only two are among the ten states with the highest percent of potential LTC users receiving HCBS.
Of the states that scored highest (top five) with respect to potential LTC users receiving HCBS, two scored in the top five on the percent of Medicaid LTC spending going toward HCBS and an additional two scored in the top ten. Similarly, of the states that scored among the top five in providing HCBS to potential LTC users, all but one were among the top ten highest scorers with respect to percent of LTC users receiving HCBS and the one state not among the top ten ranked number 11.
The patterns evident in states rankings on a range of performance indicators tell a story about how state LTC policies differ with respect to priorities and trade-offs. Some states choose to provide HCBS to large numbers of eligible and potentially eligible LTC users but they spend comparatively low amounts per HCBS user. Indeed, they may have comparatively fewer nursing home residents but spend a great deal more per user on ILTC than per user of HCBS. This pattern has certain logic if state policymakers have reason to believe that the comparatively small number of ILTC users are, on average, much more severely disabled than HCBS users. In contrast, some states have comparatively large numbers of institutional residents (perhaps because they over-invested in nursing home bed capacity compared to other states several decades ago) and provide comparatively fewer LTC users with HCBS but HCBS benefits are comparatively generous (closer to the average amount spent per user on ILTC). Again, there is a certain logic to this approach if HCBS is being targeted toward beneficiaries who are considered to be at high risk of nursing home admission and in need of generous benefits in order for HCBS to serve as an effective substitute for nursing home care.
Subgroup Differences in Balance. Our review of LTC balance measures by subgroup indicated that differences between aged and disabled enrollees, as well as people with ID/DD, were widespread across the states. Although our previous study based on 2002 MAX files (Wenzlow, Schmitz, and Shepperson,2008) concluded that states with LTC systems most balanced toward HCBS among all LTC users were those with services most balanced among the aged, we found more exceptions to this pattern in 2006.
The subgroup analyses indicate that it has proved much more difficult for states to overcome the "institutional bias" with respect to LTC spending and services for the elderly than for LTC users under age 65, especially those with ID/DD. Only a handful of states among the 40 included in this study stand out as having been much more successful than others in re-balancing their LTC systems for the elderly. These states are Alaska, New Mexico, California, and Washington. These four states all spend 46 percent or more of their Medicaid LTC dollars for the elderly on HCBS. In addition, they all provide HCBS to 63 percent or more of elderly LTC users. With the exception of Alaska (an exceptionally high cost state), all of these states spend less per elderly ILTC user than the national average. New Mexico and Washington spend slightly more per elderly HCBS user than the national average California is unique in also spending less per elderly HCBS user than the national average. California also has a ratio of spending per elderly HCBS user to spending per elderly ILTC user lower than the national average, whereas the ratio of per-user spending on HCBS for the elderly relative to per-user spending on ILTC is higher than the national average in the other three states. Washington -- but especially California and Alaska -- all serve far higher percentages of potential elderly HCBS users than the national average, whereas this is not the case for New Mexico. Of these four states, California appears to have arrived at the most cost effective formula for re-balancing Medicaid LTC spending and use patterns among the elderly toward HCBS while also providing HCBS to more potential users.
The states that stand out as having maximized spending for and use of HCBS among the under 65 Medicaid LTC population are, in addition to Alaska and New Mexico (the states that rank highly on re-balancing measures for all populations), Vermont, Colorado, Hawaii, and Wyoming. These states all spend more than 90 percent of their Medicaid LTC dollars for the ID/DD subgroup on HCBS and all of them. Alaska, Colorado, and Vermont have fewer than 2 percent of LTC users with ID/DD in institutional care. Alaska and Vermont now spend more per user on HCBS than on ILTC (although this indicator may no longer mean much since these two states have virtually eliminated ILTC for people with ID/DD). All of these states -- with the exception of Colorado -- spend well above the national average on HCBS per user with ID/DD. Colorado spends slightly above the national average on HCBS per user with ID/DD. Wyoming provides HCBS to 20.8 percent of potential users with ID/DD (compared to a national average of 8.7 percent); Hawaii, Colorado, Vermont, and New Mexico all exceed the national average with respect to providing HCBS to potential users with ID/DD. However, Alaska falls below the national average in providing HCBS to potential users with ID/DD. It should perhaps also be noted that quite a few states (California, Connecticut, North Dakota, Nebraska, Iowa, Maryland, New York, and South Dakota) provide HCBS to percentages of potentially eligible users with ID/DD well above the national average, but do not perform as well as Alaska, Hawaii, Colorado, Vermont and New Mexico with respect to the percent of spending on HCBS or percent of users receiving HCBS measures for people with ID/DD. Looking at the pattern of rankings across indicators, Colorado, among the 40 states in our study, appears to be doing the best job of "re-balancing" its LTC system for people with ID/DD toward reliance on HCBS rather than ILTC.
For LTC users under age 65 with physical disabilities, Alaska and New Mexico, again rank in the top five with respect to LTC spending on HCBS and LTC users receiving HCBS. However, the other states that appear to do the best job of promoting access to HCBS over ILTC for the non-elderly with physical disabilities are Kansas, Missouri, and Vermont for the percent of LTC spending on HCBS measure. North Carolina, California, and Kansas perform best on the percent of LTC users receiving HCBS measure. In addition to Alaska (but not New Mexico), the states that rank highest in providing HCBS to potential LTC users under age 65 with physical disabilities are Kansas, New York, California, and Vermont. Three states that rank among the top five with respect to the ratio of per-user spending on HCBS to per-user spending also rank in the top five on one or more of the other performance indicators (Missouri, Kansas, and Vermont) whereas two do not (Tennessee and Indiana). However, California, Kansas, and Vermont rank, respectively, below, similar to, or only slightly above the national average in per-user spending on HCBS for the under 65 population with physical disabilities. In Kansas, Vermont, and Missouri, the ratios of HCBS to ILTC spending per user for this population is comparatively high because these states spend considerably under the national average on ILTC spending per user for the ILTC population. Taking all systems performance measures into consideration, Kansas seems to be doing the best overall job of re-orienting its Medicaid LTC system away from ILTC toward HCBS for low-income persons under age 65 with physical disabilities.
The only "state policy" variable measured in this study that predicted high ranking with respect to the percent of LTC spending dedicated to HCBS for the under 65 population with physical disabilities was availability of consumer direction. Kansas has a long-standing tradition of using Independent Living Centers to facilitate consumer-directed personal care services (funded by HCBS waivers). All of the other states that ranked in the top five on performance indicators for this subpopulation (except the ratio of HCBS spending per user to ILTC spending per user) are states where consumer direction is widely available and, in California, Vermont, Alaska, and Missouri, the use of consumer-directed services, especially among the subgroup of Medicaid LTC beneficiaries under 65 with physical disabilities, was much more prevalent in 2006 than in the nation as a whole (Sciegaj and Selkow 2011). We identified the most significant differences in measures of LTC balance by population age group and service delivery system. A comparison of our results with those reported in Wenzlow et al. (2008) for 34 states included in both studies indicated that HCBS spending as a share of all LTC spending increased by at least five percentage points in all age groups between 2002 and 2006 for those enrolled in Medicaid. That is, system transformations over those years appear to have increased use of HCBS across all ages. Our analysis also shows that, although Medicaid systems appear to be least balanced toward HCBS for the aged and most balanced for enrollees with ID/DD, a relatively small share of those with ID/DD potentially eligible for Medicaid LTC services actually receive them. Monitoring the larger population with ID/DD and their needs will be critical for better understanding Medicaid system performance.
Correlates of Long-Term Care System Performance. We conducted an exploratory analysis of the bivariate association between constraints and LTC policies and three indicators of system performance. Our results suggest that two types of state factors are associated with systems less balanced toward HCBS: (1) poor weather conditions that may make it more challenging to serve enrollees with LTC needs in their homes; and (2) the size of the workforce needed to provide adequate HCBS. We also found that availability of consumer-directed services and personal care coverage were positively associated with HCBS use and expenditures, but not for those enrollees with ID/DD. Other factors, not subject to much state control appear to be related to the progress in re-balancing LTC for this population, most notably the availability of resources (per-capita income and availability of taxable resources). Finally, state SSI supplements for people living in the community were associated with more balanced systems among the aged and people with ID/DD. We cannot infer causal relationships from these findings, but rather note that they point the way toward possibly fruitful work in the future.
Our analyses of factors that predicted higher ratios of LTC spending on HCBS for the ID/DD population was particularly weak and failed to yield much useful information with respect to state policy variables, except for a negative relationship between nursing home bed supply and higher percentages of LTC spending for the ID/DD population on HCBS. Future research will need to explore indicators that are more specific to the ID/DD service delivery system, which, as earlier noted, tends to be quite separate from the services system for the elderly and younger adults (under age 65) with physical disabilities. One promising avenue is to look at the percentages of publicly-funded ID/DD beneficiaries receiving services in any institutions with 16 or more beds, in state-run ICFs/IID, and in residential settings with six or fewer residents. It is noteworthy, that of 37 states in our study for which it was possible to measure the percentage of total Medicaid LTC spending for the ID/DD population spent on HCBS, five had closed all of their public ICFs/IID prior to 2006. Four of these states were among the top five highest ranking states with respect to the percent of LTC spending on HCBS for the ID/DD population. Colorado (ranking #3) was the only state that had not closed all of its public ICFs/IID and Colorado had only 2.3 percent of persons with ID/DD receiving public benefits residing in public ICFs/IID (compared to a national average of 8 percent. Moreover, Colorado had no beneficiaries with ID/DD residing in private ICFs/IID, whereas nationally 14.4 percent of all beneficiaries with ID/DD resided in either public or private ICFs/IID in 2006 (RTCL/UMN 2012). In all these states 90 percent or more of public beneficiaries with ID/DD in out-of-home placements resided in settings with fewer than six or more residents with ID/DD. In contrast, the states that ranked lowest in terms of percent Medicaid LTC spending on HCBS for the ID/DD population (Mississippi, Louisiana, Illinois, Arkansas, and North Carolina) all had percentages of beneficiaries with ID/DD residing in state-run institutions in excess of the national average. For example, Mississippi had 40.3 percent of state residents with ID/DD receiving public benefits in state-run ICFs/IID (all with 16 or more beds) in 2006 (RTCL/UMN 2012).