An Investigation of Interstate Variation in Medicaid Long-Term Care Use and Expenditures Across 40 States in 2006. C. Subgroup Differences

07/01/2013

Progress toward re-balancing varies substantially among the different types of users of LTC services. HCBS and ILTC service use and expenditures were far more balanced among young enrollees with disabilities than among enrollees age 65 or older. The percentages of total LTC spending accounted for by HCBS ranged from 66 for people under age 21, 37 for people between ages 65 and 74, and 18 for those age 85 and older (Table II.3). HCBS accounted for an average of 26 percent of LTC spending among all enrollees age 65 and older, compared with 56 percent for those under 65.

Compared to estimates for 2002 reported in Wenzlow et al. (2008), the current results suggest that the spending share for HCBS increased by about five percentage points or more in each age group (under 65, 65-74, 75-84, and 85 or older) since 2002 in the 34 states included in both studies.10 This suggests that gains have been made in the provision of HCBS to people of all ages over the four-year period.

People with ID/DD are primarily under age 65 and thus make up a large portion of enrollees under 65 with disabilities. In 2006, they accounted for 14 percent of LTC users but 31 percent of Medicaid LTC spending and almost half (46 percent) of HCBS spending. More than 85 percent of these enrollees used HCBS (compared to 64 percent overall and 77 of those with physical disabilities) and HCBS accounted for almost 61 percent of the Medicaid LTC spending used for them (compared to 41 percent overall and 46 percent for those with physical disabilities). Per-user spending on HCBS for people with ID/DD was almost $41,000 in 2006, higher than for any other subgroup shown in Table II.3. However, Medicaid spent only 35 cents per user on HCBS for every dollar spent for persons using costly ICFS/IID care. Other enrollee characteristics associated with use of HCBS included Hispanic ethnicity, male gender, and enrollment in only Medicaid (not Medicare) -- all factors associated with age.

TABLE II.3. Expenditure and Utilization-Based Measures of the Balance of LTC Among Enrollees Who Were Aged or Had Disabilities and Were Eligible for Full Medicaid Benefits in 2006, by Population Subgroup
Subgroup Total
LTC $
  Percentage of Medicaid  
LTC $ Allocated
to HCBS
Total
  LTC Users  
Percentage of LTC
  Users Receiving HCBS  
Per-User
  $ on HCBS  
  Ratio of Per-User  
$ on HCBS
Relative to ILTC
Total   76,879,134,892   40.8 2,904,883 63.8 16,914 0.46
Enrolled all year 69,320,813,194 43.2 2,332,924 69.2 18,549 0.40
Under age 21 4,337,490,156 66.1 173,192 87.6 18,902 0.34
21-44 years 15,614,619,795 62.6 417,912 85.0 27,515 0.37
45-64 years 18,385,690,043 47.6 680,016 74.4 17,309 0.39
65-74 years 9,092,379,504 36.6 440,951 66.9 11,282 0.34
75-84 years 14,724,011,450 27.0 641,408 53.5 11,610 0.37
85 years and older 14,705,784,900 17.8 547,883 36.2 13,160 0.42
Unknown Age 19,159,044 83.2 3,521 98.4 4,598 0.11
Aged (65 & older) 38,970,178,862 25.8 1,648,932 51.4 11,875 0.38
Enrollees with disabilities (under 65) 37,908,956,030 56.1 1,255,951 80.0 21,167 0.40
Enrollees Under 65, excluding people with ID/DDa 13,754,095,189 46.2 802,200 77.2 10,262 0.31
Enrollees Under 65 with ID/DDa 22,407,952,989 60.8 391,061 85.2 40,895 0.35
Non-Hispanic White 51,188,363,483 37.6 1,794,052 56.9 18,832 0.53
Black 14,848,124,996 39.8 595,584 67.4 14,710 0.38
Hispanic 5,312,278,986 55.6 252,993 82.3 14,197 0.35
Other or missing race 5,530,367,427 58.5 262,254 84.4 14,620 0.33
Female 45,076,528,382 38.0 1,835,846 62.5 14,941 0.43
Male 31,800,667,824 44.7 1,068,944 66.0 20,120 0.48
Not dually enrolled in Medicare & Medicaid 18,728,364,922 54.4 732,506 80.8 17,205 0.36
Sometimes a dual-eligible 2,385,590,593 37.0 146,291 60.1 10,027 0.47
Always a dual-eligible 55,765,179,377 36.3 2,026,086 57.9 17,284 0.48
SOURCE: Mathematica analysis of 2006 MAX data for 39 states and the District of Columbia with representative FFS LTC data (excludes data from Arizona, Maine, Massachusetts, Michigan, Minnesota, Montana, New Hampshire, Oregon, Pennsylvania, Rhode Island, and Texas).
NOTES: Excludes enrollees in managed LTC and those eligible for restricted Medicaid benefits only. HCBS include 1915(c) waiver services and state plan services for personal care, residential care, home health, adult day care, and private duty nursing. ILTC includes services provided in nursing homes, ICFs/IID, mental hospitals for the aged, and inpatient psychiatric facilities for people under age 21.
  1. Excludes data from District of Columbia, Washington, and Wisconsin (people with ID/DD could not be distinguished from other enrollees in these states).

1. Long-Term Care System Performance Indicators by Basis of Eligibility

As in the overall population, system performance varied across states within subgroup. Tables II.4-II.6 summarize, for each of the three indicators, the scores and ranks for the overall population, those 65 and older, and those under 65 eligible on the basis of disability. (See Appendix Table D.2, Table D.3 and Table D.4 to view information in Table II.4, Table II.5 and Table II.6, ordered alphabetically by state.) Each portion of the table is ordered by the overall balance of LTC for that measure across all subgroups.

In some cases, states with the highest HCBS spending or use overall, were those with the highest HCBS or use among subpopulations. Alaska and New Mexico were ranked among the top five for HCBS spending as a percentage of overall HCBS spending overall and in each subgroup (aged, all LTC users under age 65, all ID/DD users under age 65, and non-ID/DD LTC users under age 65). The other top three overall scorers on this measure failed to score in the top five for one or more subpopulations. Vermont and Wyoming, top scorers overall and for LTC users under age 65, ranked 15th and 27th respectively with respect to the percentage of HCBS spending relative to ILTC spending among LTC users age 65 and older. Wisconsin's rank in the top quartile (#10) the expenditure share measure (Table II.4) appears to be driven primarily by its high ranking among young enrollees (#9) since it ranks much lower (#21) on this measure among the elderly. (Data for Wisconsin could not be disaggregated for the under 65 subgroups with ID/DD and physically disabilities.) In contrast, the high ranks of California (#6) and New York (#9) on the overall HCBS relative to ILTC spending and service user ratios appear driven by HCBS use among the aged. The District of Columbia ranked 2nd to last overall despite ranking 9thfor aged enrollees. When measured in terms of the HCBS and institutional care expenditure ratio (Table II.6), Kansas ranked in the top ten overall despite ranking 25th for people over 65 with disabilities. Yet Kansas ranked comparatively highly on the percentage of HCBS/LTC spending on people with disabilities under age 65 (#7) and the percentage of LTC users under age 65 receiving HCBS. Kansas' high rank on the HCBS/ ILTC spending ratio per user overall appears to reflect its comparatively high rank on this measure for the elderly population (#10); yet, here again, whereas Kansas scored high in terms of percentage of total LTC spending on HCBS for the elderly, it scored much lower (#20) on the percentage of elderly LTC users receiving HCBS.

TABLE II.4. Percentage of LTC Expenditures Allocated to HCBS in 2006, Overall and by BOE
State Overall Aged (65+) Enrollees with Disabilities
(<65, including ID/DD)
Total LTC $   % HCBS     Rank     % HCBS     Rank     % HCBS     Rank  
Alaska 284,916,040 72.7 1 59.0 1 84.0 4
New Mexico 687,375,842 70.3 2 48.0 2 86.5 2
Washington 1,510,683,980 65.2 3 46.4 4 85.2 3
Vermont 257,050,002 57.8 4 20.3 15 91.5 1
Wyoming 176,243,168 57.0 5 14.8 27 83.5 5
California 9,878,514,101 54.7 6 46.7 3 63.0 15
Kansas 840,599,103 52.5 7 22.7 10 77.3 7
Colorado 1,019,876,958 50.7 8 22.3 11 78.5 6
New York 17,776,758,555 45.3 9 36.3 5 53.9 27
Wisconsin 1,764,144,875 44.5 10 17.0 21 72.0 9
North Carolina 2,701,905,573 43.3 11 32.0 6 55.6 25
Nevada 306,338,277 43.3 12 27.0 8 59.6 18
Maryland 1,768,700,598 42.8 13 16.0 23 69.2 10
Idaho 371,132,820 42.6 14 27.1 7 55.7 23
Virginia 1,421,468,659 42.6 15 19.4 16 67.3 11
All 40 states   76,879,134,892   40.8   25.8   56.1  
Missouri 1,466,773,653 40.7 16 19.3 17 63.9 13
Oklahoma 1,012,058,004 40.5 17 20.9 14 58.1 21
Utah 334,796,035 38.9 18 8.6 36 53.2 29
Hawaii 329,343,209 38.5 19 14.1 28 73.8 8
West Virginia 734,425,562 38.0 20 13.6 29 65.8 12
Nebraska 562,110,501 37.4 21 16.6 22 57.5 22
Iowa 1,157,728,242 37.2 22 21.0 13 49.7 30
Tennessee 1,854,934,959 37.0 23 12.5 30 59.7 17
South Dakota 251,692,447 35.9 24 8.7 35 63.7 14
South Carolina 909,136,545 34.6 25 14.9 26 54.8 26
Delaware 301,695,573 34.0 26 10.2 31 59.2 19
Ohio 4,884,852,294 33.5 27 19.0 18 49.5 31
Connecticut 2,238,931,231 31.6 28 15.6 24 53.7 28
New Jersey 3,447,275,904 31.2 29 21.4 12 43.1 35
Florida 3,747,337,138 31.1 30 9.7 33 58.6 20
Illinois 3,176,627,446 30.8 31 18.9 19 39.1 36
Georgia 1,493,201,190 28.4 32 10.0 32 55.7 24
Louisiana 1,525,871,254 27.5 33 15.6 25 35.8 37
Indiana 1,828,498,633 27.3 34 5.9 40 47.4 33
Alabama 1,130,404,702 27.2 35 8.9 34 60.3 16
Kentucky 1,209,161,974 25.8 36 8.0 37 47.4 34
North Dakota 305,327,011 25.8 37 7.6 38 47.6 32
Arkansas 858,715,978 24.9 38 17.1 20 35.2 38
District of Columbia 315,228,327 23.2 39 25.9 9 19.6 39
Mississippi 1,037,298,529 11.1 40 7.1 39 16.9 40
SOURCE: Mathematica analysis of 2006 MAX data for 39 states and the District of Columbia with representative FFS LTC data (excludes data from Arizona, Maine, Massachusetts, Michigan, Minnesota, Montana, New Hampshire, Oregon, Pennsylvania, Rhode Island, and Texas).
NOTE: Excludes enrollees in managed LTC and those eligible for restricted Medicaid benefits only. HCBS include 1915(c) waiver services and state plan services for personal care, residential care, home health, adult day care, and private duty nursing. ILTC includes services provided in nursing homes, ICFs/IID, mental hospitals for the aged, and inpatient psychiatric facilities for people under age 21.

TABLE II.5. Percentage of LTC Users Receiving HCBS in 2006, Overall and by BOE
State Overall Aged (65+) Enrollees with Disabilities
(<65, including ID/DD)
Total LTC Users   % HCBS     Rank     % HCBS     Rank     % HCBS     Rank  
Alaska 7,591 87.0 1 82.2 1 91.1 3
California 578,611 82.5 2 77.5 2 89.2 8
Washington 75,694 78.5 3 69.9 4 89.0 9
Kentucky 50,373 77.5 4 73.5 3 82.5 16
New Mexico 24,595 76.4 5 63.5 5 91.1 2
Idaho 17,227 72.9 6 61.4 6 82.9 14
North Carolina 145,432 72.2 7 61.1 7 86.0 11
Colorado 42,632 69.8 8 51.6 12 89.4 5
Vermont 9,493 68.2 9 46.3 16 91.5 1
New York 385,991 68.2 10 55.5 8 84.3 12
Iowa 51,128 68.1 11 54.7 11 83.8 13
Missouri 90,743 66.4 12 55.3 10 81.6 18
Virginia 52,361 65.6 13 50.6 13 89.3 7
Kansas 40,507 65.4 14 42.4 20 89.3 6
Nevada 12,164 64.2 15 55.3 9 76.1 25
All 40 states   2,904,883   63.8   51.4   80.0  
Wyoming 6,059 63.6 16 33.5 29 89.5 4
Alabama 59,526 61.6 17 40.1 23 86.9 10
South Carolina 43,085 60.2 18 42.1 21 82.4 17
Oklahoma 50,793 58.0 19 47.6 15 73.0 31
New Jersey 99,441 57.7 20 48.1 14 75.0 27
West Virginia 25,825 57.1 21 36.4 26 80.1 21
Maryland 52,081 56.7 22 29.7 31 81.3 19
Hawaii 9,711 56.3 23 38.2 25 82.6 15
Ohio 163,699 55.3 24 45.4 18 68.9 33
Connecticut 56,805 53.1 25 39.9 24 75.2 26
Illinois 153,120 52.2 26 40.8 22 64.9 35
Arkansas 40,947 51.4 27 45.7 17 61.8 37
Nebraska 21,186 51.1 28 35.6 27 73.3 29
Florida 153,416 50.9 29 33.6 28 76.3 24
Utah 11,264 49.8 30 22.2 37 68.6 34
Wisconsin 61,721 48.4 31 26.5 32 80.2 20
Delaware 6,662 47.9 32 26.3 33 77.1 23
South Dakota 10,327 46.2 33 24.9 35 77.8 22
District of Columbia 7,841 45.3 34 42.5 19 49.4 40
North Dakota 9,380 44.3 35 26.0 34 73.4 28
Georgia 66,667 43.4 36 24.9 36 73.2 30
Mississippi 39,336 41.0 37 33.3 30 54.2 39
Louisiana 60,275 38.1 38 20.2 38 55.1 38
Tennessee 51,989 35.2 39 10.7 40 69.9 32
Indiana 59,185 32.5 40 10.9 39 62.0 36
SOURCE: Mathematica analysis of 2006 MAX data for 39 states and the District of Columbia with representative FFS LTC data (excludes data from Arizona, Maine, Massachusetts, Michigan, Minnesota, Montana, New Hampshire, Oregon, Pennsylvania, Rhode Island, and Texas).
NOTE: Excludes enrollees in managed LTC and those eligible for restricted Medicaid benefits only. HCBS include 1915(c) waiver services and state plan services for personal care, residential care, home health, adult day care, and private duty nursing. ILTC includes services provided in nursing homes, ICFs/IID, mental hospitals for the aged, and inpatient psychiatric facilities for people under age 21.

TABLE II.6. Ratio of Per-User Expenditures on HCBS Relative to Per-User Expenditures on Institutional Care in 2006, Overall and by BOE
State Overall Aged (65+) Enrollees with Disabilities
(<65, including ID/DD)
  Per-User  
HCBS $
  Ratio     Rank     Ratio     Rank     Ratio     Rank  
Tennessee 37,521 1.112 1 1.220 1 0.684 8
Wisconsin 26,260 0.927 2 0.602 3 0.783 3
Wyoming 26,045 0.838 3 0.373 13 0.729 7
New Mexico 25,725 0.827 4 0.593 4 0.732 6
Indiana 25,979 0.815 5 0.530 6 0.614 9
Vermont 22,928 0.765 6 0.349 15 1.343 1
South Dakota 18,956 0.707 7 0.305 28 0.577 14
Utah 23,234 0.689 8 0.347 16 0.573 15
Nebraska 19,410 0.649 9 0.405 12 0.583 11
Kansas 16,645 0.644 10 0.434 10 0.493 25
Louisiana 18,253 0.644 11 0.737 2 0.501 23
Washington 16,570 0.644 12 0.459 9 0.967 2
Delaware 32,215 0.608 13 0.340 19 0.501 22
Maryland 25,675 0.604 14 0.466 8 0.577 13
Virginia 17,618 0.589 15 0.327 22 0.608 10
Alaska 31,371 0.561 16 0.432 11 0.737 5
Oklahoma 13,902 0.551 17 0.324 23 0.582 12
Georgia 14,636 0.537 18 0.347 18 0.497 24
Hawaii 23,187 0.531 19 0.282 32 0.751 4
Colorado 17,375 0.514 20 0.305 27 0.554 16
New York 30,580 0.498 21 0.564 5 0.321 32
Ohio 18,044 0.497 22 0.347 17 0.538 18
West Virginia 18,914 0.494 23 0.288 30 0.551 17
Connecticut 23,454 0.483 24 0.319 24 0.511 20
North Dakota 18,943 0.478 25 0.253 35 0.383 29
Nevada 16,978 0.473 26 0.329 21 0.527 19
Illinois 12,256 0.472 27 0.364 14 0.454 27
Florida 14,924 0.462 28 0.226 37 0.467 26
All 40 states 16,914 0.458   0.378   0.397  
Missouri 9,908 0.442 29 0.248 36 0.501 21
District of Columbia 20,620 0.398 30 0.514 7 0.275 36
South Carolina 12,107 0.374 31 0.257 34 0.287 33
Iowa 12,375 0.364 32 0.294 29 0.236 39
New Jersey 18,755 0.359 33 0.313 25 0.281 34
Idaho 12,601 0.348 34 0.283 31 0.351 31
Arkansas 10,165 0.342 35 0.270 33 0.360 30
North Carolina 11,151 0.337 36 0.340 20 0.244 38
California 11,325 0.312 37 0.306 26 0.259 37
Alabama 8,385 0.260 38 0.160 39 0.280 35
Kentucky 7,991 0.253 39 0.087 40 0.387 28
Mississippi 7,115 0.191 40 0.163 38 0.184 40
SOURCE: Mathematica analysis of 2006 MAX data for 39 states and the District of Columbia with representative FFS LTC data (excludes data from Arizona, Maine, Massachusetts, Michigan, Minnesota, Montana, New Hampshire, Oregon, Pennsylvania, Rhode Island, and Texas).
NOTE: Excludes enrollees in managed LTC and those eligible for restricted Medicaid benefits only. HCBS include 1915(c) waiver services and state plan services for personal care, residential care, home health, adult day care, and private duty nursing. ILTC includes services provided in nursing homes, ICFs/IID, mental hospitals for the aged, and inpatient psychiatric facilities for people under age 21.

2. Long-Term Care System Performance Indicators by System Type (Aged, ID/DD, and non-ID/DD)

Medicaid beneficiaries using LTC are served by distinct "systems" (program administration and service delivery infrastructures depending on whether their disabilities are ID/DD or not. The administrative and service delivery infrastructures for the elderly and younger disabled individuals (particularly adults under age 65) are the same or closely linked in many states. Both the elderly and younger adults with disabilities other than ID/DD receive ILTC in nursing homes. Whereas skilled home health care delivered by home health agencies (HHAs) may be provided to all subgroups, state plan personal care services are provided almost entirely to the elderly and younger adults without ID/DD, although a small percentage of adults and children with ID/DD may receive this benefit if they also have severe physical disabilities. In some states, the elderly and younger adults with disabilities but without ID/DD are served in the same HCBS waiver programs; however, in a number of states there are separate HCBS waiver programs for people with non-ID/DD according to their age (i.e., either under or over age 60 or age 65).11 Insofar as states have created distinct and separate service systems for subgroups of LTC users, it is not surprising that there is variation across "re-balancing" or other system performance measures by subgroup. Table II.7, Table II.8 and Table II.9 display, for each of the three balance measures, the scores and ranks for each subgroup. These tables replicate those shown earlier by BOE except that, for people under 65, performance indicators are displayed separately for people using ID/DD system services and those not using such services.12 As above, the states in each table are ordered by the overall balance of LTC for that measure across all subgroups. (See Appendix Table D.5 and Table D.6 for a summary of performance indicators by state for people with physical disabilities and those with ID/DD.)

In almost all states, the lowest HCBS spending and user share is evident among the aged and the highest among Medicaid enrollees using ID/DD services. Among people with ID/DD, either the spending or utilization share allocated to HCBS exceeded that for people with physical disabilities in all but five states (Idaho, Iowa, Mississippi, Missouri, and North Carolina).

Although these results suggest that the Medicaid service system is far more balanced for people with ID/DD than for those who are aged or have physical disabilities, we note several caveats. Some enrollees with ID/DD may be using nursing home or other institutional care services, especially in states where ICFs/IID have closed. We were unable to identify such enrollees using MAX data. Analyses of Nursing Home Minimum Data Set assessment data indicate that only 2.4 percent of NF residents have ID/DD as well as other medical conditions and physical disabilities that justify their nursing home placement (Martin et al. 2011). However, other research evidence indicates that Medicaid beneficiaries who are not living in ICFs/IID but are also not living at home with family members or in supported housing (e.g., shared apartments) are living in "group homes" of limited bed size (16 beds or fewer) where the room and board costs are covered by Social Security and/or Supplemental Security Income (SSI) and related state supplemental cash assistance payments and the service costs are covered under HCBS waiver programs. (O'Keeffe et al. 2010). One definitional issue is that some states certify similar facilities as "small" ICFs/IID so that Medicaid reimbursement will also be available for room and board. To the degree that similar small residential settings for people with ID/DD are Medicaid-covered institutions (ICFs/IID) in some states but classified as non-institutional residential setting ineligible for Medicaid reimbursement for the entirety of their costs in other states, the performance indicators presented here will be biased.

Our measures capture Medicaid system performance among Medicaid enrollees only. To test whether states differ in the extent to which they cover other low-income people with disabilities we employed a combination of MAX and ACS data. The ACS measures age and income in a conceptually clear manner, and low-income was defined to include SSI recipients and other individuals with income up to 300 percent of the SSI limit. However, the ACS disability questions are relatively primitive measures of LTC disability, especially when compared with the detailed assessments that are typically performed to satisfy the medical criteria for Medicaid LTC services. Nevertheless, they provide a consistent measure across states, and are a useful gauge of the extent to which Medicaid programs serve broadly-defined groups of low-income people with disabilities.

We compared Medicaid coverage of the elderly and the non-elderly using two ACS disability questions: (1) Does the person have a physical, mental, or emotional condition lasting six months or more resulting either in difficulties conducting ADLs (dressing, bathing, or getting around inside the home)? and (2) Does the person have a physical, mental, or emotional condition lasting six months or more resulting in difficulties in learning, remembering, or concentrating? To compare the elderly to the non-elderly we considered a person disabled if the response was yes to either question. Our calculations based on combined ACS and MAX data suggest that only 15 percent of people under 65 potentially eligible for Medicaid LTC actually used Medicaid HCBS, compared to 22 percent of potential aged eligibles.13 (See Appendix Table D.5 and Table D.3 for state-level detail.) On this indicator as well as the others previously discussed, we see considerable interstate variation in where states rank with respect to the percentage of potential LTC users receiving HCBS by subgroup. California ranks highest overall and for the elderly and is second for LTC users under age 65. The other states in the top five overall include New York, Iowa, Alaska, and Vermont. The other states in the top five for the elderly include Alaska, Washington, New York, and Iowa. New York ranks first for the under 65 population, followed by California, Kansas, Iowa, and Vermont.

For the 37 states where we could separate those with ID/DD from other HCBS users under age 65, we tailored the definition of disability more closely to the characteristics of the population. We considered a person to be disabled with ID/DD if they had difficulties in learning, remembering, or concentrating. We considered a person to have a disability other than ID/DD if they had difficulties conducting ADLs, since most people which did not have ID/DD probably had a physical limitation that restricted performance of ADLs.

On this indicator as well as the others previously discussed, we see considerable interstate variation in where states rank with respect to the percentage of potential LTC users receiving HCBS by subgroup. For people with ID/DD, the top five states had very similar scores. They include Wyoming, South Dakota, Iowa, North Dakota, and New York. For people with physical disabilities, the top five states also had very similar scores. They include Kansas, Alaska, New York, California, and Vermont. (See Appendix Table D.5 and Table D.6).

It is not surprising that the states that serve the highest percent of aged potential eligibles also tend to do well in serving those under 65 with physical disabilities, since state service delivery and program administration infrastructure for LTC tend to be quite different for Medicaid beneficiaries with ID/DD and the aged/disabled (those with disabled conditions other than ID/DD). In effect, state LTC systems for these groups are almost entirely separate. In contrast, the infrastructure serving the elderly and people with non-ID/DD are often combined.

View full report

Preview
Download

"40State.pdf" (pdf, 787.2Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®