Personal assistance benefits and federal health care reforms: who is eligible on the basis of ADL assistance criteria?
The World Institute on Disability defines personal assistance services (PAS) as assistance, under maximum feasible user control, with tasks aimed at maintaining well-being, personal appearance, comfort, safety, and interactions within the community and society as a whole. These may include personal maintenance and hygiene tasks, mobility tasks, household maintenance tasks, infant and child care related tasks, cognitive tasks, security related services and communication services (Litvak, Heumann, and Zukas, 1987), although available services in existing state PAS programs tend to be much more circumscribed (Kennedy, 1992). Disability advocates argue that the provision of adequate and affordable PAS is an essential component of community integration and participation guaranteed by the Americans with Disabilities Act (Heumann, 1991), and have devoted considerable political energies toward the expansion of federal funding for PAS.
In an effort to target services and contain program costs, recent legislative proposals to expand PAS (from the 1987 LongTerm Home Care Catastrophic Protection Act to the 1993 Health Security Act) have defined program eligibility in terms of need for assistance with ADLs. These eligibility criteria measures, which are certainly familiar to rehabilitation professionals, assess the capacity to perform basic tasks of self maintenance (in the case of proposed federal legislation, those tasks are bathing, dressing, transferring, toileting, and eating). Such indices have become a ubiquitous feature of aging and disability research as well as rehabilitative practice. ADLs are of particular interest to policymakers, mainly because they have proven to be robust predictors of nursing home admissions (Branch & Jette, 1982; McFall & Miller, 1992; Shapiro & Tate, 1985: Williams et. al., 1994), which are heavily subsidized.
As ADLs assume a more central role in national policy, the need for accurate national estimates across populations has become pressing. Yet the published literature in this area has been surprisingly limited. A handful of analyses looked at the number of persons who would be eligible for a federal PAS benefit under various activity limitation criteria (Jackson, Burwell, Clark & Harahan, 1992; Spector, 1991; Spillman & Kemper, 1992; Stone & Murtaugh, 1990), but focused only on the proportion of the population aged 65 or older. This paper uses national survey data from the U.S. Census Bureau's 1990 and 1991 Surveys of Income and Program Participation (SIPP) to explore the size and characteristics of the total adult population eligible for services under different ADL, age and income criteria. It is based on an analysis recently completed by the Disability Statistics Rehabilitation Research and Training Center (Kennedy & LaPlante, 1997).
Development and Political Justification of Personal Assistance Programs
While the policy debate regarding the relative merits of "outdoor" or home and community-based support versus "indoor" or institution-based disability services began with the Industrial Revolution (Applebaum, 1993; Benjamin, 1993), the modern development of personal assistance programs in the U.S. can be linked to the passage of Titles XIX and XX of the Social Security Act (Litvak, 1991). This federal funding allowed more progressive state governments to experiment with alternatives to institutional placement for some populations with disabilities. Using a variety of age, impairment, and functional limitation criteria, these programs typically offered modest levels of paid assistance in the home.
Aside from increased government funding, political and demographic factors contributed to the research interest and policy development of personal assistance programs in the 1970s. Due to increases in longevity and decreases in family size, the number of elders seeking institutional placement began to rise (a phenomenon which is continuing). At the same time, nursing home costs began to increase at several times the rate of inflation. A supply-side solution to rising nursing home expenditures was building restrictions for new facilities, which placed added pressure on existing beds. These factors led to a concerted effort on the part of researchers and service providers to identify less costly alternatives to institutional placement. Political pressure for expanded community services came from leaders of the increasingly powerful independent living movement (Scotch, 1989) and the "gray lobby" (Wallace, Williamson, Lung & Powell, 1991).
By the mid-80s, however, a growing body of evaluation research was calling into question one of the most appealing rationales for government-funded PAS, namely that these services would be essentially self-financing, as persons with disabilities were diverted from costly nursing home placements to lower cost community-based services. A number of studies indicated that, while personal assistance services allowed some recipients to postpone or prevent institutionalization, most service beneficiaries would have managed to avoid such placement in any case (Weissert, Cready & Pawelak, 1988). Under rigid cost-benefit criteria used by evaluators, the provision of services to persons who would not have been placed in nursing homes was viewed as a net loss. The culmination of this line of research was the U.S. Department of Health and Human Services' multi-million dollar Channeling Study, which randomly assigned extremely impaired seniors to treatment and control groups and assessed the impact of case management and modest amounts of personal assistance. Again, researchers showed that personal assistance had little effect on nursing home admissions (Kemper, 1988).
These research findings were consistent with the new Reagan zeitgeist of fiscal conservatism, which also influenced policy initiatives of the period. If PAS did not inevitably lead to decreased nursing home expenditures, then they would be made to do so by legislative fiat. The only new source of federal funding established in the 1980s was the 1981 Medicaid Home and Community-based Service Waivers, which allowed states to redirect federal nursing home funds to home and community-based alternatives, but made program expenditures contingent on a decline in nursing home admissions. The main effect of the Medicaid Waiver legislation and subsequent modifications was the creation of a number of small state programs targeted to very specific populations, primarily persons with mental retardation and technology dependent children (Miller, 1991). States also continued to expand provision of PAS to larger populations through existing finance mechanisms, notably the Medicaid Personal Care Option (Litvak & Kennedy, 1991).
Despite heightened cost concerns on the part of policymakers, the themes of self-reliance, efficiency, and communitarianism, which figure so prominently in PAS advocacy efforts, still resonate across the political spectrum in the 1990s. Moreover, longterm care expenditures continue to increase, along with population pressures on existing service delivery systems. Community-based services still hold out the promise of cost-containment, so long as they are sufficiently targeted. Targeting of services has therefore emerged as the major concern of policymakers and researchers. This is apparent in recent federal PAS proposals which limit program eligibility only to persons who need assistance with two or three of five basic ADLs.
The legislative appeal of ADLs is clear. The fact that they reliably predict nursing home admissions suggests that they are an appropriate targeting mechanism. They also perform an important triage function, by selecting a relatively small and therefore fiscally manageable subset of the entire population with disabilities. The face validity of ADLs is also politically useful for consensus building. We all eat, dress, use the toilet, bathe, and get in and out of chairs and beds. We can readily imagine how much more difficult life would be if we needed another person to do these activities, and are likely to agree that individuals in such circumstances merit some sort of public assistance.
If ADL criteria are to be effectively employed in eligibility and/or needs assessment decisions, it is important for policymakers and programmers to understand the size and composition of the target population. The following sections will briefly describe the number and type of adults living in the community who require ADL assistance and the proportion eligible for PAS benefits under different criteria, using 1990-91 SIPP data.
ADL Assistance Prevalence in the Adult Population
According to the 1990-91 SIPP, an estimated 3.7 million adults, or 1.9 percent of the noninstitutionalized population aged 15 or older, said that they need the assistance of another person to perform one or more of the five basic activities of daily living. Women were more likely than men to require personal assistance-2.2 percent versus 1.5 percent. Differences in longevity are probably a key reason for the disparity, since women have significantly greater life expectancies than men (i.e., the average adult age for the SIPP respondents was 44 for women, but only 42 for men). In all, women accounted for 61.5 percent of persons needing personal assistance in 1 or more ADLs.
The likelihood of needing ADL assistance increased markedly with age, from 0.3 percent at ages 15-24 to 22.9 percent of those aged 85 and above. Although only 15.7 percent of the adult non-institutional population was 65 or older, this group accounted for 61.0 percent of those needing personal assistance. A comparison of racial/ethnic groups indicated that African Americans had the highest rate of assistance needs, 2.8 percent. Non-Hispanic whites had a rate of 1.9 percent. The lowest rates were observed among those of Hispanic origin (1.3 percent) and other racial groups (0.8 percent). Again, differences in the age distributions of these populations may account for some of these disparities in ADL assistance rates.
As shown in table 2, the primary health condition identified by respondents as causing a need for ADL assistance was arthritis or rheumatism (17%), followed by an aggregate "other" category (a mix of low incidence and nonspecific health conditions). Back and spine problems accounted for roughly 12% of cases requiring ADL assistance, and heart disease and stroke each accounted for over 8%. Paralysis and head or spinal cord injury together accounted for less than 7% of all cases.
TABLE 2 Health Conditions Causing Need for ADL Assistance
Health Conditions Reported Needs Help with as Cause of ADL Assistance 1 or more ADLs N(1000s) % Adults Needing ADL Assistance 3,676 100.0 Arthritis or Rheumatism 627 17.1 Other(*) 616 16.7 Back or Spine Problems 432 11.8 Heart Trouble (e.g. Heart Attack, 309 8.4 Hardening of the Arteries) Stroke 305 8.3 Lung or Respiratory Trouble 209 5.7 Paralysis of any kind 152 4.1 Senility/Dementia/ 125 3.4 Alzheimer's Disease Cancer 113 3.1 Broken Bone/Fracture 111 3.0 Stiffness or Deformity of Foot, 106 2.9 Leg, Arm or Hand Diabetes 106 2.9 Mental Retardation 95 2.6 Head or Spinal Cord Injury 93 2.5 Blindness or Vision Problems 78 2.1 Mental or Emotional Disorder 62 1.7 High Blood Pressure 49 1.3 (Hypertension) Missing Legs, Feet, Arms, 49 1.3 Hands, or Fingers Kidney Stones or 39 1.1 Chronic Kidney Trouble
(*) Includes non-specified conditions as well as the following low incidence conditions (listed alphabetically): AIDS or ARC, Alcohol or Drug Problems, Cerebral Palsy, Deafness, Epilepsy, Hernia or Rupture, Learning Disability, Mental or Emotional Disorder, Speech Disorder, Stomach Trouble, Thyroid Trouble or Goiter, Tumor, Cyst, or Growth
Of course the entire population requiring ADL assistance is not eligible for a federal PAS benefit under any of the legislative proposals discussed earlier in this article. Policy debate has centered primarily on the appropriate ADL thresholds (i.e., need for assistance with at least 2 of 5 ADLs or need for assistance with 3 of 5 ADLs). The next section will discuss the size of these populations. both independently and in conjunction with income and age criteria.
ADL Assistance and Program Eligibility
According to the 1990-91 SIPP estimates, if eligibility for a federal PAS benefit is based solely on need for assistance with at least 2 of 5 ADLs, approximately 2.2 million adults would have been potentially eligible for services. Raising the limitation level to 3 of 5 ADLs, as the Health Security Act proposed, reduced the eligible population by 36 percent, to 1.4 million (see Table 3). If eligibility was limited to older people (over age 64) who need assistance with 2 of 5 ADLs, approximately 1.3 million people would have been eligible for services. Raising the limitation level to 3 or more ADLs would drop the population to 900,000 seniors. Table 3: ADL, Age, and Income Eligibility Criteria
Eligibility Criteria Needs Help w/2 or more ADLS N(1000s) % Adults Needing ADL Assistance 2,153 1.1 Adults Aged 15-64 819 0.4 Adults Aged 65 or Older 1,334 0.7 * [is less than] Poverty Level 531 0.3 * 100 - 150% Poverty Level 294 0.2 * 150 - 200% Poverty Level 335 0.2 * = or [is greater than] 200% Poverty Level 993 0.5 Eligibility Criteria Needs Help w/3 or more ADLs N(1000s) % Adults Needing ADL Assistance 1,387 0.7 Adults Aged 15-64 486 0.2 Adults Aged 65 or Older 901 0.5 * [is less than] Poverty Level 305 0.2 * 100 - 150% Poverty Level 216 0.1 * 150 - 200% Poverty Level 212 0.1 * = or [is greater than] 200% Poverty Level 654 0.3
It is also worth noting the other side of these estimates - 35% of adults who needed assistance with 3 or more ADLs were under age 65. The younger population with significant ADL limitations may have distinct support service needs, particularly with regards to work and childcare (Dejong, Batavia & Griss, 1989). If the absence of PAS constitutes a serious impedance to workforce participation, as some analysts suggest (Nosek & Howland, 1993), then there may be good reason to establish a system of support services specifically for young people with disabilities who are currently employed or seeking employment. This is a different policy goal than the prevention of institutionalization, and may require a different type of program. The Health Security Act, for example, had a separate provision regarding tax credits for PAS users who are working or seeking work (Kennedy, 1994).
If eligibility is linked to low income as well as ADLs, the size of the eligible population would be dramatically reduced. Requiring family incomes to be at or below the poverty level on top of needing assistance with at least 2 of 5 ADLs, would reduce the eligible population to approximately 530,000. Only 304,000 people have incomes below poverty level and need assistance with 3 or more ADLs. Note however, that less stringent income criteria yield larger eligible populations. Among those needing help with 3 or more ADLs, 520,000 have family incomes at or below 150 percent of poverty level, and 730,000 have family incomes at or below 200 percent of poverty level. This suggests that a large portion of this population is "near poor".
ADL Measurement Issues and Need for Assistance
Of course any cross-sectional estimates are likely to be misleading for programming purposes, simply because a significant portion of activity limitations are transitory. For example, table 4 shows that 20.3 percent of those adults who reported needing ADL assistance said they had needed that assistance for less than one year. Because PAS and other long-term support services are premised on the need for support of substantial duration and magnitude, cross-sectional data like that reported here will not accurately identify long-term users of PAS. Researchers have also noted that self-reported ADL data is prone to measurement error. Mathiowitz and Lair (1994) found significant instability in rates of ADL limitation in their analysis of the 1987 National Medical Expenditure Survey, with unexpectedly large portions of the elderly population with disabilities reporting improvement or decline in a one-year retest of ADL status.
The precise phrasing of the ADL items is another source of concern. While some surveys ask specifically about supervision or standby assistance with ADLs, the SEPP merely asks whether the respondent needs the assistance of another person. This may lead to bias by not identifying people with cognitive or psychiatric disabilities (Kane et. al., 1991; Kasper, 1990; Spector, 1991), and it almost certainly leads to an undercounting of need for assistance (Stone & Murtaugh, 1990). One method proposed for including this population in a federal PAS program is to expand the number and/or type of activities counted, including higher-order tasks such as instrumental activities of daily living or IADLs (Lawton and Brody, 1969). Continuing research in this area should identify key populations with disabilities which may not be captured by standard ADL criteria but who would benefit from the proposed support services. TABLE 4 Duration of ADL Assistance Need
Period Assistance has Been Needed N (1000s) % * [is less than] 6 months 415 11.4 * 6 - 11 months 322 8.9 * 1 - 2 years 980 27.0 * 3 - 5 years 790 21.7 * 5+ years 1,128 31.0
Aside from the various measurement issues surrounding the need for ADL assistance, some critics worry about programmatic incentives which would distort ADL reporting and therefore increase costs. Wilensky (1993), for example, predicts "staggering" cost overruns if PAS benefits like those contained in the Health Security Act are eventually implemented:
In an aging population, the potential for seeking home
care and receiving some medical benefits from home
care is enormous. While the Administration has
attempted to limit the eligible population, by including
certain restrictions on daily activities as a requirement
for eligibility and specifying that only those who are at
risk of institutionalization are eligible, our ability to
predict who is at risk for institutionalization is
notoriously poor. Once you provide an incentives for
being classified as having one or more limitations on
daily activity, the number of people so identified can
be expected to increase exponentially (p. 16).
This scenario posits a phenomenon akin to "DRG creep" (where hospital condition codes submitted for Medicare reimbursement rise to higher levels of severity to increase federal payment levels). The issue here is not so much fakery on the part of persons with disabilities as it is client advocacy and possibly self-interest on the part of service providers. But there are administrative solutions to this type of problem, i.e., independent evaluation of the disability status of program applicants. No legislators or advocates are suggesting that benefits be offered to applicants solely on the basis of self-reported ADL status. Independent measures of the type and magnitude of individual assistance needed will be crucial to the implementation and administration of any new federal benefit. Rehabilitation professionals, with their extensive assessment skills, would presumably play a significant role in eligibility and service allocation decisions.
A critical issue for advocates is not whether the ADL criteria are too broad, but whether the proposed federal benefits are too narrow. The problem is that the constant framing of need for assistance in terms of the most rudimentary acts of survival may have the insidious effect of truncating the perceived range of activities with which assistance is needed, thereby limiting the utility of the program benefits for a substantial portion of the population technically eligible for services. The Health Security Act, for example, would have required states to provide only the most basic hands-on and standby assistance with ADLs, although they could have chosen to provide a more extensive range of services if desired (Kennedy, 1994).
Policy-makers and analysts should recognize that people with ADL assistance needs have the same diverse preferences, interests, aspirations, and inclinations as everyone else. As White (1991) observed, "life is more than simply the sum of a list of activities of daily living." This means that people who need help toileting will probably also need help going to work, or to church, or to the weekly poker game, and a federal PAS program should reflect this in its service menu and allocation levels. If PAS benefits are to fulfill their rehabilitative potential as a vehicle of community integration, they must be appropriately designed and adequately funded, as well as accurately targeted. TABLE 1: Number, proportion, and type of adults needing ADL assistance
Sociodemographic Need Help with Group Total 1 or more ADLs N (1,000s) N (1,000s) % Total Adult Population 194,861 3,676 1.9 Gender * Male 93,541 1,420 1.5 * Female 101,320 2,256 2.2 Age Group * 15-24 34,667 90 0.3 * 25-34 42,646 226 0.5 * 35-44 39,358 342 0.9 * 45-54 26,620 319 1.2 * 55-64 21,032 454 2.2 * 65-74 18,280 740 4.1 * 75-84 9,824 946 9.6 * 85+ 2,434 556 22.9 Race and Ethnicity * Hispanic 10,570 142 1.3 * Non-hispanic White 157,256 2,925 1.9 * Non-hispanic Black 20,187 555 2.8 * Non-hispanic Other 6,848 54 0.8
Applebaum, R. (1993). Stuck in adolescence: will home care ever come of age? The Gerontologist, 33(2), 278-280.
Benjamin, A. (1993). An historical perspective on home care policy. Milbank Quarterly, 71(1), 129-138.
Branch, L., & Jette, A. (1982). A prospective study of long-term care institutionalization among the aged. American Journal of Public Health, 72(12), 1373-1379.
DeJong, G., Batavia, A. I., & Griss, R. (1989). America's neglected health minority: Working-age persons with disabilities. Milbank Quarterly, 67, 311-351.
Heumann, J. (1991, July 25). Testimony before the Senate Committee on Labor and Human Resources. Washington, DC.
Jackson, M., Burwell, B., Clark, R., & Harahan, M. (1992). Eligibility for publicly financed home care. American Journal of Public Health, 82(6), 853-6.
Kane, R., Saslow, M., & Brundage, T. (1991). Using ADLs to establish eligibility for long-term care among the cognitively impaired. Gerontologist, 31(1), 60-6.
Kasper, J. (1990). Cognitive impairment among functionally limited elderly people in the community: future considerations for long-term care policy. Milbank Quarterly, 68(1), 81-109.
Kemper, P. (1988). The evaluation of the NLTCD: overview of the findings. Health Services Research, 23 (1), 161-174.
Kennedy, J. (1993). Policy and program issues in providing personal assistance services. Joumal of Rehabilitation, 3, 17-23.
Kennedy, J. (1994). An analysis of home and community based services proposed in the Clinton administration's Health Security Act. Disability Studies Quarterly, 14(3), 20-26.
Kennedy, J. & LaPlante, M. (1997). A profile of adults needing assistance with activities of daily living, Disability Statistics Report 9.
Lawton, M., & Brody, E. (1969). Assessment of older people: self-maintaining and instrumental activities. Gerontologist, 9, 179-186.
Litvak, S., Heumann, J. & Zukas, H. (1987). Attending to America: personal assistance for independent living. Oakland, CA: World Institute on Disability.
Litvak, S. & Kennedy, J. (1991). Policy issues affecting the Medicaid personal care services optional benefit. Oakland: World Institute on Disability.
Litvak, S. (1991 ). "State and federal funding sources for PAS" in Personal Assistance Services: A Guide for Policy and Action. Oakland, CA: World Institute on Disability.
Mathiowitz, N., & Lair, T. (1994). Getting Better?: Changes or errors in the measurement of functional limitations. Paper presented at the annual meeting of the American Statistical Association, Chicago, IL.
McFall, S., & Miller, B. (1992). Caregiver burden and nursing home admission of frail elderly persons. Journal of Gerontology, 47 (2), 73-79.
Miller, N. (1991). Medicaid Home and Community Based Service Waivers: 1982-1989. Paper presented at the annual meeting if the American Public Health Association. Washington, DC: Health Care Finance Administration.
Nosek, M. & Howland, C. A. (1993). Personal assistance services: the hub of the policy wheel for community integration of people with severe physical disabilities. Policy Studies Journal, 21(4), 789-812.
Scotch, R. (1989). Politics and policy in the history of the disability rights movement. Milbank Quarterly, 67, 380-400.
Shapiro, E., & Tate, R. (1985). Predictors of long-term care facility use among the elderly. Canadian Journal on Aging, 4, 11-19.
Spector, W. (1991). Cognitive impairment and disruptive behaviors among community-based elderly persons: implications for targeting long-term care. Gerontologist, 31(1), 51-59.
Spillman, B., & Kemper, P. (1992). Long-term care arrangements for elderly persons with disabilities: private and public roles. Home Health Care Services Quarterly, 13, 5-34.
Stone, R. & Murtaugh, C. (1990). The elderly population with chronic functional disability: implications for home care eligibility. Gerontologist, 30(4), 491-6.
Wallace, P., Williamson, J., Lung, R., & Powell, L. (1991). A lamb in wolf's clothing?: the reality of senior power and social policy. In M. Minkler & C. Estes (Eds.), Critical Perspectives on Aging: The Political and Moral Economy of Growing Old. Amityville, NY: Baywood Publishing Company.
Weissert, W., Cready, C., & Pawelak, J. (1988). The past and future of home and community-based long-term care. Milbank Quarterly, 66(2), 309-380.
White, H. (1991). Disabled elderly: you can't run, but you can hide. Gerontologist, 31(2), 278, 1991.
Williams, B., Fries, B., Foley, W., Schneider, D., & Gavazzi, M. (1994). Activities of daily living and costs in nursing homes. Health Care Finance Review, 15(4), 117-135.
Wilensky, G. (1993). The Clintons' tooth-fairy financing. Journal of American Health Policy 3(6), 14-19.