Because long-term care has evolved since an of Institute of Medicine report, this research report examines the characteristics of adult long-term care users. In analyzing the care, the report distinguishes between community-based and institutional care and also the age of recipients.
The great diversity of long-term care users and the increasing level of disability of the elderly long-term care population documented suggest that it will remain difficult and expensive to ensure access to long-term care and meet the needs of this population. Both nursing home and home care expenditures in Medicare and Medicaid are projected to double from 1995 to 2005, and efforts to reduce public costs need to be monitored carefully to ensure that this population's needs are met.
Print copies of The Characteristics of Long-term Care Users (AHRQ Publication No. 00-0049) are available from the AHRQ Publications Clearinghouse (1-800-358-9295).
By William D. Spector, Ph.D., and John A. Fleishman, Ph.D., Agency for Healthcare Research and Quality (AHRQ); Liliana E. Pezzin, Ph.D., Department of Medicine, Johns Hopkins University; Brenda C. Spillman, Ph.D., Urban Institute. This paper was commissioned by the Committee on Improving Quality in Long-term Care, Institute of Medicine.
In 1998 the Institute of Medicine (IOM) established a new committee to document changes in the long-term care population and to report on the state of quality in long-term care. This committee follows the 1986 IOM Committee on Nursing Home Regulation, whose report, "Improving the Quality of Care in Nursing Homes," laid the foundations for major changes in the nursing home quality regulatory system. Many of the recommendations for changes made by the IOM Committee were enacted into law as part of the Omnibus Budget Reconciliation Act of 1987 (OBRA-87), which was implemented in 1990 and 1991.
The present Committee on Improving Quality in Long-term Care was given a broad mandate to examine all long-term care settings, not just nursing homes. The mandate of the new committee reflects the evolution of long-term care since the prior IOM committee's report was issued. Long-term care is no longer synonymous with nursing home care. Considerable growth has occurred at both the low and high ends of the care spectrum. At the low end of the spectrum, the use of non-institutional long-term care alternatives, such as home care, personal care homes, and assisted living facilities, has increased. At the high end, growth in the use of sub-acute care has occurred.
Moreover, other broad policy changes have occurred over the past decade that have affected the long-term care population either directly or indirectly. For example, the increase in the use of pre-admission screening for nursing homes, the expanded role of Medicaid home and community-based waivers, the introduction of Medicare and Medicaid managed care programs, and the general trend toward prospective payment and more rapid discharge from hospitals may also be altering patterns of long-term care utilization. Such changes suggest that the time is appropriate to reexamine the characteristics of long-term care users.
In recognition of this broader mandate, we examine the characteristics of adult long-term care users, defined by type of assistance received rather than the setting in which it is received. Specifically, we define long-term care users as those who report receiving human assistance (hands on, supervision, or standby help) with activities of daily living (ADLs) (Katz et al., 1963) or instrumental activities of daily living (IADLs) (Lawton and Brody, 1969) because of a health problem.
ADLs represent the primary activities necessary to carry out basic functions, such as eating, dressing, and bathing: IADLs include tasks necessary for independent community living, such as shopping, managing finances, and house cleaning.
We limit long-term care users to those who receive human help. We exclude persons who perform these activities with only mechanical or non-human help. We include all persons living in nursing homes regardless of ADL or IADL criteria since the small number who do not have ADL limitations may reasonably be assumed to receive IADL assistance.
This broader definition of long-term care necessitates examining a wide range of care settings and care recipients, not merely elderly nursing home residents. When analyzing care settings, we distinguish between community-based and institutional care. A second distinction concerns the age of long-term care recipients. Most attention has focused on the elderly because of their relatively high use of long-term care. Younger persons, however, may also require ADL or IADL assistance, due to injuries, developmental disabilities, or other factors not necessarily associated with aging. Although they are less at risk of receiving long-term care, non-elderly adults nevertheless represent a large proportion of the long-term care population (U.S. Bureau of the Census, 1996). A comprehensive picture of long-term care users must, therefore, include persons younger than age 65.
We structure the description of the characteristics of long-term care users around the two distinctions of:
In particular, for each of these four segments, we provide estimates of the number of long-term care users and describe their demographic and health characteristics. For both elderly and non-elderly adults, we compare users with nonusers and users of community care with users of institutional care. For both younger and older care recipients, we examine factors associated with the type of care received (informal, formal, and institutional care). For the elderly, we also assess changes in characteristics of long-term care users between 1984 and 1994, for both community and institutionalized persons, and include descriptions of the characteristics of persons in housing with supportive services and informal caregivers. Taken together, these findings present a comprehensive picture of persons using long-term care in community and institutional settings.
No single data set provides national estimates for elderly and non-elderly long-term care users in both community and institutional settings. Consequently, this paper relies on several nationally representative data sets to provide a comprehensive overview of the characteristics of long-term care users.
The 1994 Disability Supplement to the National Health Interview Survey (NHIS-D) Phase I is a household survey of non-elderly community-dwelling individuals. We use these data to examine the characteristics of community-based users and nonusers of long-term care, as well as to compare non-elderly and elderly community care users. This survey collected basic data on disability in 1994 and 1995. The NHIS-D provides the most comprehensive information to date about the prevalence of disabilities in the U.S. population living in the community and the health and demographic characteristics of the community population receiving long-term care. We do not present information about changes in the characteristics of the non-elderly adults living in the community, because there are no comparable data for prior years.
The National Long-term Care Survey (NLTCS) is a periodic nationally representative survey of persons aged 65 and older designed to identify elderly persons who are chronically disabled in one or more basic or instrumental activities of daily living (ADLs/IADLs). Beginning in 1984, data were collected for both community and institutionalized elderly persons. The NLTCS includes detailed information on demographic, health, and disability-related characteristics, including receipt of help for community residents and functional assessment information for institutionalized persons. In this paper, we use the 1994 data to describe the characteristics of the long-term care users aged 65 or older. We compare data from the 1984 and 1994 surveys to examine changes over time. The 1994 survey is also used to examine the characteristics of caregivers of elderly long-term care users.
Two additional data sets are used in this paper. The Asset and Health Dynamics Among the Oldest Old (AHEAD) is an ongoing nationally representative sample of persons aged 70 and older that began in 1993 and provides a wide range of information on health, family, and economic resources. Because wave one of this survey includes information on housing characteristics including services offered to residents, we use these data to describe elderly persons living in housing with supportive services.
The Medical Expenditure Panel Survey Nursing Home Component (MEPS NHC) is a nationally representative survey of nursing homes and their residents in 1996 that gathers detailed information on demographic characteristics, residence history, health and functional status, use of services, and health care expenditures. Current residents on January 1, 1996, are used in this paper to describe the nursing home population aged 18 to 64. It should be noted that we do not analyze data for persons aged 18 to 64 in other institutions, such as State mental hospitals or large intermediate care facilities for the mentally retarded (ICF-MR). These individuals are not included in any national survey frame. Estimates of the size of this population are presented in the discussion.
The definition of long-term care use is largely comparable across the data used in this study:
Because the number of IADLs included in these surveys was not the same, long-term care use is based on a different number of IADLs in the two surveys:
It is important to note that we do not include heavy housework in our definition of long-term care users. We made the judgment that persons only receiving human help with heavy housework (e.g., scrubbing floors, washing windows, and doing heavy yard work) were not sufficiently disabled to be deemed long-term care users.
We also examine the relationship between health, disability, and demographic characteristics and the types of care received by long-term care users. We distinguish three types of care: formal and informal care, each provided in the community, and institutional care.
Formal care is defined in this paper as community care paid for directly by public programs or private individuals or provided by agencies. Informal care is defined as help provided by relatives or friends, regardless of payment. While it is certainly possible for relatives to be paid caregivers, the NLTCS does not ask if relatives are paid; we used a consistent definition that relatives are always informal caregivers in both the NLTCS and the NHIS-D. In addition, in the NHIS-D it was not possible to distinguish caregivers who were private unrelated unpaid individuals from representatives of charitable agencies, so for this survey all unrelated non-resident caregivers are classified as informal.
All estimates presented in tables are weighted using sampling weights that adjust for disproportionate selection probabilities. Unless otherwise noted, only statistically significant differences at the 5 percent level are discussed in the text. Estimates that may be unreliable because of high sampling variability (relative standard errors greater than 30 percent) are also noted in the tables. For the NHIS-D and MEPS NHC, standard errors are adjusted for complex survey design using SUDAAN (Cox and Cohen, 1985). Calculation of standard errors for estimates based on the NLTCS use the generalized variance function method described in the Bureau of the Census "Source and Accuracy Statement." Weighted standard errors are used for hypothesis testing of estimates based on the AHEAD survey.
Table 1 (Text Version) presents estimates of the prevalence of receipt of community-based long-term care among non-elderly adults. Among the more than 158 million people aged 18 to 64 residing in the community in 1994, 2.1 percent (over 3.3 million) received assistance with ADL or IADL tasks. The proportion receiving ADL or IADL assistance was consistently higher in older age groups, rising from a low of 1.1 percent among those aged 18-29 to 4.8 percent in the 60-64 year-old group. A higher proportion of women than men used community-based long-term care (2.4 percent versus 1.8 percent), and the proportion of users among the widowed (5.5 percent) was greater than any other marital status group. Finally, 3.3 percent of black respondents received community-based long-term care, which was higher than the rates for whites (2.0 percent) or for other racial groups (1.8 percent).
Table 2 (Text Version) presents distributions of selected characteristics within each of three non-elderly groups:
In terms of sociodemographic characteristics, age, gender, race, and marital status were related to the likelihood of receiving long-term care. The mean age of those receiving community-based care (45 years) was higher than nonusers (38 years), and the mean age of persons in nursing homes was higher still (51 years). Users of community-based long-term care were more likely to be female than either nonusers or nursing-home residents; the latter two groups were similar. Community care users and nursing home residents were each more likely than nonusers to be black (19 percent versus 12 percent).
Community-based users were less likely than nonusers to be married (47 percent versus 66 percent), and nursing home residents were the least likely to be married (16.4 percent). The modal marital status among non-elderly nursing home residents was never married, while in the other two groups the modal status was married.
Differences among the three groups were also apparent in terms of education, employment, and poverty status. Compared with nonusers, users of community-based long-term care had less education, and non-elderly nursing home residents had even less education. Community-based long-term care users were less likely than nonusers to be employed (18.7 percent versus 74.9 percent). Although the employment rate among community care users was relatively low, it is notable that almost one-fifth of community-based users had a job, despite receiving help with ADLs or IADLs. Community-based care users were also more likely than nonusers to be poor: over one-quarter of users were below the poverty threshold, compared with one-tenth of nonusers. (Employment and poverty status were not assessed for nursing home residents.)
Turning to aspects of health status, the majority of community-based long-term care users (55.9 percent) received help with only IADL tasks, and only 17.4 percent received help with three or more ADL tasks. As expected, a very high proportion (75.5 percent) of nursing home residents required help with three to six ADLs. Nearly 8 percent of non-elderly adult nursing home residents, however, received help with only IADLs.
The data also reveal the importance of lower-body functional impairments, defined as having difficulty lifting 10 pounds, walking up 10 steps, walking a quarter of a mile, standing for about 20 minutes, or bending down to pick up an object.
Nearly three-quarters of users reported difficulty in one or more of these activities, while less than 40 percent reported difficulty with upper-body movements (reaching up over the head, picking up a glass, holding a pen or pencil). Less than 2 percent of users reported upper body limitations in the absence of lower body limitations. It is noteworthy that over one quarter of users of community long-term care (26.2 percent) reported no limitations in either upper or lower body functioning, while over 7 percent of nonusers reported some functional limitation.
The key role played by lower body functional limitations is also revealed by the findings that one-quarter of community-based long-term care recipients used mobility equipment, such as a cane, crutch, or walker, and 12.7 percent used a wheelchair. In contrast, less than 1 percent of those not receiving long-term care used mobility aids or wheelchairs. Users were also substantially more likely to have difficulty communicating, difficulty understanding others, and mental retardation compared with nonusers.
Non-elderly nursing home residents were more impaired than community-based care recipients. Roughly two-thirds of these nursing home residents used a wheelchair (65.2 percent), but fewer used other mobility aids, compared with community care users (13.0 percent versus 25.0 percent), reflecting the greater immobility of the nursing home population. Nearly a third of non-elderly adult nursing home residents had difficulty seeing (30.2 percent); this rate was notably higher than for community care users. The proportions of nursing home residents with difficulty hearing, communicating, or understanding others were slightly larger than corresponding proportions among community care users.
Table 3 (Text Version) presents information on the type of care—formal, informal, or both—received by community-dwelling non-elderly adults. Nearly three-quarters of non-elderly adult users of long-term care received only informal assistance. The proportions receiving only formal care and both formal and informal care were much lower and roughly equal (5.5 percent and 5.8 percent, respectively). The type of care could not be determined for 17.7 percent of NHIS-D respondents, who were sometimes unsure if the person providing help with ADL and IADL tasks was a relative or whether the helper was paid. Thus, for a number of respondents the nature of the help provided could not be ascertained, resulting in their being assigned to a "missing" category.
A series of cross tabular analyses examined factors that could be related to the type of care a person received. The interpretation of these analyses was complicated by the presence of the substantial number of persons for whom type of care could not be determined. Because differences in rates of missing information make interpretation of the results difficult, as a first step, for each variable, the proportions with missing data were compared for each pair of categories. Significant differences in the proportions of cases with missing data were obtained only for age, marital status, and dizziness. Results for these variables are not discussed.
Among sociodemographic variables, only education was related to the type of care received. Persons with less than 12 years of education were least likely to receive any formal care (by itself or in combination with informal care), while those with more than a high-school education were most likely to use any formal care. Gender was unrelated to the type of care received. Compared with whites, a smaller proportion of blacks received both formal and informal care, but a larger proportion received only formal care.
Persons in the community receiving help with three or more ADLs had the lowest proportion receiving only informal help, significantly lower than those with IADLs only (62.5 percent versus 73.7 percent); on the other hand, this group was more likely to receive combined formal and informal help than either those with IADLs only or those with one to two ADLs. These results suggest that care requirements for persons with three or more ADLs are sufficiently complex and demanding as to make informal help alone insufficient.
The pattern of functional limitations also influenced the type of long-term care received. In particular, functional limitations involving the lower body were associated with increased probability of receipt of formal long-term care. Persons with only upper body limitations overwhelmingly received only informal care; no cases occurred in which someone with only upper body limitations received formal care alone. Persons with both upper and lower body limitations were most likely to receive formal care, either by itself or in conjunction with informal care. Similarly, using a wheelchair or mobility equipment, which also reflect lower-body limitations, were each associated with a greater likelihood of receiving formal care (both by itself and combined with informal care).
Other variables, including employment status and difficulties with vision, hearing, communication, and understanding, were not significantly related to type of long-term care received by community dwelling adults aged 18 to 64.
In 1994, nearly 17 percent of the 33 million persons aged 65 or older were receiving long-term care (Table 4, Text Version). As has been true in the past, the likelihood of using long-term care increased rapidly with age. Only 6.5 percent of those aged 65 to 69 received assistance, compared with 27 percent of 80 to 84 year-old persons and 80 percent of those aged 95 or older. The likelihood of being in a nursing home also rose with age, from less than 1 percent for those aged 65 to 69 to more than 43 percent of those aged 95 or older. Most persons of all ages received their care in the community. Only persons aged 95 or older appear to have been more likely to be receiving care in a nursing home than in the community, but the difference is not statistically significant.
Women, blacks, and those who were widowed or never married were more likely to receive long-term care. Women, widows, and those who never married also were more likely to be in institutions than were men and those who were married. Finally, although blacks were more likely than whites to receive care in the community, they were less likely to be in institutions.
Between 1984 and 1994, the number of elderly persons receiving long-term care remained essentially constant at 5.5 million persons (Table 5, Text Version), while the prevalence of long-term care use declined from 19.7 percent to 16.7 percent. In fact, the prevalence of any long-term care use declined for all age groups among the elderly (not shown).
Accompanying the decline in prevalence, however, was a striking increase in the level of disability among those who received help. The percent receiving help with three to six ADLs increased from 35.4 percent to 42.9 percent, while the percent of care recipients receiving only IADL help dropped nearly 10 points and accounted for virtually all the decline in the receipt of long-term care. The prevalence of help with ADLs among the total elderly population remained virtually constant at about 12 percent (not shown).
The prevalence of cognitive impairment rose over the decade. In contrast, the prevalence of functional limitations (upper and lower body difficulties) and difficulty with seeing and communicating remained virtually unchanged.
Consistent with the increase in the level of disability among long-term care users, there was an increase in the use of both formal care in the community and institutional care. The increase in formal care use, however, occurred primarily among those also receiving informal care; the proportion using formal care only was essentially unchanged. The percent receiving only informal care from family or friends declined substantially, from 51.2 percent in 1984 to 40.1 percent in 1994.
Although the number of elderly community long-term care users dropped, the number in institutions actually rose from 1.4 million to over 1.6 million, yielding an unchanged institutional prevalence of about 5 percent (Table 6, Text Version). As with the combined population, mean age in each setting rose by about a year, but other demographic characteristics generally were stable. The exceptions were a small increase in the proportion of community long-term care users who were women, a modest increase in the proportion of community long-term care users who were divorced or separated, and a larger decrease in the proportion of institutional residents who were never married.
The increase in disability level seen in the combined population also occurred within both community and institutional settings. An 11 percentage point decrease in IADL disability in the community (from 49.7 percent to 38.9 percent) was accompanied by almost equal increases in the proportions of persons receiving help with one to two ADLs and three to six ADLs. Among institutionalized persons, there was no significant drop in IADL disability, but level of disability among those receiving ADL assistance increased. In 1994, nearly three-quarters of nursing home residents received help with three to six ADLs, up from about two-thirds in 1984. The prevalence of cognitive impairment also rose in both the community and institutional long-term care populations.