Review of Literature
An estimated 4.32 million Americans in 2000 had an intellectual or developmental disability (I/DD) and approximately 10%, or 433,799 of these persons, lived in supervised residential settings (Braddock et al., 2000; Fujiura, 1998). Since the 1970s most states have aggressively reduced their reliance on institutional settings for persons with I/DD in favor of smaller, community-based options. However, in 2000, 116,527 persons with I/DD still lived in congregate settings -- 47,374 people in state institutions, 34,410 people in private institutions, and 34,743 persons in nursing facilities (NF). Nursing facility utilization by individuals with I/DD varies greatly among the states. The proportion of each state's total residential service system served in nursing facilities ranged from less than 1% in Alaska, Idaho, and Kansas, to over 20% in Alabama, Georgia, Kentucky, and Oklahoma (Braddock, Hemp, Rizzolo, Parish & Pomeranz, 2002).
The positive association between community placement from congregate facilities and the acquisition of adaptive skills by persons with I/DD has been frequently documented ( Conroy, Efthimiou & Lemanowicz, 1982; Eastwood & Fisher, 1988; Larson & Lakin, 1989). Increased health benefits have also been realized (Heller, Factor & Hahn, 1999; Heller, Factor, Hsieh & Hahn, 1998). Additionally, smaller community settings are usually preferred by self-advocates (Nelis, 1995; Nelis & Ward, 1995).
Evolving ideological beliefs hold that "large segregated institutional settings [can] not provide proper care and support for individuals with intellectual disabilities (Heller, 2002, p. 504). This is particularly true for nursing facilities as they are designed to provide nursing care as opposed to habilitation or active treatment (Mitchell & Braddock, 1990; Uehara, Silverstein, Davis & Geron, 1991) and the majority of individuals with I/DD living in nursing homes have been found to not require round the clock nursing care (Lakin, Hill & Anderson, 1991). Additionally, an institutional environment may predispose nursing facility staff, in an effort to save time, to do tasks for persons with disabilities that they could otherwise do independently (Kane, 1996). Researchers have also documented increased medication usage in these settings, which supports a possible tendency of nursing facility staff to treat individuals with medical interventions as opposed to behavioral ones (Spreat & Conroy, 1998).
Though the advantages of community integration for persons with I/DD from both functional and theoretical perspectives have been frequently noted in the literature, every state in the U.S. continues to utilize nursing facilities as a residential option for persons with I/DD. Since the federal government does not dictate precisely the settings in which states must provide residential services to people with I/DD (Braddock, 1987) -- states have great latitude in deciding whether to utilize institutional or community services (Braddock, Hemp, Parish & Rizzolo, 2000). This latitude may increase as the federal government continues to devolve authority to the states.
There is no existing research on 1) the factors associated with a state's decision to utilize nursing facilities as a significant part of its service delivery system or to opt for more community based facilities or 2) on the extent of declines in utilization since the passage of the OBRA '87 legislation; and 3) and if so, the predictors of change in the states.
Data Collection and Measures
The first variable, state wealth, will be operationalized as personal income per capita. Personal income is a good indicator of a state's economic capacity to fund programs and services. Some economists believe that states with greater amounts of "slack resources" will have more resources available to fund social programs such as community integration.
The second variable is a state's political culture. Many researchers assert that differing policy outcomes in the states can be attributed to historical and cultural variables (Gray, 1996; Hanson, 1991; Wildavsky, 1985). The study will utilize Sharkansky's index (1969), a measurable scale based on a political typology developed by Elazar (1966), which defined each state's political culture as moralistic, individualistic, or traditionalistic. Political culture has been found to be a significant predictor of community integration for juvenile offenders (Downs, 1976) and persons with I/DD (Sigelman et al., 1981). Sigelman and his colleagues found that "traditional" states were less likely to move their citizens with I/DD from state institutions than were "individualistic" or "moralistic" states. Sigelman et al. (1981) theorized that traditionalistic cultures were opposed to deinstitutionalization because it involved a redistribution of money and individuals, upsetting the status quo.
The third variable in the model is a state's legislative professionalism. Highly professional legislatures are characterized by full time staff members, highly paid members, and substantial staff support (Patterson, 1996Legislative professionalism, which has been measured using the Citizens Conference on State Legislatures (CCSL) index has been found to be a significant predictor of state deinstitutionalization of persons with I/DD(Sigelman et al., 1981). A recent update of the CCSL index (Squire, 1992a, 1992b, 2000) will be utilized.
The fourth variable, advocacy, will be operationalized as per capita membership in the state Arc chapter. Braddock et al. (2000) documented the active role state advocacy organizations have played in litigation regarding the expansion of community services in the states. Advocacy has been found to be a significant predictor of state spending for community LTC for persons with I/DD (Braddock & Fujiura, 1991; Braddock et al., 1998, 2000) and has long been recognized as an important means of restructuring states' service delivery.
The fifth variable will be state participation in the DD Home and Community Based Services (HCBS) waiver program, which since 1981 has allowed states to increase the federal support for community services and may reduce nursing facility use (Doty, 2000). Participation in the Waiver has been shown to be a significant predictor of community LTC spending for persons with I/DD (Braddock, 2002a, 2002b; Braddock & Fujiura, 1987, 1991; Braddock et al., 1998), physical disabilities, and mental illness (Braddock, 2002a, 2002b).
The last variable, nursing home utilization for the general population, has not been studied previously. In 2001, 1,311,465 persons resided in certified nursing facilities (Harrington et al., 2002).
A second goal will be to examine characteristics of the persons with I/DD served in nursing homes in each state using descriptive MDS data supplied by the University of Minnesota's Institute on Community Integration. The OBRA '87 legislation included a mandate that all facilities receiving Medicare and Medicaid funding conduct periodic evaluations on each of their residents. As a result, the Health Care Financing Administration, now the Centers for Medicare and Medicaid Services (CMS), produced a standardized assessment instrument, the Resident Assessment Instrument (RAI) that includes the Minimum Data Set (MDS).
Information from this dataset for persons with I/DD in each of the states will be obtained on the average age, presence of Alzheimer's, presence of severe mobility limitations, presence of frequent or total incontinence, and use of a feeding tube. These data will be collected by University of Minnesota researchers and provided to the proposed study's research team in aggregate form so as to protect confidentiality.
Tools Being Developed
Tools Being Adapted
Data Collection Status
Data collection complete.
No project modifications.
Project Progress to Date
To test the first five research questions, I constructed a political-socioeconomic model with the following variables: personal income, political culture, legislative professionalism, advocacy (measured as membership in the state Arc chapter), state spending for the HCBS Waiver for persons with developmental disabilities, and utilization of nursing facilities for the elderly. The variables were entered into a hierarchical regression model.
1. What variables are associated with state utilization of nursing facilities for persons with developmental disabilities in 1991?
The results indicate that in 1991, the full model with all six independent variables entered accounted for 29% of the variance. Twelve percent of the variance was explained by political culture. More traditionalistic states were more likely to utilize nursing facilities for persons with developmental disabilities. This may be due to traditionalistic states’ reluctance to reallocate wealth or services, as it may result in a fundamental change in the status quo. More moralistic states, on the other hand, are more likely to support “deinstitutionalization,” possibly because it has the potential to reduce unnecessary expenditures, serve a greater number of persons, promote independence and choice, and create a more equitable service system. Five states used nursing facilities for 30% or more of their DD residential service system in 1991. Each of these states had a traditionalistic political culture.
2. What variables are associated with state utilization of nursing facilities for persons with developmental disabilities in 2000?
In 2000, the full model explained 34% of the variance in utilization of nursing homes for persons with developmental disabilities. Political culture explained 22% of the variance while nursing home utilization for the elderly explained an additional 8%. Again, more traditionalistic states used nursing homes more, while more moralistic states utilized this option to a lesser degree. Four states used nursing homes for 20% or more of their service system. Each of these states (Alabama, Georgia, Kentucky, and Oklahoma) were more traditionalistic. Nine additional states used nursing facilities for 10-20% of their system. Of these nine, three were individualistic, and six were traditionalistic. No predominantly moralistic state utilized nursing homes for more than 9% of their DD residential system in 2000. Nursing home utilization for persons with DD was also positively associated with nursing home utilization for the elderly.
3. What factors are associated with changes in state utilization of nursing facilities for persons with developmental disabilities since passage of OBRA ’87?
To determine the predictors of change in utilization of nursing facilities for persons with developmental disabilities since passage of OBRA ‘87, I controlled for the initial level of utilization in 1991, the baseline year. This took into account the inherent correlation between the level of utilization in 2000 and the level of utilization in 1991 (Bohrnstedt, 1969; Markus, 1979). As predicted, utilization of nursing homes in 2000 was positively associated with nursing home utilization in 1991 (accounting for 73% of the variance). Political culture also reached statistical significance, though it only accounted for an additional 3% of the variance. More traditionalistic states used nursing homes to a greater degree than more individualistic or more moralistic states, even after taking into account the utilization rates in 1991.
4. What variables are associated with state utilization of state institutions for persons with developmental disabilities in 1991?
In 1991, the full model explained 48% of the variance between the states in utilization of state-operated institutions for persons with developmental disabilities. 38% of the variance was attributed to political culture while 8% was attributed to spending on the HCBS Waiver. More traditionalistic states were associated with greater use of state institutions than were more individualistic or more moralistic states. State spending on the HCBS Waiver was positively associated with state institution use.
Income served as a suppressor variable. Though not significant in the bivariate analysis, it became significant after the Waiver was introduced into the model. Higher personal income was associated with greater use of state-operated institutions. Legislative professionalism also became significant after the Waiver variable was added. Less professional legislatures, as in the Sigelman et al. (1981) model, were associated with increased utilization of state-operated institutions in FY 1991. The majority of states that utilized state institutions for 30% or more of their DD residential service system in FY 1991 were traditionalistic (11 of the 14 states fell into this category). In addition, these 11 traditionalistic states were less professional and spent less on the HCBS Waiver for persons with developmental disabilities. South Carolina and Mississippi utilized state institutions to the greatest extent—for 50% or more of each of their systems in FY 1991. These states scored low on Squire's professionalism index, and had not begun utilizing the HCBS Waiver.
5. What variables are associated with state utilization of state institutions for persons with developmental disabilities in 2000?
In 2000, the model explained 57% of the variance. Political culture explained 44%, while the waiver explained an additional 9%. Again, income, political culture and the Waiver reached statistical significance. Legislative professionalism was no longer significant as it had been in the 1991 model. In 2000, ten states utilized state institutions for 20% or more of their DD residential system. All of these states, with the exception of New Jersey, had predominantly traditionalistic political cultures. Furthermore, these states spent less on the Waiver than most states ($8.09 per capita versus $15.51 nationally).
These findings illustrate the robustness of the political culture variable. Political culture has been found to be a significant predictor of numerous public policies, including deinstitutionalization of state-operated institutions in 1977 (Sigelman et al., 1981) and facilities for juvenile offenders (Downs, 1967).
6. What are the characteristics of persons with developmental disabilities in nursing facilities in the states?
The average age of persons without developmental disabilities in nursing facilities across the US in 2000 was 80 years. The average age of individuals with developmental disabilities in nursing facilities that same year was significantly lower—64 years. The average age for persons with developmental disabilities in these settings ranged from 57 years in Oregon to 73 years in Rhode Island.
In 2000, approximately 13% of all persons without a developmental disability living in nursing facilities across the nation were diagnosed with this form of dementia. The population of persons with developmental disabilities with Alzheimer’s in these facilities, however, was significantly lower: less than 6%.
The low rate of Alzheimer’s disease in the population of persons with developmental disabilities in nursing homes found in the present study may be explained by the reduced age of this group compared to the general nursing home population.
Given the difficulty in diagnosing dementia in persons with cognitive disabilities, it is possible that many individuals are not being properly diagnosed.
Possibly a form of diagnostic overshadowing?
In 2000, almost 7% of persons without developmental disabilities in nursing facilities were bedfast all or most of the time. The proportion of persons with developmental disabilities who were bedfast in these settings was 6%.
Results from the MDS indicate that almost half of all individuals with developmental disabilities in nursing facilities in the states are occasionally, frequently or completely incontinent. For persons without developmental disabilities in nursing facilities, approximately 42% are incontinent.
We can assume, however, from the extensive literature on staff shortages and staff burden that many individuals who could benefit from continence training may not be receiving it. The time and effort required may be too taxing for an already overburdened staff.
Unknown what proportion of incontinence is due to medical conditions, side effects of medication, etc.
Results from the MDS indicate that the proportion of persons with developmental disabilities requiring a feeding tube in nursing facilities in the states is significantly higher than the proportion of persons without developmental disabilities in these settings—11.9% compared to 7.2%, respectively.