Assisted Living and Alternate Family Care Evaluation
EXECUTIVE SUMMARY
INTRODUCTION
Assisted Living (AL) facilities and Alternate Family Care (AFC) homes offer residential care in nonnursing home settings. These settings allow seniors to age in place by altertng the level of assistance, as the persons' needs change. A Medicaid waiver, approved in 1996 and renewed for 5 years in 1999, created 1,500 Medicaid-funded slots in these types of settings in New Jersey. Currently, about 1,000 of these slots are utilized; program staff are projecting the 1,500 slots will be full by late fall/early winter 2001.
Toe Center for State Health Policy (CSHP) is working collaboratively with the Division of Con-sumer Support to assist with these efforts. One potential way to encourage facilities to utilize more Medicaid-funded slots is by restructurtng the payment methodology. Currently, the DHSS pays a flat fee based on the type of setting. This method is in contrast to industry standards whereby facilities either charge a flat fee based on the client's assessed needs, or charge for the actual services received throughout a billing period. Under consideration is a plan to adopt a model that reimburses facilities based on the assessed-need category of the client. This model would place seniors into a set number of tiered categories. To consider this type of change in payment policy, information must be obtained about the current population, methodologies that might be used to develop these categories of need, and the potential impact of alternative payment strategies upon the New Jersey Medicaid-funded slots.
METHODS
To inform the development of this needs-based tiered system, we begin with a brief background describing Assisted Living and literature on long-term care reimbursement methodologies. Then, we review reimbursement initiatives used in other states, and summarize our discussions with representatives from New Jersey's Assisted Living industry. Finally, we conclude with a recommended methodology for a New Jersey needs-based tiered system.
RESULTS
Assisted Living and Long-term Care Reimbursement Methodologies
Assisted living provides a mixture of housing, personalized support services and health care in a noninstitutional setting for both senior citizens and portions of the younger disabled population. Although the assisted living industry is relatively new, with more than half of assisted living facilities having been in business for less than ten years and one-third in business for less than five, these facilities generate about $15 billion per year with fees ranging from as low as $1,000 a month to as high as $5,000 a month (Hawes et al., 1999) with two-thirds of the rooms paid for privately.
In general, assisted living facilities care for elderly persons who either do not want to live in their own home or are no longer capable of living in their own home. The majority of services they need include medication management, meals, recreation, and help with activities of daily living
(ADLs) and instrumental activities of daily living (IADLs).
Since most AL facilities cater to private paying residents, these facilities are able to set strict admission criteria. Admission and retention policies are of particular importance to a state's reimbursement strategy, as some individuals may not be admitted. For instance, many people with diseases such as Alzheimer's, diabetes, or arthritis may need assistance with ADLs such as bathing, dressing, preparing meals, and taking medications, but may not be accepted into AL facilities because of their higher and sometimes more complex needs (Brown, 2000). The assisted living industry is caught in a bind where it is not fmancially beneficial to accept individuals who lie on the borderline of the aging-in-place environment, yet require more attention than the average resident.
State governments find the idea of using public funds for assisted living costs attractive because reimbursement rates in these alternative long-term care facilities are usually less than the cost of traditional nursing home care. Although some seniors can afford the monthly charges and sometimes high entrance fees which many facilities charge, those with less means have to either spend down their assets and apply as a Medicaid-funded resident, or are simply not able to afford assisted living. However, government funding of AL programs could prove costly for states unless they are cautious about reimbursement rates, facility regulations regarding services provided, and admission and retention policies.
Review of Literature on Long-term Care Reimbursement Methodologies
Although needs-based systems are frequently used by AL facilities to determine their own charges, most of the literature regarding needs-based structures is derived from the nursing home literature on case-mix reimbursement strategies.
Case-mix systems can be classified as either resident or facility-level systems. Facility-level systems use measures such as age, payer, and average length of stay distributions within a facility. Resident-level systems use resident variables such as functional status often measured by ADLs, health status, behavior/ cognition, and use of special services. The trend for case-mix systems in nursing homes is exemplified by states such as Maryland, Minnesota, and New York. Each of these states has a long-standing program of using ADLs, special care services, and behavioral status. As these same measures are often used by AL facilities to structure their own needs-based tiered system; these factors are critical to include in a needs-based tiered system for Medicaidfunded residents.
To assess other states' reimbursement strategies, CSHP conducted a telephone and mail survey of all 50 states (response rate= 100%). Eleven (22%) states reported using case-mix (n=9), or a modified case-mix (n=2) reimbursement mechanism for AL. Specific information was collected for 9 states (Arizona, Arkansas, Delaware, Maryland, Minnesota, New York, Oregon, Vermont, and Washington).
Common factors across states were evident. For example, most states use 3 payment levels, all of these states used ADLs, and a majority of states used medication management, special services, and cognitive status. These factors are also consistent with a trend toward Federal and State case-mix payment systems for nursing homes. These criteria were also mentioned in our discussions with representatives from New Jersey's AL facilities.
Recommended Variables for the New Jersey Needs-based Tiered System
Based upon our review of the AL and long-term care reimbursement systems literature, the survey of other states, and our discussions with the New Jersey Assisted Living Association (NJALA), we recommend using ADLs/IADLs, cognition/behavior, and special services. Since New Jersey currently collects these variables (on the Comprehensive Assessment Tool-CAT), to assess the medical and social needs of Medicaid (and Medicaid-eligible) persons, we will use these data to develop the needs-based tiered system. Data analysis will be used to refine which variables will be included in the recommended model. The following discussion describes these recommendations, as well as the analytic approaches that will be used.
Activities of Daily Living
Activities of daily living are commonly used to determine reimbursement and staffing levels
(Lazaridis, Rudberg, Furner, and Cassel, 1994). One important factor to consider, however, is how ADLs are measured. ADL scores are most often used in three ways: as simple counts, ordered, or hierarchical. Each of these approaches has its own advantages and disadvantages.
Simple Counts
It is common to find simplified ADL scoring (Bennett, 1999). A cut-off is used below which individuals are rated as dependent (e.g., as used by the Pepper Commission), and then dependence in each ADL is counted equally. This method creates two problems. First, some ADLs such as bathing and toileting require more personal assistance or staff time than others such as dressing. Second, determining dependence can be confusing because different assessors/researchers use different thresholds (Jette, 1994).
Ordered ADLs
It is often noted that decline and recovery from illness in later life follows development in childhood. That is, the least complex tasks are acquired first and retained the longest; whereas, the most complex were acquired later and lost sooner (Travis and McAuley, 1990). ADLs follow this same pattern with the least complex task being feeding and the most complex bathing. For this reason they are often used in an ordered format.
Hierarchical ADLs
Katz et al. (1963) developed a hierarchy (also known as a Guttman scale) of ADLs (Lazaridis, Rudberg, Furner, and Cassel. 1994). This method weights or scores certain ADLs so that they contribute more significantly to a cumulative score.
To improve the effectiveness of hierarchical models to measure the need for care, different weighting schemes have·been developed. Finch et al. (1995) developed a model whereby the weight assigned to each component is self-contained. Thus, individual ADL items can be omitted. Spector, Katz, Murphy, and Fulton (1987) developed a combined ADL-IADL1 scale. This scale has a broad range of levels, especially when including the non-institutionalized populations. The combined scale consists of the IADLs shopping and transportation and the AD Ls bathing, dressing, transferring, and feeding.
Although ADLs have been universally used, they have also been criticized (Bennett, 1999; Fried et al., 1996; Siu, Reuben and Hays, 1990). In spite of these concerns, ADL measures are key in most payment methodologies. Thus, we propose to test various models (Katz; Siu, Reuben and Hays; Finch et al.; Spector Katz, Murphy, and Fulton) with the CAT data.
Cognition
People with dementia often fall through the cracks, especially with aging-in-place initiatives. One major thrust of policy makers and the assisted living industry is to provide dementia care. Clearly, some dementia residents do not need nursing home care. But, most patients with dementia do require a safe environment with skilled staff (Pieffer, 1997). The ability to care for some dementia residents further reinforces a role for assisted living. Therefore, a "fair" payment system should account for these residents.
To appropriately account for dementia we will determine whether this measure is required to account for the care time needed for these residents, and then determine the most appropriate metric for cognition.
Special Services
Seniors may also receive special services to maintain or improve their health status. These services may include such factors as pain management, medication by injection, and oxygen therapy. These services are also not necessarily accounted for in ADL scales or by cognition problems, and in some cases can be associated with higher levels of caregiver time. We will first determine whether this measure is required to account for the care time needed for these residents, and then determine the most appropriate metric for these services.
CONCLUSIONS
In general, we found that ADLs, cognition, and special care needs are especially relevant for a needs-based tiered system. These elements were found not only in the long-term care literature regarding nursing home case-mix reimbursement schemes, but also in our own survey of state AL reimbursement mechanisms. Additionally, these factors seem to be consistent with those factors used by the AL industry in determining tiered rates, or reimbursement criteria. Moreover, representatives from New Jersey AL facilities verified these factors as critical in developing a needsbased tiered system.
Therefore, we will include ADLs, cognition, and special services in our analyses. Since these factors are highly associated with IADLs and continence, and the ADL measurement literature includes these as intervening factors, we will also include these two variables:
In a subsequent report, we will be refining how these factors will be used in our models. Additionally, we·will be analyzing data from current assisted living residents as well as a population of seniors who were screened for Medicaid long-term care services using these factors.