The Effect of Licensure Type on the Policies, Practices, and Resident Composition of Florida Assisted Living Facilities

Abstract

Purpose: Most assisted living facility (ALF) residents are White widows in their mid- to late 80s who need assistance with activities of daily living (ADLs) because of frailty or cognitive decline. Yet, ALFs also serve younger individuals with physical disabilities, traumatic brain injury, or serious mental illness. We compare Florida ALFs with different licensure profiles by admission–discharge policies and resident population characteristics. Design and Methods: We use state administrative data and facility survey data from the Florida Study of Assisted Living (FSAL) to classify ALFs by licensure type and to determine how licensure influences ALF policies, practices, and resident population profiles. Results: Standard-licensed traditional ALFs primarily serve elderly White women with physical care needs and typically retain residents when their physical health deteriorates. Some ALFs that hold specialty licenses (extended congregate care and limited nursing services) offer extra physical care services and serve an older, more physically frail population with greater physical and cognitive challenges. ALFs with limited mental health (LMH) licenses serve clientele who are more racially and ethnically diverse, younger, and more likely to be men and single. LMH facilities also have a significant proportion of frail elder residents who live alongside these younger residents, including some who exhibit behavioral problems. LMH facilities also employ discharge policies that make it more difficult for frail elderly residents to age in place. Implications: These differences by facility type raise important quality of life issues for both the frail elderly individuals and assisted living residents who do not fit the conventional demographic profile.

The majority of assisted living facility (ALF) residents nationwide are White widowed women in their mid- to late 80s who need assistance with activities of daily living (ADLs) because of physical frailty or cognitive decline (Ball et al., 2004; Chou, Bouldy, & Lee, 2003; Hawes, Rose, & Phillips, 1999). Assisted living (AL) residency often represents a preferred alternative to nursing home (NH) placement for this vulnerable population (Ball et al., 2000; Bishop, 1999; Street & Quadagno, 2004). To date, most research on ALFs has focused almost exclusively on facilities that serve the dominant client group, frail elders (Golant, 2004; Hawes et al., 1999; Zimmerman et al., 2003), to the relative neglect of research on special populations (Hernandez & Newcomer, 2007) or younger individuals with disabilities (but see Heumann, 1996; Salmon, Hyer, Hedgecock, Zayac, & Engh, 2004). This overriding focus on the “conventional” AL resident profile, although understandable, has contributed to a knowledge gap, compounded across long-term care settings (DeJong, Palsbo, & Beatty, 2002). Further complicating the picture is how AL has been transformed since its inception from a particular vision of residential care for elderly individuals into a care setting with many variants (Kane, Chan, & Kane, 2007).

Although the proportion of nonelderly AL residents nationwide currently is relatively small (estimates range from 10% to 20%), the proportion of nonelderly residents in NHs has been growing (Centers for Medicare and Medicaid Services, 2005; Miller &Weinstein, 2007). For a variety of reasons we explore in the following, ALFs are likely to experience a similar increase in their younger resident populations. Yet, little is known about the policies and practices of facilities that include large numbers of nonelderly residents and how those conditions may be shaped by the residential circumstances created by state regulations.

In fact, the range of different types of residential facilities and programs that serve frail elders and other disabled individuals varies substantially across the country. Facilities that provide such care and services are variously called ALFs, adult family care homes, adult foster homes, homes for the aged, residential care facilities, and myriad others, depending on the state. Throughout the country, such adult residences with services serve not only a dominantly elderly population but also sizeable adult populations with other physical and mental care needs. Further variation across the states is expressed in how each uses Medicaid resources to compensate providers for care. Most, but not all, states have Medicaid home- and community-based services waivers, aged and disabled waivers, or AL for elderly waivers that pay for some services in AL (National Center for Assisted Living [NCAL], 2008; Chapin & Dobbs-Kepper, 2001). Unlike federal standards applied nationwide to NH care, AL regulations have developed in state-specific contests over time. These regulations reflect conditions relating to residential care for elders and other disabled persons in particular states at the time that regulations were enacted, and later adoptions of regulations often mirror policy developments in other states’ best practices in AL regulations (Frank, 2002). Like Florida, many states have different types of licensure statuses, usually reflecting the intensity of services offered to different aged resident populations. In contrast to Florida, few states have specially licensed facilities catering to the needs of mentally ill adults, with Wisconsin, New Mexico, and Michigan as notable exceptions. In fact, most states have little or no mention of mental illness in their AL regulations at all, aside from a few states’ regulations that prohibit the admission of residents who are a danger to themselves or others, or who are “incompatible” with other residents (NCAL, 2008).

Florida is an ideal site for comparing facilities that serve mainly traditional elderly populations with facilities that meet more specialized resident needs. Florida is one of three states, the others being California and Pennsylvania, that account for 33% of all ALF units nationwide (Mollica, Johnson-LaMarche, & O’Keeffe, 2005). Florida also has a diverse AL industry. In Florida, AL is broadly defined as a residential environment that provides a minimum bundle of services including meals, supportive services, and 24-hr on-site supervision at a particular location. Within this general definition, Florida has four separate licensure categories: a standard license category for ALFs that offer basic services and three specialty licenses that enable ALFs to offer extra services beyond those covered by the standard license.

Facilities holding a standard license provide housing, meals, and personal care services, including supervision of ADLs and self-administered medication. Facilities with limited nursing services (LNS) licensure provide all standard services and may provide additional nursing services such as changing routine dressings, conducting passive range of motion exercises, applying heat and ice caps, cutting toenails, and so forth. Facilities with extended congregate care (ECC) licensure can offer extra physical care services beyond the routinely offered personal care. These extra services allow residents of ECC-licensed facilities to age in place despite physical limitations that might disqualify them from residency in a facility holding a standard or an LNS license. Finally, facilities with limited mental health (LMH) licenses can provide programs of specialized behavioral care for individuals diagnosed with a mental illness (such as schizophrenia or bipolar disorder) or for individuals with age-related organic brain disorders (such as Alzheimer’s and some dementias). Regardless of licensure type, ALFs are prohibited from offering most of the skilled medical services that are provided in an NH. Table 1 summarizes the Florida AL licensure categories.

Table 1.

Florida Assisted Living Licensure Categories

Licensure type	Requirements	Analytic category
Standard	Can provide housing, meals, and personal care services including supervision of activities of daily living and self-administered medication	Traditional
Limited nursing services (LNS)	May provide all standard license services and additional nursing services such as changing routine dressings, conducting passive range of motion exercises, applying heat and ice, cutting toenails, etc.	High frailty
Extended congregate care (ECC)	May provide all standard and LNS services. ECC license allows residents to age in place despite physical limitations that might otherwise disqualify them from residency under a standard or an LNS license
Limited mental health	Assisted living facilities may serve 3 or more mental health residents defined as individuals with Social Security disability income or SSI due to mental disorder	Behavioral

Note: SSI = Supplemental Security Income.

Licensure type, because it determines the kinds of services offered, can have a significant effect on an ALF’s institutional practices and on the characteristics of its resident population. Most of Florida’s ALFs hold either just the standard license or the standard license combined with an LNS or ECC license. These facilities are able to meet the physical care needs of a conventional aging population of AL residents. To the extent that nontraditional residents—such as younger residents with disabilities who require assistance with personal care—need the same kinds of services, they can be adequately served in such facilities. Assisted living facilities holding LMH licenses are capable of serving both older residents and younger nontraditional residents in a setting that is desirable, both because it is less restrictive than a mental hospital or an NH and because it costs less (Heumann, 1996). Thus, the various types of AL licenses enable facilities to serve a range of resident needs.

Although the predominant AL population consists of frail elderly individuals, the proportion of younger residents in AL has been growing due to broad changes in the long-term care environment as a whole. One important factor contributing to this trend has been efforts by states to reduce Medicaid expenditures, which have been increasing rapidly and squeezing out funds for other social needs. In 2003, Medicaid constituted 16% of state budgets on average (Smith et al., 2004). In response to rising costs, many states applied for 1915c waivers, which allowed them to shift funds from providing institutional care in NHs to home- and community-based services (HCBS). Under a waiver, a state can provide services that are not covered by the state Medicaid plan to participants who meet that state’s NH level of care criteria. Specifically, a waiver allows a state to provide a greater amount, duration, and scope of services as well as the ability to set eligibility at up to 300% of the federal supplemental security income (SSI) benefit. Although states cannot use Medicaid funds to pay for basic housing expenses or food in AL, they can reimburse an ALF for medically necessary assistive services provided to qualified residents (Mollica et al., 2005). The result has been an increase in HCBS spending in ALFs through Medicaid waivers, with much of the growth being in services for people with mental retardation and developmental disabilities. States may also pay for services for this population through the state Medicaid plan where NH eligibility is not required but where the maximum income eligibility level is 100% of federal poverty level or the state’s medically needy standard. In 2004, 29 states were providing services through waivers only, 6 states through the state plan only, and 8, including Florida, through a combination of the two (Mollica et al., 2005). Florida has an Assisted Living for the Elderly (ALE) waiver and Assistive Care, the state Medicaid program that provides necessary services for many very low-income AL residents (http://www.floridaaffordableassistedliving.org/consumer/assistivecare.html).

Another factor that has contributed to the increase in younger AL residents is the decline in mentally ill residents in NHs. During the 1970s, NHs had become a “dumping ground” for the mentally ill due to the closing of many mental hospitals in the aftermath of deinstitutionalization (Mechanic & McAlpine, 2000). This practice was brought to a halt following the 1999 Supreme Court Olmstead decision that recognized the right of people with disabilities to receive care in the most integrated community setting that could meet their needs. Subsequent federal regulations associated with the Omnibus Budget Reconciliation Act of 1987 required preadmission screening to prevent inappropriate placement of mentally ill persons in NHs that lacked appropriate services and prohibited state Medicaid programs from reimbursing NHs for care provided to nonelderly residents whose primary diagnosis is mental illness. (Individuals with comorbidities including a primary physical diagnosis and a secondary mental diagnosis can be covered in NHs.) Because AL is considered a community setting, as opposed to an institutional setting, AL became a more appropriate placement—perhaps even replacing NHs as the place of “last refuge” for seriously mentally ill adults (Mechanic & McAlpine, 2000)—giving states an additional incentive to move nonelderly clients from NHs to ALFs.

Further, NHs cannot be certified as Medicaid-eligible facilities if more than 50% of their residents have a primary diagnosis of mental illness. If NHs fail to meet these criteria, then all the direct care they provide becomes ineligible for Medicaid reimbursement. As a result, there are few uncertified NHs specializing in treating mental illness operating in any state. In contrast, Medicaid regulations do not prohibit NHs from serving elderly residents with dementia or Alzheimer’s, who are classified as cognitively impaired with “organic brain disorders” and not as mentally ill (Street, 2006).

Overall, a combination of factors has made AL an attractive alternative to NHs. Whereas NHs are highly regulated and financed mainly with public funds, ALFs depend primarily on private payment and are significantly less regulated (Golant, 2004; Hawes, et al., 2000; Hawes, et al., 2003). Whereas NHs operate under federal standards and must abide by age/diagnostic restrictions on the types of residents eligible for admission (and eventual Medicaid entitlement), ALFs do not (Hawes & Phillips, 2007). This greater flexibility makes it feasible for states to accommodate younger long-term care individuals with a range of mental and physical care needs in ALFs. Not only is placement in AL for younger disabled individuals (whether physically or mentally disabled) less expensive than NH placement, it also helps NHs avoid the risk of triggering facility decertification (which ends Medicaid and Medicare reimbursements) or an Olmstead appropriate placement challenge, as often happens in NHs. Consequently, although frail elders represent the largest client group in most long-term care settings, whether NHs or ALFs, ALFs have a greater capacity to also serve younger residents with physical and mental disabilities who are unable to live independently and who qualify for waiver services or the state Medicaid plan. Because of the eligibility criteria, AL residents who are funded under a 1915c waiver or state Medicaid plan are generally low income or poor, unlike the typical private-pay resident.

Differences among facilities in resident composition—whether based on age, physical health, or mental status—have profound implications for the quality of life of both younger residents and the frail elderly individuals who reside in ALFs with mixed populations (Heumann, 1996). Prior research has shown that frail elders who reside in housing with a substantial population of nontraditional residents report lower quality of life, fewer opportunities for socialization with other residents, and lower levels of involvement with nonresident family and friends (Burby & Rohe, 1990; Cohen, Bearison, & Muller, 1987; Heumann, 1996; Kellam, 1992; Street, Quadagno, & Burge, 2005; Street, et al., 2007). Yet, only recently has ALF-specific research begun to focus on specialized types of facilities or “nontypical” ALF subpopulations, or both (see, e.g., Cummings, Chapin, Dobbs, & Hayes, 2004; Hedrick et al., 2007).

Research Questions

States’ efforts to reduce long-term care expenditures and adequately serve residents with a variety of needs, coupled with preferences of the frail elderly individuals for AL over NH care, have expanded and diversified the AL industry. In the context of these trends, it is important to understand how ALFs make decisions about admitting and discharging residents and how such decisions, in turn, shape population characteristics and ultimately residents’ quality of life. Our research questions are motivated by these issues. First, how do Florida facilities with LMH licenses differ from standard and other specialty licensed (ECC and LNS) ALFs in their admission–discharge policies and practices? Second, what implications do these differences have for resident demographic composition? Finally, does the opportunity to age in place vary by facility type based on discharge policies? This article uses state administrative data and facility survey data from the Florida Study of Assisted Living (FSAL) to compare ALFs with different licensure profiles in terms of admission–discharge policies and resident population characteristics.

In our analysis, we classify ALFs according to licensure type. Traditional facilities hold only a standard license, high-frailty facilities have either an ECC or an LNS license, and behavioral facilities are those with an LMH license. We hypothesize that high-frailty facilities will have the most liberal physical care admission policies and the least restrictive physical care discharge criteria compared with either traditional or behavioral facilities. We pose this hypothesis because as a matter of explicit policy, high-frailty facilities are expected to offer frail elderly residents the opportunity to age in place. Compared with behavioral facilities, we expect traditional facilities to have more liberal physical care admission policies but more restrictive behavioral care admission policies. We also expect that traditional facilities will be more likely than LMH licensed facilities to discharge residents with behavioral problems but more likely to retain residents with higher levels of physical impairment. Consequently, we expect that behavioral facilities will be least likely to offer frail elderly residents opportunities to age in place. Finally, we expect that the differences in admission and discharge policies across high-frailty, traditional, and behavioral facilities will shape resident population profiles in unique and important ways. Specifically, we expect traditional and high-frailty facilities to be dominated by elderly, White widowed women, whereas we expect that behavioral facilities will serve a much more diverse population with a larger proportion of men, racial or ethnic minorities, and residents who require behavioral management.

Methods

Variables and Data Analysis

Data are from the FSAL surveys completed by facility administrators (N = 463) in spring 2004, representing approximately 24% of facilities in operation during the study period (Street et al., 2005). The FSAL facility sample was broadly representative of all Florida ALs at the time of the study, having a very similar distribution of facility sizes and geographic distribution of ALFs throughout the 11 planning and service areas in the state. With respect to the proportion of residential capacity represented by particular facility licensure types, among all licensed Florida ALFs listed in the state administrative data, 43.2% of licensed beds were in traditional, 36.8% high-frailty, and 10.9% behavioral ALFs. In the FSAL sample, 41.2% of licensed beds were in traditional facilities, 44.5% in high-frailty, and 8.5% in behavioral ALFs. Thus, high-frailty facilities are somewhat overrepresented and behavioral facilities slightly underrepresented in the FSAL data. Fifteen facilities from the FSAL study could not be classified in the three-way facility type categorization we used (see Table 1) and are excluded for analytic clarity. Thus, the total number of facilities for which we report findings is 448. (The 15 facilities excluded from the analysis had both an LMH license and either an ECC or an LNS license and were too few to support reliable analysis. These “mixed” license facilities represent about 5.8% of all licensed AL beds in the original sample and about 7.3% of all licensed AL beds in Florida, according to the state-provided licensure administrative data.)

Data from the FSAL facility survey include admission and discharge policies, a demographic profile for each facility’s resident population (including health and behavioral measures), and patterns of resident admission and discharge. The sampling frame for the FSAL, consisting of all ALFs licensed in the state, was derived from ALF administrative data provided by the Florida Agency for Health Care Administration. State administrative records included facility address, size, and licensure data, which we used to assess the representativeness of the sample and to designate facilities into three empirically derived categories: traditional (standard license), frail (ECC and LNS licenses), and behavioral (LMH licenses).

Facility Type.—

Table 2 shows the distribution of ALFs by licensure type grouped according to our classification criteria: ALFs licensed mainly to serve traditional populations of elderly residents needing personal assistance and supervision (standard license), high-frailty facilities licensed to care for more physically frail residents (LNS and/or ECC licenses), and behavioral facilities licensed to both meet standard/traditional physical care needs and serve residents with behavioral care needs (LMH licenses). Traditional ALFs have lower occupancy rates than either high-frailty or behavioral facilities. They are also dominated by private-pay residents, being least likely to receive Medicaid reimbursement for their residents. Compared with traditional facilities, the percentage of residents depending on Medicaid is double in high-frailty facilities and more than four times as great in behavioral ALFs.

Table 2.

Facilities by Type From the Florida Study of Assisted Living, N = 448

	Traditional	High-frailty	Behavioral
Number of facilities	215	160	73
Number of licensed beds	7,224	7,802	1,499
Occupancy rate (%)	79.4	82.3	86.6
Percent of residents using Medicaid	13.7	26.5	60.3

Variables From the FSAL Facility Survey.—

Admission and discharge policies are considered in the context of both physical and mental functioning characteristics of facility resident populations. Policies related to residents’ physical function include whether the facility will admit residents who use wheelchairs, need transfer assistance, or experience bladder or bowel incontinence they cannot manage themselves. Policies related to residents’ behavioral characteristics include a facility’s willingness to admit residents with a history of mental illness or residents who wander, are verbally or physically aggressive, exhibit socially inappropriate behavior or severe memory/judgment problems, or refuse needed services.

For admission policies related to both residents’ physical function and behavioral characteristics, we measure the percentage of facilities responding that they would admit residents with each previously listed physical and behavioral characteristic. Responses were based on a 3-point scale (would admit, admission would depend on individual circumstances, would not admit). We use a conservative measure of admission: only unequivocal agreement that a resident with particular characteristics would routinely be admitted. Discharge policies are responses to items that asked whether the facility would discharge current residents if the same physical and behavioral characteristics developed during residency. Facilities that responded they would discharge “depending on individual circumstances” or that they would routinely discharge residents who develop additional physical care needs or behavioral characteristics are reported as using such particular discharge criteria.

Resident population characteristics are the percentages of residents in each facility reported by standard demographic categories (gender, marital status, race/ethnicity, age group). We also include measures of residents’ physical function (incontinent, chair bound, wheelchair bound), mental function (Alzheimer’s, mental illness, developmental disability), and need for assistance with ADLs (bathe, dress, eat, transfer, toilet). Other resident characteristics distinguished by facility type are patterns of admission (percent of residents admitted from home, hospital, NH/rehab center, a different ALF, or some other place) and discharge destination (percent of residents discharged to home, hospital, psychiatric hospital, NH, a different ALF, unknown location, or who died).

Data Analysis

In the analyses, we compare traditional ALFs, high-frailty ALFs, and behavioral facilities on patterns of admission and discharge policies and resident population profiles. We use analysis of variance to determine statistical differences in means across comparison groups of facilities, as well as Bonferroni’s, Scheffe’s, and Sidak’s tests of multiple comparisons to determine significant differences between specially licensed (behavioral and high-frailty) facilities and traditional (standard license only) facilities. Statistical differences between means are noted with asterisks, whereas differences between specially licensed and traditional facilities are noted with bold print.

Results

Admission Policies

Table 3 shows that traditional and high-frailty ALFs use significantly different resident physical functioning criteria than behavioral ALFs to determine admission. Behavioral facilities are significantly less likely than either traditional or high-frailty ALFs to admit residents who have a high physical level of care need—residents who use wheelchairs, need assistance transferring, or experience bladder or bowel incontinence. In three-way comparisons of admission policies, there is even greater variation across resident behavior functions by facility type. Compared with traditional ALFs, high-frailty facilities are significantly more likely to admit residents with a history of mental illness, who wander, exhibit socially inappropriate behavior, or have severe memory/judgment problems. Compared with traditional ALFs, behavioral ALFs are significantly more likely to admit residents with a history of mental illness, who are verbally aggressive, who are physically aggressive, who have severe memory/judgment problems, and who refuse needed services. Although both the high-frailty and the behavioral ALFs are more likely than traditional ALFs to admit residents with a history of mental illness or severe memory/judgment problems, high-frailty ALFs have a behavior function profile that is most typically associated with age-related dementias and cognitive decline, whereas behavioral ALFs admit more residents with classic symptoms/behaviors associated with mental illness. Behavioral ALFs are more likely to admit physically aggressive residents (8.6%) than traditional (1.9%) or high-frailty (3.9%) ALFs.

Table 3.

Admission and Discharge Policies by Facility Type (Florida Study of Assisted Living facility survey N = 448)

	Traditional (%)	High-frailty (%)	Behavioral (%)	Significance
Would admit
Physical function
Uses wheelchair	76.5	84.0	55.5	***
Needs transfer assistance	72.9	79.2	40.3	***
Cannot manage bladder incontinence	85.6	91.7	49.3	***
Cannot manage bowel incontinence	88.3	92.9	72.2	***
Behavioral function
History of mental illness	27.9	43.0	84.3	***
Wanders	26.0	43.0	34.8	**
Verbally aggressive	10.8	20.4	34.3	***
Physically aggressive	1.9	3.9	8.6	*
Socially inappropriate behavior	14.6	25.5	25.4	*
Severe memory/judgment problems	36.2	61.6	55.6	***
Refuses needed services	15.2	19.6	30.4	*
Would discharge
Physical function
Uses wheelchair	15.8	10.4	45.9	***
Needs transfer assistance	22.6	16.2	61.4	***
Cannot manage bladder incontinence	13.5	6.9	50.0	***
Cannot manage bowel incontinence	27.4	15.2	60.7	***
Behavioral function
History of mental illness	67.0	52.3	21.0	***
Wanders	72.1	58.8	74.2	*
Verbally aggressive	88.2	83.0	75.8
Physically aggressive	96.9	95.8	93.8
Socially inappropriate behavior	84.2	70.9	78.5	*
Severe memory/judgment problems	57.9	36.8	47.6	***
Refuses needed services	83.4	77.3	66.7	*

Note: Bolded values significant at p < .05. *p < .05. **p < .01. ***p < .001.

Discharge Policies

Behavioral ALFs also differ from traditional and high-frailty ALFs in how their discharge policies treat declines in resident physical functioning. For example, only a small minority of traditional (16%) and high-frailty (10%) ALFs would discharge residents who eventually needed to use a wheelchair compared with nearly 46% of behavioral facilities. For residents needing transfer assistance, the percentages of traditional and high-frailty ALFs reporting that they would (or would consider) discharging such residents were relatively low, whereas 61.4% of behavioral ALFs would discharge such residents. The differences between traditional and high-frailty facilities and behavioral ALFs are even more stark when incontinence is considered. Residents in behavioral facilities are three to four times more likely to be discharged for incontinence than residents in other types of facilities.

In contrast, discharge policies related to changes in residents’ behavioral functions do not yield such crisp distinctions between facility types. There are no statistically significant differences among the three facility types in terms of discharging residents who become verbally or physically aggressive. Such behavior would likely trigger discharge regardless of facility type. High-frailty ALFs are significantly less likely than traditional and behavioral ALFs to report that they would discharge residents who wander, who exhibit socially inappropriate behavior, or who have severe memory/judgment problems. They are also the most willing to retain residents exhibiting behaviors that are often associated with dementias and cognitive decline in elderly populations. In contrast to those particular age-related resident behavioral manifestations, behavioral functions that are not associated with aging processes—having a history of mental illness and refusing needed services—are significantly less likely to trigger discharge from behavioral ALFs than from either of the other two types.

Resident Population Characteristics

Given differences across facility type in admission and discharge policies, distinctive resident population profiles by facility type are to be expected (shown in Table 4). Residents of traditional and high-frailty ALFs have fairly similar demographic characteristics; the majority of residents are White widowed women, most older than 75 years. High-frailty ALFs have somewhat older resident population age profiles and greater race or ethnic diversity than traditional facilities, but the differences are minimal. In contrast, behavioral ALFs have significantly greater percentages of residents who are men and who are divorced or never married. They are much more racially and ethnically diverse than the other facility types and have significantly more nonelderly residents. Despite the significant differences in the age profiles, it is important to note that behavioral ALFs nonetheless serve a considerable number of elderly residents with high levels of physical care needs, alongside relatively younger ones.

Table 4.

Resident Population Profiles by Facility Type (Florida Study of Assisted Living facility survey N = 448)

	Traditional (%)	High-frailty (%)	Behavioral (%)	Significance
Resident demographics
Gender
Female	70.8	76.3	43.0	***
Male	29.3	23.7	57.0	***
Marital status
Married	9.0	10.1	4.0	**
Widowed	68.7	72.1	16.3	***
Divorced	8.8	9.1	25.6	***
Never married	12.3	6.5	50.5	***
Unspecified	1.3	2.3	3.6
Race/ethnicity
White	90.2	81.0	62.1	***
Black	4.4	5.0	19.8	***
Hispanic	4.3	12.7	15.7	***
Other/unspecified	1.2	1.2	2.4
Age (years)
<64	11.3	5.3	67.1	***
65–74	10.2	9.6	17.2	**
75–84	31.1	36.2	9.6	***
85+	47.4	48.8	6.1	***
Resident characteristics
Resident physical characteristic
Incontinent	28.3	41.8	10.9	***
Chair bound	4.8	6.8	1.2	**
Wheelchair bound	16.3	19.9	3.7	***
Resident mental characteristic
Alzheimer’s	27.6	39.7	7.8	***
Mental illness	7.1	5.8	71.1	***
Developmental disability	5.5	2.1	10.5	**
Resident activities of daily living characteristics
Bathe	62.8	74.5	28.6	***
Dress	52.9	62.9	22.0	***
Eat	21.6	28.1	11.9	**
Transfer	31.3	38.7	14.4	***
Toilet	38.9	49.5	15.1	***

Note: Bolded values significant at p < .05. *p < .05. **p < .01. ***p < .001.

In terms of characteristics based on resident functioning, high-frailty facilities have much higher percentages of incontinent residents than either traditional or behavioral facilities. Behavioral facilities report significantly lower percentages of residents who are either chair or wheelchair bound than either traditional or high-frailty facilities. Assisted living facility resident populations also vary according to mental function. Compared with traditional or high-frailty ALFs, behavioral ALFs have significantly higher percentages of residents who have a diagnosis of mental illness or who are developmentally disabled, and significantly lower percentages of residents with Alzheimer’s. Although high-frailty and traditional ALFs are similar in the percentage of residents with mental illness or developmental disability, high-frailty facilities have significantly higher percentages of residents with Alzheimer’s.

As might also be expected, given licensure type and dominant resident population demographic characteristics, both traditional and high-frailty facilities have significantly greater percentages of residents who require assistance with ADLs (including bathing, dressing, eating, transferring, and toileting) than behavioral facilities. However, the percentage of residents in high-frailty ALFs requiring assistance with bathing, dressing, and using the toilet is greater than in standard-licensed traditional facilities. This is also to be expected because they hold specialty licenses that allow them to offer a broader package of physical care services.

Patterns of residents’ admission and discharge also differ significantly by facility type (see Table 5). In traditional and high-frailty facilities, the patterns of admission are very similar, except that high-frailty ALF residents are slightly more likely to have been admitted from an NH or a rehabilitation center. This is consistent with the extra services that specialty-licensed high-frailty ALFs can offer residents with relatively high levels of physical care needs. Patterns of admission for behavioral facilities differ significantly from both traditional and high-frailty ALFs. Behavioral ALF residents are less likely to have been admitted from their own homes or from an NH and much more likely to have been admitted from a hospital (often a mental hospital or an acute psych ward in a general hospital) or from another ALF.

Table 5.

Resident Admission–Discharge by Facility Type (Florida Study of Assisted Living facility survey N = 448)

	Traditional (%)	High-frailty (%)	Behavioral (%)	Significance
Residents admitted from
Home	50.5	48.5	25.6	***
Hospital	11.2	10.8	36.0	***
Nursing home/rehab center	7.7	10.4	4.1	*
A different assisted living facility	16.4	16.9	24.0	*
Other	14.2	13.3	10.3
Residents discharged to
Home	16.7	11.0	18.7	*
Hospital	20.2	22.9	16.1
Psychiatric hospital	2.9	1.4	9.7	***
Nursing home	17.3	20.4	8.9	**
A different assisted living facility	13.0	8.6	23.8	***
Moved out (unknown location)	9.3	6.0	14.4
Died	20.6	29.7	8.5	***

Note: Bolded values significant at p < .05. *p < .05. **p < .01. ***p < .001.

Compared with residents of either behavioral or traditional ALFs, residents in high-frailty ALFs are significantly more likely to have died and less likely to be discharged to homes in the community, a finding consistent with higher levels of resident frailty. Residents of behavioral ALFs are less likely to have been discharged to NHs than residents from the other two types of ALFs. This is not surprising because AL residency for many residents of behavioral ALFs was less likely caused by deteriorating physical health circumstances that required medical or skilled nursing care. A pattern of movement between ALFs is much more apparent in behavioral facilities, with residents significantly more likely to be discharged to another ALF than in either of the other two ALF types. This notable pattern indicates churning that occurs in behavioral ALFs, where there is substantial resident movement between different facilities. We suspect this is due to a combination of push and pull factors: behavioral ALF administrators’ efforts to discharge “problem” residents combined with residents’ attempts to find ALFs they liked and in which they wanted to live.

Discussion

Our research reflects the variation and ambiguity of AL identity nationwide, highlighted by other researchers (Golant, 2004) by showing how facility licensure characteristics influence a range of outcomes, from admission and discharge policies to resident composition. These practices, in turn, create an organizational context that undoubtedly shapes resident experiences in important ways across the three types of Florida ALFs we considered. Although ALFs serve distinctive populations depending on the type of license they hold, these differences are seldom reported in the research literature on AL.

Assisted living facilities in Florida serve a variety of residents in keeping with their licensure type. Traditional and high-frailty ALFs share many similarities and have somewhat overlapping resident population characteristics. They also differ, however, on several key admission and discharge practices and in some resident characteristics. These distinctions are driven in large measure by policies that reflect the different packages of services that ALFs can deliver, depending on whether they hold specialty licenses beyond standard ones. This difference in licensure status distinguishes the almost entirely custodial functions of traditional ALFs from the more specialized package of services that can be assembled and offered to residents with higher levels of frailty in more specialized ALFs. Although both ALF types provide services for the most typical AL population—elderly individuals—each also provides services to at least some residents who do not fit that age profile.

Behavioral ALFs, in contrast, have distinctively different policies and practices and resident profiles that little resemble what most would consider a conventional AL population. Although about one third of residents in behavioral ALFs are elderly individuals, many residents are younger and substantially more diverse on several key demographic characteristics. Just as high-frailty facilities have carved out a niche by offering specialized care services for conventional populations, behavioral ALFs have taken a different track, offering care services for a less typical AL population.

These differences by facility type raise important quality-of-life issues for both frail elderly individuals and residents who do not necessarily fit that conventional demographic profile. Traditional and high-frailty ALFs serve primarily frail, relatively affluent, elderly White women with physical care needs who pay privately for services and housing. Such ALFs seem committed to admitting residents with greater initial care needs and to retaining these residents as their health or functioning deteriorates, prompting greater need for more intensive physical care. The vast majority of traditional and high-frailty ALFs would not discharge residents who use a wheelchair, who need help transferring, and who cannot manage their own incontinence. These practices make it possible for residents to age in place. However, such facilities may not provide as compatible a residential environment for residents who do not conform to the frail elder model.

Behavioral ALFs serve a more diverse population that includes more young people, more men, and more divorced and never married individuals—and significantly, a population with much lower incomes, whose expenses are often paid in part by the ALE program or by Medicaid Assistive Care. These ALFs are less willing to admit residents with high levels of physical care needs or to keep residents when their physical care needs increase. By contrast, they are more willing to admit and retain residents with behavioral problems often associated with mental illness, such as being verbally aggressive or refusing needed services. Nonetheless, many frail elders and residents who are racial or ethnic minority individuals reside in behavioral ALFs. Behavioral ALFs may offer satisfactory care to one segment of their target population but perhaps provide a less appropriate living arrangement for frail elders who share housing with residents exhibiting behavioral problems or who have less opportunity to age in place because of more rigid discharge policies.

Most elderly residents are admitted to AL either from their own homes or from NHs/rehab centers, and nearly half are discharged to NHs or die. Residents in behavioral ALFs, however, tend to enter AL from hospitals or from other ALFs and more often are discharged to medical or psychiatric hospitals or to another ALF. Thus, behavioral and more conventional (traditional and high-frailty) ALFs draw from substantially different pools of potential residents, with distinct sets of care needs and resources, whether familial or fiscal, to meet them. Residents of behavioral facilities often hopscotch through a series of ALF admissions and discharges before settling into a suitable ALF. The residential churning is not unusual as residents and facilities alike try to “make matches” that will work in an underresourced placement environment for a population with high and challenging needs.

Although facility differences in the Florida AL industry are significant, our findings should be interpreted conservatively for several reasons. First, the data are 4 years old. However, there have been no major changes in Florida’s AL regulatory or policy landscape, which gives us confidence that conditions observed in the data continue to be relevant to current experiences. Second, the 24% response rate seems on the low side; however, the response rate is higher than many recent private-sector business surveys, and comparison of the analytic sample and state administrative data show similar spatial, facility size, and licensure type distributions, alleviating concerns that the sample risks being unrepresentative. Third, FSAL data are from a single state with its own state-specific AL regulatory framework. In some states, AL is a separate category based on a consumer-centered service philosophy, private apartments, and an array of services that support aging in place. In other states, AL includes this newer model as well as board and care homes, multiunit housing for the aged, congregate housing, and even adult foster care (Mollica et al., 2005). Most states’ AL policy regimes include licensure for facilities serving dominantly physically frail or elderly populations, and only a few others have regulations for facilities or services designed to serve mentally ill residents. Most other states lack the specialty licensure categories that create the high-frailty and behavioral niches we find in Florida, reflecting patterns of conventional AL care that are more consistent with a picture of mostly custodial placement and supervision. Many other states’ AL regulatory mechanisms are more consistent with what Florida’s traditional facilities offer, the picture of assisted living that seems to dominate public perceptions of AL. Although the extent of state-specific differences of licensure types and definitions of AL limits the generalizability of our findings, three other states have specialized mental health regulations that may generate similar outcomes to our findings and many others may have de facto behavioral ALFs, albeit not recognized in their current regulations. This may be the case because the modal category for mental illness in state regulations is no mention of it at all, despite a sizeable likely population needing residential care and services in every state. It is important to recognize that the issues facing Florida are similar in all states and that all states are experimenting with waivers and with their Medicaid plans to serve AL clients in the least restrictive setting as required by federal regulations. Finally, the nature of the AL industry (whether in Florida or in other states around the country) is continually evolving, as states refine their regulatory frameworks and experiment with substitution of AL for NH care (and the usually accompanying Medicaid AL waivers and adjustments), and as the AL industry adapts to changing market conditions and niche opportunities. Because FSAL data are from a particular period in one state, conclusions are suggestive rather than definitive.

Despite these limitations, this study offers insights into some of the complexities that are often overlooked in research on AL. Assisted living, as commonly understood, is regarded as a residential alternative to NH for frail elderly individuals who are no longer able to live in their own homes. Yet, in Florida (and many other states), AL also represents a much more diversified residential care setting than such a characterization suggests. Assisted living facilities provide for residents with needs ranging from care and services required by younger individuals with developmental or physical disabilities, to domicile for adults with serious mental illness, to havens for the old and frail. To the extent that policies and practices designed with frail elders in mind overlook or fail to address the particular care needs of “nontraditional” subpopulations, they risk poorly serving these residents. Not only do these different types of residents have different care needs but also diverse interests. As a result, activities designed to interest one group may have little appeal for the other, a reality seemingly recognized in some states’ regulatory admonition that “incompatible” residents should not be housed together (NCAL, 2008). The presence of a mixed population also raises policy issues about maintaining the quality of life for frail elders and allowing them to age in place to the maximum extent possible.

Conclusions

Licensure type results in substantial variation in admission and discharge policies and in resident population characteristics in Florida ALFs. Although many Florida ALFs conform to the generalized impression of AL for frail elders, closer examination shows that ALFs with different licensure types serve distinctive populations. The AL industry is evolving in a market and regulatory climate shaped by complex social and demographic circumstances. Forces that have fueled growth in the AL industry include state efforts to rein in public expenditure for Medicaid long-term care, judicial rulings that permit individuals to receive services in the least restrictive environments possible, the increasing affluence of segments of the elderly population, and the desire of most individuals to avoid NH at all costs (Mehdizadeh, et al., 2006).

Given the confluence of factors that are leading to the expansion of the AL industry in Florida and around the country, the elaboration of AL types that serve different population subgroups is not surprising. The unanswered question is how differences in admission and discharge policies and the resulting resident population characteristics affect the daily lives of those residing in these settings. Our results suggest that as the AL industry has grown, it has become socially stratified. Residents who have the resources to pay privately for care appear to be able to age in place in a compatible environment, whereas lower income individuals and racial or ethnic minorities are more likely to reside in facilities that have a larger proportion of younger residents and are at greater risk of being discharged as they become more frail. Such stratification is likely a result both of the growth of ALFs willing to accept residents whose services are paid by Medicaid funds and of states’ need to license existing facilities as they seek proper placements for these clients. The challenge for all states is to devise regulations that adequately serve the needs of sick and frail elderly residents, people with dementia, and younger adults with developmental disabilities or mental illness. This is an important avenue for future research that can help inform policies for a rapidly changing industry.

Funding

Support for this study was provided by the National Institute on Aging, the Florida Agency for Health Care Administration, and a Florida State University Program Enhancement Grant.