| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
Department of Geography, University of Florida, Gainesville.
Address correspondence to Stephen M. Golant, Department of Geography, University of Florida, 3117 Turlington Hall, PO Box 117315, Gainesville, FL 32611. E-mail: golant{at}geog.ufl.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Information on study design and six indicators of the occupancy patterns of older persons in ALFs were abstracted from six national studies. The collected data were based on reports by the administrators of ALFs.
Results. The six reviewed studies had several methodologic weaknesses, resulting in different statistical populations of ALFs, samples with very different numerical and attribute properties, and findings based on disparate indicators. The older residents in ALFs were less physically and cognitively impaired than those in nursing homes. ALF facilities were more likely to admit or retain frail older persons when they had relatively minor or less serious physical or cognitive impairment or health care needs.
Discussion. ALFs are currently serving older residents who require less nursing care and who are less functionally and cognitively impaired than those found in nursing homes. The more restrictive admitting and discharge criteria of a substantial share of ALFs guarantee their less frail occupant profile. This is, however, an extraordinarily diverse shelter and care alternative, and very frail older persons with serious chronic health problems can be found in ALFs. Average duration of stays may be as long as 3 years. Researchers must conduct more carefully executed studies with replicable methodologies that produce unbiased and generalized findings.
THE nursing home persists as the most institutionalized, long-term care setting in the United States with the highest prevalence of older persons who are chronically ill and who have difficulty functioning because of their cognitive and physical disabilities (American Health Care Association, 2001
; Hawes, 2001; Sahyoun, Pratt, Lentzner, Dey, & Robinson, 2001; Zimmerman et al., 2003). Most long-term care, however, is now provided outside the nursing home in conventional dwellings and increasingly in group residential care settings (Bishop, 1999; Quadagno & Stahl, 2003). The assisted living facility (ALF) has emerged as the most important of these shelter and care alternatives (Mollica, 2002; Mollica & Jenkens, 2001). About 32 states and the District of Columbia have an "assisted living" licensing category or statute. Over 40 states have Medicaid programs that subsidize the services that low-income seniors receive in ALFs (Mollica, 2002).
The ALF is often compared with the nursing home, and its identity as a shelter and care setting often rests on these distinctions. Two sets of attributes are usually the basis for the comparison. First, the ALF is viewed as a more residential-like and less institutionalized setting than the nursing home that can provide personal care and supportive services and sometimes health-related and skilled nursing services to meet the scheduled and unscheduled needs of frail older persons. This is accomplished in a setting that "promotes the resident's dignity, autonomy, independence, and quality of life" (Assisted Living Workgroup Steering Committee, 2003, p. 12) and, unlike the nursing home, does not have an overly regulated, inflexible environment and medical-like qualities. Thus, in this first sense, the emergence of ALFs is construed as a reaction to a host of negative features attributed to the nursing home. Second, the types of services, the amount of care offered by an ALF, and its tolerated resident frailty thresholds are often expressed in relative terms, that is, by their similarity to the nursing home. Thus, in this second sense, the ALF care mission is defined in reference to a nursing home standard. This is illustrated by the oft-made observation that ALFs can now accommodate those less impaired older persons who once would have had to inappropriately occupy a skilled nursing home (Bishop, 1999). It is also embodied in the conclusions of a highly respected group of researchers who argued that the ALF "may be best considered a discrete node that services residents similar to those in [nursing homes] and in a similar fashion" (Zimmerman et al., 2003, p. 108). Yet another example is offered by a consortium of assisted living stakeholders who recently emphasized that assisted living "does not generally provide ongoing, 24-hour skilled nursing care" like the nursing home and questioned the desirability of such a mission in the absence of more highly regulated ALFs (Assisted Living Workgroup Steering Committee, 2003, p. 12).
In practice, the identity of the ALF is often ambiguous because it is not a uniform and clearly formulated product. Its physical infrastructure, operating features, standards and levels of care, and tolerated frailty thresholds can differ substantially. Indeed, consensus on how to define the ALF is now more elusive than ever (Assisted Living Workgroup Steering Committee, 2003).
This is because absent federal guidelines, ALFs largely reflect the philosophies of individual state governments. Their regulatory environments largely determine the extent to which ALFs can accept and retain physically and cognitively frail older persons with health care needs (Mollica, 2002). Regulations in some states, for example, allow for a high degree of resident impairment, whereas others have more restrictive and inflexible tenant frailty policies (Kane & Wilson, 2001; Mollica, 2002; Mollica & Jenkens, 2001). Even within the same state, however, providers may impose more restrictive admitting and discharge policies that result in their ALFs having tenants who are less frail than state regulations would tolerate (Mollica & Jenkens, 2001). Another source of variation arises when Medicaid funding makes the services found in private-pay ALFs affordable to low-income residents. Depending on state Medicaid policies, the residents in these facilities may have the same impairment levels as those in nursing homes (Mollica, 2002).
A number of stakeholders are concerned about the level of care offered by ALFs and the frailty profiles of their residents. Elderly consumers (and their family members) who sometimes must look outside their conventional dwellings to deal with their declines and care needs view nursing homes as an undesirable "last resort" alternative. When considering an ALF, however, they must determine if it will admit residents as frail as themselves and retain them if they decline further. Thus, they have to judge whether the ALF will be their last home or merely a temporary way station from which they will have to relocate when they have more demanding long-term care needs (Zimmerman et al., 2003). They must then decide whether even the most desirable ALF experience can justify the considerable time, energy, and psychic costs that they can experience if they have to move again (Chapin & Dobbs-Kepper, 2001; Lieberman, 1991). They must also sometimes consider if the anticipated stress linked to this moving effort might be tempered if the ALF were physically connected to a nursing home. State governments charged with planning long-term care alternatives have also looked favorably on the ALF alternative, because they view it as a more affordable long-term care solution for impaired low-income seniors that can slow their Medicaid nursing home expenditures (National Governors' Association, 2000; Polivka, 1997). Not all experts, however, agree that states will save money in the aggregate (Manard, 1999). ALF advocates fear that if state governments use Medicaid subsidies to make this alternative affordable to the very impaired poor, this will invite overly strict federal regulatory intervention (Kane & Wilson, 2001). State regulators are concerned with the level of care and type of services the ALF offers because they are responsible for quality assurances. Quality-of-care issues are also paramount to a host of health and long-term care professionals, such as nurses, medical directors, pharmacists, social workers, psychologists, occupational therapists, and nutritionists, who want assurances that only ALFs with appropriate staffing and operation standards will accommodate frail residents. Nursing homes have an obvious interest in the impairment profile of ALF residents because although they see the ALF as a feeder for their facilities (Zimmerman et al., 2003), they also must be concerned about whether they and ALFs are competing for the same market share. Last, to the extent that ALFs serve a frail and vulnerable group of consumers, they have come under the radar screen of insurance companies because of liability issues (Polivka, Salmon, Hyer, Johnson, & Hedgecock, 2003).
In light of the diversity of ALFs, the extent to which older residents with physical and cognitive disabilities and health care problems occupy them becomes an empirical question. To answer it, this article compares the findings and methodologic integrity of six national studies of ALF facilities in the United States.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
), whereas professional organizations representing ALF providers funded four studies (ASHA: American Seniors Housing Association, 2001; ALFA: Assisted Living Federation of America, 2001; NCAL: National Center for Assisted Living, 2001; NIC: National Investment Conference, 1998). A federal oversight agency (GAO: U.S. General Accounting Office, 1999) funded the sixth study. The GAO study focused on only four states (California, Florida, Ohio, and Oregon), but its findings were based on a relatively large sample of ALFs in four distinct regions, and its results were widely disseminated, covered by the national press, and reviewed in government hearings and in the scientific and professional literature because they raised concerns about the quality of ALFs (Frytak, Kane, Finch, & Maude-Griffin, 2001; Kane & Wilson, 2001; U.S. Senate, Special Committee on Aging, 1999). The review excluded two other ALF studies because they focused on facilities in only a few selected states and did not examine comparable occupancy indicators (Bernard, Zimmerman, & Eckert, 2001; Chapin & Dobbs-Kepper, 2001).
The Statistical Population of ALFs
Whereas most states now use the ALF terminology, others use labels such as residential care homes, personal care homes, shelter care facilities, adult foster care, and boarding homes (Mollica, 2002). Some states have multiple designations to distinguish ALFs regulated under different licensure categories because they are intended for older persons with different levels of impairment (Mollica, 2002). The ALF designation is also sometimes reserved for properties with a minimum number of beds or operated by professional staff (Hawes et al., 1999; Morgan, Eckert, & Lyons, 1995; National Investment Conference, 1998; Zimmerman et al., 2003). ALFs are also distinguished by whether they are freestanding or physically attached to another building or wing that houses a nursing facility and/or an adult congregate care facility (all three shelter and care levels are usually available in continuing care retirement communities [CCRCs]). This latter distinction is important because this analysis will reveal that freestanding and multilevel facilities have different admitting and retention policies (Hawes et al., 1999).
Altogether, the variation in labeling conventions makes the identification of the statistical population of ALFs far from clear-cut. These definitional ambiguities help explain why there is no single agreed-upon database of ALF facilities in the United States. Only one professional trade group (Promatura Group, 2000) has published a statistical enumeration of ALFs that both corresponds to the collection periods of the six ALF surveys and distinguishes ALFs by the number of their properties, beds, and units and whether they are freestanding or multilevel properties. It categorized senior housing properties as assisted living as long as they primarily served older adults and provided 24-hour protective oversight, at least two meals per day, and any "personal care services" (assistance with eating, bathing, toileting, transferring/mobility, or grooming) for their residents. The included properties, however, did not have to be licensed by their respective states. The Promatura Group also distinguished large properties with 
30 units from small properties with 1–29 units. (A unit referred to a room or a conventional apartment occupied by one or more beds or residents.) Findings regarding ALF bed and unit statistics are similar, and thus this analysis will report only on the numerical tabulations of ALF beds. These data will offer one basis on which to judge the validity of the statistical populations identified in the six studies.
Mollica (2002) provided the most detailed descriptions of ALFs based on information he obtained from state surveys of ALFs operating under state regulations or statutes. His summary tabulations, however, do not distinguish the size (properties with small and larger number of beds) and freestanding status of properties. The two sources, however, do offer somewhat similar supply estimates. The Promatura Group (2000) estimated there were just over 27,000 properties containing about 778,000 ALF beds in the United States by 1999; Mollica (2002) estimated there were about 33,000 facilities containing about 795,000 beds by the year 2000. The discrepancy partly reflects the 1-year difference in the time of collection but also reflects the different ALF definitions used by the two enumeration efforts.
Statistical Population Selection and Sampling Procedures of the Six Studies
The validity of the generalizations drawn from the findings of the six studies will depend on the extent to which they were conducted at roughly the same time, used the same types of data collection, defined the same statistical population of ALFs, and employed comparable and reliable statistical sampling procedures. The state coverage by these studies will also be important, given that the different ALF regulatory environments of states help explain the variation in the impairment profiles of their residents (Mollica, 2002). To measure these sources of validity, this analysis examined the following study-specific information: survey date and data collection method; sources of ALF listings, criteria for inclusion, and size of provider statistical population; sampling methodology; sample size of eligible ALFs; returned number (surveys) of sampled ALF properties; sample response rate; number of ALF beds in surveyed sample and average property size; percentage of facilities that are freestanding in surveyed sample; and number of states covered by surveyed sample.
The Impairments and Health Care Needs of ALF Older Occupants
Based on ALF industry and scientific conventions (Hawes et al., 1999; Kane & Wilson, 2001; National Investment Conference, 1998), the analysis defines six indicators that reveal whether older residents with physical and cognitive disabilities and health care problems are likely to occupy ALFs: (a) percentage of ALF residents requiring assistance with "activities of daily living" (ADLs) or who had cognitive impairments; (b) average length of stay at ALFs; (c) reasons for leaving ALFs; (d) percentage of ALFs admitting residents with specific care and impairment needs; (e) percentage of ALFs retaining residents with specific care and impairment needs; and (f) percentage of freestanding and multilevel ALFs with different admitting and care patterns.
The resident impairment (ADL or cognitive) profile of ALFs is one indicator of whether vulnerable older residents have entered and remained in facilities for whatever reasons (e.g., individual care preferences, prospects of adequate care, facilitative state regulations, or flexible facility care policies). Four of the six studies (ALFA, NCAL, NIC, and GAO) measured the extent to which their residents required assistance performing at least five of the following six ADLs: bathing, dressing, toileting, mobility or locomotion, transferring, and eating. Three studies (ALFA, NCAL, and NIC) measured the mean number of ADLs for which residents required assistance. All but one study (ASHA) inquired about the cognitive impairments of the residents. Provider estimates and not resident self-assessments were the source of the ADL and cognitive assessments. Some response variability will reflect small variations in question wording (Guralnik & Simonsick, 1993). None of the cognitive impairment questions inquired about whether the residents suffered from depression or ascertained if they were mentally retarded or developmentally disabled. As a standard basis for comparison, the analysis reports on the ADL impairments of elderly nursing home residents (American Health Care Association, 2001). These data are based on reports by nursing facilities during their standard surveys for Medicare and Medicaid certification.
Average length of stay by residents in ALFs is the simplest indicator of whether residents have occupied ALFs over a sustained period, even as they are possibly afflicted with physical impairments, cognitive deficits, and health care problems. Four of the six studies (ASHA, ALFA, NCAL, and NIC) reported this information.
Reasons for leaving ALFs reveal whether the ALF older residents have relocated to a potentially more supportive setting. These allow some inferences about whether their current ALFs are unwilling or unable to retain occupants who have higher levels of frailty or more serious health conditions.
Providers' reports of their resident admission and discharge (retention) policies and their differences reveal the maximum seriousness of the disabilities and chronic health problems that will be tolerated in ALF facilities, irrespective of whether these frailty thresholds reflect state regulations or how facilities comply with them (Mollica & Jenkens, 2001). A comparison of admitting and discharge policies specifically reveals whether ALFs are likely to tolerate residents when their conditions worsen. Only three of the six national studies (HAWES, NIC, and GAO) reported on these policies, although they did not consistently define or measure the same resident conditions. The analysis calculates and then ranks the weighted average percentage of facilities admitting residents with specific conditions. Admitting and retention patterns are compared. Only one study provided standard errors of its reported percentages (HAWES).
A final analysis distinguishes the admission and care policies of freestanding and multilevel facilities. This distinction is investigated because of arguments that the physical and organizational linkages of an ALF may influence whether it accepts or retains frail residents. ALFs in multilevel facilities often are connected physically and organizationally with a nursing home, and thus they may be less inclined to accept or retain frail residents but rather transfer them to the higher-care facility (National Investment Conference, 1998). Alternatively, ALFs in multilevel facilities are more likely than freestanding facilities to have access to the physical plant, assistive devices, and staffing that may give them the capabilities to accept frail seniors as occupants (Hawes et al., 1999).
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
The six studies used different criteria to identify the eligible statistical population of ALFs. Two studies included all ALF facilities that were on their provider lists (ASHA and ALFA). The other four studies (HAWES, NCAL, NIC, and GAO) excluded providers from their lists that did not offer some designated set of services or level of personal or health care (see Table 2). Two studies (HAWES and NIC) further required that their facilities have a minimum number of beds: in one instance >10 beds, in the other >7 units. Only three of the six studies (HAWES, NCAL, and GAO) reported their estimated statistical populations (of properties) from which they drew their samples, and these ranged in size from 2,652 to 12,000. These reported estimates, based, as they were, on different resident eligibility criteria, varied substantially from the published estimates (see Table 1).
The six studies also used very different sampling methodologies (see Table 2), and two did not collect a statistical sample (ASHA and ALFA). Two studies relied on simple random samples to select their facilities (NCAL and NIC). Two relied on a hierarchical sampling design: one initially stratifying by the county concentrations of ALFs in 34 states (HAWES), the other stratifying states by the relative size of their age 65-and-over populations (NIC).
The sample sizes of eligible ALFs ranged from a low of 328 (NIC) to a high of 2,893 (NCAL), with two studies not reporting this information (ASHA and ALFA). The returned number of sampled ALF properties (completed surveys) ranged from a low of 39 (ASHA) to a high of 1,251 (HAWES). The sample response rates ranged from a low of 10.5% (NCAL) to a high of 81.0% (HAWES), with two studies not reporting (ASHA and ALFA). The number of the ALF beds in the surveyed (completed) samples ranged from a low of 3,753 beds (ASHA) to a high of 66,678 (HAWES), with one study not reporting (NCAL). The average number of beds per facility ranged from 96.2 beds (ASHA) to 53.3 beds (HAWES; see Table 2).
The percentages of the surveyed ALF properties that were freestanding as opposed to multilevel facilities ranged from a low of 15.3% (ALFA) to a high of 100.0% (ASHA). Among the five studies claiming a national focus, the geographic coverage ranged from a low of 34 (HAWES) to a high of 48 (NIC) states.
After reviewing the procedures for population selection and sampling used in the six studies, the advisability of employing standard tests of statistical significance appeared questionable. Only three of the six studies reported their statistical populations, and these varied greatly in size. The ALF property inclusion standards and the geographic coverage (states) used to select the statistical population also varied greatly among the six studies, thereby increasing the likelihood of systematic selection bias and other nonrandom errors. Two studies did not conduct a sample (ASHA and ALFA), four others had wide-ranging sampling response rates (HAWES, NCAL, NIC, and GAO), and only one (HAWES) reported standard errors of some (but not all) of its reported percentages. In several instances, studies defined and measured different and difficult-to-compare indicators or attributes. This methodologic environment is clearly unsatisfactory but is the state of the art for these national assessments.
Variation in the Impairment Profiles of Elderly Residents
Four studies (see Table 3) reported on both the physical and the cognitive impairments of their older residents (ALFA, NCAL, NIC, and GAO) and one only on cognitive impairments (HAWES). The percentage difference between the highest and lowest estimates of ADL limitations was >20% for bathing, dressing, and transferring, 16–17% for toileting and eating, and 18% for cognitive impairments. All studies agreed, however, that their ALF residents were most likely to require assistance with their bathing and dressing and to a much lesser extent with their toileting and locomotion. The residents were least likely to require assistance with their eating and transferring (e.g., getting in and out of a chair or bed). Overall, the ALF residents in all four studies were less likely to be physically and cognitively impaired than nursing home residents. The ALF study with the most impaired residents (ALFA) reported that they needed assistance with an average of 2.8 ADLs, but this was still lower than the comparable 3.75 ADLs reported for nursing home residents (American Health Care Association, 2001). The study with the least impaired residents (NIC) reported their residents needed assistance with an average of only 1.3 ADLs. The mean number of residents with three or more ADL impairments would better reveal the level of frailty tolerated by ALFs relative to nursing homes, but this measure was not available. The highest average proportion of ALF residents with cognitive impairments was just over 52% (ALFA), but this was still less than the comparable nursing home estimate of just over 56%. Two studies (HAWES and GAO) reported relatively small proportions of their residents (34.0% and 35.3%) with cognitive impairments.
|
).
|
, p. 40) reported that their focus group interviews with ALF residents revealed that many "were dissatisfied with being discharged to a hospital or nursing home whenever they needed any daily nursing care or monitoring." Thus, these researchers interpreted these departures as a failure of some facilities "to retain residents who needed nursing care or monitoring" (p. 40). Interpreting the implications of the 18% (ALFA) or 28% (NCAL) of the residents who departed for other alternatives (such as moves to retirement homes, another ALF, or a child's home) is even more difficult. It is impossible to determine whether such adjustments occurred because their current ALFs would not retain them or because their health and impairment status improved and they no longer needed their current and more supportive care environment.
These findings are difficult to interpret for another reason. Studies compute departure distributions for facilities with possibly very different resident turnover rates, that is, the relative size of the pool of residents that leave ALFs for any reason in any given period. Thus, the departure distributions of two studies may not be comparable. Assume, for example, that two studies, each reporting on a facility with 100 residents, both report that 30% of their annually "leaving" residents depart to a nursing home. (This is the type of finding presented in Table 4.) In the first study, 40% of the ALF residents have left their facility for any reason (the turnover rate) in a given year, whereas in the second study, 60% of the ALF residents have left their facility for any reason over the same period. The 30% "leaving" departure measures are obviously not equivalent because they are being applied to very different-sized pools of residents. In the first study, 12 residents (.3 x 40) will have left for a nursing home, but in the second study, 18 residents (.3 x 60) will have left for a nursing home. Thus, even as both studies had an at-risk population of 100 residents, in the first, 12% of the original ALF residents entered a nursing home, whereas in the second, 18% of the original ALF residents entered a nursing home. It is evident that generalizations are difficult to make from studies that do not document the turnover rates of their ALFs. Comparing the departure patterns of comparable longitudinal samples or panels of at-risk ALF residents eliminates this ambiguity. The destination patterns are also difficult to interpret because the studies do not report on the length of their recording period. Resident turnover rates will obviously depend on the length of time that residents are at risk of leaving their facilities.
Resident Admitting and Retaining Policies
Three of the ALF studies (HAWES, NIC, and GAO) reported on their resident admitting and retaining (discharge) policies, but generalizations are difficult because they measured different impairment or care indicators (see Table 5). One study (HAWES), for example, asked about admitting policies for residents with urinary incontinence (mean admitting rate, 61.4%), whereas the other two studies (NIC and GAO) asked about residents who could self-manage their urinary incontinence (weighted mean admitting rate, 93.0%) as opposed to those who required assistance with their urinary incontinence (weighted mean admitting rate, 71.3%). Only one of the three studies (GAO) inquired about admitting and retaining policies for residents requiring intravenous medication, tube feeding, use of an indwelling urinary catheter, and colostomy care.
|
Interpretations must be made with care for another reason. The unit of analysis in all three studies is the "facility," but one study (HAWES) reports a lower average number of facility beds (53) than the other two (63). Thus, its admitting rate policy will generally impact fewer older residents. Furthermore, none of the studies reports whether their smaller and larger bed facilities have the same admitting and retaining rates. Thus, the more lenient admitting rates of only a few facilities with a disproportionately large number of beds may get "lost" in the statistics, if a relatively large number of facilities with a relatively few beds report more stringent admitting rates.
A related deficiency is that the studies do not survey the facilities on how frequently they actually apply these retaining or admitting policies. For example, both the NIC and the GAO studies report that similar average percentages of their facilities will retain older residents who use a wheelchair (92.3% versus 90.2%). Assume (without any empirical evidence) that in the NIC study, facilities have to apply this retaining rule an average of once a month, whereas in the GAO study, facilities have to apply this rule an average of twice a month. Facilities in the GAO study are obviously having a greater impact on the ALF occupancy behavior of older persons in wheelchairs. None of the reviewed studies reports information on these policy practices.
These methodologic qualifications aside, generalizations are possible. ALF facilities were more likely to admit frail older persons when they had relatively minor or less serious physical or cognitive impairments or health care needs. These included residents who were less likely to demand hands-on care by unskilled aides (e.g., were not bedfast, moved around the facility without assistance, did not need help transferring or assistance with their locomotion); when they had care needs that did not demand skilled nursing procedures (e.g., the use of a catheter, intravenous medication, tube feeding, or ventilator, or did not need colostomy or ileostomy care, or could self-manage urinary or bowel incontinence); and when they did not require ongoing supervision or monitoring (e.g., suffered only from mild or moderate confusion, did not wander, or did not have memory, judgment, or behavioral problems).
Less than 50% (weighted average) of the ALF facilities would admit older persons if they were afflicted with any of 14 health conditions (see Table 5): colostomy or ileostomy care required, transfer help needed or required, indwelling urinary catheter needed or required, moderate to severe cognitive impairments or behavioral problems or symptoms, ongoing need for nursing care, ventilators used or required, intravenous medication or therapy required, tube feeding on an ongoing basis, or bedfast. Between one half and two thirds of the ALFs would admit residents with the following health conditions: electric cart used, locomotion help received, urinary incontinence, catheter or ostomies, and bowel incontinence assistance needed. Very similar generalizations can be made about the resident-retaining policies of ALFs. Less than 50% (weighted average) of the ALF facilities would retain older persons if they were afflicted with any of 10 health conditions.
In two of the three reporting studies (NIC and GAO), most of the resident discharge or retaining policies were substantially more lenient than the admitting policies. This allows for the tentative conclusion that most ALFs will let some share of their older residents remain in their facilities even as their conditions worsen. For example, 63% of the facilities in one study (NIC) would admit residents with a self-manageable bowel incontinence problem, but 73% of the facilities would retain such older residents. In the same study, just over 71% of the facilities would admit moderately confused older persons, but over 82% of the facilities would retain such residents. Similarly, in another study (GAO), about 4% of the facilities would admit bedfast older persons, but 23% of the facilities would retain them. Whereas the facilities in two studies (NIC and GAO) reported these discordant resident admission and retaining patterns, the third study (HAWES) found that very similar percentages of their facilities both admitted and retained older persons with the same impairment—a conclusion emphasized by its authors (Hawes et al., 1999).
Variations in the Admitting, Retaining, and Staffing Characteristics of Freestanding and Multilevel Facilities
ALFs in multilevel facilities were significantly more likely to admit and retain more seriously impaired older residents who needed nursing care than freestanding facilities (HAWES; see Table 6). Multilevel ALF facilities were also significantly more likely to have full- or part-time nursing staff and to offer nursing services.
|
). The implication is this: When nursing homes (both alone or part of a CCRC) are proximate to an ALF facility, high-care residents are more likely to be admitted to the nursing homes than to the ALFs. It is impossible to reconcile these opposite findings. Whatever the correct direction of the relationship, however, the findings emphasize that the admitting and retaining patterns attributed to ALFs overall will be significantly influenced by the mix of freestanding and multilevel facilities in their study's sample. It was earlier shown that 85% of all ALF properties are freestanding (see Table 1). If the smallest (under 30 units) freestanding ALF properties are excluded, then 48% of ALFs are freestanding. In comparison, the percentage of freestanding ALFs in the six studies ranged on average from 15% to 100% (see Table 2). Thus, some variation in the findings reported by the six studies will be systematically biased by their sample allocations of freestanding and multilevel facilities.
The statistical differences found in the admitting and characteristics of freestanding and multilevel facilities may also be spurious; that is, they may merely reflect a bed size differential. Freestanding ALF facilities are generally smaller and thus are more likely to look and function like mom-and-pop residences (board and care or foster homes) than ALFs in multilevel facilities (Morgan et al., 1995). As Table 1 shows, over 19,000 of about 23,000 freestanding assisted living properties consisted of 1–29 units.
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Two weaknesses of the data analyses found in all six studies deserve comment. First, the studies do not assess if there are state differences in the prevalence of frail residents in ALFs or in their admitting/discharge policies. This is an unexplainable omission, given that state regulatory environments so substantially explain why ALFs differently tolerate functionally disabled older persons.
Second, all the studies use the "property" or "facility" as their data collection unit rather than "beds." This is important because facilities differ greatly by bed size. This unfairly results in a larger facility of 80 beds having the same statistical "weight" as a smaller facility of 40 beds. This leaves open the possibility that the frailty-tolerant admitting policies of a 100-bed ALF facility are statistically equivalent to the frailty-intolerant admitting policies of a 20-bed facility even as the larger facility's policies impact five times as many seniors. These interpretation problems may be compounded because the inconsistent statistical populations and sampling procedures of these six studies also resulted in their ALF bed size averages ranging from 53 to 96 (see Table 2).
ALFs are dissimilar in many ways (Mollica, 2002). The findings of this analysis specifically emphasized the importance of distinguishing freestanding from multilevel ALF facilities. Findings supported the conclusion that facilities in these two groups have different admitting and discharge policies. Other research, however, will have to clarify the reasons for this relationship and if it persists after controlling for other ALF differences such as facility size.
The cross-sectional or point-in-time research methodologies of all six studies also produced resident length-of-occupancy and departure pattern findings that were difficult to interpret. Future national studies of ALFs would benefit greatly (despite the implementation expense) if they incorporated a longitudinal or panel methodologic design whereupon a national sample of ALF residents in different facilities could be followed through time not only to assess the trajectories of their impairment changes and adjustments, but also to track how facilities differently respond to their changing needs. Only very small ALF studies have employed such methodologies (Frytak et al., 2001; Pruchno & Rose, 2000).
The absence of such longitudinal research designs also explains why it is difficult for these six studies to assess whether older persons have aged in place in their ALFs. "Aging in place" refers to more than just older persons remaining in the same place for a sustained duration, a criterion used to define long-term care settings (Kane, Kane, & Ladd, 1998). Rather, aging in place also involves a proactive response from the older person's residential setting (whether initiated by the older person, significant others, other occupants, or providers) that in some way introduces solutions that respond to or compensate for an individual's functional or physical health declines. In the instance of ALFs, providers would have to adjust their "service provision and level of care criteria to meet residents' changing needs and to avoid having to discharge individuals to a higher level care prematurely" (Chapin & Dobbs-Kepper, 2001, p. 43). The six reviewed studies do not offer insights about these types of changes. The finding that ALF retention/discharge policies are more tolerant of higher levels of frailty than their admitting policies is the only evidence of aging-in-place practices in ALFs.
What is established is that more physically and vulnerable older persons are less likely to occupy ALFs than nursing homes, a conclusion also reached by studies focused on smaller ALF samples (Kane & Wilson, 2001; Pruchno & Rose, 2000; Zimmerman et al., 2003). Comparisons distinguishing severely impaired older persons (e.g., limitations in three or more ADLs) would help strengthen this generalization. ALF facilities were also more likely to admit or retain frail older persons when they had relatively minor or less serious physical or cognitive impairments or health care needs. Less than half of the ALF facilities would admit older persons who suffered from any of 14 health or disability problems or would retain older persons if they were afflicted with any of 10 health conditions.
Overall, these findings distinguish ALFs from nursing homes with their "more encompassing admission policies" (Zimmerman et al., 2003, p. 114). They are also consistent with the conclusion reported by Phillips and colleagues (2000) that the main reason why residents leave an ALF is because they need more care. On the other hand, older residents who are admitted to ALFs may enjoy relatively long stays, on average as much as 3 years, according to one of the six reviewed studies. The ability of older residents to delay their relocation to a nursing home for this length of time may be justification enough for them to enter an ALF, even though it will not be their last home.
The ALF alternative is here to stay. Intermediate projections estimate a 41% growth rate of this shelter and care option over the next 20 years (National Investment Conference, 2001). Over the last 5 to 10 years, state governments have tended to amend their ALF regulatory environments to make it more possible for this alternative to accommodate the physically and cognitively vulnerable old (Mollica, 2002). Many stakeholders have a strong vested interest in making the ALF a viable alternative to the nursing home, specifically for low-income seniors. Given the importance of the ALF as a long-term care alternative, stakeholders require more methodologically sound investigations to document its ever-changing potential.
|
Footnotes |
|---|
Received for publication March 24, 2003. Accepted for publication October 1, 2003.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. G. Kelsey, S. B. Laditka, and J. N. Laditka Dementia and Transitioning From Assisted Living to Memory Care Units: Perspectives of Administrators in Three Facility Types Gerontologist, July 24, 2009; (2009) gnp115v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Munn, D. Dobbs, A. Meier, C. S. Williams, H. Biola, and S. Zimmerman The End-of-Life Experience in Long-Term Care: Five Themes Identified From Focus Groups With Residents, Family Members, and Staff Gerontologist, August 1, 2008; 48(4): 485 - 494. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Giuliani, A. L. Gruber-Baldini, N. S. Park, L. A. Schrodt, F. Rokoske, P. D. Sloane, and S. Zimmerman Physical Performance Characteristics of Assisted Living Residents and Risk for Adverse Health Outcomes Gerontologist, April 1, 2008; 48(2): 203 - 212. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Brennan and S. J. Bally Psychosocial Adaptations to Dual Sensory Loss in Middle and Late Adulthood Trends in Amplification, December 1, 2007; 11(4): 281 - 300. [Abstract] [PDF]
|
|
|
|
|
|
R. I. Stone and S. C. Reinhard The Place of Assisted Living in Long-Term Care and Related Service Systems Gerontologist, December 1, 2007; 47(suppl_1): 23 - 32. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Zimmerman and P. D. Sloane Definition and Classification of Assisted Living Gerontologist, December 1, 2007; 47(suppl_1): 33 - 39. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Hyde, R. Perez, and B. Forester Dementia and Assisted Living Gerontologist, December 1, 2007; 47(suppl_1): 51 - 67. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hedrick, M. Guihan, M. Chapko, L. Manheim, J. Sullivan, M. Thomas, S. Barry, and A. Zhou Characteristics of Residents and Providers in the Assisted Living Pilot Program Gerontologist, June 1, 2007; 47(3): 365 - 377. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. S. Park, S. Zimmerman, P. D. Sloane, A. L. Gruber-Baldini, and J. K. Eckert An Empirical Typology of Residential Care/Assisted Living Based on a Four-State Study Gerontologist, April 1, 2006; 46(2): 238 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Zimmerman, C. S. Williams, P. S. Reed, M. Boustani, J. S. Preisser, E. Heck, and P. D. Sloane Attitudes, Stress, and Satisfaction of Staff Who Care for Residents With Dementia Gerontologist, October 1, 2005; 45(suppl_1): 96 - 105. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
