| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
Correspondence: Address correspondence to Sheryl Zimmerman, PhD, University of North Carolina, Cecil G. Sheps Center for Health Services Research, 725 Martin Luther King, Jr. Boulevard, Chapel Hill, NC 27514. E-mail: sheryl_zimmerman{at}unc.edu
| Abstract |
|---|
|
|
|---|
Key Words: Typology Subtypes Residential Care Senior Housing
In response to this latter concern, the U.S. Senate Special Committee on Aging convened an AL workgroup of 49 national organizations to develop guidelines for policy, regulation, and operations. The workgroup's final report to Congress proposed a three-part definition of AL related to (a) services and regulation, (b) private units, (c) and levels of care. Specifically, the workgroup offered a definition of AL as residences that "provide or coordinate oversight and services to meet individualized scheduled needs, based on assessments and service plans, and unscheduled needs" and noted that services include 24-hr awake staff, provision and oversight of personal and supportive services, health-related services (excluding 24-hr nursing care), social services, recreational activities, meals, housekeeping, laundry, and transportation (Assisted Living Workgroup, 2003, p. 12). The workgroup further recommended that AL units be private occupancy, shared only by choice, and that there be at least two levels of care in accordance with what the residence can accommodate in regard to staffing, special care, health care, and fire safety (Assisted Living Workgroup, 2003).
However, only a minority of the workgroup attendees supported any of the three elements of the definition without qualification, and only one supported all three as a unified definition. Participants took issue with the point related to the provision of health care, for example, because many felt that it remained too general. The requirement for private rooms was met with dissent because limiting AL to such units could, in fact, lessen consumer choice. Requiring at least two levels of care raised concern about how such levels would affect affordability. In addition, many questioned whether certain core principles—such as individualized care and maximizing of resident control—should be integral to AL, because such principles are aspirations rather than standard practice.
Given the outcome of this recent national effort, one can only conclude that the definition and classification of AL remains elusive. This point notwithstanding, it is worth noting that in developing regulations for AL, states have created de facto definitions of what is and is not AL. For example, Florida requires that a residence providing care to terminally ill residents have hospice coordinate those services (and meet other criteria) if it is to be considered AL (Office of Disability, Aging, and Long-Term Care Policy [DALTCP], 2004a). This is by no means a requirement in all states, however.
Thus, the search for definition and classification continues, and for understandable reasons. Clarity and accord could help researchers generalize their findings, policy makers work toward equitable service provision, practitioners provide quality care, and consumers exercise informed choice in selecting long-term care. An important dialectic is implicit in this quest, however: A low-bar definition (i.e., one with few requirements) would maximize affordability and choice but perhaps do so at the price of quality, whereas a high-bar definition (i.e., one with many requirements) would maximize services and quality but might cater to a select few—presumably those private-pay individuals for whom a common definition of AL is less necessary. Consequently, if a definition of AL is to be purposively promoted, it likely will address the minimum of what AL should be, much like the definition promoted by Kane and colleagues (1993) a decade ago.
This lack of consensus of a definition reflects the diversity of settings and the regulations under which they operate. For this reason, it is valuable to seek to differentiate various subtypes of AL. Such differentiation is important for consumer choice, regulatory oversight, developing and sharing of best practices, census tracking, marketing, and research. Furthermore, it is relevant regardless of the level at which the definition bar is set. This article addresses the overall definition of AL, describes and contrasts the nine typologies of AL that have been identified in the literature to date, discusses the conceptual issues inherent in the attempt to define and classify AL, and provides suggestions to further guide this effort.
| The Evolution of AL and AL Typologies |
|---|
|
|
|---|
Given that the term assisted living is applied quite broadly to a variety of facility types, some researchers have pursued an alternative strategy to focus the field, albeit with similar lack of consensus: developing a typology of AL. A typology, or a "systematic classification or study of types" (Webster's Unabridged Dictionary, 2001), is a potentially helpful effort in reductionism, identifying common elements notwithstanding noted differences. Grouping residences into subtypes can lead to an understanding of the interrelationships among and between variables and, when properly constructed, results in within-group homogeneity and between-group heterogeneity (Gold, Sloane, Mathew, Bledsoe, & Konanc, 1991). Experts have put forth multiple typologies relative to AL, embracing to various degrees the structure of AL, the process of AL, the population of AL, and the philosophy of AL care. All have been developed around the type of AL in existence at the time.
A typology put forth in 1992, and based on the information and types of facilities available up to that time, classified AL into three types: public housing, units in continuing care retirement communities, and freestanding facilities (that may or may not be on the campus of a nursing facility). Although clearly focused on macrolevel environmental/structural differences, each type differed in its target populations, funding, and organization and delivery of services (Manard et al., 1992). Of note, even this earliest typology reflected that AL served both publicly subsidized populations and others. Six years later, the National Investment Conference and the Assisted Living Federation of America (1998) identified similar types based on the facilities then evidenced: freestanding facilities, units in nursing homes, units in continuing care retirement communities, and units within congregate communities (National Investment Conference, 1998). Another typology categorized facilities that provided dementia programs. It also focused on these dimensions, noting subtypes to include small independently operated homes; small homes affiliated with other homes; larger, all-dementia facilities; programs in larger AL facilities; and programs in multilevel facilities (Davis et al., 2000).
Other typologies have approached the matter differently, focusing less on the macro-, structural elements, and more on some of the components that are consistent with the AL philosophy. Heumann and Boldy (1993) addressed predisposing conditions and environmental dimensions in their typology. Predisposing factors included the social values that underlie AL (e.g., self-actualization), the type of AL ownership, and the degree of governmental support and services (e.g., visiting or onsite services); environmental dimensions referred to matters such as the residence's integration with the neighborhood and the degree of communal space. Also using the core AL philosophy, MacDonald, Remus, and Laing (1994) suggested a typology that considers AL residences in terms of the degree to which they are constant (i.e., an environment that does not change over time) or accommodating (i.e., an environment that changes over time as residents age in place). Hawes and Phillips (2000) divided the universe of AL residences into four subtypes that expressly reflected consumer preference (Kane, Baker, Salmon, & Veazie, 1998). Their two dimensions, environmental (i.e., privacy) and service, resulted in the following four AL types: low/minimal privacy and service, high privacy and low service, high service and low privacy, and high privacy and high service.
In the mid-1990s, enough states had begun regulating AL that it was possible to use these regulations to deductively create a typology. Based on a review of state regulations, Mollica, Wilson, Ryther, and Lamarche (1995) identified three AL types as institutional board and care, housing and services, and purely service oriented. The first subtype subsumed generic residences, those in which shared rooms were common and few services were provided; the second recognized AL as a new model of residential long-term care that required apartments, provided more services, and espoused aging in place; and the third viewed AL as a service rather than a particular environmental configuration and allowed services to be provided in a variety of settings. In noting that models were not exclusive and could be combined, Mollica (1998) also identified umbrella models that use one set of regulations to cover more than one type of arrangement, and multiple levels of licensing within a single residence.
Finally, Zimmerman, Sloane, Eckert, and colleagues addressed the matter of typology development in an empirical manner and developed two typologies to classify AL. The first (Zimmerman et al., 2001, 2003) incorporated the dimensions of size and services, resulting in three subtypes: smaller facilities (operationalized as <16 beds), traditional facilities, and new-model facilities. These subtypes were presumed, and later found, to reflect differences in AL. Zimmerman and colleagues defined the latter two types using an empirical exercise that differentiated those known by expert judgment to reflect the purpose-built, new-model of AL care: (a) built in or after 1987; and (b) having two or more private-pay rates, at least 20% of residents who required assistance in transfer, at least 25% of residents who were incontinent, or a registered nurse or licensed practical nurse on duty at all times. Later, Park, Zimmerman, Sloane, Gruber-Baldini and Eckert (2006) undertook a cluster analysis of the 193 AL residences in the same sample and, using a comprehensive list of 26 structure, process, and case-mix variables, identified five AL types: impairment oriented, dementia oriented, mental illness oriented, housing oriented, and a mixed model. This typology included elements of structure, philosophy, and regulation. Table 1 summarizes all of these typologies and their respective components.
|
| Discussion |
|---|
|
|
|---|
What, then, are the relative merits of the typologies identified to date? It seems a foregone conclusion that a field that ranges from housing with services to nursing homes needs something more than macro-, structure-based characteristics to differentiate the types. The AL philosophy-based typologies hold an intuitive appeal, in that one would assume that in so being, the typology might differentiate AL from other types of housing with services. However, because a wide range of enumerated values could guide this effort (Assisted Living Quality Coalition, 1998); the time needed to obtain valid and reliable data to type them would be significant; they represent more of an ideal than a reality; and the philosophies as embodied in practice may change over time; this strategy may have more intuitive than practical appeal. Still, this approach would appeal to consumers—especially those who have sufficient resources and flexibility to choose among residences.
In discussing a philosophy-based typology, it is important to note that AL values are not the sole province of AL. The culture change movement espoused by the Pioneer Network (http://www.pioneernetwork.net) is working to put a new face on nursing home care by focusing more on resident quality of life with the goal of improving day-to-day experiences. The most well-known example of this movement is the Eden Alternative, which incorporates organizational and architectural changes, pets, plants, community involvement, and self-directed work teams into the nursing home experience (Coleman et al., 2002). More recently, it is evident in the Green House model, which promotes small houses with private rooms and full bathrooms and shared communal dining and cooking space in a family-style atmosphere (Rabig, Thomas, Kane, Cutler, & McAlilly, 2006). Thus, the only clear difference between nursing homes and AL is that only nursing homes are required to provide nursing services (although many AL facilities do so, as well; Zimmerman et al., 2005).
Perhaps, then, the issue of regulation should drive AL typologies. In fact, this is already occurring. Florida, for example, has four typologies of AL related to the type and intensity of care provision: standard, limited nursing services, extended congregate care, and limited mental health license (DALTCP, 2004a). Maryland, alternatively, has three typologies related to the level of care provided, labeled simply Level I, II, and III (DALTCP, 2004b). New Jersey has AL residences, comprehensive personal care homes, and AL programs, which speaks more to differences in structure and financing than type and intensity of care provision (DALTCP, 2004c). North Carolina, in contrast, has adult care homes, family care homes, and multiunit AL with services, the first two of which are differentiated by size and the third of which is not licensed (DALTCP, 2004d). Given the diversity that is AL, and the fact that states have developed their own typologies, it seems both unavoidable and appropriate that these regulatory types help define the field. In this strategy, state-based typologies would be identified and empirical methods used to determine national commonalities. One obvious benefit of doing this is that individuals typically look within a state for their housing with services, and so it is relevant within-state differences that are likely to matter to consumers.
The matter is not quite that simple, however, as even within a state, actual practices may well exceed or otherwise differ from regulation; thus, creating types based on categories will not fully capture reality. Another limitation is that state-based typologies complicate national dialogue on the topic. Among the multistate typologies, Mollica and colleagues' (1995) typologies were developed from state regulations but seem too broad to capture relevant differences related to financing and intensity of services, for example. Park and colleagues' (2006) typology, having been developed empirically using structure, process, and case-mix information from four states (Florida, Maryland, New Jersey, and North Carolina), goes further in identifying relevant characteristics that are not state bound. An especially important component of this typology is that in identifying case-mix differences across types, the likelihood of a residence to provide care to individuals with different needs, which has implications for transitions and relocations, becomes evident. Thus, replicating this effort by obtaining relevant structure, process/philosophy, and case-mix information from a representative sample of facilities across the country, and then conducting a cluster analysis as Park and colleagues did, might well yield some interesting and practical results of national relevance.
The Office of the Assistant Secretary for Planning and Evaluation, the National Center for Health Statistics, and the Agency for Healthcare Research and Quality plan to obtain national data in 2007 that will be used to describe AL settings. This effort is not intended to develop a typology, but it is conceivable that the resulting data could be examined for such a purpose. Although this is clearly a worthwhile exercise, one must consider its certain limitations. First and foremost, the typology will be restricted to the residences included in the sample, and so eligibility criteria (necessary for research and to focus any such effort) will in and of themselves set parameters for the typology. Examples of such limiting criteria are sampling frames that predefine the size of the eligible residence or the characteristics of the clientele. Also, there is risk that the very existence of a typology will give the illusion of more consensus and certainty than is justifiable. The typologies overviewed in Table 1 variously reflect the importance of structural variables such as size, of process and philosophy variables such as service, and of policy variables such as government funding. All differ in what they consider to be especially salient, as is likely to happen with any effort that is not driven by a key, definitive variable. Even if such a variable were to be selected, those familiar with statistical methods appreciate that regression techniques (which will likely be used in typology development) can lead to misleading conclusions when the potential predictor variables are intercorrelated—as is the case with those under study. Furthermore, unless a small group of variables can be identified that are especially important prognostic indicators, then the concepts on which the typology is built may not reflect truly important differences. Another limitation of typologies is that once they are developed, they likely will be promulgated, which will reduce consumer choice and stifle innovation. This is an important point, as consumer choice was and remains a driving force in the field of housing with services. Finally, regardless of the amount of effort, some residences are likely to elude classification, and others may belong in more than one category; depending on the purpose toward which the typology will be put, such omission or misclassification may be costly. For example, the cluster analysis undertaken by Park and colleagues (2006) results in slightly different typologies depending on which criterion variables one chooses to drive the clusters; among other differences, one strategy derives a cluster of resource rich residences that is not evident using a different strategy.
Table 2 summarizes the benefits and limitations of classifying AL residences. The potential benefit related to assessing quality is especially relevant given many of the quality-of-care concerns that have plagued the field of long-term care. To the extent that even the simple typology of (a) smaller residences (with <16 beds), (b) traditional residences, and (c) new-model residences has identified differences in care and outcomes across types (Zimmerman et al., 2003, 2005), this is clearly an effort than can provide information to inform practice and policy.
|
The preceding discussion begs certain considerations for future efforts to create a typology to classify AL (see Table 3). First and foremost, and as affirmed by many who attended the U.S. Department of Health and Human Services, National Center for Health Statistics, Expert Meeting on Typology of Long-Term Care Residential Places (January 2004), the consumer should drive a typology of AL; that is, the typology should reflect the places in which older adults who need assistance actually live. This perspective suggests that the resident be the sampling unit, which is an admittedly more complicated enumeration strategy than one in which the residence is the sampling unit. Thus, at least initially, a strategy that uses both a person-based and provider-based approach might be indicated, with one validating the other and both indicating what the other does not capture.
|
| Footnotes |
|---|
1 Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill. ![]()
2 School of Social Work, University of North Carolina at Chapel Hill. ![]()
3 Department of Family Medicine, University of North Carolina at Chapel Hill. ![]()
Decision Editor: Susan L. Hughes, DSW
Received for publication March 21, 2006. Accepted for publication February 21, 2007.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|