Defining Quality in Assisted Living: Comparing Apples, Oranges, and Broccoli

Transcription

1 The Gerontologist Vol. 47, Special Issue III, Copyright 2007 by The Gerontological Society of America Defining Quality in Assisted Living: Comparing Apples, Oranges, and Broccoli Purpose: The purpose of this article is to discuss and describe various measures of quality, quality indicators, and uses of information on quality with specific reference to the role or purpose of assisted living. Design and Methods: We reviewed a variety of major studies of assisted living quality. We elaborated models of assisted living based on differing goals or claims made for it by providers and policy makers. We then searched for available quality measures that might indicate whether assisted living was meeting those goals or living up to those claims. Results: Any meaningful concept of quality must embrace a variety of dimensions, including quality of care, quality of life, the physical environment, and resident rights. The ability to use a multidimensional concept of quality is complicated by the lack of consensus, confusion, and disagreement among consumers, providers, and regulators about the role of assisted living. This disagreement significantly confounds the task of comparing quality among assisted living settings and between assisted living and other types of long-term care. Implications: We propose ways that researchers may compare quality along dimensions claimed to be intrinsic to assisted living as part of an effort to inform consumer information systems, quality monitoring and assurance systems, and policy-relevant research. Such comparisons would vary, depending on the intended use of the indicators and role defined for assisted living. We wish to acknowledge support from Grant RO1-HS from the Agency for Healthcare Research and Quality (AHRQ). In addition, a conference grant from AHRQ to Dr. Rosalie Kane supported this effort. The article s contents are solely our own responsibility and do not necessarily represent the official views of AHRQ. We also wish to acknowledge our theft of part of an article title from Sheryl Zimmerman and colleagues (2003), who first articulated the idea that comparing nursing homes to assisted living could be like comparing apples and oranges; and from Susan Hughes (1985), who was first to use fruit metaphors to illustrate substantive differences. Address correspondence to Catherine Hawes, PhD, Regents Professor, Department of Health Policy and Management, School of Rural Public Health, Texas A&M HSC, 1266 TAMU, College Station, TX hawes@srph.tamhsc.edu 1 Department of Health Policy and Management, School of Rural Public Health, Texas A&M University, College Station. Catherine Hawes, PhD, 1 and Charles D. Phillips, PhD, MPH 1 However, all uses contain structural, process, and outcome quality indicators, including direct feedback from interviews with residents that go beyond satisfaction measures. Key Words: Assisted living, Residential care, Quality, Quality indicators, Long-term care The last decades of the 20th century saw the emergence and growth of a new industry known as assisted living (AL). Consumer demand, concerns about nursing home quality, pressure from providers, and the availability of capital for construction and conversion of facilities have combined with states interest in containing nursing home use to produce dramatic growth in this industry. These pressures have dramatically increased over the past few years, with heightened pressure on the states to implement the Olmstead decision (requiring that individuals with disabilities be cared for in the least restrictive environment possible; Olmstead v. L. C., 1999), reduce use of nursing homes and state mental institutions, respond to initiatives sponsored by the Bush Administration to reduce nursing home use (e.g., the New Freedom Initiative, Executive Order 13217, and Real Choice Systems Change grants, cms.hhs.gov/realchoice/), and address continued efforts to shrink Medicaid spending on nursing homes (Kasper & O Malley, 2006; Summer, 2005). Some observers have argued that the development of AL represents a fundamental change for long-term care (LTC; Kane & Wilson, 1993). Advocates assert that AL represents a promising new model of residential longterm care, one that blurs the sharp and invidious distinction between receiving long-term care in one s own home and in an institution (Hawes, Phillips, & Rose, 2000, p. 1). Initially, the development of AL was largely an unregulated market response to both demographic trends and consumer preferences. More recently, however, state involvement in setting standards and developing Medicaid payment policies and waiver programs for AL has expanded exponentially 40 The Gerontologist

2 (Mollica, 2002; Mollica & Johnson-LaMarche, 2005). As of 2002, 32 states had a licensure category or a statute that used the term assisted living. In addition, states reported growth in supply a 30% increase in the number of AL facilities (ALFs) between 1998 and 2000 and 14.5% growth in the number of ALFs between 2000 and 2002 (Mollica, 2002). In part, this growth represented a combination of new ALF construction and conversion of existing facilities to AL. In some instances, growth resulted from the reclassification or renaming of some type of already available residential LTC facilities, such as personal care homes, into AL, either voluntarily by the provider or as part of a state regulatory name change. Growth in the industry has slowed more recently (i.e., 3% growth in units between 2002 and 2004 and no growth in the number of facilities), but AL remains a significant sector in LTC (Mollica & Johnson-LaMarche, 2005). There is no federal regulation of AL and no mandatory definition of what constitutes AL. Thus, no reliable national list or count is available. Because of this, estimates of the total number of facilities vary, in large measure because of the differences in definitions across states and among facilities concerning what constitutes an ALF. As a result, studies use different definitions and produce very different numbers. For example, a national study by Hawes and colleagues used a standard definition of AL that was independent of various state categorizations and estimated 11,500 ALFs in 1998 with slightly more than 611,000 beds that provided some level assistance with activities of daily living (ADLs) and medication (Hawes, Phillips, Rose, Holan, & Sherman, 2003). Based on state licensure data, which allowed each state to individually define AL in 2004 there were 36,451 licensed residential care facilities with 937,601 units/beds, some of which would be more accurately described as board and care with minimal services and shared accommodations (Mollica & Johnson-LaMarche, 2005). As a point of comparison, there were an estimated 17,000 nursing homes with 1.6 million beds in 1996 (Krauss et al., 1997). In any event, no matter what definition one uses, AL now constitutes a major segment of the LTC services sector. Despite this growth and the current size of the AL industry, researchers and policymakers know relatively little about ALFs and their residents, much less about the quality of life and care in these facilities. This lack of information is particularly troubling because there is tremendous variability among facilities known as AL (Burdick et al., 2005; Hawes, Rose, & Phillips, 2000; Hawes et al., 2003; Hedrick et al., 2003; Morgan, Eckert, Gruber-Baldini, & Zimmerman, 2004; Zimmerman et al., 2003). Moreover, this dearth of information is especially problematic for public officials and policymakers. They must face the rapid growth of the AL industry; its increasing role in providing LTC for frail elders; Vol. 47, Special Issue III, the largely uncritical enthusiasm for AL that dominated the policy process during the past decade; and emerging concerns about quality, including poor medication management and underrecognition and treatment of various conditions (Ball et al., 2004; Beattie, Song, & LaGore, 2005; Gray et al., 2006; Gruber-Baldini, Boustani, Sloane, & Zimmerman, 2004; Magsi & Malloy, 2005; Sloane et al., 2004; U.S. General Accounting Office, 1999). For the purposes of the academic community, more information on AL quality would be helpful and interesting. For policy makers faced with the circumstances noted above, such information is crucial. It is difficult to see how AL can continue for long in its current niche, or develop along the lines desired by policy makers, without facing increased public scrutiny of its performance. No matter how vociferously the industry might oppose it, some form of quality assurance process or processes will eventually emerge from the current jumble of legislation. There are several reasons for expecting that a quality assurance process will be imposed. The AL industry serves an increasingly vulnerable population. It is an industry that provides services without any internally developed or professionally mandated standards of quality or of care. It is also an industry beginning to take more public funds in payment for its services, and there are growing concerns about quality. These conditions constitute a clear prescription for the development of standards of care and of an array of quality indicators that individual consumers and public programs can use to differentiate among ALFs. These same basic conditions apply to nursing homes, and these factors were the impetus for the elaborate structure of regulation that exists in the nursing home sector today (Phillips, 2000). Even if only some of these conditions apply to AL, it seems very unlikely that the industry will long escape more regularized scrutiny. This is especially likely to be the case if the AL industry follows the poor example of the nursing home industry and tries to eschew all responsibility for ensuring the quality of care provided by its members. Methods The focus of this article is on developing links between models and goals of AL and potential quality indicators. Drawing out the implications of these models and the philosophy of AL and then determining which quality indicators might serve those interested in each model is our goal. In light of that charge, we reviewed what we considered major examples of research focusing on quality in AL. We also reviewed the publications of providers and provider-oriented groups. Finally, we married all of this information with our own experience with

3 quality measurement and quality assurance in nursing homes and other LTC settings. The findings we present first delineate the basic disagreements among consumers, providers, and policy makers about exactly what one can expect of AL. We next clarify the most common concept of AL and the AL philosophy and draw out their implications for quality measurement and the choice of quality indicators. We then discuss how other models of AL might drive one toward different approaches to quality measurement in AL. We next discuss consumer attitudes and beliefs about what constitutes quality in AL. Finally, we discuss the uses to which quality indicators might be put. Findings Disagreement Concerning the Role of AL Despite the growing need for quality measures for AL, achieving agreement on what they should be is challenging. For nursing homes, there is general agreement that a valid concept of quality must embrace quality of care; quality of life, including aspects of the environment; and resident rights (Institute of Medicine, 1986; Kane, 2001). Indeed, this view of what quality comprises is so widely accepted that these domains of quality form the framework for nursing home regulatory reform adopted under the 1987 Omnibus Budget Reconciliation Act. There is less agreement on key domains for AL. This lack of consensus grows out of confusion and, in some cases, out-and-out disagreement about the role of AL among consumers, providers, regulators, and policy makers. There appears to be some consensus about the philosophy of AL. Indeed, these philosophical principles have been the basis of the advocates position that AL represents a new model of residential LTC. A useful definition of AL that incorporates this philosophy was articulated by the Assisted Living Quality Coalition, a group of consumer and provider groups including the Alzheimer s Association, AARP, the American Association of Homes and Services for the Aging, the Assisted Living Federation of America, the American Seniors Housing Association, and the American Health Care Association/ National Center for Assisted Living: A congregate residential setting that provides or coordinates personal services, 24-hour supervision and assistance (scheduled and unscheduled), activities, and health related services; designed to minimize the need to move, designed to accommodate individual residents changing needs and preferences, designed to maximize residents dignity, autonomy, privacy, independence, and safety; and designed to encourage family and community involvement. (Assisted Living Quality Coalition, 1998, p. 2) Despite agreement on this philosophical concept of AL, there has been disagreement about two fundamental issues related to the role of AL. The first divergence of opinion focuses on whether AL should be expected to address the health care needs of residents. Similarly, there is disagreement over the term residents, with some arguing that tenant is a more appropriate appellation for individuals who live in ALFs. This view essentially espouses the position that consumers are expected to make decisions and arrangements about any needed health care services independently and select and direct personal care services as well, according to their preferences. Advocates of this view of AL believe such independence is part of the core of consumer autonomy and makes AL consistent with a consumerdirected model of care. As a practical matter, this view can be seen operationally in what are referred to as negotiated risk contracts, and it is embraced by many ALF providers. For example, half (50%) of a national probability sample of administrators argued that it was the responsibility of residents and their families to recognize and address changing health care needs (Hawes et al., 2000). This concept of the role of AL also asserts that it is fundamentally a social model that focuses on environmental amenities, honors resident independence and autonomy, and eschews what it regards as the more hierarchical and overly directive medical model found in nursing homes. Another concept of the role of AL asserts that such facilities should be able to address residents changing health care needs, even within a social model. Indeed, part of the public philosophy of AL has been that ALFs should meet the scheduled and unscheduled needs of residents and allow residents to age in place by offering or arranging services that meet changes in resident needs and preferences over time (Assisted Living Quality Coalition, 1998; Kane & Wilson, 1993). For some, this means that AL must, to some degree, recognize and help address residents functional limitations and health care needs. Over the past few years, there has been some erosion of support for the idea that the role of AL is to facilitate agingin place among residents, as providers have recognized the challenges inherent in such an undertaking. Nonetheless, some members of what came to be known as the Assisted Living Workgroup strongly endorsed the idea. The Workgroup was an assemblage of stakeholders including consumer advocates, providers, regulators, and health care professionals, such as geriatricians initially brought together through the efforts of the U.S. Senate Special Committee on Aging. A key goal for the group was to make recommendations about how to assure quality in AL. Within this group there were serious divisions around issues of health care and the responsibility of ALFs, as well as other issues (Assisted Living Workgroup, 2003). Clearly, this 42 The Gerontologist

4 disagreement about the role of AL vis-à-vis health care issues had a significant impact on views of what constitutes quality and how it should be measured. The second area of disagreement among members of the AL industry relates to features of the environment. Surveys of consumers have demonstrated that they overwhelmingly prefer privacy of accommodations, including bathrooms (Jenkens, 1997; Kane, Baker, & Veazie, 1998). However, although privacy is much more common in places known as ALFs than in nursing homes, a substantial proportion of the industry still offers shared accommodations. Indeed, one of the major provider associations, the Assisted Living Federation of America, was formed by the combination of two groups. One group comprised mainly the newer model of AL, including many purpose-built facilities, whereas the other group, the National Association of Residential Facilities, represented traditional board and care homes. Thus, the Assisted Living Federation of America does not endorse single-unit occupancy as a key element of AL, and one can also see this split in many states in which licensure regulations have classified all residential care facilities as AL. This divergence of opinion among consumers and providers about the role of AL and about what is and what is not AL is further complicated by the disparate roles assigned to ALFs by state policy makers. Oregon implemented the first licensure regulation specifically directed toward AL in By 1992, fewer than 10 states had such regulations in place. However, as noted, by 2004, most states had expanded their definition of residential care to include a specific licensure category known as AL or had simply incorporated these facilities into their traditional concept of residential care, and most states provided some coverage for the service component of AL through the Medicaid program or Medicaid waivers (Mollica & Johnson- LaMarche, 2005). Despite the phenomenal growth in both public and private support for AL, no consensus emerged on the appropriate regulatory model. Models have varied from state policies, like Oregon s, that seek to create AL as a unique LTC arrangement with distinctive service requirements and environmental features (e.g., services to meet scheduled and unscheduled needs, requiring apartments with kitchens and privacy in accommodations) to regulatory models that basically allow the types of shared rooms and limited services typically found in traditional board and care homes. In addition, states have differed on whether ALFs should be required to, allowed to, or prohibited from providing any nursing care and monitoring, as well as whether regulation should be limited to only the service component, in effect treating AL as a kind of home health service (Mollica, 1998, 2002). The Medicaid waiver programs that include AL demonstrate even more Vol. 47, Special Issue III, diversity, covering, as they do, only persons eligible for a nursing home level of care and introducing the role of ALFs as a substitute for nursing homes. Disagreements among consumer advocates, providers, and regulatory bodies concerning the role of AL and what comprises its essential features complicate the task of defining quality. At the same time, the need for information on quality is especially pressing as the industry continues to expand and policymakers seek to define a meaningful role for AL within the portfolio of LTC services. Thus, in this article, we discuss a variety of potential uses of quality indicators and a variety of users of information on quality. These include consumer information systems, quality monitoring and assurance systems, and policy-relevant research. These potential users need quality measures that can facilitate relevant comparisons among different models or types of ALFs, between AL and traditional board and care or residential care homes, and between ALFs and nursing homes. Adherence to the Philosophical Principles of AL Comparisons among ALFs might reasonably take several forms. For example, because some observers view AL as a new model of LTC, one might wish to determine the degree to which the universe of places that call themselves ALFs adhere to the philosophy that underpins this new model. The potential quality indicators tied to this philosophy imply a fairly comprehensive role for AL and, thus, a reasonably inclusive concept of quality. The key domains include a range of structure, process, and outcome quality indicators and focus on the key philosophical tenets of AL. These tenets were derived from various sources that sought to articulate the key features of AL. According to the definition propounded by the Assisted Living Quality Coalition (1998, p. 2), the key features of AL include the following: Provision of key services, including: personal services and oversight available 24 hr a day (provided or arranged/coordinated by the ALF); services to meet residents scheduled and unscheduled needs, including health-related services; and activities. Provision of services designed so as to minimize the residents need to move; accommodate individual residents changing needs and preferences; and maximize residents dignity, autonomy, independence, and safety. Provision of a setting that maximizes autonomy, privacy, and safety. Provision of an environment that encourages family and community involvement. The translation of this philosophy into operational definitions and measures is both challenging and fraught with potentially disparate concepts of

5 Table 1. Potential Indicators of an Assisted Living Facility s Level of Adherence to Key Principles Included in the Philosophy of Assisted Living Environmental/Facility Characteristics: Illustrative Structural and Process Quality Measures Principle: Provide resident autonomy in the environment Private vs. shared rooms, apartments Private vs. shared bathrooms (full and half) Ability to bring and arrange furniture, control individual unit temperature, have pet, lock door, store and prepare food Availability of adequate public space Accessibility of outdoor space Principle: Meet scheduled and unscheduled needs Availability of nursing care Direct care staffing levels Staff training and knowledge Availability of basic personal care, medication assistance, transportation, activities Availability of special services, such as help with transfers, incontinence care, help with toileting Principle: Facilitate aging in place Policies on discharge Availability of supportive devices Policies on use of Medicaid and charity care Policies on use of wheelchairs, walkers Ability to manage residents with increasing cognitive impairment or behaviors (e.g., safe environment, special staff training, appropriate activities, provision of cueing and supervision) Principle: Promote resident independence and autonomy Availability of health promotion activities, exercise, strength and stamina training Provide care and services so as to promote maximum independence Policies and practices that provide choices (e.g., activities, meal time, menus, service provision) Availability of meaningful negotiated risk a Outcomes or Effects on Residents Resident rating/perception of choices and control environment, services, personal assistance, food, staffing, safety, cleanliness, whether needs are met Resident perception of whether treated with respect and dignity Resident confidence that future needs can be met by the facility Residents unmet needs Residents length of stay Reasons for which, and location to which, residents are discharged Resident involvement with the community Resident level of social interaction Resident involvement/social interaction with family Note: a An individual assisted living facility resident formally accepts any risk involved in some activity (e.g., walking outside when he or she has a history of falls) through a negotiated agreement signed by the resident and facility management. such issues as which structures and processes facilitate aging in place or minimize a resident s need to move; what consumer autonomy means in practice, particularly for consumers with different levels of physical and cognitive functioning; and what unscheduled needs actually are. Having said that, there is some general agreement of what the philosophy might mean in practice. Table 1 provides an illustrative list of potential indicators. The service components of the philosophy are typically translated into an operational definition of an ALF as a facility that provides or arranges at least the following: 24-hr staff, housekeeping, at least two meals a day (and usually three), and help with at least two ADLs or one or more ADLs and medication assistance. ALFs may provide these services directly with their own staff or arrange for the provision of services through an external provider, such as a home health agency. However, there has been some erosion of agreement about whether aging in place is part of AL, and there is no consensus about whether licensed nursing services are essential to meeting scheduled and unscheduled needs or whether such services facilitate aging in place. However, if one focuses on the key philosophical tenets, quality indicators should include availability of those key services, as well as indicators of process quality related to whether services are arranged and provided so as to maximize a resident s independence. Furthermore, this concept of the role of AL may imply the need for measurement of policies that are intended to promote autonomy, dignity, and choice. For example, some ALFs restrict the use of wheelchairs and walkers in the public spaces of the facility. Although such policies may meet the preferences of physically intact residents, they may limit the ability of residents to age in place. 44 The Gerontologist

6 Experts should develop a similar set of indicators to assess the features of the setting and the environment within the facilities. For example, although there is some disagreement among the providers about whether private accommodations and bathrooms are essential components of AL, privacy is typically considered a key element of AL (Jenkens, 1997; Kane et al., 1998). However, privacy is not the only aspect of the environment that may contribute to resident autonomy and independence. Focus groups with residents and family members suggest that environmental autonomy might also include residents having the ability to personalize their private space by bringing in and arranging furniture; having adequate storage for clothing and other possessions; and being able to store and prepare food (usually small meals or snacks), control the environment, lock the door, and, in some cases, have a pet. Some also suggested that the availability of adequate public spaces for dining, visiting with friends, and engaging in activities, as well as attractive and accessible outdoor spaces, were also important parts of environmental quality from the residents perspective (Fralich & Hawes, 2004; Greene, Hawes, Wood, & Woodsong, 1997/1998). Finally, this expansive concept of AL and its role in meeting the needs and preferences of frail elders could include a more direct examination of what AL actually accomplishes, rather than merely its capacity to achieve philosophical goals. This would necessitate the development of process and outcome indicators that attempt to capture the goals of AL at the resident level, assessing residents day-to-day experience living in an ALF and the outcomes of this arrangement. Asking residents directly about their experiences and their assessment of the facility s performance is one key aspect of such an undertaking. Rather than merely asking about satisfaction, the indicators might assess the residents views on the choices and control they have, how they are treated by staff, the level of social interaction and community involvement, whether activities meet their preferences, and so on. In addition, if one accepts the premise that ALFs should meet residents scheduled and unscheduled needs, one might look not only at policies and service availability but also at residents unmet needs. Furthermore, length of stay, discharge location, and reasons for exit may be relevant indicators of whether residents are able to age in place or, in conjunction with indicators of unmet needs, at least the degree to which residents are able to have changing needs and preferences met in the facility. Comparisons Among Different Models of AL For Medicaid waiver programs, the underlying issue, by regulation, is whether AL can reduce nursing home use. In addition, states wishing to Vol. 47, Special Issue III, contain nursing home use and costs are increasingly interested in the potential of various types of residential care settings to delay or prevent nursing home care for some proportion of residents or potential residents (Mollica, 1998, 2002). Indeed, some researchers have suggested that the use of AL for persons with Alzheimer s disease and other dementias could reduce nursing home utilization and Medicaid costs by an estimated $2,000 per person per month as of the late 1990s (Leon, Cheng, & Neumann, 1998). The Olmstead decision has also encouraged states to examine noninstitutional settings for persons with disabilities and to expand the use of residential LTC settings (U.S. Government Accountability Office, 2001). Despite this growth in state policy involvement, states continue to vary in the role they define for AL, as noted previously. Some states have redefined all residential care facilities as AL, retaining a traditional board and care orientation. Other states have structured their licensure regulations and Medicaid waiver policies to divert frail elders from nursing homes. Some states have allowed ALFs to provide daily nursing care, whereas others explicitly prohibit such services. Some states require private accommodations, whereas others allow semiprivate and even ward-type rooms (Mollica, 2002; Mollica & Johnson-LaMarche, 2005). In addition, even within the constraints imposed by licensure, individual facilities and multistate chains have selected models with varying degrees of privacy and accommodations (Hawes et al., 2003). As a result, there is considerable variation among places known as AL in ownership, auspice, size, target resident population, accommodations, services, staffing, and price. Thus, both public and private policies have contributed to the emergence of different types of ALFs across the country. The result can be seen in national and multistate studies (Hawes et al., 2000, 2003; Hedrick et al., 2003; Zimmerman et al., 2003). These different models suggest a variety of research issues and the need for diverse quality indicators. Issues related to this diversity among ALFs include the following: Are differences in outcomes, such as prevented or delayed nursing home use, associated with different types or models of AL? Can ALFs substitute for nursing homes? Do some models or features of ALFs produce greater satisfaction among consumers? Are ALFs capable of providing appropriate care for an increasingly frail and disabled population of residents, and does this care differ across various types of ALFs? What should the eligibility criteria be for admission and retention/discharge from an ALF? The answers to these questions require empirical data. Researchers must better understand the characteristics of the ALFs (and potentially other

7 types of residential care facilities), particularly along critical dimensions that might affect resident outcomes. Thus, experts need a wide variety of structural, process, and outcome quality indicators to make relevant comparisons among ALFs, between different types of ALFs (e.g., between what Zimmerman and colleagues, 2003, called newmodel ALFs and traditional residential care/al facilities [RC/AL facilities]), between ALFs and other categories of board and care homes, and between ALFs and nursing homes. There are a variety of audiences interested in such issues. For example, sophisticated consumers and their advocates need such information to make informed choices among facilities in order to maximize the services, environment, and outcomes that the individual consumer desires. Policymakers also need measures that assess the effect of different types of models of ALFs. For example, Mollica s (1998, 2002) surveys suggested that states have envisioned different roles for AL in their portfolio of LTC services. However, research is needed to determine whether one model of AL is more or less likely to achieve specific goals. Such research can inform licensure policy and Medicaid payment and coverage policies. The only study that has produced national estimates about the universe of ALFs across the country was the one conducted for the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. This study was restricted to places that were self-proclaimed ALFs or that provided a specified range of basic services common to most ALFs (i.e., assistance with 2 ADL or 1 ADL and medications). Despite this restriction, the study still found tremendous variation among the facilities. For example, fewer than half of the resident units were apartments; most accommodations were bedrooms. The vast majority of these accommodations (73%) were private, but a sizable proportion of the apartments or rooms (27%) were shared by two or more unrelated individuals, and more than one third (35%) of the accommodations had a shared bathroom. The study found similar variability with respect to services, particularly in terms of whether ALFs had a registered nurse on staff and offered nursing care with their own staff (Hawes et al., 2003). Indeed, the study found what one might describe as four different models of AL, depending on the mix of services and privacy each offered. Although this study was largely descriptive, it did find variation across these four different types of ALFs in such features as their policies on admission and retention of residents, service availability, the level of resident functional impairment, length of stay, and likelihood of discharge to a nursing home (Hawes et al., 2003; Phillips et al., 2003). All of these might reasonably be considered structural and outcome-oriented measures of quality and might well be of interest to consumers and policy makers. In a smaller but well-defined longitudinal study that compared care among nursing homes and three types of RC/AL facilities in four states, the researchers also found significant differences among the RC/AL facilities in these states (Zimmerman et al., 2003). In this study, Zimmerman and colleagues examined variation in a large set of process quality measures among RC/AL facilities that varied in size and whether they embodied elements of traditional residential care or new-model AL. They found differences across these facilities in 10 different process quality indicators, such as the amount of health services they offered, their admission policies, privacy, policy clarity, resident control, and individual freedom or autonomy, as well as the average level of resident impairment (Zimmerman et al., 2003). These studies raise important questions about the effect of different models or types of AL and suggest the need for the development and use of (a) a variety of structural measures that adequately capture key features of the facilities and (b) process and outcome quality indicators to assess the effects of these different facility features. Effect of Different Models for Providing Needed Health Care Services When assessing the performance of ALFs, one could examine many areas of health care monitoring, oversight, or provision. However, the most visible difference among ALFs relates to nursing services, in particular whether facilities offer nursing care with their own staff, arrange nursing care, or do not offer such care or monitoring. Although variation in staffing among ALFs appears to be largely a result of choice by facilities, this is clearly an issue that is amenable to regulation (Mollica, 2002). Moreover, it is also clear that states differ on the issue of staffing type and staffing levels in AL, and this was also an area of contention among members of the Assisted Living Workgroup (2003): Whatever their position on whether to require licensed nurses on ALF staff, many states have allowed such residential care facilities to house residents with greater levels of impairment. By the mid-1990s, the majority of state licensing agencies allowed ALFs to house residents who were chairfast because of health problems or who used wheelchairs to get around inside the facility. One third of the licensing agencies allowed such facilities to retain residents who were bedfast (Hawes, Wildfire, & Lux, 1992). As a result of a variety of factors, including Olmstead and federal policies designed to increase the use of community-based services, many states have embarked on more aggressive strategies for expanding the potential role of ALFs (Kane & Wilson, 1993; 46 The Gerontologist

8 Table 2. Illustrative Indicators Related to Assisted Living Facility Differences in Health Care/Nurse Staffing Models Structural Indicators Describing the Assisted Living Facility Model Facility size Registered nurse staffing level Licensed practical/vocational nurse staffing level Policy on providing nursing care with staff Staffing levels for direct care staff Policy on arranging nursing care from outside Environmental features (e.g., privacy, type of accommodations) Indicators Effects of Interest Structural quality indicators Policies on admission and discharge Availability of health-related services Staff training Process quality indicators Adequacy of medication management; medication errors Staff knowledge of medication management and key care issues Comprehensiveness of service plan; individualized service plan Accuracy and completeness of health care records, assessment Provision of assistance so as to maximize functional and cognitive independence Resident-focused, outcome-oriented indicators Length of stay Discharge location Reasons for exit from assisted living facility Unmet care needs Satisfaction/resident ratings of staffing adequacy and ability to meet needs Confidence that change in status will be recognized and addressed Willingness to recommend assisted living facility to a friend Changes in status over time Physical functioning in activities of daily living Cognitive function Urinary continence Resident s use of hospital, emergency room Resident s use of nursing home, home health Vol. 47, Special Issue III, Mollica, 1998, 2002; Mollica & Johnson-LaMarche, 2005; Newcomer, Lee, & Wilson, 1997; Reinhard, Young, Kane, & Quinn, 2006; U.S. Government Accountability Office, 2001). These strategies include permitting the provision of daily or intermittent nursing care (including skilled care) and hospice care in these facilities, allowing retention of residents with greater levels of impairment, and modifying nurse practice acts to allow nonlicensed personnel to provide certain services independently or under supervision. As a result, there is variation within and across states and facilities in the types of residents they serve and in the way they approach meeting residents health needs. Examining the effect of differences in staffing type could inform public policy as well as programmatic decisions by ALFs. Clearly, one challenge inherent in evaluating the impact of such differences is to conceptualize the types of effects such a structural feature might reasonably be expected to produce. However, from a policy perspective, one would certainly wish to know how staffing differences affect an ALF s ability to provide appropriate quality of care and quality of life, meet resident preferences, minimize the need for residents to move, promote aging in place, and prevent or delay nursing home use. (It is important to note that some proponents of AL might view these measures as inconsistent with a social model of AL.) A related set of policy concerns arises from variation across states in the role assigned to AL by ALFs themselves and by policy makers. These embrace concerns about the capacity of ALFs to care for more impaired residents and issues of eligibility and level of care in ALFs and other residential care facilities. Increasingly, state regulators are expressing some apprehension that the combination of public policies and consumer preferences for aging in place in a noninstitutional environment may present ALFs with a mix of residents whose health-related needs cannot be adequately met. Furthermore, there is some evidence to support such concerns. For example, the national study of AL found that although relatively few residents overall needed hands-on assistance with toileting, more than one quarter (26%) of those who did need such help reported needing more assistance than they were receiving (e.g., wetting themselves while waiting too long for requested assistance; Hawes et al., 2000). Similarly, Zimmerman and colleagues (2003) found notable differences in process quality and resident outcomes across different models of RC/AL facilities. In addition, a four-state study by the U.S. Government Accountability Office (the U.S. General Accounting Office at the time of this report; 1999) raised serious concerns about quality of care in ALFs and other residential care facilities. Initial results suggested some types of ALFs produce better process quality. However, there are inconsistent results about the effect of registered nurses on discharge to nursing homes and length of stay (Aud & Rantz, 2005; Phillips et al., 2003; Zimmerman et al., 2003). Thus, there are reasonable arguments for including a variety of health outcomes among quality measures when examining some aspects of AL (Table 2 shows an illustrative list). Effect of Different Models of ALF Accommodations Another dimension on which ALFs differ is accommodations. In examining a national probability sample of ALFs, Hawes and colleagues (2003) found, for example, that nearly half (48%) of the ALFs

9 offered apartments, whereas slightly more than half (52%) offered rooms, most though not all with an attached bathroom or half bath. In terms of privacy of accommodations, there was even greater variability. A total of 31% of all ALFs had privacy for at least 80% of their accommodations, with 27% providing a setting in which all units were private. However, 40% of ALFs had semiprivate accommodations for more than one fifth of their units (in most, fewer than half of units were private). Surprisingly, an estimated 28% had one or more rooms shared by three or more persons (Hawes et al., 2000). One might expect such variations to affect residents satisfaction with their accommodations, as well as their ability to control their living space in a variety of ways, which are displayed in Table 3. What is less expected is that variations in privacy in one study were associated with statistically significant differences in admission and retention policies, with high-privacy ALFs having much more restrictive policies. ALFs that offered high privacy in their accommodations were less likely to admit or retain residents who needed help with locomotion, who were incontinent, who had behavioral symptoms, or who needed nursing care (Aud & Rantz, 2005; Hawes et al., 2000, 2003). As a result of this finding, policy makers who seek to provide privacy to consumers but want ALFs to serve a nursing-home-eligible population or to prevent or delay nursing home admission may want to assess the impact of privacy on the types of residents ALFs serve, the effect on aging in place, and whether policies can be developed and implemented that encourage both high privacy and greater flexibility on admission and discharge policies. At the same time, it is important to consider not only the appropriateness of different quality measures that is, validity and the conceptual relationship to the features being examined but also whether comparisons between very different models of AL are appropriate. Zimmerman and colleagues (2003) found differences in the environment and accommodations offered by RC/AL facilities in four states, classifying them as small (,16 beds), traditional, or new-model RC/AL facilities. In addition to finding differences in process quality related to care (e.g., services offered) and the level of resident impairment in physical and cognitive function, Zimmerman and colleagues found significant differences across these different types in terms of accommodations, privacy, resident control, and individual freedom. They also found that small facilities did poorly on most of the indicators the investigators used to compare the different types of facilities. However, the authors argued that an assessment of process quality that included a different set of indicators, such as the nature of staff/resident interactions or a measure of the degree to which a facility seemed like a real home, might produce different results (Zimmerman et al., 2003, p. 114). Table 3. Illustrative Indicators Comparing the Environmental Features of Assisted Living Facilities Descriptive Indicators Accommodation type (room or apartment) Private or shared unit Bathroom type: private/shared and full/half baths Facility size Indicators Policies or Characteristics of Interest Structural and process quality indicators Policies on admission and discharge Policies on use of supportive aids and assistive devices, such as walkers, wheelchairs Ability to control individual unit temperature Ability to lock door Ability to have a pet Ability to bring and arrange furniture Ability to store and prepare food Availability of storage space Availability of outdoor areas Availability of space for activities, gatherings Accessibility and attractiveness of outdoor space Cleanliness and maintenance of public and private spaces, indoors and outdoors Resident s use of public spaces How homelike private and public spaces are Resident-focused indicators Resident case mix in the assisted living facility (physical and cognitive function, behaviors, continence) Length of stay Discharge location Reasons for exit from assisted living facility Involvement in activities, social isolation Resident ratings of private space Resident ratings of public spaces Quality: Hearing the Consumers Voice Studies have suggested that consumers typically spend little time evaluating different LTC settings before they are faced with needing residential LTC and needing it pretty immediately (Castle, 2003). Often, even with AL, the decision is made under the pressure of an imminent discharge from an acute care setting (Fralich & Hawes, 2004). Interviews and focus groups with residents and family members indicated that factors such as location and price were central to the decision of most consumers (Fralich & Hawes, 2004). For example, 85% of a national probability sample reported that location had been a critical element in their decision, and 89% said price was an important consideration (data from Hawes et al., 2000, that were reanalyzed for this article). Although price and location were clearly significant for consumers, when asked to define quality, residents and family members were clear that, from their perspective, quality was a multidimensional concept (Fralich & Hawes, 2004; Greene et al., 1997/1998). For example, many consumers reported that the physical appearance of the facility and 48 The Gerontologist

10 accommodations were often central to their decisions, particularly when selecting a facility. Family members stressed the importance of loved ones having as homelike an environment as possible, focusing on privacy of accommodations and the attractiveness of public spaces of the facility, as well as the ability to furnish the unit with the resident s possessions. Interviews revealed similar preferences. For example, in a survey of a national probability sample of residents after they had left an ALF, 92% reported that when they selected and moved into an ALF, having a private room and bathroom was important to them; 87% cited the attractiveness and amenities of the indoor public spaces; and 76% the attractiveness of the outdoor areas (unreported data from Phillips et al., 2003). This is consistent with other studies that found that residents overwhelmingly prefer private accommodations in AL (Jenkens, 1997; Kane et al., 1998). Residents and family members also cited the importance of staffing type, staffing levels, and the availability of needed services. Ninety percent of residents said that the quality of the direct care staff (including staff knowledge, training, and staffing levels) were important to them when they entered an ALF, and nearly three quarters (73%) of the residents felt that the facility s having a registered nurse on staff was important. Similarly, residents mentioned the ability of the facility to provide monitoring and help with medications (91%) and to provide more or different services as their needs changed (73%) as important features of ALFs (Fralich & Hawes, 2004). In focus group interviews with current residents and with family members of ALF residents who had Alzheimer s disease, participants reported that issues of staffing and adequacy of services became more important over time (Fralich & Hawes, 2004; Greene et al., 1997/1998). Family members also discussed the importance of staff treating residents with courtesy, respect, and affection. In addition, family members of residents with dementia discussed the importance of communication communication with family members about the resident s health and well-being, and the nature and appropriateness of communication between staff and residents (Greene et al., 1997/1998). Nearly three quarters (73%) of residents interviewed asserted that the types of activities available was an important consideration in selecting an ALF, and most family members also reported that activities were a key aspect of quality in AL (Greene et al., 1997/1998; Hawes, Greene, Wood, & Woodsong, 1996). Other aspects of AL quality mentioned by consumers included the quality of the food and meal service, assistance with medications, and the availability of an attractive outdoor area (Curtis, Sales, Sullivan, Gray, & Hedrick, 2005; Greene et al., 1997/ 1998). Residents and family members also reported that their major concerns related to the ability of the ALF to meet their future needs (Curtis et al., 2005). Vol. 47, Special Issue III, More studies of consumer preferences and concepts of quality are needed, but the initial findings are important. Most of these have been discussed here as aspects of quality that might be examined in the context of comparing different models of AL and assessing their effects on potential public policy goals. However, in all of these endeavors that involve defining and measuring quality, it is important to keep consumer preferences and needs at the core of the attempts. This consumer focus, after all, is the essence of the philosophy of AL and the very feature that, if widely emulated, could transform all LTC. Conclusion The struggle to define, measure, and assure quality in LTC has something of a Sisyphean cast to it. Achieving this with respect to nursing home quality has been both a conceptual and a regulatory challenge for decades. And now, as one might feel that there has been substantial progress with nursing home care, at least at a conceptual level, the rock slips and one is then faced with the uphill challenge of considering quality s meaning in a new and different care modality AL. This is particularly true given the diversity in the AL industry and the variety of goals and expectations held for AL by different consumers, providers, and policy makers. For the industry and some researchers, such a commitment means that the formidable superstructure of regulation and monitoring that has come to characterize nursing homes must not be replicated in the world of AL. These analysts take what they see as the failures of quality assurance in nursing homes to mean that comprehensive regulation will not work in AL and that another path must be sought. In this view, creativity and responsiveness to consumer and market pressures are important engines of quality, and quality is likely to be harmed, if not killed, by the rigidity of regulatory requirements. Also, given the tender age of the AL industry, these observers argue that the interests of all are best served by allowing the industry to mature more fully before any consideration is given to regulatory standards that may stifle innovation. From this perspective, quality measurement and quality indicators are best aimed at meeting the internal requirements of the industry. Quality measures will be based on specialized assessment tools that vary from facility to facility or that focus solely on consumer satisfaction measures. The instruments and indicators will be adequate to the degree that they meet what providers consider their needs. Providers will use this information internally or as part of an advertising campaign asserting the special place occupied by a particular facility in this burgeoning industry. At the same time, many consumers residents and their families and regulators are increasingly