Recommendations: Research Issues

Older individuals and their families enter the health care system with their own unique set of problems and conditions. How well these characteristics are addressed significantly influences the individuals’ ultimate health outcomes. In the research world, these intrinsic and extrinsic characteristics can be thought of as independent variables. In the day-to-day world of older individuals and their families, they can be thought of as the circumstances in which people find themselves. A basic challenge of health outcomes research is discovering how to draw generalized conclusions about the relationship of causes (e.g., treatments, policy decisions) and effects or outcomes while recognizing that each person’s circumstances are unique and that some variation is to be expected. People have different preferences for different outcomes, and those differences need to be identified and, whenever possible, honored.

To support those efforts, the committee recommends that the following 10 areas receive concentrated attention over the next 10 years:

  • Health-Related Quality of Life

  • Satisfaction with Care

  • “Toolbox” of Outcomes Measures and a Core Set of Outcomes Measures

  • Practice Patterns and Interventions and Their Effect on Outcomes

  • Involvement in Care Decisionmaking

  • Education and Dissemination of Outcomes Information

  • Financing Systems

  • Service Delivery and Utilization

  • Transitions

  • Quality Assessment and Improvement



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade Recommendations: Research Issues Older individuals and their families enter the health care system with their own unique set of problems and conditions. How well these characteristics are addressed significantly influences the individuals’ ultimate health outcomes. In the research world, these intrinsic and extrinsic characteristics can be thought of as independent variables. In the day-to-day world of older individuals and their families, they can be thought of as the circumstances in which people find themselves. A basic challenge of health outcomes research is discovering how to draw generalized conclusions about the relationship of causes (e.g., treatments, policy decisions) and effects or outcomes while recognizing that each person’s circumstances are unique and that some variation is to be expected. People have different preferences for different outcomes, and those differences need to be identified and, whenever possible, honored. To support those efforts, the committee recommends that the following 10 areas receive concentrated attention over the next 10 years: Health-Related Quality of Life Satisfaction with Care “Toolbox” of Outcomes Measures and a Core Set of Outcomes Measures Practice Patterns and Interventions and Their Effect on Outcomes Involvement in Care Decisionmaking Education and Dissemination of Outcomes Information Financing Systems Service Delivery and Utilization Transitions Quality Assessment and Improvement

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade HEALTH-RELATED QUALITY OF LIFE Health-related quality of life should be a major focus of health outcomes research for older individuals. This research should continue to examine how to define and measure those dimensions of health-related quality of life that are particularly important to older individuals. In particular, more and better global and targeted functional measures should be developed for older individuals to describe and assess functional outcomes (a) in general, (b) for distinct clinical conditions, and (c) within specific settings. Rationale Health-related quality of life is perhaps the most important dimension of health outcomes for older individuals. Traditionally, the goal of health care has been to treat acute illness and delay or avert death. Increasingly, however, the goal of many older individuals and their families is not simply to add “years to life” but also to add “life to years.” Although the importance of assessing health-related quality of life for older individuals is increasingly being recognized, it remains a difficult concept to measure validly or reliably, especially among the older segments of these populations and for those with cognitive or communicative impairments. Several instruments have been developed to measure health-related quality of life, but more needs to be known about how accurately they take into account older peoples’ values, perceptions, and preferences and how contextual factors affect measurements. Functional health status is a significant dimension of health-related quality of life for older individuals. Maintenance of function or the slowing of deterioration in function may be the primary goal of treatment of many older people, especially those with multiple health conditions. Targeted measures are needed to compare the efficacy and effectiveness of new interventions and to assess the impact of particular diseases and impairments (e.g., measuring the ability of arthritis patients’ to walk without pain or the ability of patients with dental disease to chew food). The relative importance of certain aspects of functional status varies by setting; indeed, functional status is a key factor determining a person’s setting (e.g., where not having adequate cognitive functioning to live independently in the community requires residing in an institution). Background For the vast majority of older individuals, quality of life may have nothing to do with health care delivery. However, many quality-of-life issues do relate

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade to health care and many health outcomes are influenced by an individual’s overall quality of life. This committee has chosen to define health-related quality of life as “a personal perception characterizing the way an individual feels about his or her health status, including physical, psychological, religious, and social domains of health status.” This definition incorporates or overlaps with several other types of outcomes, including functional health status and satisfaction with care. It is also meant to capture the effect of a variety of other perceptions, including autonomy, vitality, vulnerability, well-being, and sense of isolation or marginalization. Contextual factors have an effect on health-related quality of life. These refer to an older individual’s environment and include a constellation of factors such as the availability of and access to appropriate health care, transportation, caregiving arrangements, culture, and social and community support. They also refer to the type of setting an individual is in, whether it be a hospital, nursing facility, assisted living facility, or single-family home. The relative importance of the various dimensions of health-related quality of life often differ depending on the context or setting. For example, among older persons in nursing homes, health-related quality of life may include the ability to make decisions, have control over their daily schedule, and maintain a sense of dignity and privacy. Although few people question the importance of quality of life, most researchers outside the field of health outcomes research still find this a rather abstract and subjective concept. Despite the increase in the use of the term “health-related quality of life” and a proliferation of measures that attempt to capture this domain, no theoretical framework fully explains why and how disease and disability influence quality of life. Most articles addressing quality of life give only briefer superficial definitions of the construct, if any at all. Finally, assessment of health-related quality of life needs to be based on the values and preferences of older individuals. Some widely used instruments, such as the Quality of Well-Being Scale, rate death as the worst functional outcome. Patrick and colleagues (1994) point out, however, that many patients believe there are some health states worse than death, such as being in a persistive vegetative state. In those cases, death—especially death with dignity—might actually be the individual’s preferred health status outcome. Some instruments, such as the Health Utility Index, do incorporate health states worse than death and are beginning to be more widely used. Functional Health Status Functional health status is an important dimension of health-related quality of life. It refers to an individual’s ability to perform certain types of activities.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade Assessing functional health status involves an evaluation of three areas: physical functioning, psychological functioning (including cognitive functioning, mental health, coping ability, and sense of autonomy), and social functioning. As individuals age, their likelihood of developing a pathology, that is, a specific disease, that interrupts or interferes with normal bodily processes or structures increases. Indeed, the older a person becomes the more likely he or she is to develop several such conditions (often referred to as comorbidities), particularly ones that are chronic or long lasting. Additionally, as an individual ages he or she is more likely to experience sensory and other impairments, such as poor vision or hearing. These diseases or impairments can lead to functional limitations that restrict an individual’s ability to perform an action or activity in the manner or within the range that is considered normal. The cumulative effect of this process results in disability: the inability or limitation in performing socially defined activities and roles expected of individuals within a social and physical environment (IOM, 1991). The following conceptualization depicts this process. Although it is not depicted here, the environment also plays a key role in determining the severity of this disabling process. Pathology→Impairment→Functional Limitation→Disability A great deal of medical research has examined how the first two components of this conceptualization—pathology and impairment—affect older individuals’ health outcomes. The committee believes this research is vital and ultimately should be conducted for all the diseases and impairments that commonly affect older individuals. The committee did not, however, attempt to prioritize which specific clinical conditions should be examined within the next 10 years. Such prioritization has been done before, however. Although not specifically focused on aging issues, the Institute of Medicine developed a model process for picking clinical conditions and technologies to be given priority research funding in a series of reports on effectiveness research (IOM, 1989, 1990a, b, d). Additionally, the federal Task Force on Aging Research (1995) recently released a comprehensive agenda for aging research, including a prioritized listing of research needed on specific diseases, disabilities, and mental disorders (albeit not necessarily with a focus on outcomes). In general, the committee supports the type of basic biomedical research called for in those reports, with the caveat that older individuals need to be included in related clinical trials. Older people are often deliberately excluded from clinical research. Even when older individuals are included in clinical trials, interpreting the findings can be difficult. Variability in outcomes may be

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade inadequately explained even after age differences within the older population are taken into account. Often as individuals age, however, understanding a particular disease or impairment becomes less important than understanding how it affects an individual’s ability to function, as portrayed in the last two components of the above conceptualization—functional limitation and disability. Naturally, the severity and limitation or burden imposed by these diseases and impairments varies by individual and, in part, determines the extent to which the diseases or impairment become a disability. It is also important to understand the relative contribution to the effect of each individual disease or impairment on function. The committee considers a focus on overall functional outcomes essential to the everyday work of practitioners and care providers when assessing the merits of medical interventions and quality of medical care for older individuals. Yet, important measurement issues exist. For example, most current measurement tools are used to assess a patient’s functioning at a single point in time; the focus needs to be redirected to how functioning changes over time. These concerns must be addressed before practitioners can begin widespread use of these measurement and assessment tools. Examples of Needed Research Construct a better theoretical model of the dimensions of health-related quality of life most important to older persons, include a consideration of how these dimensions might change in different settings. Develop a compendium of health-related quality-of-life measurement instruments specifically for older people (see also the recommendation on a “toolbox” of outcomes measures). Support more longitudinal research on older individuals’ health-related quality of life. Investigate the major predictors of functional decline or improvement and the relationship between functional limitation and disability. Conduct longitudinal research on function change over time. Develop models to better predict future functioning. Analyze how functional health outcomes are affected when an individual has multiple diseases and impairments. Examine how current and new measures of functional health status help differentiate health care plans and practitioners. Compile measures of functional health status and evaluate their appropriateness for use with older individuals.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade SATISFACTION WITH CARE Improvements should be made in the way older individuals’ satisfaction with care is measured and interpreted. Rationale Satisfaction with care is a significant dimension of health-related quality of life for older individuals. Satisfaction is increasingly being measured as a way to judge quality—and determine payment—for both individual practitioners and health plans. The relationship between satisfaction with care and other outcomes of interest is not well known. Despite progress, available measurement tools and analytic methods of interpreting satisfaction with care can lead to spurious interpretations and conclusions. Background Measuring Satisfaction Several instruments already exist to measure overall ratings of care (i.e., asking patients to rate their care as either “poor to excellent” or to say whether they are “very dissatisfied” or “very satisfied”). These instruments are widely used throughout the health care industry. However, a variety of methodological issues need to be addressed. These include issues of question phrasing, item placement, appropriate response choices, and whether the care being rated is the last visit versus all visits in the past year. The satisfaction scores from some instruments are often skewed toward the “very good” to “excellent” responses; this has been shown to be particularly true when older individuals are surveyed. Yet the real difference between rating care as “very good” rather than “excellent” might be underestimated. Getting a “very good” response may be an early warning sign of problems and provide an opportunity for further examination into ways to improve care. Moreover, lack of a clear understanding of these distinctions makes it difficult to establish a credible benchmark of satisfaction against which to measure improvement. Some cognitive research of survey respondents is already being undertaken to understand better the process by which patients respond to questions about the rating of medical care.1 Such research is important. It is equally important 1   For example, this type of research is being undertaken as part of a recently funded study of consumer assessments of health plans. The Agency for Health Care Policy and Research is the sponsoring agency, and the study is being conducted by a consortium of researchers from Harvard University, the RAND Corporation, and the Research Triangle Institute.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade that conceptual models of the cognitive process used in determining whether one is satisfied be expanded or further developed. To date, researchers have drawn on job satisfaction and marketing research models to develop similar models in health care. In terms of reports of specific care events, further refinement and use of patient surveys to capture actual health experiences, not just “satisfaction” with the experiences, is needed. The Picker-Commonwealth Foundation (Cleary et al., 1991) developed a 61-item survey about patients’ experience of hospital care that was based solely on patient reports of specific events (e.g., “Were you told of the purpose of your medications in a way that you could understand?”). This survey, which examines hospital care from the perspectives of seven domains (including respect for patients’ values, preferences, and expressed needs; coordination and integration of care; physical comfort; transitions; and continuity of care) was developed based on focus groups and expert panel review. Use of such reports minimizes the influence of confounding factors such as patients’ expectations, personal relationships, practitioners’ style, gratitude, or response tendencies related to gender, class, or ethnicity. The committee believes it is important to measure both patients’ overall ratings of satisfaction and their reports of care events to understand more accurately the factors that influence an individual’s sense of satisfaction. For example, a patient and his or her family could report quite factually that a physician discussed whether life-sustaining treatment should be utilized. Yet, their rating of satisfaction with that conversation may be quite different depending on the physician’s skill and sensitivity. Interpreting Satisfaction Interpreting the results of satisfaction surveys will involve gaining a keener understanding of patients’ characteristics and their expectations of care. For example, Greenfield and colleagues (1996) examined the patient satisfaction scores of a large sample of diabetic patients. Although the scores were uniformly rather high, further analysis showed that a significant amount of the variance between scores could be attributed to the passivity of the patients and the severity of their conditions; more passive and more severely ill people were less likely to express dissatisfaction. Likewise, there appear to be generational differences in people’s ratings of satisfaction. Today’s 70- and 80-year-old individuals are more likely than younger people to have been socialized to avoid dissent and criticism of practitioners and the health care system. An adjustment that accounts for salient patient characteristics and illness severity seems appropriate. Similarly, Desbiens et al. (forthcoming) noted that seriously ill patients will report both severe pain and high levels of satisfaction with their pain control. A better understanding of people’s expectations might help explain this seeming

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade incongruence. Do dying patients expect to be in pain? Would patients continue to rate their satisfaction so highly after seeing an educational video that informed them that, in the majority of situations, patients do not need to suffer severe pain while dying? Greater sophistication in measuring and interpreting satisfaction reports and ratings holds the promise of improving the way practitioners and health plans deliver care. Until this sophistication is achieved, however, reports and rates of older people’s satisfaction should be regarded cautiously. Examples of Needed Research Develop standard measures of satisfaction, specifically tested for use by older individuals, that use both reports and ratings. Satisfaction measures should include both global and disease- or setting-specific assessments. Establish the advantages and disadvantages inherent in the use of ratings and reports of satisfaction. Utilize focus groups and other qualitative methods to illuminate how patient expectations influence satisfaction. Determine how satisfaction rates might change with age and other characteristics such as cultural differences. Establish purposes and methods for using family reports and ratings. Explore how the perspectives of patients and family members differ in both reports and ratings of health care. Examine how current and new measures of satisfaction help differentiate health care plans and practitioners. Examine how mild to moderate cognitive impairment affects older people’s responses to satisfaction surveys. Examine older individuals’ patterns of enrollment and disenrollment in health plans and how these patterns relate to satisfaction with care.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade “TOOLBOX” OF OUTCOMES MEASURES AND A CORE SET OF OUTCOMES MEASURES A well-defined set or “toolbox” of outcomes measures for older individuals should be developed and further refined. The toolbox should allow easy and appropriate application of these measurement tools by nonresearchers in a variety of settings. A core set of outcomes measures should be developed for use by practitioners, health care plans, and organizations serving older individuals and populations. Methods of making outcomes data more useful clinically should be developed. Rationale A variety of health outcomes measures exist, but many may not have been tested specifically for older people, particularly for those with cognitive or communicative impairments. Use of a core set of outcomes measures can provide the basis for making clinically relevant care planning decisions, whether by hospital, home health agency, nursing home staff, primary care practitioner, patient, or family caregiver. These data can facilitate the transfer of information across organizational boundaries, creating a common language of assessment, as well as facilitate communication among clinical staff from different disciplines (e.g., physician, nurse, dentist, physical therapist) located within the same organization but in different settings. These data can be used to characterize and monitor the performance of individual practitioners, provider organizations, and integrated networks of health care organizations. Background Although significant progress has been made in objectively measuring and assessing a wide spectrum of health outcomes, more and better measurement tools and analysis are needed. Once these are developed, the committee recommends that they be assembled in a way that allows easy and appropriate application by nonresearchers in a variety of settings. Various methodologic questions must be addressed during the development of this toolbox. These include (a) the applicability of using existing outcomes measures developed for generally healthy, younger adults for an older population; (b) the creation of instruments that will assess these health outcome domains reliably, validly, and with minimum respondent burden, among persons throughout the older age span; and (c) the availability of alternative formats (e.g., questionnaires or instruments that are self-administered by paper-based instruments or by computer-assisted or touch-screen technologies, and those that

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade are interviewer-administered (which can be face to face or by telephone) and foreign language versions. Furthermore, less is known than might be desirable about how to meet special methodologic problems of assessing health outcomes among certain groups of older people. For individuals who are cognitively impaired, as an example, more research is needed on how best to assess their health status and who should provide the information if the individuals are unable to do so themselves. Other groups that pose a methodologic challenge include individuals in various types of residential facilities, individuals from ethnic minorities for whom English is not a first language, and very old individuals. Core Set of Outcomes Measures A great deal of data about patient outcomes is already being collected. For example, Medicare and Medicaid have claims data for almost all inpatient hospital stays. In 1991 the Minimum Data Set (MDS) and Resident Assessment Instrument (RAI) were mandated as the standard assessment system for all residents in all Medicare- and Medicaid-certified nursing facilities. A similar assessment system and instrument are being developed for assisted living or board and care facilities. Rehabilitation hospitals use the Functional Impact Measure (FIM) as their standard assessment instrument. Home health agencies are just beginning to implement the Outcome Assessment Instrument Set (OASIS) to track outcomes. Unfortunately, as people move back and forth across these settings, very little information has been shared. Many of these diverse instruments and measurement traditions have been developed independently of one another. Except for activities of daily living (ADLs), few data elements and response categories in the nursing home MDS, the rehabilitation FIM, or the home care OASIS have identical meaning or scoring. Even when examining ADLs, wording differences can, in some circumstances, cause incomparability. Differences in time frames and indicators of intensity or frequency in most measures also make combining data from different measurement tools difficult. For example, using different severity measurement systems can result in very different assessments of risk, even when used on only one individual (Iezzoni, 1994). A core set of measures can help to overcome these problems. The core set might contain measures that address outcomes as diverse as cost, satisfaction with care, and health-related quality of life. Certainly, more specialized or targeted information to supplement the core set of measures should and will continue to be collected. For example, in some cases investigators might want to add some measures designed specifically for a given diagnosis, such as the ability of older individuals with arthritis to walk without pain. Similarly, a hospice organization may want to ask its patients more questions about pain than would be needed for someone in an assisted living facility. In other cases, the inclusion

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade of some information, such as asking about a family’s ability to care for the patient at home for some permanent residents of nursing homes, may not be needed. Most importantly, the measures must reflect the characteristics, concerns, and preferences of older individuals. The measures should be able to be tracked over time. For example, patients with chronic obstructive pulmonary disease should be able to compare the functional health status of similar patients within and across different health plans. Additionally, appropriate guidelines need to be developed about when the responses of family members and other proxies should be used in establishing the core set of measurements. (See “Involvement in Care Decisionmaking” for examples of research needed regarding the use of proxies.) Movement toward developing and implementing a common set of indicators is already under way. Several organizations, including JCAHO, NCQA, and the Medical Outcomes Trust, have begun to collect key indicators and measurement instruments. The technical capacity for capturing data in a variety of different ways is improving; in addition, the costs for equipment and systems as diverse as text scanning, speech recognition, and bar code reading are decreasing. Software vendors and those designing integrated information systems are focusing on portability of the clinical and financial information across traditional institutional boundaries as networks of providers develop in selected markets. All these forces bode well for the emergence of a more uniform, if not standardized, clinical data collection mechanism within and across organizations. Examples of Needed Research Develop more and better outcome measurement tools specifically tested for older people. Once developed, assemble a compendium of these measurement tools. Examine ways to ensure that data gathered from multiple sources, including proxies, are accurate. Develop broad consensus on which outcome measures should form the core. Simultaneously develop better measurement tools and analytical methods. This includes valid and reliable standard definitions, criteria, mechanisms for assessing data collection for equivalency, and data collection methods. Investigate ways to expand and promote the clinical and organizational utility of these measures and tools. Train providers and practitioners to collect and use health outcomes data (e.g., involve doctors in development of the core set of measures). Then determine whether having more clinically useful data actually improves data quality, reliability, and validity. Explore ways to minimize the data collection burden.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade percent. In general, plans are at risk for losses if they exceed the total capitated payment amount for all their enrollees, but they can keep any “savings” or un-used amounts rather than return them to the government. The concern is that capitated payments create strong incentives to under-provide services or to enroll healthier individuals who need fewer services. Despite considerable research and analysis, more work remains to devise and test payment methods that do not reward health plans or practitioners for seeking the well (i.e., promoting favorable selection) and under-treating the sick (i.e., avoiding adverse selection). Another concern is that disincentives in the payment system will lead to the failure to provide certain types of care. For example, geriatric assessment requires in-depth (and hence, expensive) evaluation of an individual’s functional abilities, but the administration and interpretation of these evaluations are typically not reimbursable. Several alternative capitation methods have been suggested. Some of these alternatives have been tested (but not yet demonstrated to be satisfactory), and others are as yet untried. Risk Adjustment Even though they tend to have more health problems than younger people, older people vary considerably in their health status and in the economic risks they pose to capitated health plans. Risk adjustment seeks to adjust payments to plans to better reflect actual patient or participant characteristics (i.e., plans would get more money for enrolling participants with complex medical needs). How this is done and precisely which demographic and functional “adjusters” should be included in a risk adjustment model has been the subject of much debate. Part of the debate revolves around technical questions related to data availability, quality, and predictive power. Other questions focus on possible unwanted effects of recognizing, and thereby encouraging, certain factors (e.g., past utilization of services) in the payment scheme. Competitive Pricing HCFA is currently exploring the use of a competitive pricing system to set its Medicare reimbursement rates. In one planned demonstration project, all health plans in a certain region will be asked to submit bids for their estimated costs of providing a standard benefit package. Because few health plans have extensive experience with serving large numbers of very old and very frail members, concerns arise about the willingness of these plans to participate and their ability to set appropriate rates. Information is also needed on the optimum mix of high- and low-risk members, as well as the mix of private and government contracts. A health plan must accurately determine these mixes in order to be competitive in terms of benefits and services. Little is known about how

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade these issues will interact in a rate-setting context or how the quality of care, health of members, or even costs will be affected. Use of Outcomes Measures to Structure Incentives Another approach that remains relatively untested is incorporating outcomes measures into payment incentives. Those organizations and practitioners that achieve good outcomes would receive more money. The issue of how to handle providers who do not achieve good outcomes is more problematic; if their payments are reduced, it might be more difficult for them to make needed improvements. Nonetheless, risk adjustment models need further refinement before adoption and implementation of a payment system adjusted for both risk and outcomes could serve as a strong motivation for providers to achieve good outcomes. Examples of Needed Research Conduct randomized controlled trials and quasi-experiments to test the incentives and costs of different models of reimbursement and their effect on integration and coordination of services, cost-effectiveness, and health outcomes of older individuals and their families. Conduct policy research to determine the impact on health outcomes of adverse and favorable selection in health plans. If appropriate, develop methods to limit favorable selection or offset the consequences of adverse selection. Refine risk adjustment models further. Identify payment methods with disincentives for providing certain types of care (e.g., not appropriately reimbursing practitioners for conducting geriatric assessments) and examine the effect on patient outcomes and costs of care.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade SERVICE DELIVERY AND UTILIZATION As alternative forms of organizing and delivering health care for older individuals are encouraged and continue to be developed, the effect on health outcomes should receive continuing and rigorous evaluation. Rationale The health care system is undergoing substantial restructuring and reorganization. Older people and their health outcomes may be particularly affected by changes in where services are provided, by whom, how often, and in what mix and intensity. Background Many of the current Medicare and Medicaid reform proposals call for beneficiaries to enroll in managed care-type health plans. However, health plans have traditionally been reluctant to become involved with older individuals in part because of fears of unlimited liability for complex care and a dearth of people willing to join (Iversen et al., 1988). Nevertheless, recent evidence from the Medicare risk program indicates that such attitudes may be changing; the number of plans entering into such contracts increased from a low of 93 in 1991 to 171 by August 1995 (GAO, 1996). More fully integrated care delivery systems offer the potential of providing new and improved services, which could, in turn, improve health outcomes—especially for people with chronic conditions. Access to care, continuity of care, and the cost-effectiveness of care may be enhanced. The extent to which reality actually meets expectations, however, requires careful study. One issue is whether, if more care is provided and organized by an array of primary care clinicians, the benefits of coordination will exceed the possible harms to older, sicker people of being treated by those with less specialized training. Another issue is whether the emphasis on health promotion and disease prevention translates into better health status for both the more healthy and the less healthy segments of the older population. Concerns continue to be raised that some health plans are limiting or rationing care to the detriment of their members’ health. For example, plans sometimes restrict the list of prescription drugs that are covered by the plan (i.e., the plan’s drug formularies). A recent study reported that this can lead to decreased quality of life, higher use of health care services, and greater health care costs (Horn et al., 1996). Likewise, utilization review policies and procedures guiding decisions about what services are medically necessary may restrict access to necessary care.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade Choice of providers and access to practitioners also demand close scrutiny, particularly in underserved areas. In some rural areas, health plans enroll participants but then reject local practitioners for reasons such as lack of board certification. Such limitation of choice and potentially of access may have an impact on health outcomes. On a related note, the correlation between the usual measures of practitioner qualifications, such as board certification, and health outcomes has not been clearly established. It is also not known if this relationship, if it exists at all, varies by setting (e.g., hospital versus ambulatory care setting). The environment in which health care organizations operate can play an important role in the way services are provided. For example, early indications from one study suggest that the probability of hospitalization for nursing home patients is substantially higher in markets with more beds per 1,000 and lower hospital occupancy rates (Mor, 1996). Similarly, the use of nursing facilities, assisted living facilities, and home health services has been shown to vary enormously by region and appears to be linked to the number and availability of these services (Delfosse, 1995). Similar variation appears to be developing as some regions begin to use hospital-based “transitional care units” or nursing home-based “subacute acute care centers” to a greater extent. The role that geography plays in determining health outcomes also requires more study. Some research has examined how factors such as distance and topographical barriers and lack of transportation affect access to care in rural or underserved urban environments, but few studies have examined the effect such factors have on health outcomes. New telecommunication technologies offer the opportunity to increase access to care for residents in these areas, but, again, little research has been done to determine if this increased access translates into improved health outcomes. Finally, patterns of enrollment and disenrollment may highlight some of the effects of changes in the financing and service delivery systems on older individuals’ health outcomes. Some argue that greater choice in selecting health plans and practitioners will allow older individuals to “vote with their feet”; if they are dissatisfied they can simply choose another plan or practitioner. This assumption needs to be evaluated carefully by determining who disenrolls, what their medical and functional characteristics are, and where they go after disenrolling. Given Medicare’s relatively open (monthly) enrollment policies, the effect of frequent disenrollment on the health plan also needs to be monitored. Examples of Needed Research Determine under what circumstances underutilization and overutilization of services have positive or adverse health and cost consequences.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade Examine the factors that have gone into health plans’ decisions to serve older populations. Similarly, examine factors that will induce older people to join managed care plans. Evaluate whether more fully integrated care delivery systems actually provide new and improved services and, thereby, improve health outcomes—especially for people with chronic conditions. Determine how older individuals’ health outcomes are affected by (a) the prevalence and nature of grievance and appeals procedures in health plans, (b) the extent to which they are used by the elderly (as contrasted with younger patients, and for different subgroups of the elderly, such as those with serious chronic illness), and (c) the deliberate or inadvertent barriers to their use. Study whether the usual measures of practitioner qualifications (e.g., board certification) correlate with good health outcomes. Determine if this possible correlation varies by setting. Demonstrate whether the use of nonphysician health care providers, improved systems of transporting patients, and telecommunication technology can improve health outcomes of rural and underserved populations without significantly increasing the costs of care.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade TRANSITIONS The impact on health outcomes when older individuals make transitions between types of care (e.g., from active treatment to palliative care), treatment settings, and health plans should be explored. Rationale Research often has been directed at patients in specific settings (e.g., living in nursing facilities versus dwelling in the community) and in discrete health states (e.g., an acute episode of cardiovascular disease requiring hospitalization). Fragmentation in the health care system means older individuals are being served in multiple settings. Older adults often experience complex health patterns, undergoing multiple stages of deterioration and improvement in their health, even during a single episode of care. Background There is a misperception that as people age they move along a linear continuum of care as their status gradually and smoothly changes from that of a well, active, and independent senior living in the community to that of a completely dependent nursing home resident. In reality, very few people move along such a linear path. Often the decision to move an older individual to an institutional setting where more care can be provided is put off until the individual is very disabled and in need of the highest level of care. Likewise, the goals of treatment for disease sometimes do not change from curing a patient to caring for a terminally ill patient until the disease is quite progressed. Even when chronic conditions are present, older individuals often continue to have acute care episodes. Traditionally, health outcomes research has relied on measures of mortality to rank quality of care. The implicit assumption is that a low mortality rate equals the best possible care. For some patients, life-sustaining treatment (i.e., treatment that reduces mortality) merely prolongs dying and suffering. With an aging population now living longer than ever before, often with multiple acute and chronic problems, the health care system must become increasingly attentive to how transitions in the goals of medical care are managed. Fundamental to both geriatric and palliative care medicine is that practitioners must learn to negotiate skillfully a transition in goals of care, going from a major emphasis on extending life to an emphasis on patient comfort, even if focusing on the latter to the exclusion of the former means a patient’s life is shortened. Often, however, the timing of this transition is made too late and results in adverse outcomes (SUPPORT, 1995). At the same time, real risks exist that the timing of this transition, particularly for older individuals, will occur too soon.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade The current health care delivery system remains fragmented. Although improvements are being made, the interface between the settings of acute, subacute, home, ambulatory, and long-term care is not well developed. Important information about an older individual’s condition or preferences may not be communicated as the individual moves between various settings. The goals of treatment and the instruments available to measure an individual’s status vary for each of these different settings. Better ways of tracking patients across settings and time are needed. Finally, improved coordination of care among a variety of practitioners and care providers is also needed. The setting of care often plays a crucial role in an individual’s health outcomes; yet the effect of making a transition from one setting to another is often not well known. For example, suppose, as is often the case, that going to a nursing home even for a relatively short stay is perceived very negatively by a hospitalized patient. Perhaps the patient’s long-term results would be better if he or she stayed in the hospital a little longer to receive rehabilitation services at a single site rather than be moved to a nursing home, and thus fragmenting care. In addition to the transitions an older individual makes along the continuum of care, changes are also occurring among providers and practitioners. These changes raise questions about whether they may disrupt patterns of care and, ultimately, affect health outcomes of older people. Examples of such changes are many: physicians retire or become employees, begin to use nurse practitioners and case managers, or become part of an interdisciplinary team; physicians sometimes do not follow their patients’ care into the nursing home; hospitals or clinics close; use of telephone triage may be implemented (which could have a particularly adverse effect on older individuals, who might have greater difficulty hearing or quickly comprehending verbal information); managed care contracts change hands; health systems, plans, and organizations merge or are acquired; senior management of health plans or organizations change. All of these changes have the potential of affecting the health outcomes of older people, both as individuals and as a population. Examples of Needed Research Support experiments and innovations in continuity of care and coordination of care. Support longitudinal research exploring the impact of transitions on health outcomes. Examine patient utilities for the timing and choice of transitions between care settings. Develop valid and reliable markers of health status to identify patients who may need to discuss a change in treatment goals.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade QUALITY ASSESSMENT AND IMPROVEMENT The performance of government regulatory agencies, private-sector accreditation organizations, and organizations’ internal programs in using outcomes-based quality assessment and improvement systems should be evaluated for effectiveness in improving health outcomes for older populations. Rationale Governments, private-sector accreditation organizations, and health care delivery organizations are increasingly using outcomes measures as a basis for assessing and improving the quality of care. Little is known about how effective these approaches are in terms of actually improving health outcomes. Depending on how they are structured and implemented, external and internal strategies for quality monitoring might contribute positively or negatively to improving desired health outcomes. Background Quality can be assessed and improved by both external and internal programs. Traditional quality assurance programs were, in general, imposed by outside agencies and entities such as the federal government or private-sector accreditation organizations. They attempted to define at least minimum standards for quality and to enforce them through a variety of approaches. These approaches classically involved evaluation and audit functions; professional practices such as credentialing, peer review, and continuing education; and various kinds of licensure, certification, and accreditation. The primary focus of these external and internal quality programs was often on using structural and process measures to identify problems, determine their causes, and discipline poor individual performance. This approach to quality assurance has, in recent years, been transformed into a quality improvement approach that focuses on how organizations can improve systems, processes, and outcomes in a continuous way. Programs using this approach are often described as “customer-” or “patient-centered,” and many apply the principles of “total quality management” and “continuous quality improvement.” These principles shift the focus on quality efforts from the identification of deficient individuals to the identification of opportunities to improve systems and processes in a way that will prevent quality problems and improve outcomes. Other principles emphasize scientifically and statistically based planning and assessment procedures, including the measurement of health outcomes and patient (i.e., customer) satisfaction; standardization of care processes; and feedback to practitioners of information on how their practices and

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade results may differ from those of their peers or from evidence-based standards for practice. In recent years, both external and internal quality programs have started to move beyond basic structure and process indicators of quality to identify whether desired outcomes are being achieved (IOM, 1990d). The Health Care Financing Administration, the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO), several peer review organizations (PROs), the National Committee for Quality Assurance (NCQA), and the Foundation for Accountability (FACCT) have all begun major initiatives to identify, refine, collect, and use outcomes measures as a way to assess and improve the quality of health care. JCAHO is doing so for use in the accreditation of a variety of health care facilities and entities such as hospitals, long-term-care facilities, home health agencies, ambulatory care settings, behavioral health care organizations, health plans, and integrated delivery systems. NCQA and FACCT are identifying and developing measures and instruments to be used in evaluating health plans. The federal government’s efforts have focused not only on services for Medicare and Medicaid but also on health care provided by the Departments of Defense and Veterans Affairs to military personnel and veterans. Much of the private-sector work has been stimulated by major employers’ initiatives to reduce their health care spending, to determine what value is being received for this spending, and to manage more closely the health services their employees receive. Although it is the committee’s overall sense that current programs of quality management are moving in the right direction, improvements can continue to be made and different models tested. Researchers need to examine how the outcomes-oriented programs are actually being implemented and what the barriers to their successful implementation may be. For example, using certain kinds of health outcomes measures—or failing to adjust outcomes for severity of illness and comorbidities—may discourage health care organizations from serving the sickest patients. Examples of Needed Research Evaluate the effects of various approaches to external and internal quality assessment programs on health outcomes for older individuals. Determine if outcomes measures can effectively be used to supplement the current standards-based on-site surveys for accreditation and regulatory purposes. Evaluate the effect on outcomes of graduated approaches of quality assessment in which organizations that have bad outcomes receive more site visits and more technical assistance and those that achieve good outcomes are relieved of some reporting or survey requirements. Study approaches to quality assessment and management to determine how they take into account the relative values that older individuals place on

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade quality of care, avoidance of harm, costs of care, and other dimensions of health outcomes. Examine how different kinds of employers use outcomes measures to make decisions about which health plans are available to employees and covered retirees. REFERENCES Batalden, P.B., E.C.Nelson, and J.S.Roberts. Linking Outcomes Measurement to Continual Improvement: The Serial “V” Way of Thinking About Improving Clinical Care. Journal on Quality Improvement 20(4):167–180, 1994. Chassin, M.R., E.L.Hannan, and B.A.DeBuono. Benefits and Hazards of Reporting Medical Outcomes Publicly. New England Journal of Medicine 334:394–398, 1996. Cleary, P.D., S.Edgman-Levitan, M.Roberts, et al. Patients Evaluate Their Hospital Care: A National Survey. Health Affairs Winter:254–267, 1991. Delfosse, R. Urban and Rural Classification of the National Health Provider Inventory Providers: United States, 1991. Advance Data. No. 266. Hyattsville, Md.: National Center for Health Statistics, 1995. Desbiens, N.A., A.W.Wu, S.K.Broste, et al. Pain and Satisfaction with Pain Control in Seriously Ill Hospitalized Adults—Findings from SUPPORT Critical Care Medicine, forthcoming. GAO (General Accounting Office). Medicare HMOs: Rapid Enrollment Growth Concentrated in Selected States. GAO/HEHS-96–20. Washington, D.C.: GAO, 1996. Greco P.J., and J.M.Eisenberg. Changing Physicians’ Practices. New England Journal of Medicine 329:1271–1274, 1993. Greenfield, S., L.M.Sullivan, K.A.Dukes, T.J.Tripp, and S.H.Kaplan. Unadjusted Patient Satisfaction Scores Used in Physician Profiling Result in Spurious Ranking. Boston: New England Medical Center, 1996. Horn, S.D., P.D.Sharkey, and J.Gassaway. Intended and Unintended Consequences of HMO Cost-Containment Strategies: Results from the Managed Care Outcomes Project. The American Journal of Managed Care 2(3):237–247, 1996. Iezzoni, L.I., ed. Risk Adjustment for Measuring Health Care Outcomes. Ann Arbor, Mich.: Health Administration Press, 1994. IOM (Institute of Medicine). Effectiveness Initiative: Setting Priorities for Clinical Conditons. Washington, D.C.: National Academy Press, 1989. IOM. Acute Myocardial Infarction: Setting Priorities for Effectiveness Research. P.H.Mattingly and K.N.Lohr, eds. Washington, D.C.: National Academy Press, 1990a. IOM. Breast Cancer: Setting Priorities for Effectiveness Research. K.N.Lohr, ed. Washington, D.C.: National Academy Press, 1990b. IOM. Clinical Practice Guidelines: Directions for a New Program. M.J.Field and K.N.Lohr, eds. Washington, D.C.: National Academy Press, 1990c. IOM. Hip Fracture: Setting Priorities for Effectiveness Research. K.A.Heithoff and K.N.Lohr, eds. Washington, D.C.: National Academy Press, 1990d. IOM. Medicare: A Strategy for Quality Assurance. K.Lohr, ed. Washington, D.C.: National Academy Press, 1990e.

OCR for page 15
Health Outcomes for Older People: Questions for the Coming Decade IOM. Disability in America: Toward a National Agenda for Prevention. A.M.Pope and A.R.Tarlov, eds. Washington, D.C.: National Academy Press, 1991. IOM. Summary of Committee Views and Workshop Examining the Feasibility of an Institute of Medicine Study of Dying, Decisionmaking, and Appropriate Care. Washington, D.C.: IOM 1994. Iversen, L.H., C.N.Oberg, and C.L.Polich. The Availability of Long-Term Care Services for Medicare Beneficiaries in Health Maintenance Organizations. Medical Care 26:918–925, 1988. Kaplan, S.H., K.A.Dukes, L.M.Sullivan, T.J.Tripp, and S.Greenfield. Is Passivity A Risk Factor for Poor Health Outcomes? Boston: New England Medical Center, 1996. Mor, V. Using Global and Targeted Outcomes Measures in Applied Settings. Presentation at the Institute of Medicine workshop “Evaluating Health Outcomes for Elderly People in a Changing Health Care Marketplace,” Washington, D.C., June 20, 1996. Patrick, D.L., H.E.Starks, K.C.Cain, R.F.Uhlmann, and R.A.Perlman. Measuring Preferences for Health States Worse Than Death. Medical Decision Making 14(1):9–18, 1994. Schneider, E.C., and A.E.Epstein. Influence of Cardiac-Surgery Performance Reports on Referral Practices and Access to Care. New England Journal of Medicine 335:251–256, 1996. SUPPORT (The SUPPORT Principal Investigators for the SUPPORT Project). A Controlled Trial to Improve Care for Seriously Ill Hospitalized Patients: The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments. Journal of the American Medical Association 274:1591–1598, 1995. Task Force on Aging Research. The Threshold of Discovery: Future Directions for Research on Aging. Administrative Document. Report of the Task Force on Aging Research. Washington, D.C.: Department of Health and Human Services, April 1995.