Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Measuring Quality of Care at the End of Life

Measuring Quality of Care at the End of Life Caring for patients at the end of life presents a series of quality-of-care problems to the health care system. In the past, concern has focused on overaggressive treatment of dying patients. Given rapid changes in the financing and delivery of care, it is time to focus on a range of quality problems and address ways to improve care and achieve outcomes desired by patients and their families. We provide a framework for conceptualizing such a task. This article addresses the purposes of measurement, definition of the patient population, timing of measurement, use of surrogates in measurement, scope of services to be evaluated, and the choice of measures. It emphasizes the necessary links between quality measurement and quality improvement.Most Americans have come to expect long and generally healthy lives. They also have come to understand that the consequence of longer life is that most people today die in old age, often of chronic and progressive diseases. Heart disease, cancer, stroke, and chronic obstructive pulmonary disease are the 4 leading causes of death among older persons. The incidence of disease and disability and the use of health care resources rise with age. Almost three quarters of the 2.3 million Americans who die each year are 65 years old or older, and about 5% of the Medicare population will die in a year.Estimates for the year 2030 show a population of more than 70 million people older than 65 years of age, 20% of the total population, compared with 13% in 1994.Dying slowly—at any age—presents a series of quality-of-care problems to the health care system, which has for the most part been designed to respond to acute illness. Throughout the course of chronic progressive illnesses, patients are likely to be examined and treated by a series of practitioners in different care settings—acute and long-term—for multiple problems that may ebb and flow in importance and in their effect on function and health-related quality of life. These circumstances tend to attenuate any one clinician's responsibility for the overall quality of a patient's care.Why should the very end of life be a subject of particular interest for the quality of health care?First, dying patients are particularly vulnerable, and the care provided to the most vulnerable in our society is a measure of the health care system itself. Second, quality-of-care problems have been documented for dying patients, including shortcomings in communication, highly aggressive treatment, and inadequate relief of pain and other symptoms.Third, in a time of rapidly changing health care financing and provision, the elderly and very ill may be affected more than others who have greater physiological, psychological, social, or financial resources.In the past, much concern has focused on overaggressive treatment of dying patients, in particular the use of life-sustaining technologies when the marginal benefits are insufficient to warrant their use and when patients themselves may not desire such care. Researchers have also documented considerable undertreatment of pain among people with advanced illness. A cost-driven shift to managed care (a simple label for a diverse set of financial and administrative practices) is further raising anxiety about underuse of service in some plans.Given these concerns, policymakers, managers of programs serving the very ill, clinicians, and others committed to improving the care of dying patients need methods to measure and evaluate the quality of care provided, thereby identifying poor care, assessing how changes in systems of care affect patients, and devising ways to correct problems. This article considers strategies for assessing and improving care at the end of life. The starting point is an examination of the concepts of quality and accountability as they relate to the care of people approaching death.CONCEPTSThe last 3 decades have seen continued advances in the conceptualization of quality of care and health outcomes, measurement tools, and strategies for improving health care and its results.More recently, those concerned about the quality of care for dying patients have begun to focus attention on how these concepts, tools, and strategies can be applied and adapted to measure, monitor, and improve care for those approaching death.QUALITY OF CAREIn 1990, an Institute of Medicine (Washington, DC) report, Medicare: A Strategy for Quality Assurance,defined quality of care as "the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge." This definition covered individuals and populations, which is consistent with increasing concerns about how different clinical interventions, programs, and resources can be deployed to the greatest social and individual advantage. Because the evidence base about what works in health care is still modest for many medical conditions, the definition acknowledged the relevance of professional knowledge, which includes experience and judgment, as well as the results of biomedical and clinical research. The definition deliberately omits resource constraints on the ground that judgments of what constitutes excellent, acceptable, or unacceptable quality should be independent of constraints on resources. Decision makers cannot, however, ignore resources in selecting the level of quality that is desired and affordable for a particular institution, population, or community. This means paying attention to quality and resources for patients covered by public and private insurance arrangements and for those excluded from such arrangements.In the context of approaching death as in other areas of health care, quality problems can be categorized in 3 broad areas: (1) overuse, (2) underuse, and (3) poor practitioner skills or performance. Examples of overuse are overuse of acute care settings, such as hospitals and intensive care; of diagnostic technologies when intervention is not warranted; of medical or surgical interventions, such as radiation, chemotherapy, and surgery; of life-prolonging or life-sustaining technologies, such as blood products, hemodialysis, total parenteral nutrition, mechanical ventilation, and tube feedings; and oversedation with pain medication. Examples of underuse are underdiagnosis and care of treatable illnesses, such as a bacterial infection; underuse of technologies and services that could improve the quality of life for patients, such as self-managed pain medication; undercontrol of symptoms, such as pain, nausea, anxiety, or depression; inadequate provision of help to family members, such as bereavement services; and the lack of involvement of patient and family in decision making. Examples of poor practitioner skills or performance that indicate inadequate technical or interpersonal quality are incorrect or inappropriately timed medicines, skin breakdown, nosocomial infection, avoidance of the patient and the family, unwillingness to answer questions, and failure to recognize and relieve symptoms, failure to provide help to a bereaved spouse, including an opportunity for the spouse to ask questions and remain in touch with the caregivers. Considered together, these categories help focus attention on whether what is done (or not done) is the proper thing to do and whether it is done properly.Death—the ultimate outcome of life—has limited value for measuring the quality of care experienced by those who are dying. More relevant measures focus on symptom management, patient and family perceptions and satisfaction, spiritual well-being, and patients' quality of life, each of which can be affected by the process of care. Health-related quality of life emphasizes health as perceived and valued by people for themselves or sometimes for others.ACCOUNTABILITYThe field of palliative care defines itself as particularly, but not exclusively, accountable for the care for those approaching death. Palliative care has been described by the World Health Organization as ". . . the active total care of patients whose disease is not responsive to curative treatment"(p11) and for whom the "control of pain, and other symptoms, and of psychological, social, and spiritual needs is paramount. The goal of palliative care is achievement of the best quality of life for patients and their families."(p11) For those who die suddenly, for example, of trauma, palliative care may apply primarily in the form of bereavement support for survivors.Some have conceptualized the shift from acute to palliative care as an abrupt transition, but this does not (or should not) reflect reality.Some aspects of palliative care, including careful management of symptoms and advance care planning, can be applied in tandem with curative, life-prolonging efforts. Just as no bright line separates palliative care from other forms of care, no bright line shields any clinician responsible for patient care from being held accountable for having the knowledge and skills to care well for the dying.Although much palliative care is rendered by family and social networks, health care professionals and organizations can provide many essential medical and social services, especially the relief of pain and other distressing symptoms. They are accountable for meeting the professional standards, skillfully using a defined base of knowledge, and performing in ways "consistent with current professional knowledge" (as emphasized in the definition of quality of care by the Institute of Medicine).At the heart of the medical enterprise is an ongoing relationship between a patient and clinician. This means that even when care is provided by a team of health care professionals, patients and their families need to know who these persons are, who is "in charge" of their care, and who can answer their questions and assist with decision making. Issues of communication among caregivers, patients, and family are complex and beyond the scope of this article; nevertheless, they are appropriate subjects for quality measurement and improvement.How broadly should accountability for quality of care for dying patients be conceptualized? The concept of palliative care extends accountability to include the patient and those close to the patient. Thus, the family experience, their participation in caring for a loved one, and their emotional and financial burdens also become a responsibility. This suggests that measurement of the quality of care likewise should extend beyond patients to include those close to them. Hospice programs consider the patient and family as the unit of care, as well as a part of the care team, but other institutions may not be so explicit about their responsibilities.Even palliative care specialists may not know whether they are fulfilling their responsibilities in the absence of systems that can measure outcomes and identify problems across care settings. When care for a dying person involves many different professionals providing care in different settings with varying institutional affiliations, no one person may be responsible for that patient's care throughout the course of a progressive, eventually fatal illness. Although integrated health care systems and managed care organizations offer the promise of accountability across settings of care, health care systems—including performance measurement programs—designed primarily for younger, healthier populations may not meet the needs of those with advanced chronic illness.MEASURING QUALITY OF CARENumerous issues quickly arise in the construction of a framework for measuring quality. These include the following: (1) the purpose of quality measurement, (2) the patient population, (3) the timing of measurement, (4) the source of patient information, and (5) the settings in which care is to be measured. The selection of quality measures depends in part on how these issues are resolved.PURPOSE OF MEASUREMENTAny attempt to implement quality of care measurement must first establish its purpose. Tools and strategies for measuring, monitoring, and improving the quality of care for people who are dying will, ideally, serve multiple purposes, including the following:Educating clinicians, policymakers, and the public about what is to be sought in the care for dying patients and families and helping them assess particular organizations and instances of care;Assisting those directly responsible for end-of-life care in continuously evaluating and improving what they are doing;Helping to identify deficiencies in systems of care for dying patients and the means of correcting them; andSupporting systematic research on the effects of different clinical, organizational, and financing options for end-of-life care.More specifically, 4 principal uses of quality measurement should be considered. These are external "inspection," internal improvement in the care process, selective contracting with providers, and informed patient and purchaser choice of plans, facilities, and practitioners. External inspection determines the suitability of a setting for patients. It stresses common definitions and measures that can be translated from 1 organization to another and can be understood by all. It generally involves static measurements that set a basic floor for performance, for example, for accreditation or reimbursement eligibility. (Performance in this sense may refer to the creation of organizational structures and procedures, to the achievement of defined outcomes, or both.)Quality assessment for internal improvement is a more dynamic idea based on the notion that care can always be improved and that such improvement ought to be continuous and based on monitoring performance and seeking opportunities to improve. It means that information must be detailed and timely enough to permit translation of data into information about how to change the care that is provided. For example, a nursing home might undertake a study of the transfers of its very ill patients to hospitals to determine whether such transfers seemed consistent with patient wishes and, indeed, whether patient preferences were documented and available to guide transfers and care after the transfers.Quality assessment for selective contracting—a purpose that reflects the current stress on market forces and competition—is sometimes described as "value purchasing." Quality assessment for informed choice is a prerequisite of a well-functioning health care market. It means that measures and the results of measurement must be important to and understandable to patients and their families and responsive to their values and preferences.Policymakers and researchers are just beginning to understand how the expectations and preferences of patients are shaped over time by their experiences, rather than being fixed attributes. Preferences (including the most fundamental—the use of life-sustaining interventions) may, for example, be affected by the quality of the caregiving. The ongoing interactions among the stated preferences, expectations, and experiences of patients means that care received may help to shape patient preferences as much as the reverse is true.Information particularly relevant to dying patients, for example, the availability of palliative care specialists, is not, however, featured in current consumer information. Similarly, although the availability of tertiary care services is typically considered in the creation of health plan networks, the services available to dying patients do not seem to figure in current contracting considerations. In fact, in a competitive health care system, a plan could be disadvantaged if it were known to provide particularly high quality care options for very ill and dying patients. Given current methods of capitating health plans, adverse selection that is disproportionately attracting sicker patients poses a serious financial risk to health plans.The purpose of measurement should be very clear because it will affect the scope, depth, and type of information collected. If intended for internal improvement, information must be quite detailed, but it does not require the attention to comparability and the level of confidence in small differences that are needed for fair comparisons between sites of care. Measurement used for internal improvement also must focus on outcomes and processes that the plan or facility could affect.PATIENT POPULATIONMany of the quality-of-care issues apply across age groups and are relevant, for example, for very premature infants, neonates with severe birth defects, people of all ages with malignant carcinomas, young adults with the acquired immunodeficiency syndrome, and middle-aged adults with degenerative diseases. Nevertheless, most of those who die in the United States of chronic terminal conditions are elderly, and that will influence some elements of quality measurement and improvement strategies in many situations.This article focuses primarily on older people with progressive, eventually fatal diseases. Many of these people, however, are not perceived by themselves or others as dying, and predicting death is an imprecise exercise. This creates difficulties because measuring quality requires identifying the population of interest—in epidemiologic terms, the denominator. It also creates difficulties for scheduling measures of patient status.TIMING OF MEASUREMENTUnlike other events in health care, such as follow-up after surgical procedures or scheduled visits for monitoring or treating chronic conditions, events near the end of life generally cannot be predicted precisely.To measure care while the patient is still alive, decisions must be made about whether measurement should be tied to specific events (eg, on admission, transfer, or when significant interventions are instituted or withdrawn) or should occur periodically (eg, at specific intervals), and if so, with what frequency. Unless data are gathered frequently (eg, daily), the unpredictability of deaths means that some patients may be examined 3 days before death, others 3 weeks before death, and others not at all. Such variation could make comparisons difficult.Identifying terminally ill patients before they have died requires some sort of marker event, such as passing a clinical threshold of severity or transition to palliative care (such as transfer to hospice care or withdrawal of life support). In some cases, it may not be clear that a patient is dying, for instance, during medical rescue efforts for a person with congestive heart failure who has survived past crises. Without a better way of recognizing when people are approaching death and when it is appropriate to focus more on comfort than on cure or life extension, organizations may miss important opportunities to improve care.Dying patients also can be identified retrospectively through death records. Retrospective review of care will be adequate if administrative claims or medical records are accurate, appropriately detailed, and available from all sites in which dying patients receive care. These sources are not likely to be sufficient for other purposes, however; for example, understanding patient or family perceptions and experiences.Other information may need to be gathered retrospectively from interviews with family and caregivers, which raises different issues about the role of surrogates.THE ROLE OF SURROGATES AS A SOURCE OF PATIENT INFORMATIONBefore death, many seriously ill and dying patients may be unable to report for themselves because of the physical or mental impairments produced by the illness. Also, as just noted, the uncertain timing of death makes direct collection of information from all patients improbable. The question thus arises whether a proxy or surrogate, someone such as a family member or a formal or informal caregiver, can provide reliable and valid information on the patient's experience. The evidence on this point is limited.Continued research is urgently needed to assess the degree to which patient and surrogate perceptions match (when both can be measured), the extent to which different surrogates provide different reports, the reasons for differences, and the bases for choosing a proxy informant. Such research is neither simple nor inexpensive because it requires not 1 but 2 data points for each variable and because a large number of patients may have to be included to allow for those who are unable at some point to communicate because of their illness or the treatment for it. Nevertheless, a better sense of the reliability and validity of surrogate data is a high priority for patients approaching death.CARE SETTINGSQuality measures should apply to all settings where patients die or receive care before they die. This includes home or other residential settings (eg, board and care facilities), hospitals (eg, intensive care or palliative care units), and nursing homes (including those with hospice units). If differences exist among settings in the quality of life for patients (as well as in cost), this is particularly useful information for patients and payers. It would also be useful to know about features of a specific setting (such as a hospice) that contribute to those differences.The kinds of interventions, staffing (ie, types of personnel and their level of training), and available technologies and labels for care (eg, hospice, palliative care, or medical care) may differ greatly, but the objectives of quality measurement should include comparing outcomes across settings and linking differences to variations in care or the environment of care (considering other factors, such as difference in patient or family characteristics). In developing such information, it is conventional to distinguish 3 broad kinds of quality measures relating to structures of care, processes of care, and outcomes of care.Table 1gives examples of structural aspects of care that might be the subject of measurement.Examples of Structural Aspects of Care That May Affect Patient CareSee table graphicThe extent and nature of links among specific health care settings, policies, and other structural factors (eg, nurse staffing ratios) and outcomes are matters of continuing dispute based on modest and sometimes conflicting research.Efforts continue to identify and establish structures and processes that can be linked to outcomes, in part to help improve care and in part to determine whether easier-to-collect structure and process data can be used as proxy measures of quality. As noted, linking care to outcomes requires that other factors that may affect outcomes—for example, how ill a patient is or how care is financed—be measured or at least defined.Two studies of hospice care attempted to make such comparisons,and they illustrate issues in quality measurement. In the first, the National Hospice Study,although hospice patients and their family caregivers were more satisfied than control patients, survival time, pain, symptom relief, or alleviation of psychological distress did not differ between hospice and nonhospice settings. Patients with cancer in a home or hospital-based hospice were compared with a control group of patients with cancer receiving conventional care. Patient outcomes (based on the report of the primary caregiver) included overall quality of life, social quality of life, pain and other symptoms, and satisfaction. Hospice patients experienced less use of aggressive or intervention therapy and diagnostic testing, but the quality of life of the patients in both settings was similar, except for pain control, which may have been better in the hospice setting. Again, hospice patients were more likely to die at home, and families were satisfied with that outcome, but otherwise no consistent differences were found.More recent studieshave questioned whether the instruments used in the National Hospice study were sufficiently sensitive to differences in care during the last 3 days before death. This suggests the importance of the timing of measurement in relation to the time of death and the need for measures that are sensitive to differences in care.TRANSITIONS ACROSS SETTINGSTransitions from 1 setting to another provide particular hazards for patients because physical, financial, informational, and other links must be planned and implemented successfully. Transitions may not be timely because they are too early or too late from the perspective of the condition of the patient; they may also be untimely if the sending or receiving organization is unprepared, as may happen when an imminently dying patient is transferred from the hospital to hospice care in a matter of a few hours. People and equipment, for example, a hospital bed for a home, may not be so quickly arranged. Salient information may not be transferred. Thus, it is important not only to measure the quality of care in the different settings, but also to consider continuity and coordination across care settings as they affect the patient's experience of care, quality of life, and other outcomes and the availability of important medical and patient preference information.MEASURES AND CRITERIAThose seeking to identify or improve tools for measuring the quality of care at the end of life have an array of potential instruments from which to choose and problems in assessing the usefulness and appropriateness of the measures. Few measures, however, have been developed with explicit attention to the concerns of dying patients and those caring for them.CATEGORIES OF OUTCOMESThe understanding of what it means to live well while dying and how to measure the quality of dying remains at an early stage. The elements of physical and mental functioning and well-being important to healthy people, people expected to recover from acute illnesses, or people expected to live for an extended period with various chronic illnesses may change in meaning for people with advanced, fatal illnesses.SurvivalIn most circumstances, survival is a good in itself—survival, for example, from surgery, after premature birth, and after organ failure have been used as markers for quality. For patients near the end of life, however, survival itself is not always viewed as the most important goal.Nevertheless, survival time, if adequately adjusted for patient preferences and for severity of illness, can be an important first screen for possible quality problems. For instance, survival time might differ for different hospices for a variety of reasons (eg, stage of illness at admission compared with success in relieving symptoms such as vomiting or dysphagia), only some of which might raise concerns about quality.Quality of LifeQuality of life is "a uniquely personal perception, denoting the way that individual patients feel about their health status and/or nonmedical aspects of their lives. . . ."(p619) Most people wish for a long and satisfying life, that is, people value how long they live andhow well they live.Health-related quality-of-life outcomes include physical, mental, social, and role functioning and sense of well-being; freedom from pain; satisfaction with health care; and an overall sense of general health. The meaning of symptoms and how they affect the quality of life of patients are intrinsically subjective, however. Some may find pain or incontinence very distressing; others may be less affected. Some may find pain less distressing than the mental haziness induced by medications. Measures of symptoms and satisfaction with care raise a variety of problems,and spiritual constructs may be particularly troublesome to define and translate into operational measures (I. Byock, MD, and M. Merriman, PhD, testimony before the Institute of Medicine Committee on Care at the End of Life, April 1996). A large literature has been developed that addresses quality-of-life measurement generically and, to some extent, on a condition-specific basis. Although self-reported quality of life during terminal care cannot be measured retrospectively (because the patient has died), with further research and refinement of measures, using proxy responses from a family member, friend, or caregiver or targeting patients ahead of death may be possible.Satisfaction With CarePatient and family satisfaction with the process and outcomes of care are important aspects of quality. Measurement of satisfaction can be made during palliative care and after death by interviews with family members. One potential difficulty with interviewing patients (and possibly family members) is that of respondent bias, which is likely if people fear that expressions of dissatisfaction with care and caregivers might make caregivers angry or neglectful. Conversely, expressions of dissatisfaction from patients or family may be expected by caregivers and thus discounted. Another problem for those truly interested in improving care is that people may tend to describe themselves as satisfied even when probing would reveal various quality concerns.In an effort to respond to some of the problems with satisfaction measures, Gerteis and colleaguesadvise asking patients and proxies not about their satisfaction with care, but about their experiences. At this time, it seems reasonable to suggest that ratings and reports be tested.Another strategy is to ask people what would improve their situation rather than ask about areas of dissatisfaction.EVALUATING MEASURESA substantial literature has developed on instruments for measuring quality of care and patient status and well-being. Many instruments exist to measure health status or quality of life. For example, 1 recent review reported 159 different instruments used in the 75 articles analyzed,ranging from narrow (eg, hand-grip strength) to comprehensive.Although some instruments have been tested and validated in end-of-life care, most have not. A few are designed for use with family members.Various groups have described the attributes of useful measures.In addition to such important attributes as validity, reliability, ease and clarity of interpretation, and flexibility in use for different populations, the suitability of measures should be considered.Suitability has several aspects, including the use for which measures are intended.Typically, measures used for internal improvement of systems must be more timely and more detailed if they are to provide valuable information for clinicians and managers. Global assessments of satisfaction are of little use, for example, to administrators and staff who need to know what to change.Suitability also depends on resources available. Care must be taken, for example, when using a crude tool because of limited resources, not to overinterpret results that may be within the range of measurement error.For patients at the end of life, a particular concern should be the measure's suitability for the stage of illness. For example, a scale of function meant for patients with chronic illness might not be appropriate for terminally ill patients who cluster at the disabled end of the scale.Suitability also depends on the physical or mental condition of the patient. Asking patients with dementia for information about their symptoms or satisfaction requires special measures. Similarly, measures of cancer pain might not be suitable for other forms of pain.Finally, the method of collection is also relevant to the suitability of a measure. For example, the level of satisfaction, function, pain, and other symptoms can be assessed by proxy interviews with family or caregivers. Although such proxies will not provide the same information as that gathered from the patient, it can augment patient reports.Are there any suitable measures of quality of care at the end of life? Many measures that are suitable for other populations, even for patients with chronic debilitating illness, have not been validated for use with those at the end of life or for use with proxy respondents. The field is embryonic, but promising work has begun, led by Joan Teno, MD, who convened a workshop to assess currently available tools and priorities for their measurement. She and her colleagues have continued collaborations with interested investigators, including a follow-up meeting held September 1997.OVERSIGHT OF FACILITIESPublic oversight of quality has been most visibly linked to mechanisms such as organizational accreditation and licensure. People who are dying are, like other patients, generally protected by standards and policies related to such matters as staff credentials, advance care planning, emergency response procedures, safe physical environments, documentation of institutional performance for other interested parties, and governing mechanisms that protect and respect patients.Because more than 70% of those who die each year are 65 years old or older and are covered by Medicare, standards set by the Health Care Financing Administration (Baltimore, Md), the federal agency that oversees the Medicare program, are particularly important. Hospitals and other health care organizations are "deemed" to have met the Conditions of Participation if they are accredited by the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO; Oakbrook Terrace, Ill), an alternative to their state's survey and certification program. In hospitals, JCAHO standards now require evidence of compliance with guidelines on cancer pain related to metastatic lung, colorectal, or female breast cancer,but do not otherwise specify measures to be used for patients at the end of life.Only standards for hospice programs focus specifically on the care of dying patients. The Health Care Financing Administration has established conditions of participation in Medicare for hospice programs, and 36 states require hospices to be state licensed. The JCAHO accredits hospices within the category of home care organizations and no longer has separate hospice accreditation standards.Section PC.5 ("Additional Applicable Standards for Hospice") identifies 3 special areas of focus in care of the dying patient: (1) treating primary and secondary symptoms, (2) appropriately and aggressively managing pain, and (3) acknowledging the psychosocial and spiritual concerns of the patient and family about dying and the expression of grief.In addition, the National Hospice Organization (Arlington, Va) publishes Standards of a Hospice Program of Care.The standards address access to care, the patient and family as the unit of care, interdisciplinary team plans of care, the scope of hospice services, palliative care services, bereavement services, coordination and continuity of care, record keeping, governance, management and administration, and quality assessment and improvement, although no guidance about specific measures is provided. The standards stipulate that a hospice quality assessment program be based on systematic collection, review, and evaluation of data. Examples of topics for evaluation include the following:Services provided by professional and volunteer staff;Outcome audits of patient charts;Reports from staff, volunteers, and clients about services;Concerns or suggestions for improvement in services;Organizational review of the hospice program;Patient and family evaluations of care; andHigh-risk, high-volume, and problem-prone activity.The National Committee for Quality Assurance (Washington, DC) is the newest organization to undertake quality measurement for elderly patients. Its Health Plan Employer Data and Information Set is a quality reporting instrument designed to produce information for purchasers about access to and the quality of care for patients in health maintenance organizations. The most recent version includes measures that address the care of elderly patients enrolled in managed-care organizations.No measures focus on the care of dying patients, although several have been proposed.FUTURE DIRECTIONSDeveloping and implementing reliable, valid, and practical quality measures for any setting or set of patients, including those at the end of life, is a daunting task. Successfully using such measures to identify and correct quality problems and to improve care is an even more formidable challenge. In part because quality measurement has many purposes and because quality improvement is a complex undertaking, progress depends on contributions from many parties, including methodologists, clinicians, administrators, payers, and funders of clinical and health services research.This article has particularly emphasized issues that require further exploration and decisions by measurement experts and providers of care with the involvement of patients and those close to them. Given the early stage of quality assessment for care at the end of life, 1 prudent path for these groups may be to focus on a few measures (eg, the overall satisfaction of the patient or surrogate or inadequately relieved pain) and evaluate the extent to which they indicate broader quality problems. Quality measurement strategies often attempt this approach, and it may be particularly useful in end-of-life care because of the difficulties we have documented in collecting extensive information from dying patients and their families. If, however, such indicators are to be used to screen for potential quality problems, follow-up procedures are required to determine whether quality problems really exist and to assure that the screening strategy does not distort care or result in harmful unintended consequences.Government (eg, the Health Care Financing Administration, Baltimore, Md, and the Agency for Health Care Policy and Research, Rockville) and foundations (eg, the John A. Hartford Foundation, New York, NY, and the Robert Wood Johnson Foundation, Princeton, NJ) have assisted—if only modestly—in funding research on quality measurement and assessment. Such assistance is also essential for improving quality measurement and assessment for end-of-life care.In addition, payers—public and private—have required and must continue to call for evidence of the adequacy and quality of care for those in their programs, in addition to costs and efficiency. Because most of those who die each year are elderly and are covered by Medicare, the federal government has a particularly central role in this regard.Federal and state governments have multiple roles in quality assessment, as purchasers, funders of research, and regulators. In recent years, regulators have moved toward strategies that emphasize internal responsibility for continuing improvement in quality.Continuous quality improvement models are generally described as a set of reinforcing principles for implementing change, and these principles seem to deserve attention by those caring for all kinds of people, well and ill.The principles tend to emphasize continuing relationships between or among the participants in health care transactions (eg, physicians, patients, and purchasers); defects in systems rather than individual deficiencies ("bad apples") as major sources of errors, and improvement activities grounded in statistical and scientific precepts and techniques; standardization of processes to reduce the opportunity for error; links among specific care processes and health outcomes; feedback of information to practitioners about how their practices may differ from those of their peers or depart from evidence-based standards for practice; and continuous improvement in contrast to merely meeting established goals or criteria.Measuring quality of care is not enough. The more difficult goal is understanding how to modify care systems to bring about effective change by drawing on the strengths of various improvement strategies, such as methods of financing, regulation, education, computer technology, feedback, and continuous quality improvement. This article is intended to encourage the development and use of measures not only to identify problems in quality, but also to improve and celebrate excellent care.RNAndersonKDKochanekSLMurphyReport of final mortality statistics, 1995.Monthly Vital Stat Rep.1997;45(suppl 2):18.FBHobbsBLDamonSixty-five plus in the U.S. population.In: Current Population Reports, Series P234-190.Washington, DC: US Census Bureau; April 1996.Institute of MedicineApproaching Death: Improving Care at the End of Life.Washington, DC: National Academy Press; 1997.The SUPPORT Principal InvestigatorsA controlled trial to improve care for seriously ill hospitalized patients: the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT).JAMA.1995;274:1591-1598.ADonabedianExplorations in Quality Assessment and Monitoring: The Definition of Quality and Approaches to Its Assessment.Ann Arbor, Mich: Health Administration Press; 1990:1.Institute of MedicineMedicare: A Strategy for Quality Assurance.Washington, DC: National Academy Press; 1990.KNLohrAdvances in health status assessment.Med Care.1992;30(suppl):MS1-MS293.DBlumenthalQuality of health care, I: quality of care—what is it?N Engl J Med.1996;335:891-894.RHBrookEAMcGlynnPDClearyQuality of health care, II: measuring quality of care.N Engl J Med.1996;335:966-970.MRChassinQuality of health care, III: improving the quality of care.N Engl J Med.1996;335:1060-1062.World Health OrganizationCancer Pain Relief and Palliative Care: Report of a WHO Expert Committee.Geneva, Switzerland: World Health Organization; 1990. Technical Report Series No. 804.JLynnEHarrell JrFCohnMBHamelNDawsonAWuDefining the ‘terminally ill': insights from SUPPORT.Duquesne Law Rev.1996;35:311-336.MSDonaldsonKNLohrInstitute of MedicineHealth Data in the Information Age: Use, Disclosure, and Privacy.Washington, DC: National Academy Press; 1994.Institute of MedicineImproving the Quality of Care in Nursing Homes.Washington, DC: National Academy Press; 1986.ASRelmanAssessment and accountability: the third revolution in medical care.N Engl J Med.1988;319:1220-1222.DLanksyThe new responsibility: measuring and reporting on quality.Jt Comm J Qual Improv.1993;19:545-551.GSWunderlichFASloanCKDavisInstitute of MedicineNursing Staff in Hospitals and Nursing Homes: Is It Adequate?Washington, DC: National Academy Press; 1996.RLKaneJWalesLBernsteinALeibowitzSKaplanA randomised controlled trial of hospice care.Lancet.1984;1:890-894.DSGreerVMorFinal Report of the National Hospice Study.Providence, RI: Brown University; 1984.DSGreerVMorJNMorrisSSherwoodDKidderHBirnbaumAn alternative in terminal care: results of the National Hospice Study.J Chronic Dis.1986;39:9-26.KAWallstonCBurgerRASmithRJBaugher,Comparing the quality of death for hospice and non-hospice cancer patients.Med Care.1988;26:177-182.MDanisDLPatrickLISoutherlandMLGreenPatients' and families' preferences for medical intensive care.JAMA.1988;260:797-802.TMGillARFeinsteinA critical appraisal of the quality of quality of life measurements.JAMA.1994;272:619-946.DLPatrickMBergnerMeasurement of health status in the 1990s.Annu Rev Public Health.1990;11:165-83.DLPatrickPEricksonHealth Status and Health Policy.New York, NY: Oxford University Press Inc; 1993.MGoldPFranksPEricksonAssessing the health of the nation: the predictive validity of a preference-based measure and self-rated health.Med Care.1996;34:163-177.VMorSMasterson-AllenHospice Care Systems: Structure, Process, Costs, and Outcome.New York, NY: Springer Publishing Co Inc; 1987.CPuchalskiSpirituality and transcendence.Presented at the conference: ‘Toolkit of Instruments to Measure End of Life Care'; August 28, 1996; Woods Hole, Mass.GHGuyattDJCookHealth status, quality of life, and the individual.JAMA.1994;272:630-631.MGerteisSEdgman-LevitanJDaleyTLDelbancoThrough the Patient's Eyes: Understanding and Promoting Patient-Centered Care.San Francisco, Calif: Jossey-Bass Publishers Inc; 1993.JTenoExecutive summary of the conference ‘Making a Toolkit of Instruments to Measure End of Life Care (TIME),' August 27-28, 1996; Woods Hole, Mass.Available at: http://www.gwu.edu/acicd/toolkit/e.htm. Accessed November 24, 1997.MBergnerRABobbittWBCarterBSGilsonThe Sickness Impact Profile: development and final revision of a health status measure.Med Care.1981;19:787-805.MJFieldKNLohrInstitute of MedicineGuidelines for Clinical Practice: From Development to Use.Washington, DC: National Academy Press; 1992.RHPalmerAGLawthersNBanksDKurlandJLHargravesLPetersonUnderstanding and Choosing Clinical Performance Measures for Quality Improvement: Development of a Typology.Rockville, Md: US Dept of Health and Human Services; 1995. AHCPR publication 95-N001.Medical Outcomes TrustInstrument Review Criteria.Med Outcomes Trust Bull.1995;3(4):I-IV.Joint Commission on Accreditation of Healthcare Organizations1996 Comprehensive Accreditation Manual for Hospitals.Oakbrook Terrace, Ill: Joint Commission on Accreditation of Healthcare Organizations; 1996.Joint Commission on Accreditation of Healthcare Organizations1995 Accreditation Manual for Home Care.Oakbrook Terrace, Ill: Joint Commission on Accreditation of Healthcare Organizations; 1994.National Hospice OrganizationStandards of a Hospice Program of Care.Arlington, Va: National Hospice Organization; 1993.National Committee for Quality AssuranceHealth Plan and Employer Data and Information Set (HEDIS 3.0).Washington, DC: National Committee for Quality Assurance; 1996.Health Care Financing AdministrationQuality of Care Information; Medicare Peer Review Organizations.In: HCFA: the Medicare and Medicaid Agency[online]. Available at: http://www.hcfa.gov. Accessed September 12, 1997.PBataldenEDBuchananIndustrial models of quality improvement.In: Goldfield N, Nash D, eds. Providing Quality Care: The Challenge to Clinicians.Philadelphia, Pa: American College of Physicians; 1989.DMBerwickContinuous improvement as an ideal in health care.N Engl J Med.1989;320:53-56.DMBerwickABGodfreyJRoessnerCuring Health Care: New Strategies for Quality Improvement.San Francisco, Calif: Jossey-Bass Publishers Inc; 1990.Accepted for publication September 25, 1997.An early draft of this article was prepared, with support from the Andrus Foundation, for the American Geriatrics Society and presented April 30, 1996, by Ms Donaldson at an invitational symposium, "A Good Dying: Shaping Health Care for the Last Months of Life," sponsored by The Center to Improve Care of the Dying, The George Washington University (Washington, DC), and the Corcoran Gallery of Art, Washington, DC, with support from the Greenwall Foundation, New York, NY.The views in this article are those of the authors and do not necessarily indicate endorsement by the American Geriatrics Society, the Center to Improve Care of the Dying, the Institute of Medicine, the National Research Council, or the supporting institutions.We thank Joanne Lynne, MD, for encouragement to undertake this project and for helpful reviews of several drafts of the manuscript; Kathleen N. Lohr, PhD, who reviewed an early draft; Joan Teno, MD, for organizing; and the Nathan Cummings Foundation, New York, NY, and the Robert Wood Johnson Foundation, Princeton, NJ, for supporting the conference "Toolkit of Instruments to Measure End of Life Care," Woods Hole, Mass, August 27-28, 1996.This article draws on the work of the Institute of Medicine Committee on Care at the End of Life and its 1997 report, Approaching Death: Improving Care at the End of Life.That study was chaired by Christine Cassel, MD, and was supported by the Open Society Institute, New York, NY; the Greenwall Foundation, New York, NY; the Charles E. Culpeper Foundation, Stanford, Conn; contract 02975 from the Robert Wood Johnson Foundation; and by contract 500-95-0026 from the Health Care Financing Administration, Baltimore, Md.Reprints: Molla S. Donaldson, Institute of Medicine, 2101 Constitution Ave NW, Washington, DC 20418 (e-mail: mdonalds@nas.edu). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JAMA Internal Medicine American Medical Association

Measuring Quality of Care at the End of Life

JAMA Internal Medicine , Volume 158 (2) – Jan 26, 1998

Loading next page...
 
/lp/american-medical-association/measuring-quality-of-care-at-the-end-of-life-VZEdoVC0Wc

References (51)

Publisher
American Medical Association
Copyright
Copyright 1998 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
ISSN
2168-6106
eISSN
2168-6114
DOI
10.1001/archinte.158.2.121
Publisher site
See Article on Publisher Site

Abstract

Caring for patients at the end of life presents a series of quality-of-care problems to the health care system. In the past, concern has focused on overaggressive treatment of dying patients. Given rapid changes in the financing and delivery of care, it is time to focus on a range of quality problems and address ways to improve care and achieve outcomes desired by patients and their families. We provide a framework for conceptualizing such a task. This article addresses the purposes of measurement, definition of the patient population, timing of measurement, use of surrogates in measurement, scope of services to be evaluated, and the choice of measures. It emphasizes the necessary links between quality measurement and quality improvement.Most Americans have come to expect long and generally healthy lives. They also have come to understand that the consequence of longer life is that most people today die in old age, often of chronic and progressive diseases. Heart disease, cancer, stroke, and chronic obstructive pulmonary disease are the 4 leading causes of death among older persons. The incidence of disease and disability and the use of health care resources rise with age. Almost three quarters of the 2.3 million Americans who die each year are 65 years old or older, and about 5% of the Medicare population will die in a year.Estimates for the year 2030 show a population of more than 70 million people older than 65 years of age, 20% of the total population, compared with 13% in 1994.Dying slowly—at any age—presents a series of quality-of-care problems to the health care system, which has for the most part been designed to respond to acute illness. Throughout the course of chronic progressive illnesses, patients are likely to be examined and treated by a series of practitioners in different care settings—acute and long-term—for multiple problems that may ebb and flow in importance and in their effect on function and health-related quality of life. These circumstances tend to attenuate any one clinician's responsibility for the overall quality of a patient's care.Why should the very end of life be a subject of particular interest for the quality of health care?First, dying patients are particularly vulnerable, and the care provided to the most vulnerable in our society is a measure of the health care system itself. Second, quality-of-care problems have been documented for dying patients, including shortcomings in communication, highly aggressive treatment, and inadequate relief of pain and other symptoms.Third, in a time of rapidly changing health care financing and provision, the elderly and very ill may be affected more than others who have greater physiological, psychological, social, or financial resources.In the past, much concern has focused on overaggressive treatment of dying patients, in particular the use of life-sustaining technologies when the marginal benefits are insufficient to warrant their use and when patients themselves may not desire such care. Researchers have also documented considerable undertreatment of pain among people with advanced illness. A cost-driven shift to managed care (a simple label for a diverse set of financial and administrative practices) is further raising anxiety about underuse of service in some plans.Given these concerns, policymakers, managers of programs serving the very ill, clinicians, and others committed to improving the care of dying patients need methods to measure and evaluate the quality of care provided, thereby identifying poor care, assessing how changes in systems of care affect patients, and devising ways to correct problems. This article considers strategies for assessing and improving care at the end of life. The starting point is an examination of the concepts of quality and accountability as they relate to the care of people approaching death.CONCEPTSThe last 3 decades have seen continued advances in the conceptualization of quality of care and health outcomes, measurement tools, and strategies for improving health care and its results.More recently, those concerned about the quality of care for dying patients have begun to focus attention on how these concepts, tools, and strategies can be applied and adapted to measure, monitor, and improve care for those approaching death.QUALITY OF CAREIn 1990, an Institute of Medicine (Washington, DC) report, Medicare: A Strategy for Quality Assurance,defined quality of care as "the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge." This definition covered individuals and populations, which is consistent with increasing concerns about how different clinical interventions, programs, and resources can be deployed to the greatest social and individual advantage. Because the evidence base about what works in health care is still modest for many medical conditions, the definition acknowledged the relevance of professional knowledge, which includes experience and judgment, as well as the results of biomedical and clinical research. The definition deliberately omits resource constraints on the ground that judgments of what constitutes excellent, acceptable, or unacceptable quality should be independent of constraints on resources. Decision makers cannot, however, ignore resources in selecting the level of quality that is desired and affordable for a particular institution, population, or community. This means paying attention to quality and resources for patients covered by public and private insurance arrangements and for those excluded from such arrangements.In the context of approaching death as in other areas of health care, quality problems can be categorized in 3 broad areas: (1) overuse, (2) underuse, and (3) poor practitioner skills or performance. Examples of overuse are overuse of acute care settings, such as hospitals and intensive care; of diagnostic technologies when intervention is not warranted; of medical or surgical interventions, such as radiation, chemotherapy, and surgery; of life-prolonging or life-sustaining technologies, such as blood products, hemodialysis, total parenteral nutrition, mechanical ventilation, and tube feedings; and oversedation with pain medication. Examples of underuse are underdiagnosis and care of treatable illnesses, such as a bacterial infection; underuse of technologies and services that could improve the quality of life for patients, such as self-managed pain medication; undercontrol of symptoms, such as pain, nausea, anxiety, or depression; inadequate provision of help to family members, such as bereavement services; and the lack of involvement of patient and family in decision making. Examples of poor practitioner skills or performance that indicate inadequate technical or interpersonal quality are incorrect or inappropriately timed medicines, skin breakdown, nosocomial infection, avoidance of the patient and the family, unwillingness to answer questions, and failure to recognize and relieve symptoms, failure to provide help to a bereaved spouse, including an opportunity for the spouse to ask questions and remain in touch with the caregivers. Considered together, these categories help focus attention on whether what is done (or not done) is the proper thing to do and whether it is done properly.Death—the ultimate outcome of life—has limited value for measuring the quality of care experienced by those who are dying. More relevant measures focus on symptom management, patient and family perceptions and satisfaction, spiritual well-being, and patients' quality of life, each of which can be affected by the process of care. Health-related quality of life emphasizes health as perceived and valued by people for themselves or sometimes for others.ACCOUNTABILITYThe field of palliative care defines itself as particularly, but not exclusively, accountable for the care for those approaching death. Palliative care has been described by the World Health Organization as ". . . the active total care of patients whose disease is not responsive to curative treatment"(p11) and for whom the "control of pain, and other symptoms, and of psychological, social, and spiritual needs is paramount. The goal of palliative care is achievement of the best quality of life for patients and their families."(p11) For those who die suddenly, for example, of trauma, palliative care may apply primarily in the form of bereavement support for survivors.Some have conceptualized the shift from acute to palliative care as an abrupt transition, but this does not (or should not) reflect reality.Some aspects of palliative care, including careful management of symptoms and advance care planning, can be applied in tandem with curative, life-prolonging efforts. Just as no bright line separates palliative care from other forms of care, no bright line shields any clinician responsible for patient care from being held accountable for having the knowledge and skills to care well for the dying.Although much palliative care is rendered by family and social networks, health care professionals and organizations can provide many essential medical and social services, especially the relief of pain and other distressing symptoms. They are accountable for meeting the professional standards, skillfully using a defined base of knowledge, and performing in ways "consistent with current professional knowledge" (as emphasized in the definition of quality of care by the Institute of Medicine).At the heart of the medical enterprise is an ongoing relationship between a patient and clinician. This means that even when care is provided by a team of health care professionals, patients and their families need to know who these persons are, who is "in charge" of their care, and who can answer their questions and assist with decision making. Issues of communication among caregivers, patients, and family are complex and beyond the scope of this article; nevertheless, they are appropriate subjects for quality measurement and improvement.How broadly should accountability for quality of care for dying patients be conceptualized? The concept of palliative care extends accountability to include the patient and those close to the patient. Thus, the family experience, their participation in caring for a loved one, and their emotional and financial burdens also become a responsibility. This suggests that measurement of the quality of care likewise should extend beyond patients to include those close to them. Hospice programs consider the patient and family as the unit of care, as well as a part of the care team, but other institutions may not be so explicit about their responsibilities.Even palliative care specialists may not know whether they are fulfilling their responsibilities in the absence of systems that can measure outcomes and identify problems across care settings. When care for a dying person involves many different professionals providing care in different settings with varying institutional affiliations, no one person may be responsible for that patient's care throughout the course of a progressive, eventually fatal illness. Although integrated health care systems and managed care organizations offer the promise of accountability across settings of care, health care systems—including performance measurement programs—designed primarily for younger, healthier populations may not meet the needs of those with advanced chronic illness.MEASURING QUALITY OF CARENumerous issues quickly arise in the construction of a framework for measuring quality. These include the following: (1) the purpose of quality measurement, (2) the patient population, (3) the timing of measurement, (4) the source of patient information, and (5) the settings in which care is to be measured. The selection of quality measures depends in part on how these issues are resolved.PURPOSE OF MEASUREMENTAny attempt to implement quality of care measurement must first establish its purpose. Tools and strategies for measuring, monitoring, and improving the quality of care for people who are dying will, ideally, serve multiple purposes, including the following:Educating clinicians, policymakers, and the public about what is to be sought in the care for dying patients and families and helping them assess particular organizations and instances of care;Assisting those directly responsible for end-of-life care in continuously evaluating and improving what they are doing;Helping to identify deficiencies in systems of care for dying patients and the means of correcting them; andSupporting systematic research on the effects of different clinical, organizational, and financing options for end-of-life care.More specifically, 4 principal uses of quality measurement should be considered. These are external "inspection," internal improvement in the care process, selective contracting with providers, and informed patient and purchaser choice of plans, facilities, and practitioners. External inspection determines the suitability of a setting for patients. It stresses common definitions and measures that can be translated from 1 organization to another and can be understood by all. It generally involves static measurements that set a basic floor for performance, for example, for accreditation or reimbursement eligibility. (Performance in this sense may refer to the creation of organizational structures and procedures, to the achievement of defined outcomes, or both.)Quality assessment for internal improvement is a more dynamic idea based on the notion that care can always be improved and that such improvement ought to be continuous and based on monitoring performance and seeking opportunities to improve. It means that information must be detailed and timely enough to permit translation of data into information about how to change the care that is provided. For example, a nursing home might undertake a study of the transfers of its very ill patients to hospitals to determine whether such transfers seemed consistent with patient wishes and, indeed, whether patient preferences were documented and available to guide transfers and care after the transfers.Quality assessment for selective contracting—a purpose that reflects the current stress on market forces and competition—is sometimes described as "value purchasing." Quality assessment for informed choice is a prerequisite of a well-functioning health care market. It means that measures and the results of measurement must be important to and understandable to patients and their families and responsive to their values and preferences.Policymakers and researchers are just beginning to understand how the expectations and preferences of patients are shaped over time by their experiences, rather than being fixed attributes. Preferences (including the most fundamental—the use of life-sustaining interventions) may, for example, be affected by the quality of the caregiving. The ongoing interactions among the stated preferences, expectations, and experiences of patients means that care received may help to shape patient preferences as much as the reverse is true.Information particularly relevant to dying patients, for example, the availability of palliative care specialists, is not, however, featured in current consumer information. Similarly, although the availability of tertiary care services is typically considered in the creation of health plan networks, the services available to dying patients do not seem to figure in current contracting considerations. In fact, in a competitive health care system, a plan could be disadvantaged if it were known to provide particularly high quality care options for very ill and dying patients. Given current methods of capitating health plans, adverse selection that is disproportionately attracting sicker patients poses a serious financial risk to health plans.The purpose of measurement should be very clear because it will affect the scope, depth, and type of information collected. If intended for internal improvement, information must be quite detailed, but it does not require the attention to comparability and the level of confidence in small differences that are needed for fair comparisons between sites of care. Measurement used for internal improvement also must focus on outcomes and processes that the plan or facility could affect.PATIENT POPULATIONMany of the quality-of-care issues apply across age groups and are relevant, for example, for very premature infants, neonates with severe birth defects, people of all ages with malignant carcinomas, young adults with the acquired immunodeficiency syndrome, and middle-aged adults with degenerative diseases. Nevertheless, most of those who die in the United States of chronic terminal conditions are elderly, and that will influence some elements of quality measurement and improvement strategies in many situations.This article focuses primarily on older people with progressive, eventually fatal diseases. Many of these people, however, are not perceived by themselves or others as dying, and predicting death is an imprecise exercise. This creates difficulties because measuring quality requires identifying the population of interest—in epidemiologic terms, the denominator. It also creates difficulties for scheduling measures of patient status.TIMING OF MEASUREMENTUnlike other events in health care, such as follow-up after surgical procedures or scheduled visits for monitoring or treating chronic conditions, events near the end of life generally cannot be predicted precisely.To measure care while the patient is still alive, decisions must be made about whether measurement should be tied to specific events (eg, on admission, transfer, or when significant interventions are instituted or withdrawn) or should occur periodically (eg, at specific intervals), and if so, with what frequency. Unless data are gathered frequently (eg, daily), the unpredictability of deaths means that some patients may be examined 3 days before death, others 3 weeks before death, and others not at all. Such variation could make comparisons difficult.Identifying terminally ill patients before they have died requires some sort of marker event, such as passing a clinical threshold of severity or transition to palliative care (such as transfer to hospice care or withdrawal of life support). In some cases, it may not be clear that a patient is dying, for instance, during medical rescue efforts for a person with congestive heart failure who has survived past crises. Without a better way of recognizing when people are approaching death and when it is appropriate to focus more on comfort than on cure or life extension, organizations may miss important opportunities to improve care.Dying patients also can be identified retrospectively through death records. Retrospective review of care will be adequate if administrative claims or medical records are accurate, appropriately detailed, and available from all sites in which dying patients receive care. These sources are not likely to be sufficient for other purposes, however; for example, understanding patient or family perceptions and experiences.Other information may need to be gathered retrospectively from interviews with family and caregivers, which raises different issues about the role of surrogates.THE ROLE OF SURROGATES AS A SOURCE OF PATIENT INFORMATIONBefore death, many seriously ill and dying patients may be unable to report for themselves because of the physical or mental impairments produced by the illness. Also, as just noted, the uncertain timing of death makes direct collection of information from all patients improbable. The question thus arises whether a proxy or surrogate, someone such as a family member or a formal or informal caregiver, can provide reliable and valid information on the patient's experience. The evidence on this point is limited.Continued research is urgently needed to assess the degree to which patient and surrogate perceptions match (when both can be measured), the extent to which different surrogates provide different reports, the reasons for differences, and the bases for choosing a proxy informant. Such research is neither simple nor inexpensive because it requires not 1 but 2 data points for each variable and because a large number of patients may have to be included to allow for those who are unable at some point to communicate because of their illness or the treatment for it. Nevertheless, a better sense of the reliability and validity of surrogate data is a high priority for patients approaching death.CARE SETTINGSQuality measures should apply to all settings where patients die or receive care before they die. This includes home or other residential settings (eg, board and care facilities), hospitals (eg, intensive care or palliative care units), and nursing homes (including those with hospice units). If differences exist among settings in the quality of life for patients (as well as in cost), this is particularly useful information for patients and payers. It would also be useful to know about features of a specific setting (such as a hospice) that contribute to those differences.The kinds of interventions, staffing (ie, types of personnel and their level of training), and available technologies and labels for care (eg, hospice, palliative care, or medical care) may differ greatly, but the objectives of quality measurement should include comparing outcomes across settings and linking differences to variations in care or the environment of care (considering other factors, such as difference in patient or family characteristics). In developing such information, it is conventional to distinguish 3 broad kinds of quality measures relating to structures of care, processes of care, and outcomes of care.Table 1gives examples of structural aspects of care that might be the subject of measurement.Examples of Structural Aspects of Care That May Affect Patient CareSee table graphicThe extent and nature of links among specific health care settings, policies, and other structural factors (eg, nurse staffing ratios) and outcomes are matters of continuing dispute based on modest and sometimes conflicting research.Efforts continue to identify and establish structures and processes that can be linked to outcomes, in part to help improve care and in part to determine whether easier-to-collect structure and process data can be used as proxy measures of quality. As noted, linking care to outcomes requires that other factors that may affect outcomes—for example, how ill a patient is or how care is financed—be measured or at least defined.Two studies of hospice care attempted to make such comparisons,and they illustrate issues in quality measurement. In the first, the National Hospice Study,although hospice patients and their family caregivers were more satisfied than control patients, survival time, pain, symptom relief, or alleviation of psychological distress did not differ between hospice and nonhospice settings. Patients with cancer in a home or hospital-based hospice were compared with a control group of patients with cancer receiving conventional care. Patient outcomes (based on the report of the primary caregiver) included overall quality of life, social quality of life, pain and other symptoms, and satisfaction. Hospice patients experienced less use of aggressive or intervention therapy and diagnostic testing, but the quality of life of the patients in both settings was similar, except for pain control, which may have been better in the hospice setting. Again, hospice patients were more likely to die at home, and families were satisfied with that outcome, but otherwise no consistent differences were found.More recent studieshave questioned whether the instruments used in the National Hospice study were sufficiently sensitive to differences in care during the last 3 days before death. This suggests the importance of the timing of measurement in relation to the time of death and the need for measures that are sensitive to differences in care.TRANSITIONS ACROSS SETTINGSTransitions from 1 setting to another provide particular hazards for patients because physical, financial, informational, and other links must be planned and implemented successfully. Transitions may not be timely because they are too early or too late from the perspective of the condition of the patient; they may also be untimely if the sending or receiving organization is unprepared, as may happen when an imminently dying patient is transferred from the hospital to hospice care in a matter of a few hours. People and equipment, for example, a hospital bed for a home, may not be so quickly arranged. Salient information may not be transferred. Thus, it is important not only to measure the quality of care in the different settings, but also to consider continuity and coordination across care settings as they affect the patient's experience of care, quality of life, and other outcomes and the availability of important medical and patient preference information.MEASURES AND CRITERIAThose seeking to identify or improve tools for measuring the quality of care at the end of life have an array of potential instruments from which to choose and problems in assessing the usefulness and appropriateness of the measures. Few measures, however, have been developed with explicit attention to the concerns of dying patients and those caring for them.CATEGORIES OF OUTCOMESThe understanding of what it means to live well while dying and how to measure the quality of dying remains at an early stage. The elements of physical and mental functioning and well-being important to healthy people, people expected to recover from acute illnesses, or people expected to live for an extended period with various chronic illnesses may change in meaning for people with advanced, fatal illnesses.SurvivalIn most circumstances, survival is a good in itself—survival, for example, from surgery, after premature birth, and after organ failure have been used as markers for quality. For patients near the end of life, however, survival itself is not always viewed as the most important goal.Nevertheless, survival time, if adequately adjusted for patient preferences and for severity of illness, can be an important first screen for possible quality problems. For instance, survival time might differ for different hospices for a variety of reasons (eg, stage of illness at admission compared with success in relieving symptoms such as vomiting or dysphagia), only some of which might raise concerns about quality.Quality of LifeQuality of life is "a uniquely personal perception, denoting the way that individual patients feel about their health status and/or nonmedical aspects of their lives. . . ."(p619) Most people wish for a long and satisfying life, that is, people value how long they live andhow well they live.Health-related quality-of-life outcomes include physical, mental, social, and role functioning and sense of well-being; freedom from pain; satisfaction with health care; and an overall sense of general health. The meaning of symptoms and how they affect the quality of life of patients are intrinsically subjective, however. Some may find pain or incontinence very distressing; others may be less affected. Some may find pain less distressing than the mental haziness induced by medications. Measures of symptoms and satisfaction with care raise a variety of problems,and spiritual constructs may be particularly troublesome to define and translate into operational measures (I. Byock, MD, and M. Merriman, PhD, testimony before the Institute of Medicine Committee on Care at the End of Life, April 1996). A large literature has been developed that addresses quality-of-life measurement generically and, to some extent, on a condition-specific basis. Although self-reported quality of life during terminal care cannot be measured retrospectively (because the patient has died), with further research and refinement of measures, using proxy responses from a family member, friend, or caregiver or targeting patients ahead of death may be possible.Satisfaction With CarePatient and family satisfaction with the process and outcomes of care are important aspects of quality. Measurement of satisfaction can be made during palliative care and after death by interviews with family members. One potential difficulty with interviewing patients (and possibly family members) is that of respondent bias, which is likely if people fear that expressions of dissatisfaction with care and caregivers might make caregivers angry or neglectful. Conversely, expressions of dissatisfaction from patients or family may be expected by caregivers and thus discounted. Another problem for those truly interested in improving care is that people may tend to describe themselves as satisfied even when probing would reveal various quality concerns.In an effort to respond to some of the problems with satisfaction measures, Gerteis and colleaguesadvise asking patients and proxies not about their satisfaction with care, but about their experiences. At this time, it seems reasonable to suggest that ratings and reports be tested.Another strategy is to ask people what would improve their situation rather than ask about areas of dissatisfaction.EVALUATING MEASURESA substantial literature has developed on instruments for measuring quality of care and patient status and well-being. Many instruments exist to measure health status or quality of life. For example, 1 recent review reported 159 different instruments used in the 75 articles analyzed,ranging from narrow (eg, hand-grip strength) to comprehensive.Although some instruments have been tested and validated in end-of-life care, most have not. A few are designed for use with family members.Various groups have described the attributes of useful measures.In addition to such important attributes as validity, reliability, ease and clarity of interpretation, and flexibility in use for different populations, the suitability of measures should be considered.Suitability has several aspects, including the use for which measures are intended.Typically, measures used for internal improvement of systems must be more timely and more detailed if they are to provide valuable information for clinicians and managers. Global assessments of satisfaction are of little use, for example, to administrators and staff who need to know what to change.Suitability also depends on resources available. Care must be taken, for example, when using a crude tool because of limited resources, not to overinterpret results that may be within the range of measurement error.For patients at the end of life, a particular concern should be the measure's suitability for the stage of illness. For example, a scale of function meant for patients with chronic illness might not be appropriate for terminally ill patients who cluster at the disabled end of the scale.Suitability also depends on the physical or mental condition of the patient. Asking patients with dementia for information about their symptoms or satisfaction requires special measures. Similarly, measures of cancer pain might not be suitable for other forms of pain.Finally, the method of collection is also relevant to the suitability of a measure. For example, the level of satisfaction, function, pain, and other symptoms can be assessed by proxy interviews with family or caregivers. Although such proxies will not provide the same information as that gathered from the patient, it can augment patient reports.Are there any suitable measures of quality of care at the end of life? Many measures that are suitable for other populations, even for patients with chronic debilitating illness, have not been validated for use with those at the end of life or for use with proxy respondents. The field is embryonic, but promising work has begun, led by Joan Teno, MD, who convened a workshop to assess currently available tools and priorities for their measurement. She and her colleagues have continued collaborations with interested investigators, including a follow-up meeting held September 1997.OVERSIGHT OF FACILITIESPublic oversight of quality has been most visibly linked to mechanisms such as organizational accreditation and licensure. People who are dying are, like other patients, generally protected by standards and policies related to such matters as staff credentials, advance care planning, emergency response procedures, safe physical environments, documentation of institutional performance for other interested parties, and governing mechanisms that protect and respect patients.Because more than 70% of those who die each year are 65 years old or older and are covered by Medicare, standards set by the Health Care Financing Administration (Baltimore, Md), the federal agency that oversees the Medicare program, are particularly important. Hospitals and other health care organizations are "deemed" to have met the Conditions of Participation if they are accredited by the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO; Oakbrook Terrace, Ill), an alternative to their state's survey and certification program. In hospitals, JCAHO standards now require evidence of compliance with guidelines on cancer pain related to metastatic lung, colorectal, or female breast cancer,but do not otherwise specify measures to be used for patients at the end of life.Only standards for hospice programs focus specifically on the care of dying patients. The Health Care Financing Administration has established conditions of participation in Medicare for hospice programs, and 36 states require hospices to be state licensed. The JCAHO accredits hospices within the category of home care organizations and no longer has separate hospice accreditation standards.Section PC.5 ("Additional Applicable Standards for Hospice") identifies 3 special areas of focus in care of the dying patient: (1) treating primary and secondary symptoms, (2) appropriately and aggressively managing pain, and (3) acknowledging the psychosocial and spiritual concerns of the patient and family about dying and the expression of grief.In addition, the National Hospice Organization (Arlington, Va) publishes Standards of a Hospice Program of Care.The standards address access to care, the patient and family as the unit of care, interdisciplinary team plans of care, the scope of hospice services, palliative care services, bereavement services, coordination and continuity of care, record keeping, governance, management and administration, and quality assessment and improvement, although no guidance about specific measures is provided. The standards stipulate that a hospice quality assessment program be based on systematic collection, review, and evaluation of data. Examples of topics for evaluation include the following:Services provided by professional and volunteer staff;Outcome audits of patient charts;Reports from staff, volunteers, and clients about services;Concerns or suggestions for improvement in services;Organizational review of the hospice program;Patient and family evaluations of care; andHigh-risk, high-volume, and problem-prone activity.The National Committee for Quality Assurance (Washington, DC) is the newest organization to undertake quality measurement for elderly patients. Its Health Plan Employer Data and Information Set is a quality reporting instrument designed to produce information for purchasers about access to and the quality of care for patients in health maintenance organizations. The most recent version includes measures that address the care of elderly patients enrolled in managed-care organizations.No measures focus on the care of dying patients, although several have been proposed.FUTURE DIRECTIONSDeveloping and implementing reliable, valid, and practical quality measures for any setting or set of patients, including those at the end of life, is a daunting task. Successfully using such measures to identify and correct quality problems and to improve care is an even more formidable challenge. In part because quality measurement has many purposes and because quality improvement is a complex undertaking, progress depends on contributions from many parties, including methodologists, clinicians, administrators, payers, and funders of clinical and health services research.This article has particularly emphasized issues that require further exploration and decisions by measurement experts and providers of care with the involvement of patients and those close to them. Given the early stage of quality assessment for care at the end of life, 1 prudent path for these groups may be to focus on a few measures (eg, the overall satisfaction of the patient or surrogate or inadequately relieved pain) and evaluate the extent to which they indicate broader quality problems. Quality measurement strategies often attempt this approach, and it may be particularly useful in end-of-life care because of the difficulties we have documented in collecting extensive information from dying patients and their families. If, however, such indicators are to be used to screen for potential quality problems, follow-up procedures are required to determine whether quality problems really exist and to assure that the screening strategy does not distort care or result in harmful unintended consequences.Government (eg, the Health Care Financing Administration, Baltimore, Md, and the Agency for Health Care Policy and Research, Rockville) and foundations (eg, the John A. Hartford Foundation, New York, NY, and the Robert Wood Johnson Foundation, Princeton, NJ) have assisted—if only modestly—in funding research on quality measurement and assessment. Such assistance is also essential for improving quality measurement and assessment for end-of-life care.In addition, payers—public and private—have required and must continue to call for evidence of the adequacy and quality of care for those in their programs, in addition to costs and efficiency. Because most of those who die each year are elderly and are covered by Medicare, the federal government has a particularly central role in this regard.Federal and state governments have multiple roles in quality assessment, as purchasers, funders of research, and regulators. In recent years, regulators have moved toward strategies that emphasize internal responsibility for continuing improvement in quality.Continuous quality improvement models are generally described as a set of reinforcing principles for implementing change, and these principles seem to deserve attention by those caring for all kinds of people, well and ill.The principles tend to emphasize continuing relationships between or among the participants in health care transactions (eg, physicians, patients, and purchasers); defects in systems rather than individual deficiencies ("bad apples") as major sources of errors, and improvement activities grounded in statistical and scientific precepts and techniques; standardization of processes to reduce the opportunity for error; links among specific care processes and health outcomes; feedback of information to practitioners about how their practices may differ from those of their peers or depart from evidence-based standards for practice; and continuous improvement in contrast to merely meeting established goals or criteria.Measuring quality of care is not enough. The more difficult goal is understanding how to modify care systems to bring about effective change by drawing on the strengths of various improvement strategies, such as methods of financing, regulation, education, computer technology, feedback, and continuous quality improvement. This article is intended to encourage the development and use of measures not only to identify problems in quality, but also to improve and celebrate excellent care.RNAndersonKDKochanekSLMurphyReport of final mortality statistics, 1995.Monthly Vital Stat Rep.1997;45(suppl 2):18.FBHobbsBLDamonSixty-five plus in the U.S. population.In: Current Population Reports, Series P234-190.Washington, DC: US Census Bureau; April 1996.Institute of MedicineApproaching Death: Improving Care at the End of Life.Washington, DC: National Academy Press; 1997.The SUPPORT Principal InvestigatorsA controlled trial to improve care for seriously ill hospitalized patients: the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT).JAMA.1995;274:1591-1598.ADonabedianExplorations in Quality Assessment and Monitoring: The Definition of Quality and Approaches to Its Assessment.Ann Arbor, Mich: Health Administration Press; 1990:1.Institute of MedicineMedicare: A Strategy for Quality Assurance.Washington, DC: National Academy Press; 1990.KNLohrAdvances in health status assessment.Med Care.1992;30(suppl):MS1-MS293.DBlumenthalQuality of health care, I: quality of care—what is it?N Engl J Med.1996;335:891-894.RHBrookEAMcGlynnPDClearyQuality of health care, II: measuring quality of care.N Engl J Med.1996;335:966-970.MRChassinQuality of health care, III: improving the quality of care.N Engl J Med.1996;335:1060-1062.World Health OrganizationCancer Pain Relief and Palliative Care: Report of a WHO Expert Committee.Geneva, Switzerland: World Health Organization; 1990. Technical Report Series No. 804.JLynnEHarrell JrFCohnMBHamelNDawsonAWuDefining the ‘terminally ill': insights from SUPPORT.Duquesne Law Rev.1996;35:311-336.MSDonaldsonKNLohrInstitute of MedicineHealth Data in the Information Age: Use, Disclosure, and Privacy.Washington, DC: National Academy Press; 1994.Institute of MedicineImproving the Quality of Care in Nursing Homes.Washington, DC: National Academy Press; 1986.ASRelmanAssessment and accountability: the third revolution in medical care.N Engl J Med.1988;319:1220-1222.DLanksyThe new responsibility: measuring and reporting on quality.Jt Comm J Qual Improv.1993;19:545-551.GSWunderlichFASloanCKDavisInstitute of MedicineNursing Staff in Hospitals and Nursing Homes: Is It Adequate?Washington, DC: National Academy Press; 1996.RLKaneJWalesLBernsteinALeibowitzSKaplanA randomised controlled trial of hospice care.Lancet.1984;1:890-894.DSGreerVMorFinal Report of the National Hospice Study.Providence, RI: Brown University; 1984.DSGreerVMorJNMorrisSSherwoodDKidderHBirnbaumAn alternative in terminal care: results of the National Hospice Study.J Chronic Dis.1986;39:9-26.KAWallstonCBurgerRASmithRJBaugher,Comparing the quality of death for hospice and non-hospice cancer patients.Med Care.1988;26:177-182.MDanisDLPatrickLISoutherlandMLGreenPatients' and families' preferences for medical intensive care.JAMA.1988;260:797-802.TMGillARFeinsteinA critical appraisal of the quality of quality of life measurements.JAMA.1994;272:619-946.DLPatrickMBergnerMeasurement of health status in the 1990s.Annu Rev Public Health.1990;11:165-83.DLPatrickPEricksonHealth Status and Health Policy.New York, NY: Oxford University Press Inc; 1993.MGoldPFranksPEricksonAssessing the health of the nation: the predictive validity of a preference-based measure and self-rated health.Med Care.1996;34:163-177.VMorSMasterson-AllenHospice Care Systems: Structure, Process, Costs, and Outcome.New York, NY: Springer Publishing Co Inc; 1987.CPuchalskiSpirituality and transcendence.Presented at the conference: ‘Toolkit of Instruments to Measure End of Life Care'; August 28, 1996; Woods Hole, Mass.GHGuyattDJCookHealth status, quality of life, and the individual.JAMA.1994;272:630-631.MGerteisSEdgman-LevitanJDaleyTLDelbancoThrough the Patient's Eyes: Understanding and Promoting Patient-Centered Care.San Francisco, Calif: Jossey-Bass Publishers Inc; 1993.JTenoExecutive summary of the conference ‘Making a Toolkit of Instruments to Measure End of Life Care (TIME),' August 27-28, 1996; Woods Hole, Mass.Available at: http://www.gwu.edu/acicd/toolkit/e.htm. Accessed November 24, 1997.MBergnerRABobbittWBCarterBSGilsonThe Sickness Impact Profile: development and final revision of a health status measure.Med Care.1981;19:787-805.MJFieldKNLohrInstitute of MedicineGuidelines for Clinical Practice: From Development to Use.Washington, DC: National Academy Press; 1992.RHPalmerAGLawthersNBanksDKurlandJLHargravesLPetersonUnderstanding and Choosing Clinical Performance Measures for Quality Improvement: Development of a Typology.Rockville, Md: US Dept of Health and Human Services; 1995. AHCPR publication 95-N001.Medical Outcomes TrustInstrument Review Criteria.Med Outcomes Trust Bull.1995;3(4):I-IV.Joint Commission on Accreditation of Healthcare Organizations1996 Comprehensive Accreditation Manual for Hospitals.Oakbrook Terrace, Ill: Joint Commission on Accreditation of Healthcare Organizations; 1996.Joint Commission on Accreditation of Healthcare Organizations1995 Accreditation Manual for Home Care.Oakbrook Terrace, Ill: Joint Commission on Accreditation of Healthcare Organizations; 1994.National Hospice OrganizationStandards of a Hospice Program of Care.Arlington, Va: National Hospice Organization; 1993.National Committee for Quality AssuranceHealth Plan and Employer Data and Information Set (HEDIS 3.0).Washington, DC: National Committee for Quality Assurance; 1996.Health Care Financing AdministrationQuality of Care Information; Medicare Peer Review Organizations.In: HCFA: the Medicare and Medicaid Agency[online]. Available at: http://www.hcfa.gov. Accessed September 12, 1997.PBataldenEDBuchananIndustrial models of quality improvement.In: Goldfield N, Nash D, eds. Providing Quality Care: The Challenge to Clinicians.Philadelphia, Pa: American College of Physicians; 1989.DMBerwickContinuous improvement as an ideal in health care.N Engl J Med.1989;320:53-56.DMBerwickABGodfreyJRoessnerCuring Health Care: New Strategies for Quality Improvement.San Francisco, Calif: Jossey-Bass Publishers Inc; 1990.Accepted for publication September 25, 1997.An early draft of this article was prepared, with support from the Andrus Foundation, for the American Geriatrics Society and presented April 30, 1996, by Ms Donaldson at an invitational symposium, "A Good Dying: Shaping Health Care for the Last Months of Life," sponsored by The Center to Improve Care of the Dying, The George Washington University (Washington, DC), and the Corcoran Gallery of Art, Washington, DC, with support from the Greenwall Foundation, New York, NY.The views in this article are those of the authors and do not necessarily indicate endorsement by the American Geriatrics Society, the Center to Improve Care of the Dying, the Institute of Medicine, the National Research Council, or the supporting institutions.We thank Joanne Lynne, MD, for encouragement to undertake this project and for helpful reviews of several drafts of the manuscript; Kathleen N. Lohr, PhD, who reviewed an early draft; Joan Teno, MD, for organizing; and the Nathan Cummings Foundation, New York, NY, and the Robert Wood Johnson Foundation, Princeton, NJ, for supporting the conference "Toolkit of Instruments to Measure End of Life Care," Woods Hole, Mass, August 27-28, 1996.This article draws on the work of the Institute of Medicine Committee on Care at the End of Life and its 1997 report, Approaching Death: Improving Care at the End of Life.That study was chaired by Christine Cassel, MD, and was supported by the Open Society Institute, New York, NY; the Greenwall Foundation, New York, NY; the Charles E. Culpeper Foundation, Stanford, Conn; contract 02975 from the Robert Wood Johnson Foundation; and by contract 500-95-0026 from the Health Care Financing Administration, Baltimore, Md.Reprints: Molla S. Donaldson, Institute of Medicine, 2101 Constitution Ave NW, Washington, DC 20418 (e-mail: mdonalds@nas.edu).

Journal

JAMA Internal MedicineAmerican Medical Association

Published: Jan 26, 1998

There are no references for this article.