Executive Summary

In the United States today, we know with reasonable certainty what people die from, how old they are when they die, and some other key pieces of information. Trends over time in these vital statistics, and associated research on factors that tend to increase or decrease the risk of death, have laid the foundation for strategies to improve public health and avoid premature deaths. But death does come—to about 2.4 million each year. Although death is largely a phenomenon of old age, deaths among younger people are significant, adding up to half a million infants, children and young adults. We have only begun to pay attention to the circumstances in which chronically ill people approach death and experience the dying process. We know relatively little about the quality, appropriateness, or costs of care they receive, or the burden on caregivers and survivors. Even less is known about deaths of young people than about deaths among middle-aged and elderly people (Seeman et al., 1989).

“Quality of care” is a subjective concept, of course, but various groups have begun to define minimum standards that can be agreed upon, as well as ideals that can be held as goals. Quality of care is not an end in itself, either for the temporarily or the fatally ill. It is one contributor to “quality of life,” regardless of the amount of time left to a life. In this report, we are concerned about describing both “quality of care” and “quality of life” near the end of life. They are distinct qualities and require different types of measurements, related either to the process and outcomes of care, in the former case, or the perceptions of the dying and those around them, in the latter.



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



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

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

OCR for page 1
Executive Summary In the United States today, we know with reasonable certainty what people die from, how old they are when they die, and some other key pieces of information. Trends over time in these vital statistics, and associated research on factors that tend to increase or decrease the risk of death, have laid the foundation for strategies to improve public health and avoid premature deaths. But death does come—to about 2.4 million each year. Although death is largely a phenomenon of old age, deaths among younger people are significant, adding up to half a million infants, children and young adults. We have only begun to pay attention to the circumstances in which chronically ill people approach death and experience the dying process. We know relatively little about the quality, appropriateness, or costs of care they receive, or the burden on caregivers and survivors. Even less is known about deaths of young people than about deaths among middle-aged and elderly people (Seeman et al., 1989). “Quality of care” is a subjective concept, of course, but various groups have begun to define minimum standards that can be agreed upon, as well as ideals that can be held as goals. Quality of care is not an end in itself, either for the temporarily or the fatally ill. It is one contributor to “quality of life,” regardless of the amount of time left to a life. In this report, we are concerned about describing both “quality of care” and “quality of life” near the end of life. They are distinct qualities and require different types of measurements, related either to the process and outcomes of care, in the former case, or the perceptions of the dying and those around them, in the latter.

OCR for page 1
The National Cancer Policy Board concluded in its July 2001 report, Improving Palliative Care for Cancer (IOM, 2001), that there is currently insufficient information to assess or improve the quality of care provided to those who die from cancer in the United States. The Board noted that we have little understanding of the particular dying experiences of most patients with cancer—where they die, who cares for them as they are dying, what the quality of such care is, whether care guidelines are being followed, and whether these things are changing over time. This lack of information hampers our ability to develop a clear policy agenda and will, in the future, impede monitoring trends to determine whether interventions are having their intended effects of improving the quality of life for individuals at the end of life. Knowing how well we are doing or whether things are getting better in end-of-life care requires some routinely collected information. It may be possible to make better use of data already being collected (including those collected for other purposes), but it also may be desirable or necessary to gather new types of data. The Board, therefore, decided to undertake an assessment of the information potential of current data sources to assess quality of life and quality of care at the end of life. Recognizing that the issue existed not only for deaths from cancer, but for all deaths, this report was undertaken in collaboration with the Board on Health Sciences Policy in this project that looks at deaths from all causes and at all ages, including childhood. This report reviews the sources of information currently available that shed light on the quality of care and quality of life at the end of life. It has uncovered wide gaps between what we would like to know and what is actually known or knowable, if existing national-level data were fully explored and analyzed. The scope of the report is limited to examining large-scale data collection through vital statistics, ongoing national surveys, administrative data from the Medicare program, and some large periodic surveys. Data from these sources, if fully exploited, can provide a framework—a skeleton—for the complete story, which can be told only through the many types of hypothesis-driven epidemiologic and health services studies that are critical to filling in the details of quality of care and quality of life of real people who are dying. Often, it is the national-level data collection efforts described in this report that are used to generate hypotheses to be tested in more focused studies. And while this report delves into the types of data needed, it does not include an analysis of current funding levels, or of the funds that would be needed to carry out the recommendations. That said, however, there is a clear theme of striving for the best use of existing data, and where changes are required— either modifications of existing systems or new efforts—proposing how to accomplish them with the fewest perturbations.

OCR for page 1
WHY WE NEED TO KNOW There are various reasons for wanting to find out more about the circumstances of people’s lives as they are approaching death. The focus of this report and its recommendations is on the quality of care and quality of life during the period leading up to death for people with progressive disease. The rationale becomes clear when viewed as a logical next report in relation to its immediate predecessor, Improving Palliative Care for Cancer (IOM, 2001), and the 1997 Institute of Medicine (IOM) report Approaching Death: Improving Care at the End of Life. The care of individuals who are approaching death is an integral and important part of health care (IOM, 1997). Furthermore, people who are nearing the end of life should be able to expect and achieve a decent or good death—one that is free from avoidable distress and suffering for patients, families, and caregivers; in general accord with patients’ and families’ wishes; and reasonably consistent with clinical, cultural, and ethical standards (IOM, 1997). We know from studies and from anecdotal information that for too many people, this expectation is not being met. But if we want to know how we’re doing as a nation, we must gather information systematically and regularly, in a way that will allow comparisons across different segments of society and over time. The experience of dying today is typically very different from what it was 100 years ago. Death at home in the care of the family has been replaced in most instances by a technological, professional, and institutional treatment of dying (IOM, 1997). A theme of the 1997 Approaching Death report was that a significant number of people experience needless suffering and distress at the end of life. The suffering and distress is caused in some cases when caregivers provide care that is clinically inappropriate or not wanted by the patient; in other cases, it results when caregivers underuse effective treatments to relieve pain or other physical or psychological symptoms (IOM, 1997). As a consequence, many Americans “have come both to fear a technologically overtreated and protracted death and to dread the prospect of abandonment and untreated physical and emotional distress” (IOM, 1997). In order to improve the quality of life and quality of care for people at the end of life, it will be necessary to develop, validate, and benchmark measures of quality for end-of-life care. As W. Edwards Deming stated, “If you don’t measure it, you can’t improve it.” Measures of the quality of life and care of dying patients are needed for the purposes of (1) public accountability (i.e., allowing policymakers, patients, families, and the public to hold organizations and individuals accountable for the quality of the care they provide to dying patients); (2) internal quality improvement efforts by clinicians and others directly responsible for end-of-life care to

OCR for page 1
evaluate and improve what they are doing on a continuing basis; and (3) further research on the effects of different clinical, organizational, and financing options for end-of-life care and on the effectiveness of alternative strategies for improving care and outcomes for patients and their families (IOM, 1997; Teno, 2001). At this point, the understanding of what it means to live well while dying and how to measure the quality of dying remains at an early stage (IOM, 1997; Teno, 2001). Basic descriptions of the dying experience and the care given to people who are dying are lacking (Teno, 2001). Clinical guidelines to help physicians, nurses, and other health professionals manage life-threatening illnesses are generally silent on the topic of managing the end stage of the disease (IOM, 2001; Teno, 2001). The field of palliative care focuses on the prevention and relief of suffering through the final stages of an illness and attends closely to the emotional, spiritual, and practical needs of patients and those close to them, but there are no quality indicators in national use that deal specifically with the quality of palliative care provided to patients with life-threatening and incurable illnesses (Teno, 2001). Strengthening accountability for dying patients’ quality of life and care will require better data and tools for evaluating outcomes important to patients and families (IOM, 1997). Patient preferences and satisfaction are important at every stage of treatment, but they have particular significance in the care of patients with life-threatening and incurable illnesses (Teno, 2001). One category of outcomes in dying patients for which there is particularly strong normative and empirical evidence to support quality measures is pain management (Teno, 2001). In 1997, the IOM reported on the American Geriatrics Society suggestions for other domains of quality of care for dying patients, among them physical and emotional symptoms, support of functioning and autonomy, advance care planning, aggressive care near death, patient and family satisfaction, and global quality of life (IOM, 1997). Only after satisfactory measures in end-of-life quality domains such as these are developed and evaluated in demonstration programs will it be possible to develop and apply standards of accountability for end-of-life care. All of this begins with knowing something about how people across the country are dying now. WHAT WE NEED TO KNOW The effort to better understand and to improve the care received by those at the end of life will require better information about dying—not just about the demographic characteristics and health conditions of those who die, but also about their quality of life as they cope with declining health, the quality of their death, and the quality of the health care pro

OCR for page 1
vided to them during this time. Nationally representative surveys, disease registries, health care billing data and vital statistics are among the possible sources of this information. To date, the focus of most analyses of these data has been on factors contributing to death, rather than on the experience of dying. To better understand dying, a very different set of topics would be of interest. These might include characteristics of the site of death, symptoms and quality of life in the last year of life for patients and their families, patient preferences for and continuity of care in the last year of life, and family burden (physical, psychological, practical, and financial). Designating a period of time as “the end of life” is a difficult but necessary first step. The conceptual model of a “terminal illness” has driven most current public policy about dying, though dying is not so easily predictable for most Americans. Death more frequently occurs during an exacerbation of an otherwise chronic illness or from a complication associated with a very long slow decline in old age (Lunney et al., 2001). In this report, the end of life is defined very broadly, to include the period during which an individual copes with declining health from an ultimately terminal illness, from a serious though perhaps chronic illness, or from the frailties associated with advanced age—even if death is not clearly imminent. Public discussions and focus group research have helped us to understand what the American public values in the care of the dying (Steinhauser et al., 2000a, b; Teno et al., 2001). Professional societies and experts in palliative and end-of-life care have also developed frameworks that describe the concerns of people who are dying, and the roles of professionals, family members, and society at large in meeting their needs. The various frameworks use different terminology, and may split into different components, but basically, they all include: The management of symptoms; Spiritual and personal growth; A familiar setting, surrounded by loved ones; Understandable information to guide decision making and planning; Confidence that one will not be a financial, emotional, or physical burden to family members; and The right of self-determination and control of treatment choices. These frameworks and the “domains” they include lead directly to specific questions that must be answered to assess and monitor how well needs are being met. We have taken the elements in the various models and developed a set of questions that would capture the information

OCR for page 1
TABLE ES-1 Questions to Assess and Monitor How Well Individuals’ End-of-Life Needs Are Being Met 1. Where are people dying and how much of the end of their lives is spent in those settings? 2. Who is providing care for them as they die? Do institutional settings support family presence at the end of life? 3. Are physical and psychological symptoms being identified and treated (including but not limited to pain)? 4. What proportion of people experience impaired cognitive function before death, to what degree, and for what period of time? 5. What proportion of people experience physical disability or social isolation before death, to what degree, and for what period of time? 6. How do patients and loved ones perceive their quality of life at various time points prior to death? 7. Are patients and loved ones achieving a sense of life closure in their relationships and spiritually? 8. Are patients and loved ones involved in decision making about treatment and care options? 9. Are patients and loved ones receiving timely and adequate information on which to base informed decisions? 10. Are patients and loved ones receiving supportive services from chaplains, therapists, health aides, and other providers? 11. Are family physical, emotional, and financial resources being heavily depleted by the care of a dying family member? 12. Are loved ones supported through the grieving process? needed to fill in important blanks (see Table ES-1). These questions are used later to query the data sets described, to determine how well each question is covered (summarized in Table ES-2, and more fully discussed in Chapter 3). SOURCES OF DATA Very few efforts to gather data specifically to illuminate the issues surrounding dying and the end of life have been made, but data are routinely collected for other purposes that can fill some of the needs. In some cases, data simply must be extracted and analyzed to prove useful, and in other cases, changes in what is collected routinely could greatly improve the usefulness of various data sources for these purposes. Chapter 3 of this report describes large datasets, including both ongoing and

OCR for page 1
periodic or one-time surveys, that contain some relevant elements. These range from the death certificate, which is completed for every person who dies in the United States and aggregated in national mortality reporting, to billing records for the Medicare program, to regular or episodic surveys of representative samples of individuals in the general population (e.g., the continuous National Health Interview Survey and the now continuous, but formerly episodic National Health and Nutrition Examination Survey, both conducted by the National Center for Health Statistics). Mortality “followback” surveys, in which proxy respondents (usually surviving spouses or other family members) are interviewed to gather information about people who died in a given year, have great potential to answer directly some of the questions identified, but the last one took place in 1993 and no new followback surveys are planned (but see Recommendations, below). In Chapter 3, the large-scale data sources with information relevant to end-of-life concerns are described and reviewed for the useful elements they contain. This information is summarized in Table ES-2 and provides the background against which recommendations described below have been developed. Methodological issues involved in studying the quality of life and care at the end of life are described in Chapter 4. CONCLUSIONS AND RECOMMENDATIONS The final chapter of the report summarizes conclusions and presents the Institute of Medicine’s recommendations. It notes that a better description of how Americans of all ages currently die and the impact on their families is much needed and must be followed by continuing efforts to track changes over time. Both the quality of dying and the quality of the health care provided at the end of life should be addressed. The IOM’s recommendations, summarized below, address supporting the use of existing data systems, improving the usefulness of existing data systems, and undertaking new data collection efforts to learn about the quality of life and care at the end of life. Recommendation #1: Support Researchers’ Use of Existing Data Systems Agencies should promote research that uses existing data resources to describe aspects of the quality of life and quality of care at the end of life, by publicizing their availability and providing funding for analysis.

OCR for page 1
TABLE ES-2 Types of Data Available from Datasets That Contain Individual-Level Information Data type   Number of deaths year[s] Cause of death Comorbidities Pain or other physical symptoms C2 AHEAD 1,900         (1993-98)   ✓ ✓ C2 LSOA I & II I: 2,900     (1984-90) II: 1,900 (1994-2000) ✓ ✓   C1 MCBS 700/yr   ✓   B MDS   ✓ ✓ B Medicare claims 1.7 million   (2000)   ✓   C1 MEPS NA2   ✓ ✓ A Death Certificates (NDI) 2.4 M (2000) ✓     C1 NHANES 50/yr estimate   ✓ ✓ C2 NHEFS 1,400     (1987-92)1 ✓ ✓ ✓ C2 NHHCS ~5,000     (2000)   ✓   C1 NHIS 55,000       (1986-94)   ✓ ✓ C2 NLTCS 17,000       (1982-86)   ✓   C2 NMFS 23,000       (1993) ✓ ✓   C1 NNHS ~1,600       (1999)   ✓   B OASIS     ✓ ✓ C1 PSID 4,300       (thru 1999)   ✓  

OCR for page 1
Physical function Psychological symptoms QoL Cognition Site of death Lived with Social support ✓ ✓ ✓ ✓   ✓ ✓ ✓       ✓ ✓ ✓ ✓     ✓   ✓ ✓ ✓ ✓   ✓   ✓ ✓   ✓ ✓   ✓   ✓     ✓   ✓     ✓   ✓   ✓   ✓ ✓   ✓   ✓   ✓ ✓   ✓   ✓   ✓ ✓   ✓   ✓ ✓ ✓     ✓ ✓ ✓   ✓         ✓   ✓ ✓   ✓   ✓   ✓         ✓  

OCR for page 1
Data type   Family care Out of pocket expenses Bereavement ADS C2 AHEAD ✓ ✓   ✓ C2 LSOA I & II ✓     ✓ C1 MCBS   ✓   ✓ B MDS   ✓   ✓ B Medicare claims   C1 MEPS ✓ ✓   A Death Certificates (NDI)   C1 NHANES C2 NHEFS C2 NHHCS ✓ ✓   C1 NHIS   ✓   C2 NLTCS ✓ ✓   ✓ C2 NMFS ✓ ✓   ✓ C1 NNHS   ✓     B OASIS ✓   C1 PSID ✓ KEY: First Column: A = Vital statistics; B = Administrative data; C = Research sample (C1 = continuous or regular sample; C2 = not continuous or regularly occurring sample); AHEAD—Asset and Health Dynamics Among the Oldest Old; LSOA—Longitudinal Survey of Aging; MCBS—Medicare Current Beneficiary Survey; MDS—nursing home Minimum Data Set; MEPS—Medicare Expenditure Panel Survey; NDI—National Death Index; NHANES—National Health and Nutrition Examination Survey; NHEFS—NHANES I Epidemiologic Followup Survey; NHHCS—National Home and Hospice Care Survey; NHIS—National Health Interview Survey; NLTCS—National Long-Term Care Survey; NMFS—National Mortality Followback Survey; NNHS— National Nursing Home Survey; OASIS—Outcome and Assessment Information Set; PSID—Panel Study of Income Dynamics.

OCR for page 1
Decision making Satisfaction Support Use of services Quality Provider   ✓   ✓   ✓ ✓ ✓ ✓     ✓ ✓     ✓   ✓   ✓ ✓ ✓ ✓ ✓   ✓     ✓     ✓ ✓     ✓ ✓       ✓ ✓   ✓   ✓ ✓ ✓     ✓   ✓ ✓ ✓   ✓   ✓   1Deaths during 5-year follow-up period total 20-year follow-up (1971-1992). 2MEPS data will be linked to the NDI when a large enough number of deaths have occurred. It is currently too early for this.

OCR for page 1
A great deal of information is recorded, for a variety of purposes, that could describe aspects of quality of life and quality of care at the end of life. It is probably not surprising that Medicare claims and other data collected for purposes other than measuring quality at the end of life are not always used to their fullest for measuring quality, given that their primary purpose is purely administrative. In addition, the full potential value of many surveys and studies of health and well-being in characterizing quality of life and care at the end of life is not realized because data are analyzed only to answer some questions but not others. The reasons for this vary, but include a general lack of funding support for further data analysis, the fact that researchers are unaware of the data resource, and in some cases, the relatively new ability to link data from different sources to facilitate meaningful answers to important questions. Currently, little support is available to researchers to make use of existing data resources in describing aspects of the quality of life and and the care provided to individuals who are dying. Research-funding agencies should provide support for using existing data resources for this purpose. Some examples of the types of studies that could be conducted with modest funding include: Making use of existing longitudinal surveys to examine the health trajectories of those who die in order to learn more about the role of suffering, disability, and chronic illnesses at the end of life. Studying patterns of costs and utilization in the years before death to more fully describe the use of home care, hospice care, in-home privately paid help, and informal care. Studying the number of care transitions in the last year of life, both in terms of settings and of providers. Comparing data from multiple settings with regard to the rates of pain and other symptom assessment, and use of opioids and other symptom-relieving interventions. Developing valid indicators of variables and constructs that are important to good end-of-life care. This is an important but slow process—and one that needs to be attended to immediately. Examining individual and institutional factors that influence racial and geographic variations in patterns of end-of-life care. Recommendation #2: Improve the Usefulness of Existing Data Systems Government and private organizations should institute training initiatives and make incremental changes to surveys to improve the

OCR for page 1
usefulness of currently collected data in describing aspects of quality of life and quality of care at the end of life. Training for researchers and incremental changes to surveys would be relatively inexpensive ways to improve the usefulness of currently collected data in describing aspects of quality of life and quality of care at the end of life. Training for Researchers More funding for analysis and better publicity about data sources will improve the use of existing information up to a point, but beyond that, researchers may need training in specific datasets or enhancement of skills to be able to understand the possibilities of existing information collection. Focused training opportunities, limited in scope, can open the way for much better use of what data currently exist. Specific recommendations for training of researchers are as follows: Government agencies that sponsor data collection should sponsor a series of training initiatives to open dialogues among researchers and health care workers to increase the reliability and validity of ongoing data collections. These should emphasize: Training professionals in various disciplines in the use of data sources. Training existing palliative care researchers to incorporate research questions in their studies, which would utilize existing data bases. Promoting existing health service researchers with expertise in these databases to collaborate with palliative care researchers. Improving the Quality of Data Recorded Professional organizations should promote increased standardization of language through open dialogue about terms that now have diverse meanings, e.g., hospice care, palliative care, and end of life. In settings where physicians complete death certificates regularly, institute training and quality control measures. The federal government should mandate that institutions and organizations providing care (hospitals, nursing homes, home health agencies, outpatient settings) conduct ongoing quality improvement efforts, including training in recording required data.

OCR for page 1
Incremental Changes to Ongoing Data Collection Efforts The information content for studying end-of-life issues can be enhanced in specific surveys by relatively minor changes. The following improvements, modifications, or supplements to ongoing data collection efforts would be relatively inexpensive ways to build upon existing efforts: Sponsors of longitudinal health surveys should insist on the routine use of “exit” surveys to capture information from the next-of-kin of participants who die between survey rounds. These supplemental surveys not only enhance the existing survey by covering a more inclusive range of outcomes but can provide rich data about end-of-life issues at relatively low cost. Improvements to the information infrastructure, such as nationwide electronic reporting of death certificates, should be made to facilitate more timely collection and analysis of vital information. Sampling frames for current surveys should be carefully reviewed for adequacy both in terms of contemporary housing arrangements and changing health care organizational structures. Survey questions regarding health care utilization should include probes that capture need for and use of the full range of supportive services. Mechanisms should be put in place for continuous refinement of large ongoing surveys, using input from a wide variety of sources. Web-based opportunities to e-mail questions and suggestions would facilitate this exchange. The Centers for Medicare and Medicaid Services (CMS)—formerly known as the Health Care Financing Administration—should consider minor increases in the data elements recorded on Medicare claims under the hospice benefit. This topic should be addressed by the Medicare Payment Advisory Commission (MedPAC), a Congressionally-mandated organization that advises CMS on the Medicare system. Ongoing efforts should be directed towards facilitating links among key types of data collection, especially between surveys and health care utilization records. Recommendation #3: A New National Mortality Followback Survey The federal government should undertake a new National Mortality Followback Survey to enhance the accuracy and richness of data collection related to quality of life and care at the end of life.

OCR for page 1
A new National Mortality Followback Survey program of regular, periodic surveys that gather comparable data over time, should be initiated. The new NMFS would be a collaborative effort between the National Center for Health Statistics (which would carry out the survey), the National Institutes of Health (NIH), and the Agency for Healthcare Research and Quality (the latter organizations would sponsor the survey). As the lead institute for end-of-life research at the NIH, the National Institute of Nursing Research could take a major role in supporting a National Mortality Followback Survey and determining content, with collaboration from other NIH institutes (in particular, the National Institute on Aging, but possibly including some of the disease-oriented institutes), and the Agency for Healthcare Research and Quality. Three specific aims should be considered for future National Mortality Followback Surveys. First, a major focus should be on determining the extent of morbidity experienced at the end of life, to aid the long-range goal of reducing unnecessary morbidity. Second, deaths among children should be oversampled in order to yield information specific to this understudied group. And third, minority populations should be over-sampled to improve understanding of ethnic and racial differences in the experience of death and dying in the United States.