Your browser doesn't support JavaScript. Please upgrade to a modern browser or enable JavaScript in your existing browser.
Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Quality Interagency Coordination (QuIC) Task Force
Return to QuIC Home
About QuIC
Steering Group
Press Releases
Related Links
Site Map
Written Statement


National Summit on Medical Errors and Patient Safety Research

Panel 3: Particular Systems Issues

Testimony of Joanne Lynn and Sarah Myers, RAND Center to Improve Care of the Dying, and Americans for Better Care of the Dying

The first National Summit on Medical Errors and Patient Safety Research was held on September 11, 2000, in Washington, DC. Sponsored by the Quality Interagency Coordination Task Force (QuIC), the Summit’s goal was to review the information needs of individuals involved in reducing medical errors and improving patient safety. More importantly, the summit set a coordinated and usable research agenda for the future to answer these identified needs.

Individuals were selected by the Agency for Healthcare Research and Quality (AHRQ) to testify at the summit as members of the witness panels. Each submitted written statements for the record before the event, documenting key issues that they confront with regard to patient safety as well as questions to be researched. Other applicants were invited to submit written statements.

Disclaimer and Copyright Statements

The RAND Center to Improve Care of the Dying is a scholarly team that does research and supports quality improvement aimed at guiding policy and practice to more effectively serve those with eventually fatal illness. RAND is a non-profit research institution that helps improve policy and decisionmaking through research and analysis.

Americans for Better Care of the Dying is a public charity that aims to ensure that all Americans can count on good care at the end of life.

Background / Summary
Research Questions / Action Steps / References

Medical errors and other lapses in patient safety are particularly likely to harm patients with serious, progressive, eventually fatal chronic illness. First, these patients have more interactions with the health care system, and we have a higher risk of encountering medical errors during that part of life. Second, this population actually has more serious effects from errors because they often cannot protect themselves from risks and they have less reserve with which to overcome the effects. Finally, the care system that serves them is rife with treatment patterns that run counter to well-substantiated evidence-based practices, thus subjecting those coming to the end of life to recurring mistakes. These medical errors make the end of life much more distressing—or even hasten death. There is evidence that we could build a care system that would reliably serve us when we need it most. Many provider organizations are undertaking efforts to improve their services to this population. The Center to Improve Care of the Dying and Americans for Better Care of the Dying have been at the center of major national collaborative quality improvement initiatives and we publish a newsletter that characterizes the best projects in end-of-life quality improvement. (Exchange, at Therefore, we have come to know the information needs of those who aim to provide more reliable and safe care for this population. We urge the Quality Interagency Coordination Task Force (QuIC) and its associated agencies to address these information needs as you develop national strategies to enhance patient safety.

Top of Page


Medication errors and other failings in patient safety are surprisingly common, often preventable, and exert an enormous toll on health care spending. The Institute of Medicine (Kohn et al., 2000) estimated that up to 98,000 deaths are caused each year by medication errors in hospitals alone and that total costs associated with preventable adverse events—medical errors that result in injury—in hospitals are between $17 billion and $29 billion annually. The IOM defined a medical error as "the failure of a planned action to be completed as intended or the use of a wrong plan to achieve an aim", and an adverse event as "an injury caused by medical management rather than the underlying condition of the patient" (Kohn et al., 2000).

Although they have the potential to affect the quality of health care provided to all Americans, medication errors and other patient safety lapses disproportionately affect the sickest patients, both in terms of their incidence and in severity of the consequences (Shelton et al., 2000; McDonald et al., 2000). Among the most vulnerable patients are those who are very sick for a long time. Mostly, serious, long-term illness occurs at the end of life—leading to more errors, more adverse effects, and more routine errors in treatment patterns.

Most Americans now spend most of our lives healthy, or living with rather minor and merely annoying chronic illnesses. However, most of us also spend a few months or years living with serious, progressive, eventually fatal illness as we come to the end of life. Although only a few percent of the population are affected at any one time, probably three-quarters of us travel this path for a few months or years before death, and probably more than half of Medicare's payments go to treatments for these patients. Furthermore, the numbers affected will more than double in the next two decades, as the Baby Boomers grow old together (Federal Interagency Forum on Aging Related Statistics, 2000).

Who is included in this especially vulnerable group? We include patients with chronic illnesses which will eventually cause death, though the timing is not clear, as well as those eligible for hospice, whose upcoming death is fairly certain. Except for cancer, we most often die within a few days or weeks of having had reasonably good likelihood of living for six months (Lynn, 1997). These patients make a coherent group because the services needed by all such patients are similar—continuity, symptom prevention and control, family support, advance care planning, and emotional support through loss and bereavement. Rescue from the jaws of death may still be important, but it is not usually as important—or as achievable—as it is for healthier patients.

Just half a century ago, most people died relatively quickly from injuries, infections, and acute conditions. In contrast, Americans are now likely to live into old age and then to live for months or years with the chronic diseases that will eventually kill them; many thousands of people are living with disabling and progressive illnesses such as heart disease, emphysema, stroke, dementia, and cancer. Nearly nine out of ten persons who die while elderly and covered by Medicare have one of these diagnoses in the year before death (Hogan et al., 2000) Even if health care and life style improvements delay the onset of disabling disease, each one of us still will die, and only a few of us will die without substantial antecedent illness.

Throughout their course, these patients tend to have substantial suffering, multiple hospitalizations, and high costs. Three-quarters of Medicare beneficiaries who die have at least one hospitalization in their last year of life, and Medicare pays an average of $26,000 per beneficiary in their last year. Furthermore, 25 percent of total Medicare outlays are spent on services during the last year of life (Hogan et al., 2000).

Sometimes health care professionals wrongly assume that patients who are already seriously ill or dying cannot be vulnerable to much harm. They may believe that those at the end of life have already received their bad news, they are only biding their time, and they are suffering already—they might seem to have little to lose. Perhaps this biased attitude gave rise to the fact that neither the IOM report (Kohn et al., 2000) nor most other recent reports on patient safety specifically address the disproportionate burden of those suffering serious chronic illness, though other vulnerable population groups such as children are targeted for special attention.

Quite the opposite, we contend that the time one has at the end of life often is especially precious. Indeed, the end of life can be very enriching—a time of enormous personal and spiritual growth, as well as a time to heal broken relationships and do the things that one always wanted to do (Lynn & Harrold, 1999). Safe, reliable, high-quality care will benefit these patients at least as much as patients with less threatening illnesses. In addition, many reformers are demonstrating that real improvement can be achieved [see, for example, the successes reported in issues of Innovations in End-of-Life Care ( and ABCD Exchange (].

Several factors make individuals at the end of life more vulnerable to medical errors and resulting adverse events:

  1. Persons living with advanced illness have substantially increased exposure to medical errors, including medication mistakes, improper procedures, and missed diagnoses.

Even as early as 25 years ago, researchers recognized that medication-associated deaths were more common "in the cases of severely or terminally ill patients treated with potentially highly toxic drugs." (Caranasos et al., 1976). We now also know that the combinations of drugs that they take can lead to other problems such as drug interactions. As the IOM Committee pointed out, "the potential for medication-related error increases as the average number of drugs administered increases." (Kohn et al., 2000) The number of drugs taken by older individuals (AARP, 1996)—especially those with a serious illness (Corcoran, 1997)—exceeds the number of drugs taken by other population groups and the drugs that they take are often more toxic (Corcoran, 1997).

Not only do they take more medications, but this vulnerable group of patients is also more likely to be in complex care settings, like hospitals and intensive care. The median patient with heart failure spends 12 percent of the last three months in a hospital bed; those with chronic obstructive pulmonary disease are in hospitals for 22 percent of that time; and those with cirrhosis are hospitalized for 20 percent of the time (Freeborne et al., 2000).

So, even if errors were spread randomly over encounters with health care, it is the very sick who would be the victims most often. McDonald et al. made the point that people who were at risk of errors were also often at high risk of dying, with or without errors (McDonald et al., 2000). While their point was that this makes it difficult to attribute causation of death to the error, they also illuminated the rather obvious fact that most errors will happen to those who are most sick.

  1. This population actually has more serious effects from errors because they often cannot protect themselves from risks and they have less reserve with which to overcome the effects.

Patients at the end of life are often too ill to monitor the care they receive. Not only do they have to navigate a fragmented, complex system of care, but also they also often must confront significant complexity and uncertainty within their own lives. Most are becoming frail—the average age at death in the nation is now 79 for women, and 74 for men (Federal Interagency Forum on Aging Related Statistics, 2000). Many have cognitive deficits—probably more than half have dementia, stroke, delirium, or coma affecting their last weeks or days (The SUPPORT Principal Investigators, 1995; Seale and Cartwright, 1994; Lynn et al., 2000b)

The Agency for Healthcare Research and Quality (AHRQ) recently recommended 20 ways that consumers can help prevent medical errors in their own care (AHRQ, 2000), including making sure that all of your doctors know about every medication you are taking and making sure your doctor knows about any allergies and adverse reactions you have had to medicines. These recommendations are not easy for many patients to follow and would be particularly difficult for seriously ill, frail, or older individuals. Patients at the end of life also deserve to be more focused on attending to their affairs and saying goodbye. They should not have to monitor their medications or keep a vigilant watch over their myriad health care providers.

Minor medication errors may go unnoticed in young, relatively healthy patients. However, individuals with progressive, chronic illness are often left terribly weak as a result of their conditions and the effects of minor errors may therefore be magnified. Patients with little physiological reserve are simply less likely to absorb and accommodate to a stress caused by an error. For example, a medication error can overwhelm a frail person with a relatively minor drug interaction, a hormonal imbalance, or another adverse effect.

  1. People at the end of life are exposed to pervasive patterns of care that run counter to well-substantiated evidence-based practices, thus subjecting those coming to the end of life to a nearly endless series of mistakes.

As recently as half a century ago, very few people lived very long with serious and progressive illness. As the United States built a health care system in the twentieth century, we mostly worried about whether people could get treatment for episodes of illness, especially operations, diagnostic procedures, and close monitoring during technically difficult treatments. With Medicare, we succeeded: beneficiaries can readily get most hospital-based care. However, no one engineered the care system to meet the needs of burgeoning numbers of people living with serious, progressive chronic illness, and it shows. (Fox, 1993). Even though good evidence shows that comprehensive, continuous care for persons with serious heart failure enhances their lives, reduces their hospitalizations, and costs less (Rich et al., 1996), that care is not generally available to Medicare beneficiaries. Though good evidence shows that people do much better if they are treated for many problems at home (Inouye et al., 1999; Leff et al., 1999), appropriate evaluation and treatment services are not commonly available at home or in nursing facilities. Only about half of the patients who need pain treatment in nursing homes get any suitable medication (Won et al., 1999). The examples are legion, and we spell out a few here in more detail.

Non-treatment of pain, delirium, and other symptoms

Symptoms such as pain (The SUPPORT Principal Investigators, 1995) and dyspnea (Dudgeon et al., 1996)—as well as patients' preferences related to symptom management—are often addressed too late or not at all, leading to much suffering and decreased quality of life. Good evidence underlies guidelines and their implementation, (AHRQ, 1992; Du Pen et al., 1999), yet we know that far too many patients still experience moderate to severe pain (Tolle et al., 2000; SUPPORT Principal Investigators, 1995), delirium (Inouye, 1999), and depression (Breitbart and Passik, 1998) at the end of life. The evidence shows that many symptoms could have been relieved quite effectively. Every practitioner would claim to aim to relieve and prevent symptoms, yet they engage in a care system that routinely ignores the opportunities to do so. The care system has practices that ensure high rates of untreated symptoms, patterns that no one would defend as intended. They persist from inertia and inattention.

Routine discontinuity within and among provider programs

Care provided to individuals at the end of life often comes from numerous providers, both at one time and over time as the patients move from home to hospital and nursing home. These providers may not communicate effectively with one another about their patients' care. This fragmentation can lead to prescriptions for medications that cause drug interactions and increased incidence in problems with medication use by the patients. One study of medication use in 111 terminally ill patients found that 90 percent of those patients had more than one health care provider who had prescribed medications for them and that 60 percent did not take all of their medications properly. Patients for whom a hospital was the main provider were less likely to use their medications properly (Zeppetella, 1999). In addition, transfers among providers commonly lead to the loss of critical information—decisions not to resuscitate, information about proxy decision-makers, and even information about past medical history. Each of these increases the risk of serious adverse events.

Location of care for provider convenience, rather than efficiency or family support

Patients in these vulnerable populations are often admitted to the hospital for lack of a better place to put them. In fact, they are often inappropriately transferred from their homes or nursing homes into hospitals despite the availability of an alternate setting for their care (Saliba, 2000). Only in hospice is care routinely available from the same team over time, with 24-hour on-call availability, and with continuity across transfers between home and inpatient care. Hospice is used by only about one-fifth of Medicare patients, for only about three weeks on average. Except for hospice care, errors arising from transfers and miscommunication are commonplace (Loxterkamp, 1996) and room for errors increases dramatically.

Lack of focus on advance care planning

Most Americans still do not receive the information that they need to plan ahead for their deaths and to do the things that make dying more likely to reflect the patient's preferences and priorities—wills, advance directives, goodbyes, etc. In particular, many are not given the opportunity to specify the approach to treatment (including types of medication, use of life-sustaining devices, and other medical interventions) that they would like to be used or avoided when they reach later stages of their illness. This lack of clarity increases the possibility that they will receive unwanted and inappropriate medications or procedures.

Top of Page


The services that are provided for people with serious, eventually fatal, chronic illnesses are systematically maladapted to their needs. Even though practitioners aim to relieve symptoms, to ensure continuity, to avoid disruption, and to make plans, they work in systems that do not support good practices. Even when good evidence instructs us as to the merits of alternatives, few adopt those alternatives. Reforms that ensure reliable and safe care will require articulated aims, reinforced intentions to make improvement, innovations, and measurement for feedback and guidance.

One example of measurement to guide change is the Assessing Care of Vulnerable Elders (ACOVE) project, which has created a quality-of-care assessment system for ill older persons (Sloss et al., 2000). This system started with identifying older people at risk of death or health deterioration, then used geriatrics experts to define guidelines for quality-of-care. Conditions and issues addressed include congestive heart failure, continuity and coordination of care, dementia, depression, and end-of-life care. The guidelines focused upon the quality of care for persons not in hospitals, including nursing homes, and sought to create the structure of a system to apply evidence-based quality indicators differentially, based on patient preference and health state. Identifying errors by applying ACOVE criteria would be an example of a way to assess improvements and compare performance in patient safety for these vulnerable elderly.

Top of Page

Research Questions

In order to affect the health care system substantially, a research agenda on preventing medical errors and increasing patient safety must include a focus on gathering data describing the needs of the most vulnerable patients. Insofar as these burdens disproportionately affect the frail and dying population, they will affect all of us eventually, first as caregivers and later as patients. Furthermore, because of the burgeoning elderly population, these errors will become much more prevalent and costly over the next few decades, unless we act now.

As examples of the information that would illuminate how to reduce medical errors and enhance patient safety for patients at the end of life, we offer the following five research questions. Each research question is followed by a brief description of our current state of knowledge and practice. Other questions will prove to be important, but answering these would be responsive to the most pressing needs that those who aim at improvement need.

Research Question 1. What are some alternative models for the current definition of "cause of death"? When should medical errors that are part of the chain of causation that ends with death be counted as a cause of death?

Our current understanding of "the cause of death" forces us to assign a single incident or diagnosis as the necessary and sufficient cause of the timing of each person's death. For younger patients with cancer or an overwhelming heart attack, that model of attributing one cause for each death works fairly well. However, the usual person now dies past age 75 and with a complex chain of causes—multiple illnesses, multiple complications, multiple medical interventions, multiple decisions as to what treatments to pursue, and multiple errors. We have very little understanding of the contributions of various elements in the causal chain and we have even less useful language with which to explain a death as having multiple causes.

This is important in work to reduce medical errors and enhance patient safety because the motivation for change and the monitoring of progress requires a method by which the contribution of errors to death can be reliably assessed.

Here are some of the information needs that would be addressed through research in this area.

  • We need to learn about better ways to monitor the patterns of care that affect the timing and manner of death, including the contribution of medication errors and other adverse events.
  • We need a better understanding of the physiology near death. Working on the contribution of medication and other errors and their role in the course would be a window into better understanding of the usual course leading to death.
  • We need to learn more about accurate modes for predicting lifespan, and more effective ways to monitor the effects of alternative treatment patterns on lifespan among those with established serious illness.

Research Question 2. What are the magnified effects of routine or "ordinary" errors (such as improper dosage or timing of medications, incorrect medications, and medication interactions) in these patients as compared to the more robust general patient population?

The QuIC (2000) points out that "Errors may be particularly difficult to recognize because variations in an individual's response to treatment is expected…medical professionals may not recognize that a particular product or procedure may have contributed to or caused the problem because the patient is already ill." Indeed, many of us will never know that we have been the victims of "small" medical errors—errors that do not result in serious adverse effects. Most of us are relatively healthy and hardy and can withstand the assault of a small dose of the wrong medication, perhaps at the wrong time, or an unsafe combination of drugs. Older patients and those at the end of life are more likely to suffer adverse consequences (Brawn and Casleton, 1990).

The same "ordinary" mistake in timing or dose of a drug that does not result in a discernable adverse effect for a healthy 25 year-old can lead to significant morbidity or even speed death in a frail 78 year-old person who is taking a complex combination of drugs to manage congestive heart failure. Indeed, medication is more likely to cause cognitive impairment (Moore and O'Keefe, 1999) in the elderly. Therefore, we must be wary of categorizing any type of error as insignificant unless we assess the impact upon the most vulnerable. Considering the small mistakes affecting the 25 year-old to be ordinary can make us inappropriately complacent about the commonplace errors by systems and providers that also treat the frail elderly.

Here are some of the information needs that would be addressed through research in this area.

  • We need to understand the mechanisms of excess morbidity and mortality and the thresholds of injury for these very vulnerable—and most at risk—people. Limiting their exposure and impact from medical error seems likely to require the best methods of reducing incidence of error as well as steps taken to buttress physiological and social reserves.
  • We need health services research on implementation of quality patterns, as well as innovation and evaluation of service delivery patterns.
  • We need to understand how error-reporting systems might allow us to better understand the disproportionate incidence of medical errors in different types of patients.

Research Question 3. Non-treatment or inappropriate treatment of pain, depression, dyspnea, and other symptoms that are common at the end of life are medical errors under the IOM report definition. What would be the effectiveness of subjecting these errors to the same kinds of system engineering, accountability, and national goal setting that is being employed for medication errors?

Research has shown that, with appropriate medication, the symptoms associated with serious and eventually fatal illness can sometimes be prevented and can virtually always be reduced. Pain can be controlled without accepting substantial sedation in up to 90 percent of terminally ill cancer patients (Portenoy et al., 1999) and depression—if detected—can also be relieved (Block, 2000). Of course, with terminal sedation, all pain can be controlled (Quill and Byock, 2000). Sadly, we know that appropriate management techniques are not often employed (Portenoy et al., 1999), or are used only after much suffering has already been endured (Lloyd-Williams et al., 1999). Relief is too often the exception rather than the rule. Others studies have pointed to differences in management of depression among elderly individuals of different races (Blazer et al., 2000).

Here are some of the information needs that would be addressed through research in this area.

  • We need to learn how to change the behavior of physicians and other health care providers so that they recognize and treat pain, depression, and other symptoms.
  • We need to learn how we can make pain and other symptom assessment a routine part of the care of our patients, as well as create a system for assuring that appropriate responses follow diagnoses.
  • We need to develop better routine practices that ensure that symptoms are prevented, assessed and treated.

Research Question 4. How can provider organizations and public health planners arrange services in a way that will provide efficient, sustainable, reliable, and safe care for this population?

The QuIC (2000) states that, "The decentralized and fragmented nature of the healthcare delivery system (some would say "nonsystem") also contributes to unsafe conditions for patients and serves as an impediment to efforts to improve safety." No single part of our healthcare system is more decentralized and fragmented than that patchwork set of services that currently is the norm for serving those at the end of life. Yet, we often have evidence that other patterns of care greatly enhance outcomes. For example, a study of Medicaid enrollees (Gill et al., 2000) found that continuity of care delivered by a single health care provider led to a decrease in the incidence and number of emergency department visits. Good palliative care programs can cut the rate of pain, anxiety, depression, family caregiver burden, shortness of breath, and hospitalization (Jenkins et al., 2000; Lynn et al., 2000a; Lynn et al., 2000b; Du Pen et al., 1999; Doyle et al., 1998).

Although in many fields the scope and shape of the problem is fairly well known, the nature and distribution of suffering at the end of life in the U.S. has never been authoritatively measured. Public policy and improvement activities are often motivated by and evaluated by anecdotes. It seems obvious that system improvement needs a background of epidemiology—population-based measures of the quality of care and the quality of life for those with serious, eventually fatal illness. The few forays into this field thus far are quite narrow or are not in the U.S.: Measures of advance care planning in La Crosse, Wisconsin (Hammes and Rooney, 1998); measures of expectations of dying in half of a county in Connecticut (Brock and Foley, 1988); measures of dying taken from a chart review in New Hampshire (Foundation for Healthy Communities), and much more comprehensive measures of symptoms and services near death in Britain (Seale and Cartwright, 1994).

In addition, while chronic disease epidemiology has been a mainstay of public health, the focus on the phase at the end of life has not. A number of state and local public health departments have come to the Center to Improve Care of the Dying seeking help with such questions as the setting in which death occurs, the presence and utility of advance directives, the range and effectiveness of measures to prevent pain and suffering, and the quality of dying for the decedent. They need the research methods and benchmarks that the work of the QuIC could generate.

Here are some of the information needs that would be addressed through research in this area.

  • In order to widely implement best practices, we need better understanding of all of the key components of system improvement: for example, the elements and design of optimal services, the social and psychological reasons for adherence to old patterns and for readiness for change; the economic and regulatory factors that shape implementation of new patterns; and effective strategies for monitoring improvement.
  • We need to understand much more about quality improvement as a method for effective social engineering and system re-design.
  • We need to learn how to engineer continuity and error-free care when very sick patients are transferred among provider programs and settings, including encouraging regional efforts to achieve reliability and high quality care.

Research Question 5. Should emergency hospitalizations of individuals at the end of life be considered a potentially preventable medical error and a significant risk factor for adverse events?

Hospitalization has been associated with a decline of health in older individuals, regardless of their baseline health status. A study of 60 functionally independent patients aged 75 and older found that, upon discharge, 75 percent were no longer independent (Lamont et al., 1983). In their study of factors associated with six-month mortality in 6784 patients aged 65 and older in the community long-term care setting, Fried et al. (1998) found hospitalization to be a predictor of death. Creditor (1993) reviewed the potential primary effects of hospitalization on the aged including deconditioning, falls, dizziness, increased fracture risk, pressure sores, and functional incontinence. He noted that because of these effects, "A high percentage of hospitalized elderly persons discharged to nursing homes never return to their homes or community."

Particularly for patients at the end of life—many of whom are aged—the hospital can be a very hazardous place. Most hospital situations are better suited to acute, curative care, with patients who can withstand long periods on hard beds, or long waits in strange settings. Very sick patients, in contrast, do best with diagnosis and treatment where they live if at all possible. Good self-care, early intervention, and comprehensive advance care planning are probably quite effective in reducing the risk of unplanned hospitalizations in this population. We need innovation, evaluation, and research to learn whether the systemic breakdowns and inappropriate decision making that lead to unplanned hospitalizations in this population can be remedied.

In the Breakthrough Series Collaboratives led by RAND's Center to Improve Care of the Dying, the Institute for Healthcare Improvement, and the Veterans Healthcare System, many provider groups reduced unnecessary hospitalizations through the use of innovative quality improvement methods. For example, nearly all of the programs that worked with heart and lung failure patients measurably cut their rate of hospitalization. One program had trained nurses available to come to the patient's home within half an hour to help with an exacerbation of symptoms. That team cut emergency room care to nearly one-tenth of the background rate (Lynn et al., 2000a). Teams that were able to reduce hospitalization rates were able to enhance the use of home care and hospice as well as build the expertise of nursing home staff in the area of palliative care.

Measures being implemented for making patient care safe during hospitalization, of course, could also be applied to non-hospital settings of care. Perhaps medication ordering systems can be deployed. Perhaps also, service delivery will have to learn to use these non-hospital settings in order to preserve the patient's daily patterns, keep the patient walking and capable of self-care, and make possible other good outcomes for patient and family.

Here are some of the information needs that would be addressed through research in this area:

  • We need to learn whether gains from evidence-based practices such as those described above are generalizable, sustainable, and replicable.
  • We need to better understand the differences in quality of life and survival outcomes between patients admitted to hospitals versus those of patients treated where they live.
Top of Page

Summary and Action Steps

Based on the information needs described above, we suggest that the QuIC include a number of relevant items on its action agenda. These items fit well with the QuIC's Compendium of Action Issues, including: research planning, learning from errors, raising the standards for health care organizations, and building purchasers' awareness of the problem. However, they add some specificity to the QuIC's points in showing how the needs of those with serious, eventually fatal chronic illness could be incorporated into an action agenda.

  1. The QuIC and its affiliated agencies should make individuals with progressive, eventually fatal, chronic illnesses a special, named, target population in initiatives to improve patient safety and reduce medical errors.
  2. AHRQ should include a focus upon these vulnerable patients in its research agenda on preventing medical errors and enhancing patient safety.
  3. The QuIC and its affiliated agencies should ensure that the merits of all proposed patient safety initiatives, including applications and measures, are specifically assessed in terms of their effects on this population.
  4. The Veterans Administration, the National Institutes of Health, the Health Resources and Services Agency, the Health Care Financing Administration, the Department of Defense and other clinical services providers or regulators should sponsor innovations in service delivery that offer the potential to engineer better practices into routine system functioning, in such areas as computer-based drug ordering, ensuring continuity, tailoring care plan to preferences, and integrating services across provider types. Once proven, they should also learn to implement evidence-based improvements system-wide.
  5. The National Institutes of Health should sponsor basic clinical research into the effects of medication errors on severely ill individuals.
  6. The patient safety practices and patient safety measurements being developed by the Quality Forum should include a set of measures and practices specifically for serving these particularly high-risk patients, modeled on the insights from ACOVE and similar projects.

Patients at the end of life are particularly vulnerable to medication errors and other lapses in patient safety. Now that we know that adequate levels of comfort and quality of life can often be maintained and even enhanced in the days and months that precede death, we would be remiss if we failed to correct the systemic issues that increase the impact of medical errors on this population. These errors can make the end of life for these patients a more distressing time and can hasten death. Attending to making the care system safe for the very sick is justified by the disproportionate burden on this population, and the opportunity to do substantial good by making care reliable for the population that most needs reliable care.

Improving care at the end of life is also in each of our best interests.Each of us faces being old, frail, and chronically ill in the months before death. We need a care system we can count on to care for us well and to reduce—rather than increase the threat of unnecessary complications.

Top of Page


AARP. Survey on prescription drug issues and usage among Americans aged 50 and older. May 1996. Online document: Accessed 8/29/00.

Agency for Healthcare Research and Quality. Acute Pain Management: Operative or Medical Procedures and Trauma: Clinical Practice Guideline No. 1. Online document: Accessed 8/31/00.

Agency for Healthcare Research and Quality. 20 Tips to Help Prevent Medical Errors. Patient Fact Sheet. AHRQ Publication No. 00-PO38, February 2000. Online document: Accessed 8/26/00.

Blazer D.G., Hybels C.F., Simonsek E.M., and Hanlon J.T. American Journal of Psychiatry 2000;157:1089-94.

Block S.D., for the ACP-ASIM End-of-Life Care Consensus Panel. Assessing and Managing Depression in the Terminally Ill Patient. Ann Intern Med 2000; 132:209-18.

Brawn L.A., Castleden C.M. Adverse drug reactions. An overview of special considerations in the management of the elderly patient. Drug Saf 1990; 5:421-35.

Breitbart W. Suicide in cancer patients. Oncology 1987;1:49-55.

Breitbart W., Passik S.D. Psychological and psychiatric interventions in pain control. In Doyle D., Hanks G.W.C., MacDonald N., editors. The Oxford Textbook of Palliative Medicine, Second Edition. Oxford: Oxford University Press; 1998.

Brock D.B., Foley, D.J. Health status and trends in the last year of life. Proceedings of the Social Sciences Section of the American Statistical Association. 1988.

Caranasos G.J., May F.E., Stewart R.B., Cluff L.E. Drug-associated deaths of medical inpatients. Arch Intern Med 1976; 136:872-5

Corcoran M. Polypharmacy in the older patient with cancer. Cancer Control 1997; 4:419-28.

Creditor M.C. Hazards of hospitalization for the elderly. Annals Intern Med 1993; 118:219-23.

Doyle D., Hanks G.W.C., MacDonald N., editors. Oxford Textbook of Palliative Medicine, Second Edition. Oxford: Oxford University Press; 1998.

Dudgeon D.J., Rosenthal S. Management of dyspnea and cough in patients with cancer. Hematol Oncol Clin North Am 1996;10:157-71.

Du Pen S.L., Du Pen A.R., Polissar N., Hansberry J., Kraybill B.M., Stillman M., Panke J., Everly R., Syrjala K. Implementing guidelines for cancer pain management: results of a randomized controlled clinical trial. J Clin Oncol 1999; 17:361-70.

Federal Interagency Forum on Aging Related Statistics. Older Americans: Key indicators of well being 2000. Online document: Accessed 8/29/00.

Foundation for Healthy Communities. End of life care project. Online Document: Accessed 8/31/00.

Fox, D. Power and illness: the failure and future of American health policy. Berkeley: University of California Press; 1993

Hammes B.J., Rooney, B.L. Death and end-of-life planning in one Midwestern community. Arch Intern Med 1998; 158:38390.

Hogan C., Lynn J., Gabel J., Lunney J., O'Mara A., Wilkinson A. Medicare beneficiaries' costs and use of care in the last year of life. Final Report to the Medicare Payment Advisory Commission. May 2000.

Inouye S.K, Bogardis S.T., Charpentier P.A., Leo-Summers L., Acampora D., Holford T.R., Cooney L.M. A multicomponent intervention to prevent delerium in hospitalized older patients. NEJM 1999; 340:669-76.

Jenkins C.A., Schulz M., Hanson J., Bruera E. Demographic, symptom, and medication profiles of cancer patients seen by a palliative care consult team in a tertiary referral hospital. J Pain Symptom Manage 2000; 19:174-84.

Kohn L.T., Corrigan J.M, Donaldson M.S., editors. To err is human: building a safer health system. Institute of Medicine. Washington: National Academy Press, 2000.

Lamont C.T., Sampson S., Matthais R., Kane R. The outcome of hospitalization for acute illness in the elderly. J Am Geriatr Soc 1983; 31:282-8.

Leape L.L., Kabcenell A.I., Gandhi T.K., Carver P., Nolan T.W., Berwick D.M. Reducing adverse drug events: lessons from a Breakthrough Series Collaborative. 2000; 26:321-31.

Leff B., Burton L., Guido S., Greenough W.B., Steinwachs D., Burton J.R. Home hospital program: a pilot study. J Am Geriatr Soc 1999; 47:697-702.

Lloyd-Williams M., Friedman T., Rudd N. A survey of antidepressant prescribing in the terminally ill. Palliat Med 1999; 13:243-8.

Loxterkamp D. Hearing voices—how should doctors respond to their calling? New Engl J Med 1996; 335:1991-3.

Lynn J., Harrell F., Cohn F., Wagner D., Connors A. Prognoses of seriously ill hospitalized patients on the days before death: implications for patient care and public policy. NewHoriz 1997; 5:56-61.

Lynn J. and Harrold J. Handbook for mortals: guidance for people facing serious illness. New York: Oxford University Press; 1999.

Lynn J., Schall M., Milne C., Nolan K., Kabcenell A. Quality improvements in end of life care: insights from two collaboratives. Joint Commission Journal on Quality Improvement 2000; 26:254-67.

Lynn J., Schuster J.L., The Center to Improve Care of the Dying, and The Institute for Healthcare Improvement. Improving Care for the End of Life: A Sourcebook for Clinicians and Managers. New York: Oxford University Press; 2000; [In press].

McDonald, C.J., Weiner, M., Hui, S.L. Deaths Due to Medical Errors Are Exaggerated in Institute of Medicine Report. JAMA 2000; 284:93-5.

Moore A.R., O'Keeffe S.T. Drug-induced cognitive impairment in the elderly. Drugs Aging 1999; 15:15-28.

Portenoy R.K., Thaler H.T., Kornblith A.B., Lepore J.M., Friedlander-Klar H., Coyle N., et al. Symptom prevalence, characteristics and distress in a cancer population. Qual Life Res 1994; 3:183-9.

Portenoy R.K. and Lesage P. Management of cancer pain. Lancet 1999; 353:1695-700.

Quill T.E. and Byock I.R. for the ACP-ASIM End-of-Life Care Consensus Panel  Responding to intractable terminal suffering: the role of terminal sedation and voluntary refusal of food and fluids.  Ann Intern Med 2000; 132:408-14.

Rich M.W., Vinson J.M., Sperry J.C., Shah A.S., Spinner L.R., Chung M.K., Davila-Roman,V. Prevention of readmission in elderly patients with congestive heart failure: results of a prospective, randomized pilot study. J Gen Intern Med 1996; 8: 585-90.

Saliba D., Kington R., Buchanan J., Bell R., Wang M., Lee M., et al. Appropriateness of the decision to transfer nursing facility residents to the hospital. J Am Geriatr Soc 2000; 48:154-63.

Seale C. and Cartwright, A. The year before death. London: Ashgate Publishing Limited; 1994.

Shelton P.S., Fritsch M.A., Scott M. Assessing medication appropriateness in the elderly: a review of available measures. Drugs and Aging 2000; 16:437-50.

Sloss E.M., Solomon D.H., Shekelle P.G., Young R.T., Saliba D., MacLean C.H., Rubenstein L.Z., Schnelle J.F., Kamberg C.J., Wenger N.S. Selecting target conditions for quality of care improvement in vulnerable older adults. J Am Geriatr Soc 2000; 48:363-9.

The SUPPORT Principal Investigators. A 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; 274:1591-98.

Tolle S.W., TildenV.P., HickmanS.E., RosenfeldA.G. Family reports of pain in dying hospitalized patients: a structured telephone survey. West J Med 2000; 172:374-7.

Ventafridda V., Ripamonti C., Tamburini M., Cassileth R.B., De Conno F. Symptom prevalence and control during cancer patients' last days of life. J Palliat Care 1990; 6:7-11.

Won A., Lapane K., Gambassi G., Bernabei R., Mor V., Lipsitz L.A. Correlates and management of nonmalignant pain in the nursing home. SAGE Study Group. Systematic assessment of geriatric drug use via epidemiology. J Am Geriatr Soc 1999; 47:936-42.

Zeppetella G. How do terminally ill patients at home take their medication? Palliat Med 1999; 13:469-75

Current as of September 2000

Return to Testimony Directory
National Summit on Medical Errors and Patient Safety
QuIC Home Page
Department of Health and Human Services

The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care