Skip to main content

Advanced Search

Advanced Search

Current Filters

Filter your query

Publication Types



Newsletter Article


Improving Care at the End of Life

Summary: As patients live longer, many states, community-based coalitions, and health care providers have begun to focus on the quality—and quantity—of medical care provided at the end of life. The resulting programs have provided physicians with techniques for delivering bad news, managing transitions to palliative care, and handling requests for therapies that are likely to be futile. They’ve also helped to elicit patient preferences, leading to lower utilization in some locations.

By Martha Hostetter and Sarah Klein

While medical advances have enabled patients to recover from once incurable diseases, they also have extended the process of dying, making it at times a protracted period of unraveling in which body, mind, and spirit give way, in no particular order. As more terminally ill patients live longer, many states, community-based coalitions, and health care providers have begun to focus on the quality—and quantity—of medical care provided at the end of life. 

California, Pennsylvania, and West Virginia have launched initiatives or funded organizations that focus on improving communication between doctors and patients about prognosis and treatments. New York has gone a step further, requiring physicians to discuss palliative care and end-of-life options with terminally ill patients.1 And most states now have in place or are developing programs to support the use of Physician Orders for Life-Sustaining Treatment (POLST), a standardized medical form that specifies the types of life-sustaining treatments that a patient does or does not want.2

These efforts have helped curtail the use of medical interventions that are not in accord with patients' wishes or offer little benefit and ensure patients are aware of the range of hospice and palliative care options. Yet there remain significant gaps between the end-of-life care most Americans say they want and the treatments they receive. For example, most Americans say they'd prefer to die at home, but 70 to 80 percent die in hospitals or other institutions.3 And while medical treatment during the last year of life accounts for a large proportion of Medicare spending as well as overall health care spending, that care is in many cases not evidence-based nor in accord with patients' desires.4

Better Communication
Part of the reason efforts to improve end-of-life care have not gained wider traction is that providers are often ill equipped to deal with the confusion, distress, anger, or other emotions that accompany discussions about illness and death. Taking cues from patients or their family members, they may accede to what's been called a "dance of denial" about a poor prognosis, until an emergency situation, such as a need for mechanical ventilation, brings the issue to the fore.

To address this, several medical schools and continuing medical education programs have developed training programs that focus on communicating the nuances of end-of-life care. The key to such programs is role-playing, says James Tulsky, M.D., director of Duke University School of Medicine's Center for Palliative Care. "The only way you learn these skills is by practicing them and getting feedback," he says. Tulsky encourages his students to use the mnemonic "SPIKES" as a roadmap for delivering serious news (Exhibit 1).

 Exhibit 1

 SET up the conversation at the right time and place
 Assess the patient's PERCEPTION of his or her condition
 Extend an INVITATION to have the conversation
 Deliver KNOWLEDGE about the condition
 Allow time for patients to EXPRESS their emotions and EXTEND empathy
 And SUMMARIZE and STRATEGIZE about next steps.

Source: Duke University School of Medicine's Center for Palliative Care

Researchers at the University of Washington, the University of Pittsburgh, and Duke University created Oncotalk, a program that has trained between 15 and 20 percent of the nation's oncology fellows over a five-year period in ways to communicate bad news, manage transitions to palliative care, and handle requests for therapies that are likely to be futile. The program involved intensive workshops, with simulated encounters between oncology fellows and actors portraying cancer patients going through various stages of treatment. Because this approach can be costly, the researchers are piloting a Web-based training tool that enables learners to record their encounters with actual patients and receive feedback on their approaches. (Patients provide informed consent before being recorded.)

In addition, videos are used to facilitate conversations between patients and their physicians about various end-of-life interventions. Angelo Volandes, M.D., M.P.H., an assistant professor at Harvard Medical School and an internist practicing at Massachusetts General Hospital, and his colleagues have created a series of videos to capture the care experiences of patients with advanced dementia, cancer, and heart disease, among many other conditions (see related story.)

Other efforts focus on helping patients and their families have these conversations. Earlier this month, the Institute for Healthcare Improvement launched The Conversation Project, a national campaign to encourage Americans to discuss end-of-life preferences with their loved ones. Its goal is to create a cultural movement akin to the birth movement in the 1970s that brought doulas and dads into the delivery room. In similar fashion, Caring Connections, a project of the National Hospice and Palliative Care Organization, encourages people to plan for their medical care before they get sick. One way to do so is through an advanced directive, or living will, which documents patients' desires for treatment at the end of life and designates a health care proxy. However, critics of advanced directives say that patients' instructions for future care have been difficult to translate into actual medical decision making, and that designated health care proxies may not know patients' wishes or may be unwilling to respect them.5

Some experts think POLST forms have greater potential since they are more specific and more closely linked to patients' actual experiences. But others caution that no single form can substitute for thoughtful conversations between providers and patients. "My bias is that there is too much emphasis on forms and checklists that don't really get at people's individuality," says Jack Freer, M.D., professor of medicine at the University at Buffalo. "The pendulum may have swung too far in the direction of autonomous decision making, leaving sick, vulnerable patients in the position of having to make complex medical decisions with inadequate medical guidance." Instead, Freer advocates a shared decision-making approach in which physicians make recommendations based on what patients say is important to them. "The best thing we can do is to sit back and listen," he says. "Then it's your job to say to the patient, 'How can we take the reality of the situation plus what you've told me that's important to you to craft a plan of care that will give you the best shot of achieving your goals? One approach I might suggest is...', and go on to describe what a coherent, comprehensive palliative plan of care would look like, and how it would meet the patient's goals of care."

Better Prognostication
Experts say another key to improving end-of-life care is to develop physicians' skills at prognostication—the ability to predict the likely progression of a disease, based on patients' clinical parameters and the evidence base. This is important because physicians may overestimate their healing powers and have a bias toward optimistic outcomes, even in the face of a patient's steady decline.6

At Sharp HealthCare system, based in San Diego, the hospice and palliative care team teaches physicians how to use "evidence-based prognostication" to help them decide when it might be time to introduce palliative care to manage symptoms. Such decisions are complicated by the gray areas in medicine, the lack of reliable outcomes data for some serious conditions, and the difficulty of applying population data to particular patients. "If you want to talk about prognosis," says Tulsky, "it's often most accurate to present ranges—days rather than weeks, months rather than years."

Palliative Care as Part of the Continuum
Better communication and prognostication lay the groundwork for key junctures in end-of-life care, including decisions to adopt palliative treatments or make referrals for hospice services.

Today, leading-edge health systems are integrating palliative approaches into the continuum of care for seriously ill patients. Palliative care professionals work alongside other clinicians to focus on patients' symptom management and quality of life—even while medical interventions to help prolong lives are ongoing. "Some people think to use palliative care services you must forgo curative care, but this is not true," says Don Schumacher, Psy.D., president and CEO of the National Hospice and Palliative Care Organization. Palliative care professionals also help patients clarify their goals for treatment—such as avoiding a feeding tube—as well as their goals for the end of life, such as spending time with friends and family members or reaching a sense of closure.

At Sharp HealthCare system, hospice care has been reimagined as "end-stage disease management," with the goal to extend the proven benefits of hospice to patients earlier in the trajectory of their disease. The program, which involves patient and caregiver education, proactive medical management, and advance health care planning, has been shown to reduce emergency department visits and hospitalizations among heart failure patients.7 Perhaps more surprising, it has also extended patients' lives.

St. Mary's Health Care System in Athens, Ga., takes a similar approach, using palliative care to improve care coordination. A nurse and social worker identify hospital patients with complex chronic conditions, such as heart failure or chronic obstructive pulmonary disease, who may be at risk of readmission. (Some are referred by hospitalists, others through a screening that occurs at admission.) Staff assess patients' immediate and ongoing needs—which may require pain management or social services––and then work closely with the patient and family to develop goals and implement a plan of care.

The program has reduced lengths of stay and readmissions. The utilization of tests and procedures often declines as well once patients fully understand their options and express their preferences, says Tanya Adcock, R.N., M.S.N., St. Mary's palliative care manager.

Sharp has seen reductions in readmissions as well. "We were seeing patients coming to hospice, particularly heart failure patients, who had been admitted to the hospital maybe 15 times during the past 12 months—clearly that was not good-quality care," says Suzi Johnson, M.P.H., R.N., vice president for Sharp's Hospice and Palliative Care. "We took our hospice expertise and interdisciplinary care team to catch patients upstream." Another benefit of the program is that, by building trust among patients and their providers, patients are often ready to accept referrals to hospice when that time comes, Johnson says.

Spreading Improvement Models
Innovative programs like this may be leading change, but it will take more for widespread progress to be made in improving end-of-life care. For that to happen, Lachlan Forrow, M.D., director of the ethics and palliative care programs at Beth Israel Deaconess Medical Center in Boston and chair of Massachusetts' Expert Panel on End-of-Life Care, says that end-of-life care needs to be defined as a core quality issue and approached through traditional quality improvement methods.

For example, health systems need to be clear about the expected results of conversations about end-of-life care, including clear designation of a health care proxy and statement of goals of care, Forrow says. He would also like to see mandatory surveys six weeks after a patient's death to ask their next of kin whether their loved one was cared for in the way he or she would have wanted, and whether he or she had the information needed to decide how they wanted to be cared for.

Others groups are hoping to promote widespread improvement in hospice care through collaborative learning approaches. The Coalition of Hospices Organized to Investigate Comparative Effectiveness (CHOICE), a network of hospice organizations from around the country, is combining data in a shared electronic platform that can be used to identify and promote best practices in hospice care.

Delivery system trends toward accountable care and global payments may encourage attention to end-of-life care, says Judy Citko, J.D., executive director of Coalition for Compassionate Care of California, a nonprofit that promotes high-quality care for people who are seriously ill. "Health systems are more interested in these issues than they used to be," she says, "and we have an opportunity to institutionalize advance care planning within the health care system." 

For example, organizations pursuing accountable care organizations will be looking to identify their high-risk populations and better manage their care and control costs; they may also want to set up systems to introduce palliative care and document consultations around end-of-life preferences.

It doesn't take long to document savings, says Flint Besecker, CEO of the Center for Hospice & Palliative Care in Cheektowaga, N.Y., which has hired 14 physicians and 10 midlevel providers to deliver palliative care services in hospitals, physician groups, nursing homes, and in patients' homes. "Most people go into the emergency department for symptom management. If you can manage that at home, it doesn't take a lot of hospital avoidances to have a big impact," he says.

Momentum for integrating palliative care earlier into the continuum initially came from palliative care professionals, says Linda Emanuel, M.D., Ph.D., director of Northwestern University's Buehler Center on Aging, Health & Society. "Then people took on the serious and politically charged issue as to whether palliative care could improve costs," she said. "After all the political debates, we have to measure the appropriateness of resource utilization by the match with patients' and families' goals for care. However, knowing what we want is not obvious. Most of us only face our end once—learning how to help someone achieve their goals for end of life is one of the main areas that still need improvement."

Improving End-of-Life Care: Selected Resources

  • Promoting Excellence in End-of-Life – clinical care tools, evaluations, and other resources for improving end-of-life care from the Robert Wood Johnson Foundation
  • The Art of Medicine Near the End of – information and tools for physicians, caregivers, and patients
  • Education in Palliative and End-of-Life Care (EPEC) – core curriculum for health care professionals in the essential clinical competencies of palliative care
  • – communication training program for oncologists
  • CHOICE: Coalition of Hospices Organized to Investigate Comparative – network of U.S. hospices sharing information to identify best practices
  • Patient Centered Care and Human Mortality: The Urgency of Health System Reforms to Ensure Respect for Patients' Wishes and Accountability for Excellence in – report and recommendations of the Massachusetts Expert Panel on End-of-Life Care



1 New York's Palliative Care Information Act of 2010 requires physicians to provide information to patients with a terminal diagnosis regarding palliative and end-of-life care options. The Palliative Care Access Act of 2011 requires health care facilities to establish policies and procedures regarding palliative care, including access to information and to palliative care consultations and services. 
2 The POLST form is completed by a health care professional, based on an explicit conversation with the patient, and signed by a physician and, in some states, the patient. It is typically printed on brightly colored paper and included first in patients' medical records.  
3 See; A 1999 Harvard Public Opinion Poll found that 71 percent of Americans would prefer to die at home. A 2002 Harris Interactive Poll found that 86 percent of Americans believe that people who have a terminal illness would most like to receive end-of-life care at home. 
4 According to CMS, the percentage of Medicare expenditures attributable to beneficiaries in their last year of life was 27.9 percent in 1999. See According to the Hastings Center, end-of-life care accounts for 10 percent of overall medical spending. See
5 H. S. Perkins, "Controlling Death: The False Promise of Advanced Directives," Annals of Internal Medicine, 247(1): 51–7. 
 N. A. Christakis and E. B. Lamont, "Extent and Determinants of Error in Doctors' Prognoses in Terminally Ill Patients: Prospective Cohort Study," British Medical Journal, Feb. 2000 320(7233):469–73. 


Publication Details