End-of-Life Issues

End-of-Life Issues

NOTE: The UW Dept. of Bioethics & Humanities is in the process of updating all Ethics in Medicine articles for attentiveness to the issues of equity, diversity, and inclusion.  Please check back soon for updates!







Topics addressed:

  • What is a "good death"? A medical perspective
  • What goals should I have in mind when working towards a decent death for my patient?
  • How do you know when someone is dying?
  • What should I know about the hospice approach?
  • What you need to understand to care for the dying
  • How do physicians who care for the dying deal with their own feelings?

As care of the dying involves so much of one's self, in this topic page I will describe my approach as an one example of how clinicians think about end-of-life care. I remember, the first time one of my patients died, feeling a chill of horror and fascination. I wasn't prepared for it. The resident yawned--a long night, then a long code. "We better go talk to the family." What in the world would we say? The dead patient, now dusky blue, looked unreal and unfamiliar. I was so wrapped up in my own feelings that I can't recall much else.

Now I find care of the dying to be one of the richest parts of my clinical life. But it is demanding in a different, more personal way, than, say, treating pneumococcal pneumonia with penicillin. Here I will describe some ways of thinking about care of the dying that have helped me figure out where I am going as I guide someone who is really sick.

Many medical students first encounter care of the dying as an unsuccessful code or a strategic withholding of CPR. Of course, an ethically sound understanding of withdrawing and withholding treatment is crucial to good care of the dying. Yet "withholding and withdrawing" only describe what we, as clinicians, decide not to do. To provide excellent care of the dying requires that we also decide what we should do. What should be the goals of medical care for people who are dying? What makes a good death?

What is a "good death"? A medical perspective

The good death is not a familiar idea in American culture. Some experts in palliative care describe the United States as a "death-defying" culture, with a mass media that spotlights only youth and beauty. Yet public interest in care of the dying is currently high. The striking public interest in physician aid-in-dying is one obvious reason. But there are other reasons: over the past 100 years, there has been a epidemiologic shift in the reasons people die. In the pre-antibiotic era, people most often died young, of infectious diseases; now, thanks to medical technology, most Americans (and others with access to this technology) live much longer, to die of degenerative, neoplastic, and even man-made diseases. Finally, there is a marked public fear that a medical death, depicted in TV shows like "ER" as an unresponsive, uncommunicative body hooked up to an array of flashing monitors, represents an irresponsible use of technology and a dishonorable way to treat a person.

Interestingly, contemporary medical literature contains little that might characterize what makes a death "good." Recently, a large, expensive empirical study of intensive care unit deaths suggested that medical care for a common type of in-hospital death is "bad" (the SUPPORT study, referenced below). In this study of dying patients, severe pain was common, decisions to withhold invasive treatments were made at the last minute, and physicians often had no knowledge of patient preferences not to have CPR. Even worse, an intervention designed to provide physicians with better prognostic information had no effect on medical decision making prior to death. While the SUPPORT authors did not actually describe these deaths as "bad," we could certainly agree that they were not "good deaths."

In caring for a person who is dying, knowing what would make the experience of dying "good" is an important goal for physicians and other members of the care team. I find it doesn't take fancy techniques-you just need to be sincere and patient and interested. Listen more and talk less. Try asking something like, "Knowing that all of us have to think about dying at some point, what would be a good death for you?" What people choose when they think about a good death for themselves is often beyond what medicine can provide-for instance, an affirmation of love, a completion of important work, or a last visit with an important person. As a physician, I can't always make those things happen. But I can help the dying person get ready-and in this way, contribute to a death that is decent.

What goals should I have in mind when working towards a decent death for my patient?

I have several working clinical goals when I am caring for someone near the end of life. I work towards:

  • Control of pain and other physical symptoms. The physical aspects of care are a prerequisite for everything that follows.
  • Involvement of people important to the patient. Death is not usually an individual experience; it occurs within a social context of family, significant others, friends, and caregivers.
  • A degree of acceptance by the patient. Acceptance doesn't mean that the patient likes what is going on, and it doesn't mean that a patient has no hopes--it just means that he can be realistic about the situation.
  • A medical understanding of the patient's disease. Most patients, families, and caregivers come to physicians in order to learn something about what is happening medically, and it is important to recognize their need for information.
  • A process of care that guides patient understanding and decision making. One great physician does not equal great care--it takes a coordinated system of providers.

How do you know when someone is dying?

This question is not as simple as it might sound. The SUPPORT study demonstrated that even for patients with a high probability of dying, it is still difficult for a clinician to predict that a particular patient is about to die. Thus it may be more useful for clinicians to give up relying on their predictive skills, and look at the common clinical paths (or trajectories) taken by dying patients, and design medical care that includes "contingency plans" for clinical problems that a person with incurable lung cancer (for example) is likely to experience. Such contingency plans might include advance directives and perhaps DNAR orders, as well as lines such as: "You will probably die from this, although we can't predict exactly when. What is really important for you in the time you have left?"

What should I know about the hospice approach?

In order to help someone towards a decent, or even good, death, the hospice framework is very helpful. Hospice started as a grassroots effort, as a view of dying that lets go of the possibility of cure. Instead, hospices emphasize symptom control and attention to psychological and spiritual issues. Pathophysiology becomes less important and personal meaning becomes more important. Thus this framework analyzes a person's medical care into four major topics, and this can be used to outline day-to-day care plans for a patient:

  • Pain - one of the things most feared by patients with life-threatening illness.
  • Symptom control - including dyspnea, nausea, confusion, delirium, skin problems, and oral care.
  • Psychological issues - especially depression, sadness, anxiety, fear, loneliness.
  • Spiritual or existential issues - including religious or non-religious beliefs about the nature of existence, the possibility of some type of afterlife.

Hospice care in Washington State is most often provided by multidisciplinary teams who go to patients' homes. This care is covered by Medicaid for patients judged to have less than six months to live. Hospice care is generally underutilized, and even though most hospice teams feel that at least six weeks of hospice care is optimal, most patients receive much less because they are either referred very late or have not wanted hospice. A major problem in connecting hospice care to acute medical care is that referral implies a "switch" from curative to palliative medicine-a model that does not fit comfortably in many illnesses.

What you need to understand to care for the dying

Another useful framework was outlined by Joanne Lynn, who was one of the principal investigators of SUPPORT. She suggests that there are four things clinicians must know to care for the dying:

  1. The patient's story - including how that person has viewed her life, the other persons important to her, and how she could bring her life to a close in a way that would be true to herself.
  2. The body - which covers the biomedical understanding of disease, and what limits and possibilities exist for that person.
  3. The medical care system available for this particular patient - knowing how you can make the system work for the patient, as well as the relevant law and ethics.
  4. Finally, you must understand yourself - because you, as a physician, can be an instrument of healing, or an instrument that does damage.

Obviously, learning how to do all this is beyond the scope of this web page--these are goals that guide a career of learning and reflection. But this framework provides guidelines for you as you develop your own approach to caring for dying patients.

How do physicians who care for the dying deal with their own feelings?

It is not hard to find physicians who are burned out - ask any nurse. What is difficult is to find for yourself a type of self-care that will enable you to develop your gifts as a physician, and continue to use them in practice. It helps to learn your strengths and weaknesses, and to actively seek whatever will nurture you - in or out of medicine. A strategy of detachment may not serve you well in the long run. There are indeed rewards for physicians who care for the dying, but as a Zen master once observed of a bingo game, "you must be present to win."

  • Lynn J. An 88-year old woman facing the end of life. JAMA 1997;277: 1633-1640.
  • Brody H, Campbell ML, Faber-Langendoen K, Ogle KS. Withdrawing intensive life-sustaining treatment--recommendations for compassionate clinical management. NEJM 1997;336:652-657.
  • Quill TE, Byock IR. Responding to intractable terminal suffering: the role of terminal sedation and voluntary refusal of food and fluids. ACP-ASIM End-of-Life Care Consensus Panel. American College of Physicians-American Society of Internal Medicine. Annals of Internal Medicine. 2000 Mar 7;132(5):408-14.
  • Arnold RM, Kellum J. Moral justifications for surrogate decision making in the intensive care unit: implications and limitations. Critical Care Medicine. 2003 May;31(5 Suppl):S347-53.
  • Cimino JE.  A clinician's understanding of ethics in palliative care: an American perspective.Critical Reviews in Oncology/Hematology. 2003 Apr;46(1):17-24. 


Skip is a 50-year-old man with metastatic nonsmall cell lung cancer. He decided to try palliative chemotherapy because "otherwise I might just as well roll over and give up." After the first cycle of carboplatin and taxol, he requires hospitalization for fever and neutropenia (a complication of the chemotherapy). You stop by for a visit, and he says he feels terrible, wonders "if the chemo is worth all this", but that he's too scared to stop.

Case Discussion

For metastatic non small cell lung cancer, palliative chemotherapy is an intervention providing, on average, a small benefit at considerable toxicity (a consideration for the Medical Indications box in a Clinical Ethics 4-box analysis). Yet for a patient who is well informed, understands the benefits and burdens, and wishes to proceed, a trial of palliative chemotherapy is justified. However, now Skip is voicing concern: the most important thing to do is hear him out. Find out what he is worried about, how he rates his quality of life, and what his goals are. This information will help you sort out what is going through his mind and help you guide him to a decision that will be the best for him.

As Skip thinks through his situation, ask him if he wants you to describe what would happen if he decides to have more chemotherapy, or stops his chemo and starts hospice care. Eventually you might ask him what a good death would be for him--he may not be able to answer immediately, but it might help him (and you) shape a care plan later. When you talk with Skip, keep in mind the goals for a decent death.

Angela is a 72-year-old woman with end stage congestive heart failure from coronary artery disease--she has had two myocardial infarctions. When her medical management is optimal, she is just able to take care of herself in her own apartment, but with any small decompensation, she ends up in the hospital. She comes in for a clinic visit, and her weight is up 2 kilograms and she is complaining of paroxysmal nocturnal dyspnea, even though she has been taking her meds as prescribed. Exasperated and discouraged, she asks, "Am I dying?". The cardiologist replies: "Well, no--this is all reversible."

What would you say?

Case Discussion

The SUPPORT study has shown us that the clinical course of dying from congestive heart failure is quite different from dying of lung cancer. Patients with lung cancer begin a visible, predictable decline several weeks before death that usually evident to experienced clinicians. Patients with congestive heart failure, however, experience periods of fairly good function alternating with decompensation right up until death, and the terminal event for these patients is often sudden. This pattern of decline is not usually labeled by patients or physicians as "dying." The unpredictable course has resulted in very few hospice referrals for patients with end-stage congestive heart failure.

The best care plan in this situation would be based on a discussion with Angela about what kinds of contingency plans should be in place if she has a severe, possibly fatal decompensation (see Advance Care Planning). Some medical centers are developing Palliative Care or Comfort Care services to try to better match the needs of patients with less predictable end-stage illnesses.