End-of-life care in the United States needs improvement, despite major social initiatives such as the hospice movement and advance directives beginning in the 1970s. These were intended to make dying less painful, more humane, and more in harmony with patients’ wishes. What is the problem in our society and in our health care system?

One factor is our society’s well-known denial of death, expressed, in part, by institutionalizing dying: unlike residents of many other countries, who die primarily at home, 75% of US residents die in health care institutions. For 60%, these institutions are acute care hospitals, and for most of them, death follows a decision to forego treatment.1

Another factor is that physicians and patients in this society do not often discuss patients’ wishes regarding the end of life, such as the kind of experience that patients want and the kinds of medical care that physicians can provide or orchestrate to respect patients’ wishes. A well-known, influential, multisite, multiyear study2 found that just before they died, half the 9,105 terminally ill patients studied spent at least 8 days comatose, in hospital intensive care units (ICUs), or in pain. There was little correlation between patients’ known preferences (conveyed by nurses to patients’ physicians—including preferences regarding resuscitation status)—and what physicians, in fact, did.

Two more recent studies3,4 measured the effects of clinical ethics consultation, both prospective and reactive, on hospitalized critically and terminally ill patients’ outcomes. Both found that these interventions, consisting of improved communication among ICU team members and between the health care team and patients/families, improved patient outcomes (reduced nonbeneficial treatments) and reduced pain and suffering for nonsurviving patients. Ethics consultants were somewhat more successful than nurses in facilitating physician–patient/family discussion of patients’ end-of-life wishes, thereby influencing treatment decisions.

There are limits, however, on what third-party intermediaries can do to increase the frequency of patient–physician discussions of patients’ wishes regarding end-of-life experience and care. Patients wait for their doctors to bring up the subject, and appreciate it when they do; doctors wait for patients to bring it up, and are happy to talk with and listen to their patients when the patients take the initiative.5

Until physicians and patients have these discussions more frequently and directly, without requiring intermediaries, physicians often will not know what even their sickest patients really want at the end of their lives. Recent evidence of this communication gap is a remarkable 2-year study6 of lung cancer patients (with a 5-year survival rate of less than 15%) admitted to medical ICUs for pulmonary problems and sepsis. It found that 64% of patients with the most lethal of lung cancers, stage IV non–small-cell lung cancer, had full-code resuscitation status. Moreover, for 38 of the study patients whose outpatient charts were available, a review of those outpatient records showed “no record of code discussion documented in any of these outpatient charts.”6

Discussion Barriers

Training and Tradition. Until recently, medical-school training in the United States neglected or marginalized death and dying, including palliative and hospice care at the end of life. That has changed during the past decade, as both have been integrated into mainstream medical practice. Nonetheless, most physicians still are not trained in palliative medicine, nor are they taught communication skills appropriate to discussing patients’ wishes and concerns about dying when the time comes.

Physicians are products of a long tradition that shapes them as agents of hope and discourages them from initiating end-of-life discussions, even when patients are close to death. In early Hippocratic writings, physicians were counseled: “Give necessary orders with cheerfulness and serenity, turning his attention away from what is being done to him; sometimes reprove sharply and emphatically, and sometimes comfort, revealing nothing of the patient’s future or present condition.”7

This counsel endured through the early part of the 20th century, as seen in the first American Medical Association code of ethics: “The life of a sick person can be shortened not only by the acts, but also by the words or manner of a physician. It is, therefore, a sacred duty to guard himself carefully in this respect, and to avoid all things which have a tendency to discourage the patient and depress his spirits.”7

Emotional Barriers. In the context of human fear of the unknown and of contemporary US death-denying culture, physicians and patients cannot be faulted for feeling uncomfortable in approaching open discussions of dying. Exploration of emotional responses to dying and death is not part of the curriculum in US schools generally, and medical schools in particular. There can also be a fear of losing control of the conversation, or of opening Pandora’s box, in having such a discussion with one’s patients.

The Managed Care Context. Primary care, specialist, and hospitalist physicians currently practice in a context of diminishing human and financial resources, and they may feel unrelenting third-party pressures to manage larger patient volumes and to move patients through the system more rapidly at all levels. The obstacle here is limited time. Even physicians who feel comfortable having these discussions—and despite the fact that such conversations need not be lengthy—may worry that they will be time-consuming, putting the physicians behind in an already packed schedule.

Cultural Barriers. A recent article8 in The Wall Street Journal illustrates and underscores the insistence in several cultural traditions (including Chinese and African American) that diagnoses of cancer are shameful, and end-of-life discussions should not occur. The article describes the concerted efforts of several hospitals in New York City to respect and accommodate this insistence. This complicates the challenge for physicians who believe that initiating end-of-life discussions with patients is important.

Inadequate Tools. In theory, advance directives, as written documents identifying surrogate decision makers in the event of decisional incapacity and expressing patients’ wishes to limit the use of life-sustaining treatments, could be effective tools for physicians to use in conducting end-of-life discussions with their patients. In practice, though, they are not.

Such directives feel artificial and negative to physicians and patients. They feel artificial because most of them go into effect when the patient is considered terminal or permanently unconscious, calling on physicians to make what are often very difficult prognoses. Actually, many patients move through the end of life by entering a trajectory of critical illness, with increasingly frequent symptom exacerbations, sometimes without entering a state clearly recognized as terminal.

Most patients are open to aggressive interventions on a time-limited basis, if these treatments have a reasonable chance of helping the patient recover, but they do not want to live in dependence on these aggressive and artificial measures, or to be sustained indefinitely in a coma. Therefore, advance directives should go into effect when a patient becomes critically ill, not terminal, and should provide guidance regarding the use or limitation of treatments during these exacerbation peaks in the trajectory.9

Advance directives can feel negative because most of them feel, to patients, like giving permission to withhold or stop medical interventions at the end. They do not allow, in other words, for patients to give maximal treatment a try during critical illness and then stop the treatment if it is not working.

In addition, current advance directives are often vague and ambiguous (for example, requesting that there be no heroic measures without defining them). They allow patients to name proxy decision makers without telling these proxies what they want and do not want, and they are frequently unavailable to physicians when decisions need to be made.9

A new generation of reformatted directives should allow patients to express what they do want, as well as what they do not want. They should require written evidence that the designated proxy has discussed the patient’s wishes with the patient. Health care organizations, and even health insurers, should be legally authorized to require their completion and to enter them into electronic registries for reliable availability, rather than requiring health care providers only to inform patients of their right to complete an advance directive. These changes would give physicians and patients a more useful tool when decisions must be made, and would encourage both parties to enter into discussions about patient preferences regarding end-of-life experiences and care.9

A New Study

In addition to time pressure and inadequate tools, the obstacles for physicians have been summarized by Gazelle,10 who wrote, “Typically, the patient’s or family’s end-of-life treatment preferences have not been ascertained. Reasons for this omission include physicians’ discomfort with the topic of death and dying, the belief that honest discussion of the prognosis would lead to a loss of hope, and an assumption that patients and their families should be the ones to initiate dialogue.”

An important new study11 published in JAMA presents data that its authors hope will help physicians move past these emotional obstacles. One surprise in the study was that 37% of 332 randomly chosen advanced-cancer patients had actually had end-of-life discussions with their physicians before baseline. My clinical experience, along with previous studies,6 would indicate that oncologists do not always have these discussions.

This study aims to reduce physicians’ concerns about inflicting psychological harm on patients and their families, presenting compelling evidence that cancer patients having end-of-life discussions with their physicians do not experience more depression than patients not having these discussions. Moreover, patients having these discussions experienced better quality of life and death, and their families experienced less pathological grief following the patient’s death. In other words, these discussions do not cause harm to patients or their families. They lead to better patient outcomes at the end of life, and they are associated with healthier bereavement by families following death. The authors hope that their evidence will motivate physicians to have end-of-life discussions with their patients because of evidence-based confidence that such discussions do measurable good and reduce measurable harm.

Clinical Applications

How might a physician initiate or facilitate end-of-life discussions with patients, in view of the obstacles mentioned earlier and the new evidence of the benefits of such discussions to patients and their families?

Possible Conversation

I want to have a conversation with you that I try to have with all my patients from time to time, to make sure that we’re on the same page about what you want from your health care and from me. I want to use my medical knowledge and skills to help you to the best of my ability, and to do that I need to make sure that I’m sensitive to what you want and to what’s important to you, now and in the future.

Tell me, then, how you see your medical condition now, what you hope for or are afraid of in the future, and what you would like me to keep in mind as I try to do what’s best for you. Some patients tell me, “Do whatever it takes to keep me living—surgeries, drugs, machines, or whatever—as long as there’s a chance of getting me through with a life that lets me do things and have some enjoyment. Try whatever might help, as long as it’s working, but do not keep going if it’s not working and is causing me more pain than good.” Other patients tell me, “I’m open to everything, except going onto a breathing machine, or having to depend on an artificial kidney machine.”

Where are your feelings or concerns along those lines? If it comes to making choices between trying to save your life and being able to keep you comfortable, which way would you like me to lean?

I’m asking you to let me know, or think about it and let me know, so that you know that I’m interested, that I’m with you for the long haul, and that I welcome your questions and your guidance to help me get to know you better.

I also want to know whom you would want to make decisions with me on your behalf if you’re ever too sick to make them yourself. Let me know, either today or when I see you next time, that person’s name and phone number. Be sure that you have a talk with that person, too, to make sure that he or she knows what you want or do not want.

First, select appropriate patients. Clinical indicators for selecting patients might include having chronic progressive health problems (for example, chronic obstructive pulmonary disease) with symptomatic exacerbations that are increasingly likely in the coming months or having a typical life expectancy of 2 years or less. Psychosocial indicators for patient selection could be lacking a family, having family members who are geographically or emotionally distant, and having family members who are likely to be in conflict if asked to respect the patient’s end-of-life preferences when the time comes to do so. Such discussions are easier to have in advance of a crisis than in the midst of one.

Next, select an appropriate framework, such as care-based routine discussions of the goals of care. The physician can introduce the discussion with a brief statement expressing honesty, concern, and commitment to working with the patient. A sample script framework (see box) illustrates a gentle approach to inviting an end-of-life discussion that opens the door to patients’ expressions of wishes without unduly raising anxieties.

Initiating these discussions does not have to be hard, and the discussions themselves do not have to be time-consuming, especially when they occur as conversations in the context of routine care. After the first such conversation, the next ones can begin with, “Well, it’s time to recheck how we’re doing with your goals. When I asked before, you told me this (summarize the patient’s preferences here). Did I get that right, and do you still feel the same way? Are there any developments or changes in emphasis that I should know about?”

Having end-of-life discussions with patients and their families does measurable good. Physicians in managed care settings should build rapport by welcoming communication of values and preferences by patients. Cultural perspectives should be respected, and vocabulary should include sensitivity to those perspectives, but patients’ culturally based misunderstandings of medical treatments need to be corrected by physicians so that decisions can be based on sound information. The JAMA study’s evidence that end-of-life discussions do no harm, and do significant good, can give physicians the confidence needed to initiate such discussions.

What Can RTs Do?

The main responsibility for having end-of-life discussions with patients and families is, of course, the physician’s. In the acute care setting, members of the team, including RTs, can help. Patients and families sometimes share their feelings with respiratory staff: concerns about their discomfort, the quality of their lives, or that their health care has veered off course, in terms of what they want. Just listening can help, as experienced RTs know.

Conveying patients’ expressions of concern or questions to the patient’s physician, when you are fairly sure that the physician is not aware of them, is important, too—both orally and through documentation in the patient’s medical record. You can convey these concerns in a nonconfrontational way, and can offer a supportive ear if the physician wants to think out loud about how to respond.

What if you’re in the middle, and you see the failure to communicate between physician and patient as risking harm or actually causing harm to the patient? If you’ve spoken up and nothing has happened, that’s when a request for the ethics committee’s assistance should be considered. You can either inform patients/families of the availability of ethics assistance in your facility and the way to tap into it, or speak with a member of the ethics team yourself.

Daniel O. Dugan, PhD, is health care ethics consultant at Swedish Covenant Hospital, Chicago, and at Emmanual Medical Center, Turlock, Calif. For further information, contact [email protected].


  1. Faber-Langendoen K, Lanken P. Dying patients in the intensive care unit: forgoing treatment, maintaining care. Ann Intern Med. 2000;133:886-893.
  2. Knaus W, Harrell FE, Lynn J, et al. The SUPPORT prognostic model. Objective estimates of survival for seriously ill hospitalized adults. Study to understand prognosis and preferences for outcomes and risks of treatments. Arch Intern Med. 1995;122:191-203.
  3. Dowdy M, Robertson C, Bander J. A study of proactive ethics consultation for critically and terminally ill patients with extended lengths of stay. Crit Care Med. 1998;26:252-9.
  4. Schneiderman L, Teetzel H, Dugan D, et al. Effect of ethics consultations on non-beneficial life-sustaining treatments in the intensive care setting: a multicenter, prospective, randomized, controlled trial. JAMA. 2003;290:1166-72.
  5. Heffner J, Barbieri C. End-of-life care preferences of patients enrolled in cardiovascular rehabilitation programs. Chest. 2000;117:1474-81.
  6. Reichner C, Thompson J, O’Brien S, Kuru T, Anderson ED. Outcome and code status of lung cancer patients admitted to the medical ICU. Chest. 2006;130:719-723.
  7. Boyle R. Communication, truthtelling, and disclosure. In: Fletcher J, Lombardo PA, Marshall MF, Miller FG, eds. Introduction to Clinical Ethics. 2nd ed. Frederick, Md: University Publishing Group; 1997:54.
  8. Lagnado L. In some cultures, cancer stirs shame. Available at: online.wsj.com/article/SB122304682088802359.html?mod=rss_whats_news_us. Accessed November 9, 2008.
  9. Abbo E, Volandes A. A forced choice: the value of requiring advance directives. J Clin Ethics. 2008;19:127-40.
  10. Gazelle G. The slow code: should anyone rush to its defense? N Engl J Med. 1998;338:467-9.
  11. Wright A, Zhang B, Ray A, et al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA. 2008;300:1665-73.