Helping patients and their families cope with end-of-life issues can be one of medicine’s greatest challenges.

A philosopher once wrote, “In the game known as life, one can take either of two paths-run and hide and perhaps be safe, or plunge headfirst into a maze of obstacles and perhaps feel lost. I suggest the latter path, for at least you will have not lost your soul along the way.”

In this article, Kathryn Koch, MD, plunges headfirst into the maze of ethical issues faced by respiratory care practitioners (RCPs). Although many articles have been written about terminal care and end-of-life issues, the reality of discontinuing life support is most poignant in the intensive care unit, where RCPs face this reality on a daily basis. We in the respiratory care world are called upon to implement withdrawal and withholding of life support. Therefore, many of us have become deeply involved with the patients and their families concerning end-of-life issues. Yet most of us were trained to aggressively manage and ward off death rather than admit that all efforts have failed and we should stop “interfering” with the dying process.

Koch addresses the five fundamental principles underlying medical ethics: autonomy, justice, beneficence, nonmaleficence, and veracity, and weaves a tale of rolling probability in which the concept of futility becomes a process rather than a single, finite point in time. This represents a unique concept in the medical ethics arena that more closely mimics real-life situations in which a patient’s medical condition may not be futile at the onset of treatment but becomes irreversible if treatment fails to improve the underlying medical condition. The examples she uses and the compassion she expresses are helpful for all of us in the respiratory care world.

Bruce P. Krieger, MD, FACP, FCCP, is President, Medical Staff, and Chief, Pulmonary Intensive Care, at Mount Sinai Medical Center, in Miami. He is also Professor of Medicine at the University of Miami, Miami State Medical Center.

One of the most difficult issues in the practice of medicine in our death-denying society is the management of patients who are dying. Ready access to advanced medical technology has changed death from an event to a process. The wide availability of mechanical ventilation and respiratory care in particular has enabled some patients to defer, if not defy, death. This has turned intensive care units (ICUs) into waiting grounds for death, with no room left for the living.

At the same time, there is a common thread of pain avoidance in our society. If the price to attempt to defer or defy death requires payment in pain and discomfort or in an unacceptable decline in the quality of life, there are those who would prefer death. The tension between these opposing desires, amplified by our inability to predict outcomes for individuals with certainty, contributes to the patient’s suffering.

Restrictions in resources available to fund treatment further influence this dilemma. An example includes the 25 percent cuts in Medicare reimbursements for home oxygen effective this past January. Such a simple treatment as oxygen delays death and improves function in patients who otherwise would have already died. It may be therapeutic for the dying if it makes them more comfortable, but it will only prolong the dying process for someone who is close to death.

Respiratory technologies and treatments-the “breath of life”-are a special concern at this border between life and death. Strong emotions and deeply held values inform the events surrounding a decision to stop these treatments and accept death.


Osler1 noted that the practice of medicine is defined as a balancing act between the art of probability and the science of uncertainty. Objective predictive models have been developed that predict with a reasonable degree of certainty what the outcome of critically ill people will be if those people are taken as a group.2-5 Prognostic models have been shown to be in serious error for a substantial proportion of individual patients.6,7

The problem-uncertainty in predicting death-has to do with the concept of rolling probability. An individual’s prognosis is constantly changing as he/she responds or fails to respond to treatment or develops new problems or complications. Even if changes in treatment are correctly chosen and implemented in a timely fashion, an individual’s luck cannot be measured. In The Study to Understand Prognoses and Preferences for Risks of Treatment,8 some patients, particularly those with cardiac diagnoses, had relatively good prognoses until right before death.

A lack of recognition of the limitations in accurate prediction itself impairs discussions with patients and family regarding prognosis. Prior experiences with other health care professionals may interfere with trust in predictions. (I will never forget one family who refused to limit treatment for their mother because they had seen another family member pronounced dead, had gone to see his body in the morgue, and found him alive! He lived for another 3 months in a nursing home.)


It has been observed that the practice of medicine is increasingly one of relationships between strangers under rescue circumstances, dominated by military and business metaphors.9 Too frequently no prior relationship has existed between the critically ill patient and his/her caregivers. The patient may have an entirely different value system than any of his/her caregivers. There may not be a strong foundation of shared experiences or mutual respect of deeply held beliefs upon which to build a plan of care, only medical facts. Too often the patient is unable to participate in his/her own medical decision-making, and the family must convey the patient’s beliefs and values to the caregivers.

The physician may not admit the limitations of medical technology or may fear that discontinuing treatment will cause death. The physician may not be comfortable with raising concerns that may lead to a decision to stop treatment and might be more comfortable with a decision to not prescribe the treatment in the first place. There is no ethical or moral difference between withholding and withdrawing treatment,10,11 but withdrawing actually makes more medical sense than withholding (the treatment has been tried and not proven successful). For many medical staff, patients, and their families, withdrawing is more difficult emotionally and spiritually. The patient and family may be reluctant to raise their concerns to the physician about continuing treatment, but will share their concerns with bedside caregivers, putting those individuals squarely in the middle of these tensions.


In facing these difficult issues, I try to develop the concept that we must together weave a plan of care. The warp of this cloth is made up of the medical facts and available technology; the weft is made up of the patient’s values and goals.

If the patient is not able to participate in the discussion, we designate an empty chair at the table to symbolize his/her presence. This reminds the family that it is their task to represent the patient’s wishes and preferences, while it is my task to determine if medicine has the ability to meet those wishes given the medical facts. Any caregiver who has spent time with the patient is welcome to participate.

The question is, can we prolong the patient’s life with value, dignity, and meaning? Can treatment meet this intended goal? The patient’s reluctance to use life-sustaining techniques would be greater in the following circumstances:

  • The patient finds no value in living persistently dependent upon a ventilator.
  • Care occurs in a context of limited access to advanced technology.
  • The patient prefers a quick death over a painful plan of care.

    On the other hand, the patient would be more likely to welcome medical intervention in these circumstances:

  • The patient values being alive, even permanently ventilator dependent.
  • Care occurs in a context of unlimited access to advanced medical technology.
  • The patient has a specific goal for which he/she is willing to take risks (eg, living long enough to see a grandson born).


Having reached some sense of the patient’s definition of quality of life, the next step is to determine the level of care to be applied. Only after the general levels of care have been defined can specific treatment options be discussed. They are:

  • Resuscitative care. All treatment, including CPR, will be administered should there be a sudden arrhythmia or other catastrophic event, recognizing that any neurologic deficit would likely be worsened should the patient survive such an event and that success rates are not very good.
  • Intensive care. All treatment, short of cardiac massage and defibrillation, including tracheostomy, intracranial pressure monitoring, dialysis, and the like, will be used if needed.
  • Supportive care. Continued routine care will be applied, but each new problem will be evaluated separately for decisions for new treatment. A tracheostomy may be performed to facilitate ongoing mechanical ventilation and antibiotics may be administered if there is a new pneumonia, but anti-arrhythmics and pressors and dialysis may be withheld if renal failure occurs.
  • Comfort care. No new treatments will be administered other than for comfort; all old treatments not required for patient comfort will be stopped, including mechanical ventilation. If it is anticipated that oxygen will prolong the dying process and increase suffering through that fact, oxygen will also be discontinued.

It must be noted that physician-assisted suicide is not one of the levels of care. This option is currently legal only in Oregon, and then only by prescription. The values informing each different preference for level of care, including requests for assistance in dying, must be understood (Table 1, page 92). It is from the patient’s values and the medical facts that the plan of care will be spun.


There are five fundamental principles underlying the decision-making process: autonomy, justice, beneficence, nonmaleficence, and veracity. These principles can be applied by any decision-maker. For any decision, and particularly ones that affect someone’s life, the application of the principles requires the education of the decision-

makers by medical staff. This is the notion of informed consent. The first task is to identify the decision-maker.

The principle of autonomy requires that the patient, if capable, be the decision-maker. The patient may appoint someone else to make the decision or may choose to have others involved in the decision. He/she may designate a surrogate or durable power of attorney for health care decisions who is empowered to make decisions if the patient is unable to.

The principle of autonomy also supports living wills, which are now available by statute in every state. A living will directs family and caregivers to follow a patient’s specific wishes in the event he/she is unable to make decisions. Living wills usually direct comfort care if the patient is terminally ill but may be modified in many ways.

The problem with most living wills is that they are boilerplate, and do not give sufficient guidance. Too frequently no discussion has occurred between patient, caregivers, and family members regarding what is important to the patient. There may even be serious questions as to whether the patient is “terminal” given the advanced state of medical technology.

My living will, for example, directs resuscitative care if there is a reasonable chance of functional survival. (I do not define reasonable, but leave that in the hands of my caregivers and loved ones.) It directs intensive care if there is a good chance of functional survival, supportive care if there is a poor chance, and comfort care if there is no chance.

The specific directive given is that I do not want aggressive care if I become quadriplegic-what is most important to me in life is that I be able to care for myself and be independent. As I cannot predict what might actually happen to me, I have not specified treatments for different hypothetical circumstances but have stuck to the idea of levels of care so that my loved ones and caregivers can adapt my care based on my wishes and the situation. I have discussed these wishes with my loved ones.

The principle of justice requires that there be just distribution of medical services. This is illustrated by the axioms underlying the American Thoracic Society’s statement12 on fair allocation of ICU resources (Table 2). They are also reflected in the statement by the Ethics Committee of the Society of Critical Care Medicine13 regarding futile and other possibly inadvisable treatments.

We must understand the difference between rationing and withholding futile treatment. Rationing is the limitation of treatment that may work but is not worth the cost. Withholding “futile” treatment is limitation of treatment that is unlikely to work and is not worth the cost. Treatment is truly futile only when it will not work (eg, administering oxygen to a dead person). Again, the question is, how likely is treatment to meet its intended goal?

The principles of beneficence and nonmaleficence are notions from the Hippocratic tradition that offer guidance. One should always do good, and failing to be able to do good, one should at least do no harm. The patient must determine what is good and what is harmful, but not in a way that his/her decision harms others. For example, a patient with gastrointestinal bleeding that is refractory to medical treatment may refuse surgery, but in doing so, may not then demand the unlimited transfusion of blood, which will be needed by other patients who have a better chance of survival.

To apply each of the aforementioned principles, one must use the principle of veracity, also an ethical virtue. It is essential to tell the truth about the situation, explore all options, define which are accessible, and discuss how each option might affect the patient’s quality of life.


As the changing practice of medical care in the United States moves physicians further away from the actual administration of care and more toward the direction and supervision of care, bedside caregivers become more important in decision-making. The bedside caregivers are the ones who are familiar with the patient and his/her experience with illness.

Their role in educating the patient and family regarding options and levels of care and providing feedback to the physician regarding an evolving issue is essential. They can guide the patient by telling him/her what is important to discuss with his/her family and doctor. They can inform the physician of any misinformation, misperceptions, or confusion the patient has about treatment options. They are the ones at the bedside and in the home.


The following cases demonstrate the unique situations one may encounter when treating patients facing life-or-death decisions.

Mr E. The nurse caring for one of my hospice patients with amyotrophic lateral sclerosis tells me that he wants assistance in dying. I thought this would come up sooner or later as his decline continued, but I never raised the issue. He had wanted very much to live as long as possible to see his baby son grow. The previous year, he had been willing to go through a difficult withdrawal of anxiolytics (originally started for his panic response to shortness of breath) in order to be more alert for his child. Escalation of narcotics controlled his panic more effectively with less confusion.

As his decline continued, however, he told his RCP that he wanted to discontinue his bilevel positive airway pressure and postural drainage because they made him more miserable than comfortable. We honored his wishes, acknowledging that his life would be shortened. His narcotics were further escalated to control his panic and bad dreams until he stabilized at approximately 4 g morphine a day.

We assembled at the bedside in his home to explore the question of assistance in dying. Despite the massive dose of narcotics he was receiving, he was awake, alert, coherent, and competent. We examined all of our mutual issues. If we achieved nothing else, we made it clear that he would not be abandoned; we would always be ready to listen to him and discuss his concerns, and we would continue to care for him.

Mrs C. The nurse and the respiratory therapist at the bedside of one of my ICU patients tell me she is ready to stop treatment. She has severe emphysema from smoking, with a major component of bronchospasm. She hates the experience of mechanical ventilation. Despite sedation and narcotics, she is miserable unless she is totally unconscious. Sufficient medication makes her hypotensive. After 2 weeks with no progress toward weaning, she indicates to her RCP during an episode of alertness that she would rather be dead.

Her family and I discuss this, but do not initially act upon it. She repeats this concern to all of her caregivers and family a number of times over the next several days. We make a decision to withhold her sedation to confirm her decision. Fully awake and alert, she indicates that she would rather take her chances without mechanical ventilation. When we discuss the issue of sedation for comfort and observe that sedation would reduce her ability to be successful if she were extubated, she indicates that she would rather have a clear mind to give herself the best chance. She acknowledges that she may die, and wants to receive sedation if she is in severe discomfort, but insists on trying.

She agrees to continue all other treatment besides mechanical ventilation, short of reintubation and resuscitation. With her family at her bedside, we extubate her from full ventilatory support, continuing oxygen and other respiratory treatments. We expect her to die. She, in fact, does much better off the ventilator. She is discharged home and enjoys a functional but oxygen- dependent existence for another year before her next episode of respiratory distress. Upon readmission, she initially refuses intubation, but as she fails to respond to aggressive treatment, she ultimately agrees to “give it another go.”

Mrs A. The nurse and the RCP at the bedside of one of my ICU patients tell me she is ready to stop treatment.

She has adenocarcinoma of the lung constricting her trachea, a staphylococcal empyema, and a bronchopleural fistula. She has experienced a respiratory arrest with subsequent neurologic deficit. With aggressive treatment, including radiation treatment while on mechanical ventilation, she has awakened but otherwise has not significantly improved. While she was obtunded, her husband decided with us to aggressively treat her. When she awakens, she agrees to put up with treatment through her course of radiation to see if she can get well enough to breathe on her own. She wants to die at home if she can.

Weaning attempts fail the day following completion of radiation. She agrees to give it one more day. The following morning she indicates to her bedside caregivers that she is ready to breathe on her own again, whether or not she is medically ready. We assemble at her bedside. After she tries and fails her T-piece trial, her husband begs her to give it one more day. She says no.

I tell her husband that she is capable of making her own decision-that this request is consistent with her previous statements. She clearly understands she is likely to die quickly if we do this. He asks me to make her try treatment for another day. I say that we must honor her wish. She wants no morphine. We tell her we will have it ready, just in case. We extubate her to room air. She asks for a cup of coffee. Although we know that she will aspirate, we honor her request. She drinks two sips, smiles, clasps her husband’s hand, and dies.


Most people face death in the same style they have used in living. Their role in the family (eg, matriarch, patriarch) will affect the dynamic of their family’s reaction to their death. Their death experience will be one of the most important experiences for them and their family. If there is any way for medicine to help them choose their own path on their trip to death, it should do so. Then we must hold their hands and the hands of their family as they proceed down that path. N

Kathryn A. Koch, MD, FCCP, FCCM, FACP, has practiced and taught critical care medicine for 20 years. She created, then chaired, the Ethics Committee at the University Medical Center in Jacksonville, Fla, for 10 years. She currently is a fellow in the Department of Ethics, Humanities, and the Law at the University of Florida, in Jacksonville.


1. Osler W. In: Bean RB, ed. Aphorisms from His Bedside Teachings and Writings. 3rd ed. Springfield, Ill: Charles C. Thomas Publ; 1961:129.

2. Knaus WA, Wagner DP, Draper EA, et al. The APACHE III prognostic system: risk prediction of hospital mortality for critically ill hospitalized adults. Chest. 1991;100:1619-1636.

3. Lemeshow S, Teres D, Klar J, et al. Mortality Probability Models (MPM II) based on an international cohort of intensive care unit patients. JAMA. 1993;270: 2478-2486.

4. Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA. 1993;270:2957-2963.

5. Knaus WA, Harrell FE, Lynn J, et al. The SUPPORT prognostic model. Objective estimates of survival for seriously ill hospitalized adults. Ann Intern Med. 1995;122:342-350.

6. Schafer JH, Maurer A, Jochimsem F, et al. Outcome prediction models on admission in a medical intensive care unit: do they predict individual outcome? Crit Care Med. 1990;18:1111-1118.

7. Iezzoni LI, Ash AS, Shwartz M, Daley J, Hughes JS, Mackiernan YD. Predicting who dies depends on how severity is measured: implications for evaluating patient outcomes. Ann Intern Med. 1995;123:763-770.

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

9. Annas GJ. Reframing the debate on health care reform by replacing our metaphors. New Engl J Med. 1995;332:744-747.

10. Bone RC, Rackow EC, Weg JC. ACCP/ SCCM consensus panel: ethical and moral guidelines for the initiation, continuation and withdrawal of intensive care. Chest. 1990; 97:949-958.

11. American Thoracic Society Medical Section of the American Lung Association. Withholding and withdrawal of life-sustaining therapy. Am Rev Respir Dis. 1991;144:726-731.

12. Fair allocation of intensive care unit resources. Official Statement of the American Thoracic Society. Am J Respir Crit Care Med. 1997;156:1282-1301.

13. The Ethics Committee of the Society of Critical Care Medicine. Consensus statement of the Society of Critical Care Medicine’s Ethics Committee regarding futile and other possibly inadvisable treatments. Crit Care Med. 1997;25:887-891.