Withholding and Withdrawing Life-Sustaining Treatment
Am Fam Physician. 2000 Oct 1;62(7):1555-1560.
Withholding or withdrawing life-sustaining therapies is ethical and medically appropriate in some circumstances. This article summarizes the American Medical Association's Education for Physicians on End-of-life Care (EPEC) curriculum module on withholding or withdrawing therapy. Before reviewing specific treatment preferences, it is useful to ask patients about their understanding of the illness and to discuss their values and general goals of care. Family physicians should feel free to provide specific advice to patients and families struggling with these decisions. Patients with decision-making capacity can opt to forego any medical intervention, including artificial nutrition/hydration and cardiopulmonary resuscitation.
Making and implementing decisions about life-sustaining treatments are essential skills for family physicians. Good care is based on an understanding of the legal and ethical issues, as well as familiarity with practical aspects of implementation (Table 1).1 Fundamental ethical principles (e.g., autonomy, beneficence and futility) and specific advance directives (e.g., a living will or durable power of attorney for health care) remain important but are not reviewed in this article.
Typical Legal and Ethical Concerns Family Physicians May Have Regarding the Withholding or Withdrawing of Potentially Life-Sustaining Treatments
Are physicians legally required to provide all life-sustaining measures possible?
No. To the contrary, patients have a right to refuse any medical treatment, even life-sustaining treatments such as mechanical ventilation, or even artificial hydration and nutrition.
Is withdrawal or withholding of treatment equivalent to euthanasia?
No. There is a strong general consensus that withdrawal or withholding of treatment is a decision that allows the disease to progress on its natural course. It is not a decision to seek death and end life.Euthanasia actively seeks to end the patient's life.
Are you killing the patient when you remove the ventilator and treat the pain?
No. The intent and sequence of actions are important, as are the means chosen. If the intent is to secure comfort, not death; if the medications are chosen for and titrated to the patient's symptoms; if the medications are not administered with the primary intent to cause death, then ventilator withdrawal and pain treatment are not euthanasia.
Can the treatment of symptoms constitute euthanasia?
No. For patients who have been using opioids for pain, it is very difficult to give such high doses that death is caused or even hastened in the absence of a disease process that is leading to imminent death. Patients tend to sleep off the effect if they receive too much medication. However, in the rare circumstance when opioids might contribute to death, provided the intent was genuinely to treat the symptoms, then opioid use is not euthanasia.
Is it illegal to prescribe large doses of opioids to relieve pain, breathlessness or other symptoms?
No. Even very large doses of opioids (thousands of milligrams per day) are permitted and appropriate, if the intent and doses given are titrated to the patient's needs.
Adapted with permission from Education for Physicians on End-of-Life Care Trainer's Guide, Module 11, withholding, withdrawing therapy. In: Emanuel LL, von Gunten C J, Ferris FD. Education for physicians on end-of-life care/Institute for Ethics at the American Medical Association. Chicago, IL: EPEC Project, The Robert Wood Johnson Foundation, 1999.
All states have laws related to withholding or withdrawing life-sustaining treatments. Institutional policies generally recommend that when in doubt, physicians should provide treatment to prolong life. However, physicians can ensure appropriate care in the hospital and across care settings, even if a patient desires not to use certain medical interventions. Unfortunately, the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment2 (SUPPORT) documented that many patients die without attention to these issues, even undergoing invasive medical treatment against previously stated wishes. On the other hand, primary care physicians in community settings may be better at communicating with patients at the end of life.3,4 These physicians have a long-term relationship with their patients, and end-of-life discussions may occur in small pieces over several months or years.
Patients may consider many life-sustaining treatments; in addition to cardiopulmonary resuscitation (CPR), options include elective intubation, mechanical ventilation, surgery, dialysis, blood transfusions, artificial nutrition and hydration, diagnostic tests, antibiotics, other medications and treatments, as well as future admissions to the hospital or to the intensive care unit. The treatment choices and the complexities increase as a patient's condition worsens. However, many patients who initially choose a do-not-resuscitate (DNR) order opt for progressively more restrictions as diseases progress. Although not using an intervention and withdrawing that intervention are ethically and legally equivalent, it is better to make these decisions ahead of time.
Family physicians should feel free to provide specific advice to patients and families wrestling with these difficult decisions.5 A thoughtful physician can furnish honest, frank advice based on medical science and personal experience. For example, it may be appropriate in one circumstance to say: “Your mother is dying. We can't change that, but we can change the way that she dies. In my opinion, inserting a feeding tube into your mother's stomach may increase the number of days she is alive, but I think she would try to pull it out, and the quality of her remaining days would be worse. I wouldn't do it.”
The Education for Physicians on End-of-life Care (EPEC) curriculum1 suggests other steps that may be helpful when deciding whether to withhold or withdraw therapy (Table 2).1 These steps ideally occur during routine office visits, although in many cases, they are not accomplished until a medical crisis occurs.
Eight-Step Protocol to Discuss Withholding or Withdrawing Therapy
Be familiar with institutional policies and state laws.
Choose an appropriate, private setting for the discussion.
Ask the patient and family what they understand.
Discuss the patient's values and general goals of care.
Establish context for the discussion.
Discuss specific treatment preferences.
Respond to emotions.
Establish and implement the plan.
Reproduced with permission from Education for Physicians on End-of-Life Care Trainer's Guide, Module 11, withholding, withdrawing therapy. In: Emanuel LL, von Gunten CJ, Ferris FD. Education for physicians on end-of-life care/Institute for Ethics at the American Medical Association. Chicago, IL: EPEC Project, The Robert Wood Johnson Foundation, 1999.
Physicians should determine what the patient, family or surrogate understands about the current health situation. One should avoid medical jargon, and it may not be helpful to provide excessive details of day-to-day management. Physicians should provide an honest opinion, expressing doubt when it exists and covering any disagreements among consultants. Some decision makers may require additional information or consultation.
The general goals of care need to be determined. For example, the physician may state, “Let me tell you what I understand you want as we plan your care.” If a patient is unable to speak, the person designated to make decisions about medical management should be encouraged to articulate the patient's values and goals of care. If family members disagree about the general goals of care, further exploration and explanation before proceeding to discussions about specific treatment decisions may be required.6
Many family members may ask the physician to “do everything.” “Everything” is much too broad and is easily misinterpreted. If a patient is terminally ill, the context may be established by saying, “Your father is dying despite all the treatment we're giving him. His death will probably occur within the next few days or weeks.” While many terminal illnesses have no specific treatment, there is always “something” to do for the patient.
Next, specific treatment preferences can be discussed. Providing information in small pieces, reinforcing the context in which the decision will apply and stopping to assess for reactions, to ask for questions and to clarify misunderstandings are all helpful. Specific treatments (e.g., dialysis, continued transfusions, vasopressors) should be reviewed in light of the general goals that were previously established.
Patients deserve a clear and complete understanding of the proposed therapies. Some patients will want to know all the details, while others will prefer to forego the details and rely solely on their physician's specific advice. Life-sustaining treatment should be discussed in simple terms: the problem the treatment will address, what the treatment or procedure involves, what is likely to happen if the patient declines treatment and the benefits and burdens of the treatment.
Responding to patient and family anxiety will acknowledge emotional content. Patients and family members may be profoundly disturbed by these discussions, and it may be helpful to ask other colleagues to assist. However, most decision makers are grateful for the advice of a trusted physician, even advice that recommends withdrawing a treatment.
Finally, a plan should be established. If the decision maker is unable to accomplish this, the plan may then be to discuss the subject again or to convene a family meeting.7 Perhaps a chaplain could be involved, or a social worker can notify a key out-of-town family member. Furthermore, hospice is available not only for patients with terminal cancer but also for patients with diseases such as dementia, heart failure or chronic lung disease. Hospice care can also be provided to patients in nursing homes.4,8–10
Application of the Protocol: Artificial Feeding and Hydration
Physicians have difficulty with the decision to initiate or continue artificial nutrition and/or hydration. Food and water are symbols of caring, so withholding artificial nutrition and hydration may be misperceived as neglect. Some physicians may find it unethical to withhold or withdraw these interventions; however, strong legal support is in place for thoughtful withholding or withdrawal of fluids and nutrition at the end of life.11 Table 31 lists situations when withdrawing/withholding artificial nutrition and hydration may be appropriate.
Considerations in Initiating, Withholding or Withdrawing Artificial Hydration and Nutrition
Indications to consider starting artificial hydration or nutrition
The patient or decision maker, after being provided appropriate counseling and options, chooses this intervention to reflect personal values.
When the primary goal of the patient is to maximize the quantity of life
When the patient is stable or improving, and the intervention has a reasonable chance of reaching the patient's goals
When the risk/benefit ratio is unclear, or the evolution of the disease is uncertain. In this case, consider a trial of the intervention with withdrawal if little or no benefit accrues
As short-term interventions in healthy patients or those with mild or moderately severe illness (e.g., acute infections, perioperative use, acute stroke)
In selected chronic conditions (e.g., patients with esophageal obstruction or massive bowel resection, where oral intake is not possible or inadequate)
When a patient is unable to swallow and remains hungry or thirsty
When delirium may be due to dehydration
To maintain life for a period, while the decision maker/family struggles with end-of-life decisions
When there is no clear decision maker, and the family cannot reach a consensus
When the quality of life is good, as defined by the patient
Potential reasons to withhold or withdraw artificial hydration or nutrition
The patient or decision maker, after being provided appropriate counseling and options, decides against this intervention to reflect personal values.
When the primary goal of the patient is palliation of symptoms
When the intervention is bound to fail (e.g., in patients with a disease that is uniformly fatal)
When the patient is dying, and the intervention is merely prolonging the dying process or causing suffering
When a patient has moderate or severe, irreversible cognitive impairment (e.g., Alzheimer's dementia or vascular dementia)
When the intervention causes complications (e.g., agitation requiring sedation or physical restraints, infection, multiple tube obstructions or withdrawal, aspiration pneumonia)
To help relieve symptoms of fluid overload (e.g., dyspnea, diarrhea, urinary frequency or bowel obstruction) in terminally ill patients
When the patient has end-stage organ failure (e.g., respiratory, cardiac, renal)
When other end-stage diseases, such as metastatic cancer or advanced acquired immunodeficiency syndrome, are present
When a patient has had a stroke, is profoundly impaired and does not recover useful swallowing function
When the risks exceed the benefits
When the quality of life is poor, as defined by the patient
Information from Education for Physicians on End-of-Life Care Trainer's Guide, Module 11, withholding, withdrawing therapy. In: Emanuel LL, von Gunten C J, Ferris FD. Education for physicians on end-of-life care/Institute for Ethics at the American Medical Association. Chicago, IL: EPEC Project, The Robert Wood Johnson Foundation, 1999.
For example, inserting a feeding tube in patients with severe dementia does not prevent aspiration, has little effect on malnutrition, does not improve survival or function, does not prevent or treat pressure sores, and may cause discomfort and other adverse effects. A conscientious program of hand feeding is usually more appropriate.12,13 Like all procedures, tube feeding is not inevitable; it is a choice.14
A first step is to evaluate the patient's values and overall goals of care.5 A review of the patient's general medical condition and an evaluation of the potential of artificial hydration and nutrition to help achieve these goals should be conducted. For example, if the patient has advanced cancer, what is the expected course of the cancer? Is anything reversible? How will intravenous or enteral tube fluids and nutrition improve the situation? Ideally, this discussion should occur before an acute medical crisis, at which time the decision maker may become overwhelmed by multiple complex decisions.14
Misperceptions are common. The family may believe that a lack of appetite and poor oral intake is causing the current disability, assuming that “if only Dad got more fluids and nutrition, he would be stronger.” Clear, simple language will focus on the true cause of the situation. For example, statements like “The cancer is taking all of your strength,” or “The fact that your heart is so weak is causing you to lose your appetite and feel so tired” will help clarify the cause of the patient's lack of appetite and oral intake. In equivocal cases, a monitored clinical trial of hydration may be warranted; however, artificial hydration may worsen physical symptoms such as edema, breathlessness, diarrhea, vomiting or incontinence. The family may feel more comfortable withdrawing intravenous hydration after a trial has shown no benefit.5
Urinary output normally declines in the patient who is dying; 300 to 500 mL per day is adequate. The large volumes (2 to 3 L per day) that physicians associate with hospitalized patients are usually the result of intravenous infusions and do not reflect normal output with oral hydration.
Providing food and water meets powerful emotional needs. Family members provide the bulk of nonmedical care to dying patients, and the physician can help the family find ways to redirect their desires to be caring.15 For example, they can take turns holding vigil, hold the patient's hand, stroke the forehead, or sing or read poetry or scripture to the patient.
Near the end of life, patients and families may naturally be concerned that the lack of fluids and food will cause suffering. While acknowledging their concerns, physicians can reassure them that dehydration is a common, natural part of the dying process. It does not affect the dying patient in the same way as a healthy person who feels thirsty on a hot day or dizzy when standing. Trying to reverse a natural trend often leads to substantial discomfort, without affecting the outcome.16–18 If a patient is not hungry or thirsty, then providing artificial fluids and nutrition will not relieve the symptoms. Some patients may even make a conscious choice to stop eating and drinking.
Application of the Protocol: Cardiopulmonary Resuscitation
Performing CPR on hospitalized patients is usually ineffective, with only 6 to 17 percent of patients surviving to discharge; many survivors die during the next few months or have a poor quality of life. Patients with malignancy, sepsis, pneumonia or renal failure have even lower survival rates. More than 90 percent of all hospital deaths now occur without performing CPR.19 Some authors recommend against routinely offering CPR to certain patient populations, such as those living in nursing homes.20
It is worth reflecting on the language that has developed around CPR and DNR orders. Too frequently, a DNR order is interpreted to imply a whole host of other decisions that, in fact, may not have been made. “Is this patient a DNR?” becomes shorthand for the more pertinent question, “What are the goals of care?”5,21
First, the physician can review the overall medical condition and the context in which CPR is being discussed. Engaging in a discussion that focuses only on CPR and the DNR status may leave the patient confused and anxious. Clear language is helpful: “If you should die despite our current therapy, do you want us to use extreme measures? Do you want CPR performed—to press on your chest and put a tube in your lungs to try to get your heart and breathing started again?” The convention of discussing CPR initially as “starting the heart” or “putting on a breathing machine” implies a false sense of reversibility, or suggests that those heart and lung functions are isolated problems. In patients with advanced illness, circulation and breathing stop because of the relentless progression of disease.
Performing CPR should not be described as “doing everything.” This implies that not performing CPR is “doing nothing.” Everyone wants to do whatever is possible to maintain health and well being. In patients with advanced progressive illness where the prognosis is limited, however, performing CPR often does not help achieve the patient's goals. The discussion of resuscitation needs to be placed into the context of other life-prolonging therapies (e.g., elective intubation and mechanical ventilation, dialysis, surgery, antibiotics, artificial fluids). Various scenarios can be described, and the physician may elicit the patient's preferences.
The discussion should conclude with an affirmation of the active interventions that will be performed. For many patients, interventions to reverse disease and sustain life are appropriate even with a DNR order in place. However, if the patient dies despite these efforts, resuscitation will not be attempted.
Finally, the physician should write an appropriate order in the record. CPR involves the establishment of a patent airway, effective breathing and cardiac output through assisted ventilation and external cardiac massage. If this is inappropriate or undesired by the patient, the family physician should write a DNR order. A DNR order does not address any aspect of care other than to prevent performing CPR.
Withholding and withdrawing therapy challenge family physicians to be excellent communicators with patients and families. Working to help patients and families to achieve their goals and to understand the limits of modern medical help can be highly rewarding. Family physicians should continue to be strong advocates for dying patients.5,22
This article is abstracted from the Education for Physicians on End-of-life Care (EPEC) curriculum, a program of the Institute for Ethics at the American Medical Association. © EPEC Project, The Robert Wood Johnson Foundation, 1999.
Special thanks are due to the EPEC team, the EPEC Advisory Group and all other contributors.
1. Education for Physicians on End-of-Life Care Trainer's Guide, Module 11, withholding, withdrawing therapy In: Emanuel LL, von Gunten CJ, Ferris FD Education for physicians on end-of-life care/Institute for Ethics at the American Medical Association Chicago, IL: EPEC Project, The Robert Wood Johnson Foundation, 1999.
2. 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).The SUPPORT Principal Investigators. JAMA. 1995;274:1591–8 [Published erratum appears in JAMA. 1996;275:1232]
3. Hanson LC, Earp JA, Garrett J, Menon M, Danis M. Community physicians who provide terminal care. Arch Intern Med. 1999;159:1133–8.
4. Ackermann RJ, Kemle KA. Death in a nursing home with active medical management. In: The annals of long-term care: the official journal of the American Medical Directors Association. American Medical Directors Association. Plainsboro, N.J.: Multimedia Healthcare, 1999;7:313–9.
5. Doukas DJ, Brody H. Care at the twilight: ethics and end-of-life care. Am Fam Physician. 1995;52:1294–6,1299.
6. Goold SD, Williams B, Arnold RM. Conflicts regarding decisions to limit treatment: a differential diagnosis. JAMA. 2000;283:909–14.
7. Streim JE, Marshall JR. The dying elderly patient. Am Fam Physician. 1988;38:175–83.
8. Keay TJ, Schonwetter RS. Hospice care in the nursing home. Am Fam Physician. 1998;57:491–4.
9. Standards and Accreditation Committee. Medical guidelines for determining prognosis in selected non-cancer diseases. 2d ed. National Hospice Organization Medical Guidelines Task Force. Arlington, VA: National Hospice Organization, 1996.
10. Volicer L, Collard A, Hurley A, Bishop C, Kern D, Karon S. Impact of special care unit for patients with advanced Alzheimer's disease on patients' discomfort and costs. J Am Geriatr Soc. 1994;42:597–603.
11. American Medical Association Council on Ethical and Judicial Affairs. Code of Medical Ethics. 1996–1997 ed. Chicago, IL: AMA, 1997.
12. Finucane TE, Christmas C, Travis K. Tube feeding in patients with advanced dementia: a review of the evidence. JAMA. 1999;282:1365–70.
13. Gillick MR. Rethinking the role of tube feeding in patients with advanced dementia. N Engl J Med. 2000;342:206–10.
14. Callahan CM, Haag KM, Buchanan NN, Nisi R. Decision-making for percutaneous endoscopic gastrostomy among older adults in a community setting. J Am Geriatr Soc. 1999;47:1105–9.
15. Emanuel EJ, Fairclough DL, Slutsman J, Alpert H, Baldwin D, Emanuel LL. Assistance from family members, friends, paid care givers, and volunteers in the care of terminally ill patients. N Engl J Med. 1999;341:956–63.
16. Vullo-Navich K, Smith S, Andrews M, Levine AM, Tischler JF, Veglia JM. Comfort and incidence of abnormal serum sodium, BUN, creatinine and osmolality in dehydration of terminal illness. Am J Hosp Palliat Care. 1998;15:77–84.
17. McCann RM, Hall WJ, Groth-Juncker A. Comfort care for terminally ill patients. The appropriate use of nutrition and hydration. JAMA. 1994;272:1263–6.
18. Billings JA. Comfort measures for the terminally ill. Is dehydration painful? J Am Geriatr Soc. 1985;33:808–10.
19. Von Gunten CF. CPR in hospitalized patients: when is it futile? Am Fam Physician. 1991;44:2130–4.
20. Finucane TE, Harper GM. Attempting resuscitation in nursing homes: policy considerations. J Am Geriatr Soc. 1999;47:1261–4.
21. Stevens MB. Withholding resuscitation. Am Fam Physician. 1986;33:207–12.
22. Reichel W. End-of-life care and family practice [Editorial]. Am Fam Physician. 1999;59:1388,1395–6,1399.
Copyright © 2000 by the American Academy of Family Physicians.
This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP. Contact firstname.lastname@example.org for copyright questions and/or permission requests.