Decisions about the use of life-sustaining treatment when a person is seriously ill and likely to be near the end of life have profound consequences for that person, for his or her family and other loved ones, and often for health care providers.2 These decisions will, to some extent, determine the time and circumstances of the person’s death. And these decisions will shape the person’s experience of remaining life—where it is lived, with whom, and with what degree of comfort or suffering. These decisions also have an inevitable social dimension. They compel us, as a society, to examine our common ethical priorities concerning the relief of suffering, what it means to live and to die, the rights of individuals, and the ultimate goals of medicine. They confront us with issues of justice, equity, and the economic constraints on the use of health care resources. The social dimensions of treatment decision-making near the end of life also include the cultural values, whether associated with religion, ethnicity, profession, or other identities and affiliations, which may inform how professionals present treatment options and how patients and surrogates make decisions.
Ethical challenges arise when the ethically appropriate course of action or range of choices is unclear, or when there are competing ethical claims that may not be reconcilable. Discussions about whether to use or forgo life-sustaining treatment frequently raise ethical concerns for health care professionals. These discussions and the resulting decisions should be informed by ethical norms and legal rights and protections concerning the health and welfare of some of the most vulnerable members of society; by health care goals that follow from these norms, rights, and protections; and by analysis based on the history of careful thinking, publication, and policy on end-of-life care. Our society must be confident that the decision-making processes used in health care institutions and by individual health care professionals are ethically sound, as well as equitable in the claims they allow any individual to make on shared health care resources.
These Guidelines are written for physicians, nurses, and other clinicians who care for patients facing decisions about life-sustaining treatment and patients nearing the end of life. These Guidelines are also intended to be useful to those responsible for the education of these clinicians, professionals who make institutional policy on care near the end of life, and public policy-makers working to improve end-of-life care. Patients, loved ones, and surrogate decision-makers may also find the Guidelines helpful.
The Function and Sources of These Ethics Guidelines
Ethics guidelines are different from clinical practice guidelines and from law, although they are related to both of these. Ethics guidelines for end-of-life care are grounded in the moral traditions of medicine and nursing and are consistent with the moral traditions of other health care professions such as clinical social work and clinical psychology. They reflect broad ethical, religious, and legal traditions of American society as well as decades of research, deliberation, and consensus-building among professionals, policy-makers, and patients and loved ones. Their claim to serious attention derives from the informed and reasoned positions they present and from the widely acknowledged standards of ethically responsible clinical practice on which they are based.
These Guidelines do not provide a formula or algorithm for making correct decisions. They are not “cookbook” ethics. They aim to guide reflection, judgment, and action in providing good care near the end of life and, in particular, to provide an ethical framework for making decisions under conditions that are challenging and sometimes psychologically distressing. They clarify the rights and responsibilities of each participant in the decision-making process. They can help ensure that these decisions are made with the appropriate deliberation, transparency, and fair process, as well as safeguards for vulnerable patients. These Guidelines can help protect patients against both undertreatment, such as failing to provide medically appropriate interventions desired by the patient, and overtreatment, in the form of interventions that are unwanted by the patient or are unduly burdensome to that patient.
In formulating these Guidelines, we have been attentive to the state of the law. However, these Guidelines do not constitute legal advice. Users of this document should determine how the law in their jurisdiction bears on these ethical recommendations. Although these Guidelines differ from law in their status and authority, they can help policy-makers, legislators, regulators, and judges identify where more explicit legal direction is needed and where existing law impedes ethical decision-making, either directly or by creating uncertainty. These Guidelines may also offer model formulations that can be translated into the language of statutes, regulations, and judicial rulings. However, not all ethics standards need to be or should be written into law.
Health care professionals in the United States worry about lawsuits and even about the possibility of criminal prosecution. Although the law serves a legitimate function in setting limits to behavior, it is a mistake to allow health care, which exists to care for the sick and to relieve suffering, to be driven primarily by liability concerns. Legal counsel to hospitals and other health care institutions, and legal advisors to individual health care providers, have a responsibility, as lawyers with expertise in health law, to help ensure that health care is driven by patient-centered goals and that health care professionals are not preoccupied by legal concerns. These Guidelines thus offer guidance to health lawyers and to health care risk managers concerning their role in institutional policy-making to support good care near the end of life.
Legal and Ethical Consensus Informing These Guidelines: Rights, Protections, and Key Philosophical Distinctions
The right of patients to forgo life-sustaining treatment has been recognized by law in the United States starting in the mid-1970s. This right is reflected in landmark cases such as Quinlan (producing the first state supreme court decision on end-of-life decision-making, in 1976) and Cruzan (producing the first U.S. Supreme Court decision on this issue, in 1990), in a large number of legal cases decided since Quinlan, and in federal and state statutes. Federal and state law recognize and authorize the use of advance directives, such as durable powers of attorney for health care decision-making (also known as proxy directives) and treatment directives (such as living wills). These allow patients to state in advance what treatments they do or do not want under different clinical circumstances, and who should act as their surrogate decision-maker, if they lose decision-making capacity. As the legal rules have been clarified, a parallel ethical consensus has developed. A broad ethical and legal consensus on much of end-of-life care is now in place. These Guidelines have been developed based on this legal and ethical consensus, which is summarized below.
Patients with decision-making capacity have a common law and constitutional right to refuse life-sustaining treatment.
The landmark Quinlan and Cruzan decisions, and many others, have established that patients have a legal right to refuse life-sustaining medical treatment. This right has been recognized both in common law rulings by judges and rulings based on the federal and state constitutions, and is broadly reflected in federal and state statutory law. (See Cited Legal Authorities.)
Patients who lack decision-making capacity have the same rights to refuse life-sustaining treatment as patients with decision-making capacity. The manner in which these rights are exercised is different, as a surrogate decision-maker must usually speak for them.
The Quinlan and Cruzan decisions, and subsequent case law and statutes, have also established that patients who lack decision-making capacity retain the right to refuse life-sustaining medical treatment. This right must usually be exercised by a surrogate decision-maker. The surrogate should decide based on the patient’s prior explicit treatment preferences or, if these are unknown, the patient’s preferences inferred from the surrogate’s knowledge of and experience with the patient. If the surrogate lacks adequate basis for inferring those preferences, then the surrogate should decide based on the patient’s best interests. Constitutional law, statutes, and judicial decisions authorize the use of advance directives and recognize surrogate decision-makers and standards for surrogate decision-making.
The right to refuse life-sustaining medical treatment does not depend on projected life expectancy, whether long or short. Patients have a basic right to be free of unwanted treatments.
Decisions about forgoing life-sustaining treatment usually arise when death is the unavoidable or probable outcome of the patient’s underlying medical condition. However, a patient does not need to be categorized as “terminally ill,” “actively dying,” or “imminently dying” for these decisions to be ethically justified. Individuals have a basic right to be free of unwanted treatment no matter what their prognosis.
There is no ethical difference between withholding and withdrawing life-sustaining medical treatment.
The basic right to be free of unwanted treatment does not depend on whether a treatment has already started. A patient’s right to refuse unwanted life-sustaining medical treatment can be exercised by opting to refrain from starting a treatment (withholding), or by opting to stop a treatment after it has been started (withdrawing). To decide to “forgo” a treatment may thus mean either deciding not to start a treatment or deciding to stop a treatment. Withholding and withdrawing may be different clinical and psychological experiences. Some health care professionals may perceive these differences as ethically challenging. However, recognizing their legal and moral equivalence is crucial. Relieving suffering, preventing burdens from unwanted or ineffective treatment, and respecting patients’ wishes and preferences or best interests may motivate decisions to withhold a treatment, withdraw a treatment, or both.
No treatment or form of care is intrinsically “ordinary” or “extraordinary.”
The benefits and burdens of a treatment from the patient’s perspective determine the appropriateness of the treatment, regardless of whether the treatment is complex and technologically sophisticated or simpler and more routine. The assessment of the benefits and burdens of medical treatment to the patient has sometimes been cast in terms of “ordinary” and “extraordinary.” This contrast has been understood in different ways, including morally required versus morally optional interventions, standard versus nonstandard care, or simple versus complex treatment. This traditional distinction and language has often confused decision-making and is now less frequently used than in the past. To avoid confusion and prevent burdens to patients resulting from unclear communication, the language of “benefits” and “burdens” is preferable. Patients have rights to refuse treatment, whether that treatment is simple or complex. The decision should hinge not on the simplicity or complexity of the treatment, but on the benefits and burdens of the treatment from the patient’s perspective.
Palliative care is integral to good health care.
The relief of suffering is integral to the role of physicians and nurses as healers and caregivers and to health care as an ethical enterprise. Efforts to cure or manage disease are also efforts to prevent or relieve suffering caused by disease. Patients may experience a failure to relieve their suffering as a harm. All patients deserve expert treatment of their pain and symptoms, and respect as whole persons and social beings during the experience of illness. This is accomplished through the integration of interdisciplinary palliative care (including palliative medicine, palliative nursing, and other palliative professions) into medical treatment and patient care. As with other treatments, the benefits and burdens of a palliative intervention from the patient’s perspective determine the appropriateness of that intervention. The Bibliography includes selections from the extensive literature on palliative care, including empirical research on the benefits of the early integration of palliative care into the treatment plan.
It is ethically acceptable to provide medication sufficient to control a patient’s pain and symptoms even in the rare circumstance in which this intervention may foreseeably hasten the patient’s death.
It is rare that medication properly administered with the intent of relieving pain and symptoms will hasten a patient’s death. Available evidence suggests that when opioids are properly administered to relieve symptoms such as dyspnea or cancer pain, the respiratory depression that may occur is unlikely to lead to respiratory failure.3 Even when medication intended to relieve pain and symptoms does have the foreseeable risk of hastening death, it is ethical for health care providers to prescribe and to administer the pain-relieving medication, following the rule of “double effect.” This rule distinguishes between practices intended to relieve a patient’s pain that may have a secondary, unintended effect of hastening death, and practices that are actually intended to bring about a patient’s death. According to the rule of double effect, an action with two possible effects—one good, one bad—is ethically acceptable if four conditions are satisfied:
- The action itself (here, administering the medication) is not morally wrong.
- The action is undertaken with the sole intention of bringing about the good effect (here, relieving pain or symptoms) even if the bad effect (here, hastening death) is foreseeable.
- The action does not bring about the good effect by means of the bad effect (here, that pain relief is not accomplished by ending life).
- The reason for undertaking an action with two possible effects is clear and urgent (here, that there is a clinical need to administer this medication, at this dose, at this time, in order to relieve pain or symptoms).
Thus, administering medication with a risk of hastening death is morally acceptable if the purpose of doing so is the relief of pain or symptoms, the medication and dose are therapeutically necessary, and, as with any treatment, the patient or surrogate consents to the risk. For selections from the literature on double effect, refer to the Bibliography.
Forgoing life-sustaining treatment is ethically and legally distinct from suicide, from euthanasia, and from physician-assisted suicide.
A patient’s or surrogate’s decision to forgo life-sustaining treatment and allow death to occur is not the ethical or legal equivalent of suicide. When health professionals implement a patient’s or surrogate’s decision to forgo life-sustaining treatment, they are not performing actions that are equivalent to killing the patient (euthanasia) or assisting the patient to end his or her own life by suicide. Although this distinction has occasioned debate, the right to refuse treatment is long settled and well established. Physician-assisted suicide and euthanasia have been far more controversial. A handful of U.S. states allow certain forms of physician-assisted suicide. No state allows euthanasia.
Changes in law and policy in three states since the mid-1990s mean that the legal and social context for these Guidelines differs somewhat from the context of the 1987 Guidelines, which stated that “[O]ur society forbids assisting suicide or active euthanasia, even if the motive is compassionate.” This is still true of euthanasia. The landscape has changed to some extent for assisted suicide.
In 1997, the U.S. Supreme Court issued rulings in two cases (Vacco v. Quill and Washington v. Glucksberg) in which the Court rejected arguments that assisted suicide was a right protected under the U.S. Constitution. However, neither ruling prohibited states from enacting their own laws to permit assisted suicide. In 1997, the state of Oregon enacted the Death with Dignity Act, which permits terminally ill state residents “to end their lives through the voluntary self-administration of lethal medications, expressly prescribed by a physician for that purpose,” thereby permitting physicians to assist in this act under these specific conditions. In 2009, the state of Washington enacted a similar Death with Dignity Act, which permits state residents with a life expectancy of less than six months “to request lethal doses of medication from medical and osteopathic physicians.” In 2010, a decision by the Supreme Court of Montana (Baxter v. Montana) concluded that a physician who assisted a patient to end his or her own life voluntarily would not be liable for prosecution under that state’s homicide statute. The evolving legal situation in Montana is related to, but distinct from, the Oregon and Washington statutes. (For citations, see Cited Legal Authorities.)
The practice authorized by statute in Oregon and Washington is known by various terms, including “assisted suicide,” “physician aid-in-dying,” “physician-assisted death,” and “physician-assisted suicide.” When referring to this practice, the authors of these Guidelines have chosen to use the term “physician-assisted suicide” (PAS) because this term seems to most clearly describe the two key features of what is authorized in these statutes, namely: 1) self-administration of a lethal dose of medication (i.e., suicide), when the medication and dose are 2) prescribed by a physician.
Avoiding confusion between the help or aid that physicians provide in the care of all dying patients and a specific practice authorized in two states has been our primary aim in making this choice of language. In our references to PAS, we have further sought to avoid language that obscures the legal and practical distinction between physician-assisted suicide and physician-administered euthanasia, which is illegal in the United States. We are mindful that some clinicians, scholars, and policy-makers use different terms (or more than one term), and that some commentators have clear preferences among these terms while others do not. We are also mindful that the word “suicide” has specific moral connotations for many members of American society. We conclude that no term is perfect, some terms are less clear or less common than others, and careful definition is crucial to informed discussion.
Published data from Oregon and Washington suggest that reported cases of physician-assisted suicide under those states’ law and protocols are uncommon. In Oregon, for example, the 71 reported deaths that occurred under the provisions of that state’s Death with Dignity Act in 2011 represented 22.5 deaths per 10,000 total deaths.4 It should be noted that Oregon and Washington analyze and tabulate only those cases of physician-assisted suicide that are reported to them as such by physicians, so there may be additional unreported cases.
Health care professionals who care for dying patients in states where PAS is legal may face special ethical challenges. Institutional policy in these states should clarify the rights of patients under state law. This policy should also clarify the rights and responsibilities of physicians, other health care professionals, and institutions with respect both to participation in physician-assisted suicide and to a refusal to participate. The statutes in Oregon and Washington have prompted the development of explicit policies for institutions. Clinician education should support professionals’ ability to obtain accurate information about relevant policies in the institutions in which they practice.
Published 2011 data from Oregon and Washington also suggest that most PAS deaths that occur in each state involve patients who are enrolled in hospice.5 Institutional policy in states where PAS is legal should therefore clarify the responsibilities of hospice workers and hospice programs concerning how to provide good care to a hospice patient who is considering physician-assisted suicide and who wishes to remain enrolled in hospice for the remainder of his or her life. Clinician education in these states should offer hospice workers opportunities to discuss and reflect on the ethical and organizational challenges they may face in providing care to patients who are considering PAS and who disclose this information to hospice caregivers.
Professionals in every state should have opportunities to learn about and reflect on the continuing debate over PAS within American medicine and American society. Clinician education should take special care to describe the Oregon and Washington statutes and policies accurately to ensure that clinicians understand what practices are (and are not) permitted in these states.
These Guidelines discuss patient-initiated inquiries about physician-assisted suicide or euthanasia in Part Two, Section 2 (“Making the Decision at Hand”). The guidelines and commentary on conscientious objection by professionals and institutions in Part Two, Section 2 (“Conflicts and Challenges Related to Treatment Decision-Making”) and Part Two, Section 6 (“Guidelines on Conflict Resolution”) are relevant to refusals to participate in physician-assisted suicide. The Bibliography includes a selection of recent literature offering a range of clinical, ethical, and legal perspectives on physician-assisted suicide and euthanasia, including empirical findings.
Notes to the Introduction
2. A “family,” as this term is used in these Guidelines, may include close friends, or an intimate partner whose relationship to the person may or may not be recognized by law, as well as individuals with a biological or legal relationship to the patient. These Guidelines also use the term “loved ones” to encompass this range of relationships.
3. Margaret Campbell, Kathryn S. Bizek, and Mary Thill, “Patient Responses During Rapid Terminal Weaning from Mechanical Ventilation: A Prospective Study,” Critical Care Medicine 27, no. 1 (1999): 73–77; Katri Elina Clemens, Ines Quednau, and Eberhard Klaschik, “Is There a Higher Risk of Respiratory Depression in Opioid-Naïve Palliative Care Patients during Symptomatic Therapy of Dyspnea with Strong Opioids?” Journal of Palliative Medicine 11, no. 2 (2008): 204–16; Susan B. LeGrand, Elias A. Khawam, Declan Walsh, and Nilo I. Rivera, “Opioids, Respiratory Function, and Dyspnea,” American Journal of Hospice and Palliative Care 20, no. 1 (2003): 57–61.
4. See Oregon Public Health Division, Oregon’s Death with Dignity Act 2011 Annual Report, pp. 1–3, at 1, 2, http://public.health.oregon.gov/ProviderPartnerResources/EvaluationResearch/DeathwithDignityAct/Pages/index.aspx. The corresponding 2011 data from Washington include seventy reported deaths that occurred after the ingestion of lethal medication. See Washington State Department of Health, Washington State Department of Health 2011 Death with Dignity Act Report: Executive Summary, pp. 1–12, at 1, http://www.doh.wa.gov/portals/1/Documents/5300/DWDA2011.pdf.
5. Of the sixty-one out of seventy-one Oregon deaths reported in 2011 for which hospice information was available, fifty-nine (96.7 percent) involved individuals enrolled in hospice, and two (3.3 percent) involved individuals not enrolled in hospice. (The hospice enrollment status of the remaining ten individuals was unknown.) Historically, 89.7 percent of all reported Oregon deaths have involved individuals confirmed as enrolled in hospice. See Oregon’s Death with Dignity Act 2011 Annual Report, Table 1, p. 1. The corresponding 2011 data from Washington include fifty-eight individuals (83 percent) enrolled in hospice and eleven individuals (16 percent) not enrolled, with one whose hospice status was unknown. See Washington State Department of Health 2011 Death with Dignity Act Report, Table 5, p. 9.