Ethics and the Legalization of Physician-Assisted Suicide: An American College of Physicians Position Paper

ethical prohibitions. have concerns about how they will die and the emphasis by medicine and society on intervention and has sometimes come at the expense of good end-of-life care. Some advocated strongly, on the basis of autonomy, that physician-assisted suicide should be a legal option at the end of life. a proponent of patient-centered care, the (ACP) is attentive to all voices, including those who speak of the desire to control when and how life will end. However, the ACP believes the ethical arguments against legalizing physician-assisted suicide remain the most compelling. On the basis of substantive ethics, clinical practice, policy, and other concerns articulated in this position paper, the ACP does not support legalization of physician-assisted suicide. It is problematic given the nature of the patient– physician relationship, affects trust in the relationship and in the profession, and fundamentally alters the medical profession's role in society. Furthermore, the principles at stake in this debate also underlie medicine's responsibilities regarding other issues and the physician's duties to provide care based on clinical judg-ment, evidence, and ethics. Society's focus at the end of life should be on efforts to address suffering and the needs of patients and families, including improving access to effective hospice and palliative care. The ACP remains committed to improving care for patients throughout and at the end of life.

Calls to legalize physician-assisted suicide have increased and public interest in the subject has grown in recent years despite ethical prohibitions. Many people have concerns about how they will die and the emphasis by medicine and society on intervention and cure has sometimes come at the expense of good endof-life care. Some have advocated strongly, on the basis of autonomy, that physician-assisted suicide should be a legal option at the end of life. As a proponent of patient-centered care, the American College of Physicians (ACP) is attentive to all voices, including those who speak of the desire to control when and how life will end. However, the ACP believes that the ethical arguments against legalizing physician-assisted suicide remain the most compelling. On the basis of substantive ethics, clinical practice, policy, and other concerns articulated in this position paper, the ACP does not support legalization of physician-assisted suicide. It is problematic given the nature of the patientphysician relationship, affects trust in the relationship and in the profession, and fundamentally alters the medical profession's role in society. Furthermore, the principles at stake in this debate also underlie medicine's responsibilities regarding other issues and the physician's duties to provide care based on clinical judgment, evidence, and ethics. Society's focus at the end of life should be on efforts to address suffering and the needs of patients and families, including improving access to effective hospice and palliative care. The ACP remains committed to improving care for patients throughout and at the end of life. agreed (2,3). The emphasis by medicine and society on intervention and cure has sometimes come at the expense of good end-of-life care. Inappropriate treatment at the end of life may be harmful and drainingphysically, emotionally, and financially-for patients and their families. Many people have concerns about death. At the end of life, some patients receive unwanted care; others do not receive needed care (4 -6). Some end-oflife concerns are outside of medicine's scope and should be addressed in other ways. Although medicine now has an unprecedented capacity to treat illness and ease the dying process, the right care in the right place at the right time has not been achieved.
Medicine and society still struggle with getting it right for all patients. Although progress has been made, the principles and practices of hospice and palliative medicine have not been fully realized (4). Revisiting these issues in 2014, the IOM's Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life reported that challenges remain in delivering quality end-of-life care to a growing and diverse elderly population, especially with regard to access to care, communication barriers, time pressures, and care coordination (7). Inadequate reimbursement and other disincentives also are barriers to palliative and hospice care.
Hospice and palliative care may ease apprehension about the dying process. Such care requires improving access to, financing of, and training in palliative care; improving hospital, nursing home, and at-home capabilities in delivering care; and encouraging advance care planning and openness to discussions about dying. Of note, 90% of U.S. adults do not know what palliative care is; however, when told the definition, more than 90% say they would want it for themselves or family members if severely ill (4).
Within this context of challenges in providing palliative and hospice care, a few U.S. jurisdictions have legalized physician-assisted suicide. This paper presents the position of the American College of Physicians (ACP) on the topic. The ACP recognizes the range of views on, the depth of feeling about, and the complexity of this issue. This executive summary is a synopsis of

METHODS
This position paper was developed from September 2015 to March 2017 on behalf of the ACP Ethics, Professionalism and Human Rights Committee (EPHRC). Committee members abide by the ACP's conflict-ofinterest policy and procedures (www.acponline.org /about-acp/who-we-are/acp-conflict-of-interest-policy -and-procedures), and appointment to and procedures of the EPHRC are governed by the ACP's bylaws (www .acponline.org/about-acp/who-we-are/acp-bylaws). After an environmental assessment to determine the scope of issues and literature reviews, the EPHRC evaluated and discussed several drafts of the paper; the paper was then reviewed by members of the ACP Board of Governors, Board of Regents, Council of Early Career Physicians, Council of Resident/Fellow Members, Council of Student Members, Council of Subspecialty Societies, Patient Partnership in Healthcare Center and Advisory Board, and other committees and experts. The paper was revised on the basis of comments from the aforementioned groups and individuals, reviewed again by the full leadership, and then revised further. Finally, the ACP Board of Regents reviewed the paper and approved it on 27 March 2017. Financial support for this project is exclusively from the ACP operating budget.

BACKGROUND AND BRIEF RATIONALE
In 2001, the ACP published a position paper opposing legalization of physician-assisted suicide (8). This issue also has been considered every few years in the American College of Physicians Ethics Manual, including the current edition (9). Given recent changes in the legal landscape, public interest in the topic, and continuing barriers to palliative and hospice care, an updated position paper is presented here. Within a framework that considers clinical practice, ethics, law, and policy, this paper provides background, discusses the role of palliative and hospice care, explores the nature of the patient-physician relationship and the distinction between refusal of life-sustaining treatment and physician-assisted suicide, and provides recommendations for responding to patient requests for physician-assisted suicide.
Medical ethics establishes the duties of physicians to patients and society, sometimes to a greater extent than the law (9). Physicians have duties to patients on the basis of the ethical principles of beneficence (that is, acting in the patient's best interest), nonmaleficence (avoiding or minimizing harm), respect for patient autonomy, and promotion of fairness and social justice (9). Medical ethics and the law strongly support a patient's right to refuse treatment, including lifesustaining treatment. The intent is to avoid or withdraw treatment that the patient judges to be inconsistent with his or her goals and preferences. Death follows naturally, after the refusal, as a result of the underlying disease (9).
Ethical arguments in support of physician-assisted suicide highlight the principle of respect for patient autonomy and a broad interpretation of a physician's duty to relieve suffering (10). Proponents view physicianassisted suicide as an act of compassion that respects patient choice and fulfills an obligation of nonabandonment (11). Opponents maintain that the profession's most consistent ethical traditions emphasize care and comfort, that physicians should not participate in intentionally ending a person's life, and that physicianassisted suicide requires physicians to breach specific prohibitions as well as the general duties of beneficence and nonmaleficence. Such breaches are viewed as inconsistent with the physician's role as healer and comforter (12,13).
Both sides agree that patient autonomy is critical and must be respected, but they also recognize that it is not absolute and must be balanced with other ethical principles (9,14). To do otherwise jeopardizes the physician's ability to practice high-value care in the best interests of the patient, in a true patient-physician partnership. Only by this balancing of ethical principles can physicians fulfill their duties, including those in more everyday encounters, such as when a physician advises against tests requested by a patient that are not medically indicated, declines to write an illegal prescription, or breaches confidentiality to protect public health. It also undergirds the physician's duty not to engage in futile care (such as care based on requests for nonindicated cardiopulmonary resuscitation or end-of-life treatment of brain-dead patients under an expansive view of patient autonomy). Physicians are members of a profession with ethical responsibilities; they are moral agents, not merely providers of services (15).
The suffering of dying patients may be great and is caused by somatic symptoms, such as pain and nausea; psychological conditions, such as depression and anxiety; interpersonal suffering due to dependency or unresolved conflict; or existential suffering based in hopelessness, indignity, or the belief that one's life has ended in a biographical sense but has not yet ended biologically. For some patients, a sense of control over the manner and timing of death brings comfort. However, is it reasonable to ask medicine to relieve all human suffering? Just as medicine cannot eliminate death, medicine cannot relieve all human suffering. Both proponents and opponents of physician-assisted suicide wish to alleviate suffering of dying patients, and physicians have an ethical duty to provide competent palliative and hospice care (9). However, is physician-assisted suicide a type of control over suffering and the dying process that is within the goals and scope of medicine?
Balancing respect for patient autonomy against other principles reflects ethical arguments about the nature of the patient-physician relationship-a relationship that is inherently unequal because of power differentials and the vulnerability of illness-physicians' duties, and the role of the medical profession in society. A fuller consideration of this ethical balance, intent and causation in acts near the end of life, medicalization versus personalization of death, and the ethics and implications of physician-assisted suicide are presented in the Appendix (16-81).

POSITION STATEMENT
The ACP affirms a professional responsibility to improve the care of dying patients and their families.
The ACP does not support the legalization of physician-assisted suicide, the practice of which raises ethical, clinical, and other concerns. The ACP and its members, including those who might lawfully participate in the practice, should ensure that all patients can rely on high-quality care through to the end of life, with prevention or relief of suffering insofar as possible, a commitment to human dignity and management of pain and other symptoms, and support for families. Physicians and patients must continue to search together for answers to the challenges posed by living with serious illness before death (9).

CONCLUSION
Society's goal should be to make dying less, not more, medical. Physician-assisted suicide is neither a therapy nor a solution to difficult questions raised at the end of life. On the basis of substantive ethics, clinical practice, policy, and other concerns, the ACP does not support legalization of physician-assisted suicide. This practice is problematic given the nature of the patientphysician relationship, affects trust in that relationship as well as in the profession, and fundamentally alters the medical profession's role in society. Furthermore, the principles at stake in this debate also underlie medicine's responsibilities on other issues and the physician's duty to provide care based on clinical judgment, evidence, and ethics. Control over the manner and timing of a person's death has not been and should not be a goal of medicine. However, through high-quality care, effective communication, compassionate support, and the right resources, physicians can help patients control many aspects of how they live out life's last chapter.
From the American College of Physicians, Philadelphia, Pennsylvania.

Acknowledgment:
The authors and the ACP Ethics, Professionalism and Human Rights Committee thank the many reviewers of the paper for helpful comments on drafts and Kathy Wynkoop for administrative assistance.

APPENDIX AND EXPANDED RATIONALE: ETHICS AND THE LEGALIZATION OF PHYSICIAN-ASSISTED SUICIDE-AN AMERICAN COLLEGE OF PHYSICIANS POSITION PAPER Framing the Issues: Care Near the End of Life
We all will die. How we die-and live at the end of life-is important, with implications for individuals, their families, and society. How we are cared for at the end of life matters.
The groundbreaking 1997 report Approaching Death: Improving Care at the End of Life, by the IOM, documented inadequate end-of-life care in the United States (1). In 2000, the SUPPORT investigators agreed (2,3). Although the cultural norm of fighting disease aggressively is the right approach in many cases, the emphasis by medicine, as well as society, on intervention and cure sometimes comes at the expense of good end-of-life care. Inappropriate treatment at the end of life may be harmful and draining-physically, emotionally, and financially-for patients and their families. Many of us have concerns or apprehensions about how we will die. Indeed, some patients receive unwanted care at the end of life, whereas others do not receive the care they need (4 -6). Although medicine now has an unprecedented capacity to treat illness and ease the dying process, the right care in the right place at the right time has not been achieved.
Medicine and society still struggle to get it right for all patients. Although progress has been made, the principles and practices of hospice and palliative medicine have not been fully realized (4). Revisiting these issues in 2014, the IOM report Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life found that challenges remain in delivering quality end-of-life care to a growing and diverse elderly population, especially regarding access to care, communication barriers, time pressures, and care coordination (7). Inadequate reimbursement and other disincentives also create barriers to palliative and hospice care.
Wide agreement exists that hospice and palliative care may ease apprehension about the dying process. Such care requires improving access to, financing of, and training in palliative care; improving hospital, nursing home, and at-home capabilities in delivering care; and encouraging advance care planning and openness to discussions about dying. Of note, 90% of U.S. adults do not know what palliative care is, but when told the definition, more than 90% say they would want it for themselves or family members if severely ill (4).
Access to state-of-the-art symptom control remains limited for all dying patients. Of particular concern, evidence of ethnic and racial disparities in access, outcomes, and communication is increasing (5, 6). Many patients fear they will not receive appropriate end-oflife care when they need it. Others are concerned about being a financial, physical, or other burden on their family, losing autonomy or control, or being placed in a long-term care facility. Some are alone or lonely; loneliness has a mortality risk similar to that of cigarette smoking, yet its health implications are underappreciated (16). Many persons approaching death are clinically depressed or have other psychiatric comorbid conditions, and some contemplate suicide (17,18). According to Wilson and colleagues, "the expression of a desire for death by a terminally ill patient should raise a suspicion about mental health problems; by itself, however, it is not definitively diagnostic of one" (17). This desire fluctuates over time (19,20) and may be related to inadequate symptom management. Medicine can and should ameliorate many of these problems; some, however, are outside the scope or goals of medicine and should be addressed in other ways.
As challenges in providing palliative and hospice care continue, a few jurisdictions have legalized physician-assisted suicide (see the Glossary for definitions and the Appendix Table for U.S. jurisdictions with physician-assisted suicide laws). The ACP recognizes the range of views, depth of feeling, and complexity of the issue of physician-assisted suicide. Appendix

Revisiting Physician-Assisted Suicide
In 2001, the ACP published a position paper opposing legalization of physician-assisted suicide (8). The issue also has been considered every few years in the American College of Physicians Ethics Manual, including the current edition (9). Given recent changes in the legal landscape, public interest in the topic, and continuing barriers to palliative and hospice care, an updated position paper is presented here. Within a framework that considers clinical practice, ethics, law, and policy, this paper provides background, discusses the role of palliative and hospice care, explores the nature of the patient-physician relationship and the distinction between refusal of life-sustaining treatment and physician-assisted suicide, and provides recommendations for responding to patient requests for physician-assisted suicide.

The Context
Physician-assisted suicide is medical help with a patient's intentional act to end his or her own life (for example, an individual taking a lethal dose of medication prescribed by a physician for that purpose). It is ethically, legally, and clinically different from patient refusal of life-sustaining treatment through the withdrawal or withholding of treatment. Physician-assisted suicide also differs from euthanasia, an act in which a physician intentionally terminates the life of a patient (such as by lethal injection), the purpose of which is to relieve pain or other suffering (8). Dictionaries define suicide as intentionally ending one's own life. Despite cultural and historical connotations, the term is neither disparaging nor a judgment. Terms for physician-assisted suicide, such as aid in dying, medical aid in dying, physicianassisted death, and hastened death, lump categories of action together, obscuring the ethics of what is at stake and making meaningful debate difficult; therefore, clarity of language is important.
Although suicide and attempted suicide have been decriminalized in the United States, assisting a suicide remains a statutory offense in most states. Euthanasia is illegal everywhere in the United States. In New Mexico, a lower-court decision authorized physician-assisted suicide, but it was struck down; like all appellate courts, the New Mexico Supreme Court ruled that there is no right to physician-assisted suicide. Elsewhere in the world, in 2015, the Parliament of the United Kingdom voted down a physician-assisted suicide bill, 330 to 118, and Canada legalized both physician-assisted suicide and euthanasia. In 2016, the Parliament of South Australia rejected a bill on euthanasia. Physicianassisted suicide and euthanasia are legal in the Netherlands, Belgium, and Luxembourg; euthanasia is legal in Colombia; and Switzerland has decriminalized assisted suicide.

Principles of Medical Ethics and Arguments, Pro and Con
Medical ethics establishes the duties of physicians to patients and society, sometimes to a greater extent than the law (9). Physicians have duties to patients based on the ethical principles of beneficence (acting in the patient's best interest), nonmaleficence (avoiding or minimizing harm), respect for patient autonomy, and promotion of fairness and social justice (9). Medical ethics and the law strongly support a patient's right to refuse treatment, including life-sustaining treatment. The intent is to avoid or withdraw treatment that the patient considers unduly burdensome and inconsistent with his or her health goals and preferences. Death follows naturally after the refusal, due to the underlying disease (9).
Ethical arguments in support of physician-assisted suicide highlight the principle of respect for patient autonomy and a broad interpretation of a physician's duty to relieve suffering. The decision to intentionally end one's life is regarded as intensely private and therefore should not be prohibited (10). Seeking physicianassisted suicide is most frequently associated with concerns about loss of autonomy and control, decreasing ability to participate in enjoyable activities, and loss of dignity, rather than pain or other symptoms (21,22). For persons who seek this type of control, palliative and hospice care are not the issue-they often are already receiving those services. In Oregon, the state with the most experience, 1327 persons have obtained prescriptions for lethal doses of medications under the law since 1997; 859 died after taking the medication. Of 105 deaths during 2014, 68% occurred in persons older than 65 years, 95% were white, 56% were men, 48% were persons with a baccalaureate degree or higher, and 69% had cancer (21). More recent justifications present physician-assisted suicide as a personal choice, avoiding discussion of important medical ethics considerations (11).
Proponents of physician-assisted suicide view it as an act of compassion that respects patient choice and fulfills an obligation of nonabandonment (11). In support of legalization, they also argue that some patients receiving a lethal prescription ultimately do not use it. In addition, some maintain that physician-assisted suicide already occurs where it is illegal (23), so legalization would result in standardization, transparency, and monitoring.
Opponents maintain that the profession's most consistent ethical traditions emphasize care and comfort and that physicians should not participate in intentionally ending a person's life (12). Physician-assisted suicide requires physicians to breach specific prohibitions as well as the general duties of beneficence and nonmaleficence. Such breaches are viewed as inconsistent with the physician's role as healer and comforter (13). Pronouncements against physician-assisted suicide date back to Hippocrates. Opponents agree that patient autonomy is critical and must be respected but recognize that it is not absolute and must be balanced with other ethical principles (9,14). To do otherwise jeopardizes the physician's ability to practice high-value care in the best interests of the patient, in a true patient-physician partnership. Only by such a balance of ethical principles can physicians fulfill their duties, including those in more everyday encounters, such as when a physician advises against tests requested by a patient that are not medically indicated, declines to write illegal prescriptions, or breaches confidentiality to protect public health. It also undergirds the duty that physicians not engage in futile care (for example, care based on requests for nonindicated cardiopulmonary resuscitation or end-of-life treatment of brain-dead patients under an expansive view of patient autonomy). Physicians are members of a profession with ethical responsibilities; they are moral agents, not merely providers of services (15).
Death certificate requirements under physicianassisted suicide laws ask physicians to list the cause of death as the underlying illness, not the new pathology caused by ingestion of a lethal dose of medicine (24), which seems inconsistent with the physician's duty of honesty. Moreover, although individual physicians may decline to participate, conscientious objection to physician-assisted suicide does not address the fundamental ethical objections to it.
The suffering of dying patients may be great; it is caused by somatic symptoms, such as pain and nausea; psychological conditions, such as depression and anxiety; interpersonal suffering due to dependency or unresolved conflict; or existential suffering based in hopelessness, indignity, or the belief that one's life has ended in a biographical sense but has not yet ended biologically. For some patients, a sense of control over the manner and timing of death brings comfort. However, is it reasonable to ask medicine to relieve all human suffering? Just as medicine cannot eliminate death, medicine cannot relieve all human suffering; attempting to do so ultimately leads to bad medical care (25). Good medicine demands compassion for the dying, but compassion also needs reason (26). Both proponents and opponents wish to alleviate suffering of dying patients, and physicians have an ethical duty to provide competent palliative and hospice care (9). However, is physician-assisted suicide a type of control over suffering and the dying process that is within the goals and scope of medicine?
Balancing respect for patient autonomy against other ethical principles reflects arguments about the nature of the patient-physician relationship, physicians' duties, and the role of the medical profession in soci-ety. In fact, one may argue that making physicians arbiters of assisted suicide is a return to paternalism and not a power physicians should want (27), that "the legalization of physician-assisted suicide does not empower patients; it empowers physicians" (28).
Legalization of physician-assisted suicide also raises social justice issues. Society and the medical profession have duties to safeguard the patient-physician relationship and human dignity. These duties apply especially to the most vulnerable members of society: the sick, the elderly, children, the disabled, the poor, minorities, and others. Some individuals might view themselves as unproductive or burdensome and, on that basis, as candidates for assisted suicide, especially if a physician raises it or validates a request. Physicianassisted suicide laws have been associated with a 6% increase in total suicides (15% in those older than 65 years) in the states where physician-assisted suicide is legal, controlling for state-specific time trends (29, 30). Although a recent study did not find vulnerable groups being pressured to accept physician-assisted suicide, it did raise questions about a lack of data on complications and on how many physicians may have assisted without reporting (31). Vulnerable communities and individuals raise strong concerns that legalization leads to attitudinal changes, subtle biases about quality of life, and judgments that some lives are not worth living (32, 33). National disability groups are opposed to physician-assisted suicide (32, 34). One article reported various opinions among focus group participants (35). Finally, advocating for physician-assisted suicide where there is no general right to health care and access to hospice and palliative care services is limited, especially in an era of health care cost containment, is ironic (8).

Ethics and the Nature of the Patient-Physician Relationship
The ACP's main concerns in this debate are ethical ones. The patient-physician relationship is inherently unequal. Physicians have specialized medical knowledge, training, experience, and prescribing powers that patients do not. Illness makes patients vulnerable (including physicians who are patients [36,37]). Patients disrobe, are examined, and disclose intimate information to their physicians. The physician must earn the patient's trust, preserve his or her confidentiality, and act as a fiduciary. Physicians publicly profess that they will act for the benefit of their patients, putting patients' welfare and best interests first and helping them cope with illness, disability, suffering, and death. The physician has a duty to respect the dignity and the cultural and spiritual uniqueness and traditions of every patient (9).
Physician-assisted suicide and euthanasia were common during the time of Hippocrates, leading to their specific prohibition in the Hippocratic Oath (38). Together with the prohibition of sexual relationships between physicians and patients and the duty to maintain patient confidentiality, the Oath provides a context for a therapeutic alliance to prevent the exploitation of patient relationships.
The Hippocratic Oath, of course, is not followed word for word today; however, it has been analyzed and applied over time in light of its fundamental principles. Acting in the best interests of the patient and recognizing the special nature of the patient-physician relationship, principles and prohibitions set ethical boundaries to prevent misunderstandings and misuse of medical authority. These boundaries encourage patients to be open and honest regarding intimate health matters in a safe space, in the context of a trusted relationship.
Physicians can influence patients, even in ways physicians may not appreciate. Patients seeking physicianassisted suicide may seek validation to end their lives. Indeed, studies have shown that socially isolated, vulnerable persons seek social support and contact through visits with their physicians (16). Physicians may influence patients based on their own fears of death and disability (39). Evidence also suggests that many physicians who participate in physician-assisted suicide are adversely affected by the experience (40). Some commentators question whether assisted suicide needs to be physician assisted and whether others might provide assistance instead (41).

The Ethics of Refusal of Treatment and Providing Symptom Control: A Closer Look at Intent and Causation
For decades, the consensus has been that after a careful weighing of patient autonomy, beneficence, nonmaleficence, and societal interests, a patient may forgo life-sustaining treatment. Although Hippocratic writings explicitly proscribe euthanasia and physicianassisted suicide, they deem treatment abatement ethically appropriate in patients who are "overmastered by disease" (42). Although some lower courts have questioned the importance of this distinction (43), the U.S. Supreme Court has distinguished the refusal of treatment from suicide (44,45). Withdrawal of treatment based on patient wishes respects the patient's bodily integrity and right to be free of unwanted treatment. Physician-assisted suicide and euthanasia are interventions done with the intent to end the patient's life (46,47). This distinction is ethically and legally important (9).
Some argue that withdrawing treatment on the basis of patient wishes-an omission, such as forgoing a mechanical ventilator in a patient with respiratory failure-and prescribing a lethal dose of medicine for the patient's use-a commission-are equivalent, because they both are acts that lead to the patient's death. However, commission (doing something) versus omission (not doing something) is not alone determinative. Withdrawing ventilator support is an act, but the act merely removes an intervention that prevented a preexisting illness from running its course. The aim of the act is not to terminate the patient's life (47). Intent and causation are critical factors in distinguishing physician-assisted suicide from withdrawal of life-sustaining treatment.
Death may be accelerated if a patient requests withdrawal of a life-sustaining treatment and that request is carried out. However, the patient could have refused the treatment when it was originally offered; therefore, he or she may request its withdrawal after it is started. If not for the intervention to which the patient consented, death would have occurred as a result of the underlying disease. As the International Association for Hospice and Palliative Care, citing the European Association for Palliative Care, stated, "Withholding or withdrawing ineffective, futile, burdensome, and unnecessary life-prolonging procedures or treatments does not constitute euthanasia or PAS [physicianassisted suicide] because it is not intended to hasten death, but rather indicate the acceptance of death as a natural consequence of the underlying disease progression" (48).
The intent of treatment refusal is freedom from an unwanted intervention. A natural death follows due to the underlying disease (in fact, imposing unwanted treatment is a bodily invasion and is considered unethical and an illegal battery). In contrast, if a person disconnects a ventilator without patient consent and the patient subsequently dies, that person has acted wrongly. In both instances, the patient dies after withdrawal of life-sustaining treatment, but in very different ways under ethics and the law. Death by medication overdose is not a natural death due to an underlying medical condition.
Research advances have introduced new lifesustaining technologies into clinical practice. For example, many patients have life-sustaining devices, such as pacemakers, implantable cardioverter-defibrillators, and ventricular assist devices. Physicians inevitably encounter patients whose underlying disease no longer is being treated effectively by the device or who have a terminal illness the device cannot treat (such as cancer). Desiring a natural death, patients or their surrogates may request withdrawal of therapies delivered by these devices. In these situations, the death that follows is due to the underlying heart disease or other comorbid conditions (49,50). Physicians should honor these requests. However, without a firm line drawn between withdrawal of life-sustaining treatment and physicianassisted suicide, or because of confusion between the two, some physicians might consider discontinuation of these therapies as intentional killing and refuse to im-plement such requests. Patients and families often, but not always, see the line.
Intent and causation also are critical factors in providing pain or symptom relief. Competent provision of symptom control is an ethical duty (9). Patients often fear the prospect of unrelieved pain. Some physicians withhold pain medication because of ungrounded concerns that higher doses may accelerate death through respiratory suppression or that the patient may become addicted to the medication. Appropriate pain relief, however, rarely results in either (51,52), and patients and families need to understand this (52). Under the rule of double effect, strong ethical support exists for increasing pain medication for terminally ill patients if the intent is to relieve pain, even if it might shorten life (9,53,54).
The rule of double effect holds that an action undertaken with the intent of achieving a benefit is morally acceptable even if it has a harmful side effect, provided that the harmful side effect is not intended, the side effect is not the cause of the benefit, and the benefit outweighs the harm. Vigorous management of pain and symptoms, such as dyspnea and nausea, at the end of life is ethical, even if the risk for shortening life is foreseeable, if the intent is to relieve those symptoms. The beneficial effects are pain and symptom control; the rare but potential harmful effect is respiratory suppression, but it is not intended. If the intent was to cause death, or to cause death to relieve pain, it would not be permissible. Likewise, it would not be in keeping with the rule of double effect to use pain control to "treat" loneliness, depression, being tired of living, or existential suffering.

Law and Ethics: U.S. Supreme Court Decisions on Assisted Suicide
Although the language of rights is sometimes invoked, there is no right to physician-assisted suicide in the United States. In fact, in landmark decisions, the U.S. Supreme Court overruled 2 lower courts that had found a constitutional right (45, 55). The lower-court rulings differed in important ways. In Compassion in Dying v. Washington (56), the U.S. Court of Appeals for the Ninth Circuit had held that persons have a right to choose how and when they die. As applied to the limited circumstance of the competent, terminally ill adult who wants a physician's prescription for a lethal dose of medication, the Washington State criminal statute banning physician-assisted suicide was found unconstitutional as a violation of the Due Process Clause of the 14th Amendment, which says a state may not "deprive any person of life, liberty, or property without due process of law." In contrast, in Quill v. Vacco (43), the U.S. Court of Appeals for the Second Circuit specifically declined to "identify a new fundamental right." It said a New York law was unconstitutional on much narrower grounds, as a violation of the Equal Protection Clause of the 14th Amendment, because competent patients at the end of life were being treated differently: Some patients could refuse life-sustaining treatment and thereby accelerate death, but others were prohibited from seeking prescriptions from physicians to hasten death. The Equal Protection Clause says that no state shall "deny to any person within its jurisdiction the equal protection of the laws." The U.S. Supreme Court found both lower-court decisions unpersuasive. Instead, it found refusal of treatment and physician-assisted suicide to be very different. Refusal of treatment, the Court concluded, means being free of the bodily invasion of unwanted medical treatment-a right to be left alone, not a right to something. This is a "negative right"-a form of right of which Americans have many-and differs from a positive right to secure assistance to kill oneself and control the manner and timing of death. Lending support to the rule of double effect, Justice Sandra Day O'Connor pointed out in her concurring opinion that vigorous pain control for the dying is ethical and available: " . . . a patient who is suffering from a terminal illness and who is experiencing great pain has no legal barriers to obtaining medication, from qualified physicians, to alleviate that suffering, even to the point of causing unconsciousness and hastening death." This would include what some refer to as palliative sedation or terminal sedation, although a more accurate term would be double-effect sedation.
The U.S. Supreme Court ruled that there is no constitutional right to assisted suicide and that states may prohibit it. However, the Court also left open the possibility that individual states could legalize it.

Slippery Slopes
Although the ACP's fundamental concerns are based on ethical principles, research suggests that a "slippery slope" exists in jurisdictions where physicianassisted suicide and euthanasia are legal. In the Netherlands, requests are granted for patients whose "medical condition" is categorized as "tired of living." Many patients report "loneliness" and "psychological suffering" as symptoms (57). One study found that persons receiving euthanasia or physician-assisted suicide in the Netherlands for psychiatric disorders were mostly women with complex and chronic psychiatric, medical, and psychosocial histories, and disagreement about patient eligibility among physicians was not unusual (58,59). In Oregon, referrals for psychiatric evaluations have been infrequent (60); in 2014, only 3 of 105 persons who died under the law were referred for formal psychiatric or psychological evaluation. In a study from Belgium, death by euthanasia increased from 2% in 2007 to 5% in 2013. Similarly, approvals of euthanasia requests increased from 55% in 2007 to 77% in 2013 (61). An editorial said these trends were "worrisome" and "require that [the slippery-slope concern] be taken very seriously" (62).
A recent review found that safeguards and controls in jurisdictions where physician-assisted suicide and euthanasia are legal are not always followed (63), and concerns have been raised about underreporting (31). Subtle long-term changes in attitudes are difficult to detect. For example, although only a small number of persons have requested physician-assisted suicide in Oregon, as noted earlier, questions arise regarding whether that fact lessens these and other concerns.
Limiting physician-assisted suicide to the terminally ill is said to be a safeguard, but prognostication raises practical concerns. Laws such as Oregon's require a consultation from a second physician to confirm the diagnosis and prognosis. However, predicting how long a terminally ill patient will live or to what extent cognitive capacity will be impaired by disease or injury often is difficult. In addition, many patients do not have longstanding relationships with physicians who know them well. Furthermore, current safeguards are likely to be challenged. Restricting physician-assisted suicide to terminally ill adults with decision-making capacity raises legal concerns about arbitrary discrimination (64). Fairness, it may be argued, would require granting access to decisionally incapable and non-terminally ill persons. Also, because some patients cannot take pills, arbitrary discrimination could be asserted, unless the practice is broadened from physician-assisted suicide to euthanasia.

Dying Well: Moving From Medicalization to Personalization of Death
Is a medicalized death a good death? Have we already gone too far down a path in which dying patients receive unwanted technology in the intensive care unit while their family members are regarded as "visitors"? Is the solution medicalization of death through medication overdose? Physician-assisted suicide is not a therapy. It runs counter to the goal of the patient rights movement to empower patients to experience a more natural death.
Medicalizing death does not address the needs of dying patients and their families. What is needed is care that emphasizes caring in the last phase of life, facilitating a natural dying process, and humanizing institutions that are used only when those settings are unavoidable. The 3 Wishes Project shows how even simple, nontechnologic approaches in the hospital intensive care unit can improve care, ease dying, enhance dignity, and give voice to patients and families while deepening the sense of vocation among clinicians (65). The 3 Wishes researchers said the project: . . . aimed to integrate palliative care and spiritual care into critical care practice. Eliciting and honoring wishes fostered a community of caring, promoting patient-and familycenteredness as a core component of palliative care. It encouraged the verbalization and realization of unmet spiritual needs, whether secular or faith-based. Our findings underscore the drive that we all have to search for meaning, memories, and closure in anticipation of death while helping to create preparedness, comfort, and connections during the dying process.
In "A Modern Ars Moriendi," a physician recounts the death of her rancher father, noting the challenges they faced trying to refuse hospital treatment. Ultimately, his wishes were met by going home and changing the "focus from life-prolonging technology to lifeenriching community" (66). Earlier hospice care, avoiding the intensive care unit in the last month of life, and experiencing death at home are associated with family perceptions of better care for cancer patients (67). Studies have found regional variations in end-oflife care, with "little relationship to patient preference," but some evidence of lower-intensity care when the primary care physician is more involved in care (68). Longitudinal relationships should be valued and supported by health care systems and payers.
Home is where most patients want to die (69), and even the discontinuation of ventilators (70) or implantable cardiac devices (71) can be done compassionately and effectively at home with hospice care. This approach is more patient centered and a better use of resources when hospital care is not truly necessary. This is the control the medical profession can and should give patients and their families. Dying well requires science and an art of caring for the dying.

Medicine's Role in a Societal Decision
The ACP recognizes that some patient cases will be medically and ethically challenging, that autonomybased arguments in support of legalization of physician-assisted suicide are compelling, and that some might find physician-assisted suicide justifiable in rare circumstances. Patients have the ultimate authority over their lives, but whether physicians should assist them in carrying out suicide is another matter.
Despite changes in the legal and political landscape, the ethical arguments against legalization of physician-assisted suicide remain the most compelling. We are mindful that ethics is not merely a matter for a vote. Majority support of a practice does not make it ethical. Medical history provides several cautionary examples of laws and practices in the United States (such as racial segregation of hospital wards) that were widely endorsed but very problematic. Furthermore, the ACP does not believe neutrality on this controversial issue is appropriate. The medical profession should not be neutral regarding matters of medical ethics (9). The ACP is not neutral on practices that affect the patient-physician relationship and trust in the profession, such as laws that restrict or mandate discussions with, or certain recommendations for, patients. According to the American College of Physicians Ethics Manual, physicians have a duty to come forward, to "clearly articulate the ethical principles that guide their behavior in clinical care, research, and teaching, or as citizens or collectively as members of the profession. It is crucial that a responsible physician perspective be heard as societal decisions are made" (9).
A few patients want to control the timing and manner of death; many more are fearful of what living the last phase of life with serious illness will be like. To the extent that the debate about legalizing physicianassisted suicide is a dilemma because of the failings of medicine to adequately provide comfort and good care to dying patients, medicine should do better. Legalized physician-assisted suicide medicalizes suicide (72). Physician-assisted suicide is not a private act but a social one, with effects on family, community, and society.

Responding to Patient Requests for Assisted Suicide
Etymologically, to be compassionate means to "suffer with" another person; remaining with a dying patient is the essence of nonabandonment (73). When the patient's suffering is interpersonal, existential, or spiritual, care coordination is necessary, and the roles of the physician are to remain present; provide compassionate care; and enlist the support of social workers, psychologists, hospice volunteers, chaplains, and family in addressing sources of suffering that are beyond the scope of medical care.
Regardless of jurisdiction, physicians may encounter patients who request physician-assisted suicide (or express fear of suffering with death). Patient concerns and reasons for the request should be discussed thoroughly. As for all patients nearing the end of life, the physician should: 1. Be present (74), listening to the patient and keeping dialogue open, exploring the reasons for the request, trying to understand its meaning and seeking alternative solutions where possible.
2. Affirm that he or she will care for and not abandon the patient, accompanying and advising the patient through the journey of end-of-life care (studies suggest "the desire to hasten death is future focused and appears to be related to fear of distress and not coping, rather than with current levels of distress or coping ability" [75]).
3. Discuss patient goals of care and the nature of curative and comfort care, explaining a both/and approach to disease-oriented and palliative care as well as an either/or approach and asking, for example, how do you hope I can help you? 4. Facilitate advance care planning and an understanding of surrogate decision making, as desired by the patient.
5. Ensure that the patient is fully informed of the right to refuse treatments and what that entails.
6. Discontinue or do not start medications and interventions that interfere with the patient's values, goals, and preferences.
7. Assess and treat the patient's pain and other distressing physical and psychological symptoms.
8. Assess and optimize patient function through a whole-patient focus.
9. Coordinate, as desired by the patient, the efforts of other members of the health care team, and use community-based resources to address financial, emotional, and spiritual burdens on the patient and family.
10. Prepare the patient and family for what they can expect as illness progresses, addressing uncertainty together and ensuring that the patient and family have informed expectations, including, for example, an understanding that advanced illness often entails a natural loss of appetite and thirst.
11. Regularly assess the patient's status and decision-making capacity.
12. Arrange hospice care at home if that is the patient's preference, being cognizant that palliative and hospice care expertise should be used as early as is indicated. Many patients in the United States receive such care too late or not at all.
Requests for physician-assisted suicide are unlikely to persist when compassionate supportive care is provided (76, 77). However, providing this care may be challenging, especially in today's time-pressured health care environment. It requires us to reflect and act on " . . . the original purpose of physicians' work: to witness others' suffering and provide comfort and care . . . the privilege at the heart of the medical profession" (78).
Physicians should consult with colleagues in caring for the patient and family but also seek support for themselves. According to Kearney and colleagues, "Self-care is an essential part of the therapeutic mandate" (79). Collegial support also reinforces better care of the patient and family. Describing a phone conversation with a colleague about the shared care of a patient, a physician reflects that it was, "A call whose sole but worthy purpose was to say, 'I feel powerless, and I know you do, too, so let's talk this over.'" Yet, it " . . . allowed two physicians to share . . . and reconcile to the inevitable. All too often, we announce our triumphs but camouflage our losses, as if the death of a patient rep-resents a personal failure. In hindsight, acknowledging the impending loss enabled appropriate palliation for the patient and timely pastoral care for her husband . . . " (80).
The need to ensure the central role of families in care; provision of consistent, high-quality care; and education, training, and support of physicians were identified as overarching themes in a series of reports on end-of-life care recently issued by the British Medical Association (81). The British Medical Association and Australian Medical Association both reaffirmed opposition to legalization of physician-assisted suicide and euthanasia in 2016.

Conclusion
The art of medicine is arguably most needed as patients live out the last phase of life. Society's goal should be to make dying less, not more, medical. The ACP affirms a professional responsibility to improve the care of dying patients and their families.
The ACP does not support the legalization of physician-assisted suicide, the practice of which raises ethical, clinical, and other concerns. The ACP and its members, including those who might lawfully participate in the practice, should ensure that all persons can rely on high-quality care through to the end of life, with prevention or relief of suffering insofar as possible, a commitment to human dignity and the management of pain and other symptoms, and support for family. Physicians and patients must continue to search together for answers to the challenges posed by living with serious illness before death (9).
Control over the manner and timing of a person's death has not been and should not be a goal of medicine. However, through high-quality care, effective communication, compassionate support, and the right resources, physicians can help patients control many aspects of how they live out life's last chapter. Throughout patients' lives, including as they face death, medicine must strive to give patients the care, respect, and comfort they deserve.