Autonomy Alone Does Not Validate Physician Assisted Suicide

However we may define autonomy, it is but one factor in judging the ethics of assisted suicide.



“Send not to know/ For whom the bell tolls,/ It tolls for thee.”

-John Donne

We appreciate Don A. Merrell, PhD’s detailed rejoinder to our 2 essays on “autonomy” in the context of physician-assisted suicide (PAS). Prof Merrell’s perspective as a philosopher is welcome, even though we do not find his conclusions persuasive. Indeed, we believe his argument rests largely on several misunderstandings of our basic claims. In addition, there are fundamental differences between the goals of philosophical analysis—useful though it is—and the urgent needs and demands of clinical medicine and medical ethics.

Before we elaborate on these issues, we want to stipulate that the principle of “autonomy” is not determinative of the essential “rightness” or “wrongness” of PAS, which must be decided on more comprehensive grounds. These include the additional principles of beneficence, malfeasance, and justice1; the historical ethos of Hippocratic medicine2; and other broad psychosocial and contextual issues. Specifically, we reject the notion that a single, highly individualistic concept of autonomy has undisputed primacy in medical ethics, to the exclusion of other equally important ethical virtues and human values.

Moreover, autonomy itself is not a unitary construct, but a diverse constellation of subtypes.3 Of particular importance in end-of-life care is the concept of relational autonomy, which “…acknowledges the central role of others in decision-making, including clinicians, who have a responsibility to engage patients' and surrogates' emotional experiences and offer clear guidance when patients are confronting serious illness.”4 We will expand on this concept presently.

Prof Merrell focuses almost exclusively on the concept of autonomy and does not indicate whether he is opposed to PAS itself. In contrast, we have repeatedly stated our opposition to PAS5 which we view as wrong in and of itself (malum in se). Indeed, we (and our colleagues) have argued that even the physician’s participation in assessments of “competency” for PAS is itself unethical.6 In short, whether or not Merrell’s position on autonomy is correct, it is in no sense dispositive regarding the moral status of PAS.

Clearing Up Some Confusion

Prof Merrell cites 2 articles in which we argue that the structure of PAS in Oregon-type models provides only pseudoautonomy for the dying patient. The thrust of Prof Merrell’s critique of this work is that we are mistaken about the nature of autonomy. He claims that our “biggest problem” is that we “…seem to accept a very naïve account of autonomy”—one that entails “omnipotence” on the part of the patient. Merrell writes that, on the contrary,

“Having autonomy does not entail omnipotence in being able to satisfy one’s wishes and desires, though this is the conception of autonomy that Pies and Geppert think is underwriting the autonomy-based defense of PAS.”

This last statement is very puzzling. It is not clear whether Merrell imputes this so-called “omnipotence” concept of autonomy to us, Pies and Geppert; or whether he is saying that we impute it to those who defend PAS. So, to be clear: nowhere in either of our articles do we use the term “omnipotence,” nor do we argue that the concept of individual autonomy entails, or ought to entail, “omnipotence” on the part of the patient. Neither do we believe—as Merrell seems to think—that individual autonomy entails the unfettered ability to do as one pleases. We recognize that the patient’s autonomy must exist within—and be limited by—a system that requires the authorization and cooperation of physicians, nurses, etc.

Autonomy of Action versus Autonomy-Based Systems

We believe Prof Merrell’s main objections to our thesis stem largely from a failure to distinguish autonomy of action (“act autonomy”) from autonomy-based systems(“systems autonomy”). We would acknowledge that whether a patient is requesting elective surgery or assisted suicide, they may sometimes be actingwith rational autonomy, as Merrell defines it: ie, the person’s actions “result from their own beliefs, values, and choices, not from someone else’s…”; and are “relevantly informed” by “an understanding of options” that pertain to the decision at hand. So far, so good.

However, the legislatively-defined systemof PAS in states like Oregon is in no meaningful sense autonomy-based. On the contrary, at every step of the PAS process, it is heteronomy—someone else’s rules—that remains the determining principle. Prof Merrell seems unaware that, at any step in the PAS process, the patient’s physician (or consultant) can change their mind regarding any component of the PAS process. For example, the physician can decide—correctly or not—that the patient is no longer mentally competent, or that the putative “terminal” illness has been mislabeled and can indeed respond to treatment. (NB: under Oregon’s PAS rules, the patient is not obligated to receive treatment, even if it could arrest or reverse the disease process; hence, patients can become “terminally ill” in virtue of their refusal of treatment). Furthermore, the very state legislature that enacted Oregon’s Death with Dignity Act in 1997 could, in its next session, vote to repeal the legislation. Thus, we stand by our characterization of this paradigm as one of pseudoautonomy.

Here is a rough analogy that may help illustrate our argument. In principle, one can decide—rationally and autonomously—to climb Mt. Everest. This is an instance of “act autonomy.” There may of course be many external hurdles that might frustrate this decision, such as the presence of an avalanche or a lack of equipment. But these hurdles alone do not vitiate rational decisional autonomy. However, if each stage of the climb requires permission from the Himalayan authorities—who could decide at any time that you are no longer permitted to make the climb—we cannot reasonably describe this “Mt. Everest system” as autonomy-based.

If Prof Merrell wants to counter with the claim that the same contingencies apply to any medical procedure—ie, they are always subject to the ongoing approval and cooperation of doctors, nurses, etc—we would readily agree. But this does not impugn our main claim. It merely illustrates that when patients enter the House of Medicine for any reason, they give up a great deal of control over specific decisions. Patients do not get to select their anesthetic for a surgical procedure, or the type of antibiotic they receive for an infection. And patients do not get to alter the physician’s orders—whereas the physician is always free to do so.

Deficiencies in DNR Orders Do Not Justify PAS

With respect to do not resuscitate (DNR) orders, Prof Merrell’s understanding of this intervention does not reflect current clinical practices in advance care planning. To be sure, taken in isolation, DNR discussions often “…fail to satisfy criteria for informed consent…[and] often omit essential information to allow patients or surrogates to make informed decisions.”7

Indeed, we would reject the physician-driven, reductionist form of DNR orders that Prof Merrell describes. In modern practice, DNR orders are merely 1 component of discussions among patients, physicians, and nurses relating to life-sustaining treatments, such as artificial nutrition, ventilator support, or cardiopulmonary resuscitation.8 This approach respects the fuller form of relational autonomy that we endorse. In any case, defects in the DNR process in no way constitute an argument favoring PAS.

A Genuinely Autonomous Act Near the End-of-Life

We would contrast the PAS process with one in which a patient in the final stages of a terminal illness decides to leave the medical system entirely and exercise true autonomy; ie, refuses further medical interventions; returns home; and elects to stop taking in fluids or food (voluntary stopping of eating and drinking, VSED). VSED is not only an act of genuine autonomy; it is also one that extricates the patient from a system that is not autonomy-based. Indeed, VSED is recognized as an ethically defensible approach to the dying patient, and—importantly—one which also appears to result in better outcomes than PAS, according to a study of hospice nurses.9

The noted psychiatrist and libertarian Thomas Szasz, MD, takes a position similar to ours in his book, Fatal Freedom. There,he characterizes physician-assisted suicide as “Bureaucratized medical killing.” Szasz pointedly notes that “… the suicide-assisting physician possesses power not only to facilitate the subject’s voluntary death, but also to foil it by incarcerating him in a mental hospital.”10 While we disagree strongly with Szasz’s views on mental illness, this claim is entirely sound. As Szasz rightly observes:

“. . . the legal definition of PAS as a procedure that only a physician can perform expands the medicalization of everyday life; extends medical control over personal conduct, especially at the end of life; and diminishes patient autonomy.”

Indeed, more than 25 years ago, law professor Giles R. Scofield, JD, argued that, “The legalization of physician-assisted suicide does not empower patients; it empowers physicians.”11

In a strictly procedural sense, this is clearly true. But on a deeper level, we would argue that the physician’s participation in PAS vitiates the exercise of their professional skill and knowledge in furthering the traditional goals of Hippocratic medicine, including the promotion of the vulnerable patient’s well-being. In this sense, PAS actually diminishes the physician’s authentic autonomy, even as it permits considerable procedural autonomy.

Relational Autonomy is Not Fostered by PAS

In recent years, the concept of relational autonomy has begun to challenge the traditional, hyper-individualistic notion of autonomy.4 While relational autonomy has been variously defined, we would define it as self-governance that involves shared decision making in the context of one’s social, familial, and cultural milieu; and consideration for the impact one’s decisions have on society and one’s significant others.

As currently structured by Oregon-type legislation, PAS is not congruent with relational autonomy. For example: the law stipulates that the physician recommend that the patient notify family of their intent to pursue PAS, yet does not require this notification. As feminist bioethicist Anne Donchin, PhD, observed, “Under some circumstances this strategy might maximize patient autonomy, but under others, it may provoke conflict that reverberates throughout the family, throwing everyone off balance and possibly jeopardizing the autonomy of family caregivers.”12

Furthermore, Donchin points out that12:

"To decide whether the patient is subject to `undue influence' the physician would need to cultivate relationships with other family members. But if the patient has declined to discuss her plans with the family, the physician is in a tight spot and lacks sufficient information to judge the patient's competence.

But even if the individual’s immediate circle of friends and family do not oppose, or even support the patient’s act, PAS ignores the detrimental impact on the wider society. As English poet John Donne wrote, “Any man’s death diminishes me, because I am involved in all mankind.” In our view, each death from PAS reduces the resilience of a society already hard-pressed in so many quarters to maintain hope.

Indeed, we believe that PAS depletes the empathy and compassion for one’s fellow human beings that is the strength of a civilized culture. We find it ironic that some PAS proponents accuse those who oppose this practice of “abandoning” suffering patients.13 We believe it is PAS that amounts to a form of abandonment, in which the dying patient is typically left to their own devices, almost always in the absence of any overseeing medical professionals. There is, for example, no requirement in most PAS legislation that the physician of record be present to monitor the patient’s final act. Moreover, PAS does not always result in a rapid and peaceful death—in some instances, death may take up to 25 hours.14 In terms of harm to society, it is also very worrying that PAS may be producing a kind of contagion effect. Specifically, recent data suggest that “assisted suicide laws lead to a substantial increase in total suicide rates and, if anything, are associated with an increase even in unassisted suicides. This effect is most pronounced amongst women.”15

From Hippocrates to Our Own Time

Our opposition to PAS is not based solely on the ethos of Hippocratic medicine, though Hippocratic physicians were notable—if not unique—for their refusal to provide lethal medication to their terminally ill patients.2 In our own time, most major medical organizations have opposed PAS, including the American Medical Association; the American College of Physicians; and the World Medical Association.16 As the authors of the position paper by the American College of Physicians put it17:

“Although respect for patient autonomy is a fundamental ethical principle, it is not the only one—it must be considered and balanced with doing no harm and acting in the patient’s best interests.

Finally, as physician and ethicist Leon Kass, MD, has argued18:

“The legalization of physician-assisted suicide [perverts] the medical profession by transforming the healer of human beings into a technical dispenser of death. For over two millennia the medical ethic . . . has held as an inviolable rule, ‘Doctors must not kill.’ The venerable Hippocratic Oath clearly rules out physician-assisted suicide. Without this taboo, medicine ceases to be a trustworthy and ethical profession . . . We must care for the dying, not make them dead.”

Dr Geppert is a professor in the Department of Psychiatry and Internal Medicine and director of ethics education at the University of New Mexico School of Medicine in Albuquerque. She is the lead ethicist for the western region and director of education, Veterans Administration National Center for Ethics in Health Care, and an adjunct professor of bioethics at the Alden March Bioethics Institute of Albany Medical College. She serves as the ethics editor for Psychiatric Times™. Dr Pies is Professor Emeritus of Psychiatry and Lecturer on Bioethics and Humanities, SUNY Upstate Medical University; Clinical Professor of Psychiatry, Tufts University School of Medicine; and Editor in Chief Emeritus of Psychiatric Times™ (2007-2010). Dr Pies is the author of several books. A collection of his works can be found on Amazon.

Acknowledgements: Thanks to John R. Peteet, MD, for providing background material on subtypes of autonomy. Thanks to our colleagues, Mark Komrad, MD, and Anne Hanson, MD, for their collaboration and contributions.


1. Beauchamp TL, Childress JF. Principles of Biomedical Ethics. Oxford University Press; 2013.

2. Lloyd GER, ed. Hippocratic Writings. Penguin Books; 1983.

3. Bishop L, Josephson A, Thielman S, Peteet J. Neutrality, autonomy and mental health: a closer look. Bull Menninger Clin. 2007;71(2):164-178.

4. Walter JK, Ross LF. Relational autonomy: moving beyond the limits of isolated individualism. Pediatrics. 2014;133 Suppl 1:S16-23.

5. Pies RW, Komrad MS, Geppert CMA, Hanson A. Against assisted suicide. Psychiatric Times. July 8, 2021.

6. Komrad MS, Pies RW, Hanson AL, Geppert CMA. Assessing competency for physician-assisted suicide is unethical. J Clin Psychiatry. 2018;79(6):18lr12566.

7. Yuen JK, Reid MC, Fetters MD. Hospital do-not-resuscitate orders: why they have failed and how to fix them. J Gen Intern Med. 2011;26(7):791-797.

8. Foglia MB, Lowery J, Sharpe VA, et al. A comprehensive approach to eliciting, documenting, and honoring patient wishes for care near the end of life: the veterans health administration's life-sustaining treatment decisions initiative. Jt Comm J Qual Patient Saf. 2019;45(1):47-56.

9. Ganzini L, Goy ER, Miller LL, et al. Nurses' experiences with hospice patients who refuse food and fluids to hasten death. N Engl J Med. 2003;349(4):359-365.

10. Szasz T. Fatal Freedom. Syracuse University Press; 1999.

11. Scofield GR. Exposing some myths about physician-assisted suicide. Seattle Univ Law Rev. 1995;18(3):473-493.

12. Donchin A. Autonomy, interdependence, and assisted suicide: respecting boundaries/crossing lines. Bioethics. 2000;14(3):187-204.

13. Strouse T. End-of-life options and the legal pathways to physician aid in dying. Journal of Community and Supportive Oncology. 2017;15(1):1-3.

14. Disability Rights Education & Defense Fund. Why assisted suicide must not be legalized. Accessed March 6, 2023.

15. Paton D, Girma S. Assisted suicide laws increase suicide rates, especially among women. VoxEu; Centre for Economic Policy Research. April 29, 2022. Accessed March 6, 2023.

16. WMA statement on physician-assisted suicide. November 5, 2019. Accessed March 6, 2023.

17. Sulmasy LS, Mueller PS; Ethics, Professionalism and Human Rights Committee of the American College of Physicians. Ethics and the legalization of physician-assisted suicide: an American College of Physicians position paper. Ann Intern Med. 2017;167(8):576-578.

18. Kass LR. Dehumanization triumphant. First Things. 1996. Accessed March 6, 2023.

Related Videos
sad girl, suicide
© 2023 MJH Life Sciences

All rights reserved.