---

Keywords: autonomy, relational autonomy, medical ethics, terminal illness, aid in dying, medical aid in dying, physician-assisted suicide

---

---

Abstract

The concept of autonomy is integral to the practice of aid in dying. This article considers the concept of relational autonomy—a more expansive and practical definition of patient autonomy—and makes initial suggestions for the more systematic adoption of the concept in patient care.

---

US in particular, arguments supporting aid in dying have focused on individual rights. However, individuals exist within a social context in which they are linked emotionally and materially with others. The concept of “autonomy” can and should encompass these relational influences. In fact, one might argue that it is unethical not to accommodate relational influences. One cannot separate the individual’s choice for assisted dying from that person’s involvement in their family and community.

Significant relationships can actually enhance autonomy by improving reflective capacity and enabling decision-making consistent with a given patient’s values. Family members can provide emotional support, bridge communication gaps, and encourage self-expression. Influential relationships therefore help patients grapple with conflict and avoid the isolation and conflict that can accompany assisted dying otherwise. Assistance in dying may be an act chosen by an individual—but this autonomy is better understood through the principle of “relational autonomy.”

---

---

organized into four pillars: non-maleficence, beneficence, autonomy, and justice. The most long-standing of these (going back to the Hippocratic oath) are beneficence, or doing “good” for patients, and non-maleficence, or not causing harm [1]. Autonomy as a core principle was originally developed as a counter to paternalism in medicine. More recently, it has been shaped by the emphasis on freedom and individual rights in our society. Justice (as fairness) has come to the forefront as clinicians offer treatments that are complex and costly and as patients face healthcare allocation challenges that have raised the specter of unequal access.

Most bioethicists now recognize a broader array of values, which can be in conflict and create ethical dilemmas. In addition to the traditional four “pillars,” ethical tensions arise in the context of values such as honesty, confidentiality (how much information we should share with families), equality (given the inherently unequal power between provider and patient), curiosity, humility, and communitarian values. As we assess ethical challenges in decision-making, we also consider the fact that there are nonrational factors that are part of every person’s reasoning. One way of synthesizing all of these into one overarching value is “respect for persons.” As with all ethical dilemmas, the principles only define the challenges: The key is still how to balance or interpret these values in a particular circumstance where they may be in tension.

For the purposes of this discussion, we define two key terms. Autonomy is an ethical principle that comes originally from Greek and means self-rule; it was political in its origins but moved into the sphere of moral philosophy. From those origins, we established our core medical principle of respect for autonomy, which affirms that every person has moral value and dignity in their own right. Respect for personal autonomy means that we recognize “the moral right of every individual to choose and follow his or her own plan of life and actions” [2]. Relational autonomy is a more recent concept that acknowledges the central role of others in individual decision-making and self-determination. Relational autonomy recognizes that individuals exist within a social context through which they are linked emotionally and materially with others [3,4]. This social context (often shortened as “family”) may include biological relatives, close friends, and members of their community, including perhaps a faith community.

---

⠀

The following two cases illustrate some challenges to the model of individual autonomy.

Case 1: Mara Buchbinder and Noah Berens describe the case of a man who requests assistance in dying in the face of prolonged illness; he has had cancer for over 20 years and is cared for in his terminal stages by a devoted wife, who struggles in her own frailty to meet his care needs. He is worried that he is becoming an undue burden on her. He has requested aid in dying, which is legal in his state. His wish is to avoid prolonging his death and further burdening his wife [4].

Case 2: A 72-year-old woman has metastatic breast cancer involving her lungs and bones. She is dependent on supplemental oxygen and has severe bone pain, which has not been controlled by pain medications. She has had two falls involving fractures of her left leg within the last four months and can no longer walk. She has a loving family which includes an attentive husband, two children, and four grandchildren. She was raised Catholic and is concerned that assisted dying is suicide and that Catholic teachings state that suffering and pain are a redemptive part of the dying experience. But she sees no joy or meaningful life in her future. The family wants to support her in whatever route to death she chooses; they see her suffering, and should she choose an assisted death, they would respect her choice.

---

---

But autonomy has become coupled with the Western emphasis on independence, such that families and other close friends who exert influence on the patient’s decisions can be seen as potentially interfering with the patient’s autonomous decision-making.

Autonomy as Relational:

Those defending the autonomous freedom of choice for individuals worry that decision-making by patients is often threatened by a mix of concerns for others, deferral to others, and the avoidance of conflict [4]. They argue that there can be many degrees of undue influence besides direct coercion—that is, explicitly forcing a person to make a decision against their wishes—and that autonomy is the absence of coercion.

Buchbinder and Berens provide some examples of these “softer” forms of coercion [5]. The order in which choices are presented or how consequences are described can shape people’s decision-making (e.g., “You will need to go into a nursing home if you don’t accept assistance in dying.”). Providers who raise the legal options of assistance in dying may be subtly intimating that this is the preferred choice for someone who is suffering from their condition. “Nudging” may take the form of a change in tone of voice or a facial expression that may be consciously or unconsciously manipulative and might be considered a subtle form of coercion. Family members who have beliefs that differ from the patient’s could convince them to adhere to the family’s values, even if the patient does not share those beliefs. For example, Catholics sometimes understand suffering to be redemptive, and they may understand this principle to prohibit choices such as assisted dying to relieve pain and suffering at the end of life.

Influence can work in the opposite way as well. Patients may perceive themselves, sometimes correctly, as a burden negatively affecting the lives of those they love most, either financially or emotionally [5,6]. Defenders of autonomous decision-making might consider these to be examples of coercive interference on the rights of individuals.

But bioethicists have increasingly pushed back against concerns about coercion and have asked us to view autonomy through a wider lens [3,4,5,7,8]. They assert that the individualistic concept of autonomy fails to recognize the complexity of decision-making for patients and contend that “autonomy” can and should encompass relational influences. Relational autonomy recognizes that individuals exist within a social context through which they are linked emotionally and materially with others. Buchbinder and Berens hypothesize that “social relationships may constitute an important source of value in end-of-life decision-making and not only a liability” [5]. This relational approach is recognized as part of the complexity in medical decision-making, challenging the myth of rational individual decision-making. Some observations focusing on the role of the family and community in decision-making at the end-of-life are notable:

Decision-making and serious illness:

There are several complex decisions patients with serious terminal illnesses face. The first is that people are very poor at predicting how their “future selves” will see life. Bullock et al. argue that evidence from behavioral economics suggests that people are unable to predict their reactions, particularly to future emotional events, or how happy or unhappy such changes will make them [9]. Sudore and Fried have reminded us of these challenges in relation to advance care planning. We are not very good at predicting how the future will feel, how much we will adapt to our condition as we near the end of life, or how past experiences may distort our ability to imagine what we would do in a similar present situation (extrapolation) [10]. Second, people change in how much agency or autonomy they want when seriously ill. It is not uncommon for patients to defer decision-making to others near the end of life. Bullock addresses the evidence that as patients become more ill, they may prefer to rescind their decision-making authority altogether [9,11].

How might this impact the process of considering aid in dying? A family that is sharing in a patient’s thoughts and decision-making process can be an important support and can understand the evolution that may occur in their loved one’s plans and wishes for the end of their life. This can happen as the patient recognizes that their life is “complete” and it is time to take control of the ending or, alternatively, that there is a new reason to try to live for a bit longer (e.g., an important family event). Relations can provide an important affirmation of the person’s decision as death nears. When a patient loses capacity on the course of their autonomous choices, the support of these relationships becomes even more important.

For example, if a patient makes the request for aid in dying earlier in the course of their illness when they clearly retain decisional capacity but then become more seriously ill, their loved one or surrogate can provide increasing input to help the patient decide if/when they want to proceed with ingestion. And since in the US, surrogates’ requests for aid in dying are not legal, those supporting a person can honor the spirit of patient’s wishes even if cognitive decline means that they lose eligibility for assisted dying. They can give voice to the patient’s autonomy by making choices on behalf of their loved one to help achieve a death consistent with their values, such as withdrawing ventilator support and assertive palliation of their pain and suffering even to the point of sedation—although requesting assistance in dying would not be a choice available to the surrogate.

Relationships and family structure always affect decision-making:

Just as in other situations of medical decision-making, patients making end-of-life decisions are acting in the context of their family and culture. Each family has its own dynamic, and that dynamic will extend into beliefs and traditions about illness and death. Some families may rely heavily on a matriarchal or patriarchal figure to make major decisions within the context of their family. Others function more democratically, taking the needs and wants of each family member into consideration.

These dynamics play out every day with seriously ill patients, and providers need to be sensitive to them. A patient may choose to undergo more treatment than they would otherwise want for the sake of their spouse, for example. Another patient may forgo treatment that they might otherwise undertake in order to protect the collective good of the family [5,4]. These choices might be driven by the desire not to bankrupt or significantly burden family and others who care for them [12]. We posit that when people make decisions that consider their families’ circumstances, they are enacting a richer form of autonomy.

All of these empiric arguments suggest that autonomy in decision-making at the end of life has complex influences and effects on community. In considering assisted dying, the outside influences could be either coercive or relational and noncoercive, but they must be recognized. Communities that surround each person considering aid in dying may include people close to the patient with strong and deeply held beliefs about the issue. These people can be either supportive of or opposed to assisted dying.

Healthcare professionals are required to ascertain whether a person requesting aid in dying is making an autonomous and voluntary choice that does not involve—or stops short of—coercion, recognizing that decision-making is influenced by many factors and many of these factors occur in the context of a family.

The duties of healthcare professionals in supporting patients who are considering aid in dying:

It is our duty as professionals to recognize that both coercive and noncoercive influences exist and to attempt to differentiate undue vs. acceptable influences. In essence, the benchmark is that one must be free to accept or reject those influences. Medical professionals already assess patient capacity for decision-making—i.e., informed consent—which is the first step of assessment. To scrutinize the effect of influences on the request for medical assistance in dying, several authors have suggested further guiding considerations [5,4]:

⠀⠀⠀⠀1)Assess the patient’s decision-making process. First, the patient must demonstrate decisional capacity in a manner similar to the way we determine whether the patient can give informed consent in any medical situation. That means that they should 1) understand relevant information, 2) appreciate the medical situation and its possible consequences, 3) be able to communicate a choice, 4) engage in rational deliberation about their values in relation to the treatment options [2].

In the context of considering a patient for assistance in dying, this can consist of more directed questions and conversations. Is the patient able to express an understanding of an adequate range of options? Can they articulate, for example, “I could continue receiving chemotherapy, move to a nursing home, or hire help.” Input from a family member—for example, who states, “You either agree to aid in dying or you’ll have to move to a chronic care facility”—must be recognized and assessed by the professional as coercive.

It is true that freedom with respect to decision-making can be challenging to assess in the context of financial limitations, homelessness, estrangement from family, or the lack of a supportive community. For such marginalized members of a community, options may be significantly narrowed, but they should still be able to discuss their choices, recognize barriers, and reflect on their reasoning.

⠀⠀⠀⠀2) Explore the manner in which decisions are made in the family. What values other than autonomy influence choices, and how are family dynamics influenced by an individual’s decisions? These dynamics may vary according to culture or religion. What conflicts are likely to develop as repercussions of your patient’s decision to consider assisted dying? [4]

⠀⠀⠀⠀3) Evaluate whether the decision is reflective of long-standing personal values. Does the decision appear “authentic”? Is it “in character”? To answer these questions, we often need to explore how a patient lived their life; we also need to be mindful that preferences do evolve or change over time—particularly at the end of life. Even changing preferences need to reflect the patient’s values and persons need to be able to reason about why they may be changing [13,14]

⠀⠀⠀⠀4) Help anticipate stresses for the patient and family. What will be the repercussions of your patient’s decisions to engage in assisted dying? Has the family been supportive and provided care willingly? Familial support of the person who requests assistance in dying can be an expression of care or a desire to be free of the burdens imposed on them, or perhaps both, and professionals should be cautious about judging the family’s ability to shoulder such burdens. Emotional support can be fraught when there is a not-uncommon component of potential financial or other gain for the family following the death of their loved one.

⠀⠀⠀⠀5) Determine whether any financial considerations might be influencing decision-making. We live in an unjust society in which options for end-of-life care may be reduced due to lack of resources. The cost of care may be prohibitive for some patients and may not be correctable [15,16]. Even those with means may not want to spend down their funds, preferring to pass their money on to their loved ones. This choice can be viewed as an autonomous decision to protect the family’s assets, as long as it comes from the patient and has not been imposed by the family.

⠀⠀⠀⠀6) Determine whether the patient has been able to explore a range of treatment options available to them. Has there been sufficient exploration of alternative means of alleviating a patient’s suffering? This is the challenge that palliative care and hospice providers confront: Can symptom relief by other means give a person more quality time as an acceptable alternative to aid in dying?

The autonomy of providers:

Prescribing providers also have the autonomy to decline to participate in assisted dying according to their personal beliefs. The opportunity to refuse to participate is written into each state law. But there is also an ethical tension between our professional obligations and our personal autonomy/beliefs. Because of the inherently unequal power relationship between patient and provider and the unique ability we have to provide or deny specific care—such as the license to prescribe—we need to think long and hard about withholding access to this choice. Providers who opt out of participation should not further impede the patient’s access to this option where it is legally available. Providers also need to ensure that they are opting out based on deeply held beliefs that they feel compel them to opt out and not merely because they perceive the process to be too burdensome for them.

---

can cause harm when it is reduced to an individualistic concept and leads to an abrogation of a healthcare professional’s responsibilities [17]. People live in communities and when autonomy is recognized in a fuller relational framework, it deserves more scrutiny and can be a richer concept within the discussion of aid in dying.

If providers recognize and examine the diverse influences that affect our patients, it becomes clear that significant relationships can enhance autonomy by improving reflective capacity and the ability to make a decision consistent with their values. Families can provide emotional support, bridge communication gaps, and encourage self-expression.

---

1. Beauchamp T, Childress J. Principles of Biomedical Ethics. 8th edition, Oxford Press, 2019.
2. Jonsen AR, Siegler M, Winslade. Clinical Ethics: A Practical Approach to Ethical Decisions in Clinical Medicine. McGraw Hill, 9th edition.2022.
3. Walter JK, Ross LF. Relational autonomy: moving beyond the limits of isolated individualism. Pediatrics. 2014;133 Suppl 1:S16-S23. doi:10.1542/peds.2013-3608D.
4. Breslin JM. Autonomy and the role of the family in making decisions at the end of life. J Clin Ethics. 2005 Spring;16(1):11-9. PMID: 15915842.
5. Buchbinder M, Berens N. Beyond coercion: reframing the influencing other in medically assisted death. J Med Ethics. 2024;50(12):841-845. Published 2024 Dec 23. doi:10.1136/jme-2023-109554.
6. Swindell JS, McGuire AL, Halpern SD. Shaping patients’ decisions. Chest. 2011 Feb;139(2):424-429. doi: 10.1378/chest.10-0605. PMID: 21285056.
7. Foo KF, Lin YP, Lin CP, Chen YC. Fostering relational autonomy in end-of-life care: a procedural approach and three-dimensional decision-making model. J Med Ethics. Published online March 25, 2024. doi:10.1136/jme-2023-109818.
8. Gómez-Vírseda C, de Maeseneer Y, Gastmans C. Relational autonomy in end-of-life care ethics: a contextualized approach to real-life complexities. BMC Med Ethics. 2020;21(1):50. Published 2020 Jun 30. doi:10.1186/s12910-020-00495-1.
9. Bullock Emma C. “Assisted Dying and the proper role of patient autonomy” in M. Cholbi and J. Varelius (eds.), New Directions in the Ethics of Assisted Suicide and Euthanasia, The International Library of Bioethics 103, Springer Nature Switzerland AG. https://doi.org/10.1007/978-3-031-25315-7_1
10. Sudore RL, Fried TR. Redefining the “planning” in advance care planning: preparing for end-of-life decision making. Ann Intern Med. 2010 Aug 17;153(4):256-61. doi: 10.7326/0003-4819-153-4-201008170-00008. PMID: 20713793; PMCID: PMC2935810.
11. Halpern J. Reluctant patients: autonomy and delegating medical decisions. J Clin Ethics. 2002 Spring;13(1):78-84. PMID: 12235688.
12. Gudat H, Ohnsorge K, Streeck N, Rehmann‐Sutter C. How palliative care patients’ feelings of being a burden to others can motivate a wish to die. Moral challenges in clinics and families. Bioethics. 2019;33:421–430. https://doi.org/10.1111/bioe.1259.
13. Dresser R. Treatment decisions and changing selves. J Med Ethics. 2015 Dec;41(12):975-6. doi: 10.1136/medethics-2014-102237. Epub 2014 Jul 8. PMID: 25005110
14. Morrison RS, Meier DE, Arnold RM. What’s Wrong With Advance Care Planning? JAMA. 2021 Oct 26;326(16):1575-1576. doi: 10.1001/jama.2021.16430. PMID: 34623373; PMCID: PMC9373875.
15. DiDanieli M. Autonomy is not a sufficient basis for analysing the choice for medical assistance in dying in unjust conditions: in favour of a dignity-based approach. J Med Ethics. 2024;50(6):421-422. Published 2024 May 22. doi:10.1136/jme-2023-109284.
16. Wiebe K, Mullin A. Choosing death in unjust conditions: hope, autonomy and harm reduction. J Med Ethics. 2024;50(6):407-412. Published 2024 May 22. doi:10.1136/jme-2022-108871.
17. Billings JA, Krakauer EL. On patient autonomy and physician responsibility in end-of-life care. Arch Intern Med. 2011 May 9;171(9):849-53. doi: 10.1001/archinternmed.2011.180. PMID: 21555664.