"But My Mom Still Blinks when I Talk to Her." Understanding the Concept of Social Death to Avoid Bias and Improve Goals-of-care Discussions at the End of Life

End-of-life decisions become ever more complex as new medical advances in life-sustaining interventions blur the conventional boundaries between life and death. These advances may prolong the dying process and create a burden of choice around “the right time” to shift goals-of-care from prolonging life to prioritising comfort and allowing natural death. One potential influence on end-of-life decisions is patients' ability to interact in a socially meaningful way. The concept of “social death” describes how people are treated differently or even dehumanised when they lose important social attributes. Understanding this concept may help clarify the role that patients' social interaction can play in decision-making and in potential disagreements regarding goals-of-care for patients near the end of life. Consider the following case. An elderly patient with multiple complex comorbidities is admitted to the intensive care unit (ICU) with acute respiratory distress syndrome in the setting of worsening pneumonia. Her hospital course has been complicated by a haemorrhagic stroke with no options for surgical intervention. She is currently intubated and sedated in the ICU. As her sedation is weaned, she continues to be minimally interactive with what the team describes to be spontaneous movements. The critical care and neurology teams believe the patient has a very low chance of significant neurologic recovery, and they are starting to feel distressed about continuing intensive care for a patient that they perceive as being noninteractive with a low likelihood of “meaningful” survival. The team starts expressing this distress on rounds, saying, for example, “We are not helping this patient” and “She has no quality of life”. The patient's family points out that their loved one is still intermittently squeezing their hands and opening her eyes in response to their voices. Thus, the family members say that they are not willing to “give up” on her and want to continue intensive care treatment. The medical team arranges for a goals-of-care meeting with the patient's family to discuss next steps in management. The above case illustrates how a clinical team and family may hold different views on appropriate goals-of-care when a patient is minimally interactive and may be near the end of life. In such cases, clinicians may focus on physiological processes and emphasise physical harms over the possible social benefits of continued life-sustaining interventions. Conversely, families, defined broadly as patient surrogates involved in decision-making, may object to withdrawing life-sustaining interventions despite a patient's poor prognosis when they view the patient as continuing to have a social presence. This perceived social presence may take the form of family members attributing meaning to certain patient actions that clinicians may view as involuntary, such as eye or hand movements. As a result, families may see signs of social life as sufficient to justify continued life-sustaining interventions despite a patient's physiological decline. Conflict on this point may be especially acute when a patient's future neurological status is uncertain. The concept of social death may help reveal sources of conflict and can be used to explore how different parties value and interpret a patient's ability to socially interact near the end of life. Understanding the influence of social death perceptions on decision-making can empower clinicians to make more reflective, patient-centred decisions, better understand family perspectives, and foster effective communication that places the humanity of the patient as central to decision-making. In this article, we first describe the concept of social death and its potential impact on medical decision-making. We then discuss how differing philosophical perspectives used by clinicians and families may explain disagreement regarding appropriate goals-of-care for minimally interactive patients near the end of life. After this, we discuss the ethical implications of clinicians consciously or unconsciously applying social death assessments to end-of-life decision-making. Lastly, we offer recommendations for more person-centred approaches to decision-making to decrease potential decisional bias stemming from assessments of social death and to improve communication with families during goals-of-care discussions. Death is defined as the end of life, typically understood as a ceasing of biological processes. With advances in somatic support and organ transplantation, the legal definition of death was specified to include the irreversible loss of circulatory or neurological functioning.1 However, debates over the specific neurological death criteria and the meaning of irreversibility continue.2, 3 These debates illustrate how defining death can be as philosophical and interpretive as it is scientific. Accepting the current legal definition of death, the question remains as to when, and what types of, losses are significant enough to impact treatment decisions for patients near the end of life. Scholars have used the concept of social death to characterise the loss of important social attributes, such as identity, agency, social interaction, social connectedness, social importance, or the ability to function as a social being.4 Social death often occurs at the same time as biological death when social interactions and physiological functioning cease simultaneously. However, when social death is perceived to occur before biological death, it may shape beliefs about others' moral obligations towards that person. Social death has been used in sociology and anthropology to describe how people dehumanise and marginalise certain vulnerable populations that are socially segregated from society, such as incarcerated persons, refugees, people with dementia and minority communities.5-8 Institutionalised settings, such as hospitals, can foster perceptions of social death due to the dehumanisation that often occurs when individuals are viewed as numbers or cases and patients lose their agency.9 In this way, social death can be self-fulfilling in that people are segregated from society, for example, by being placed in institutions or deprived of important social goods, and then treated unworthy of moral consideration due to their separateness.* In the medical context, social death is characterised by a patient's loss of social attributes paired with physical deterioration that are together perceived to forecast an irreversible trajectory towards biological death.11 It was in this medical context that the concept of social death first emerged in ethnographic studies in the 1960s to describe how clinicians in the inpatient setting treated “senile” and “comatose” patients as if they were already dead.12, 13 In his 1967 ethnography, “Passing On: The Social Organization of Dying”, sociologist David Sudnow states that “social death, within the hospital setting, is marked by that point at which a patient is treated essentially as a corpse, though perhaps still ‘clinically’ and ‘biologically’ alive”.12 For example, Sudnow observed that clinicians would start to phase out treatments, speak about patients as if they were not in the room, place increasing attention on the body over the person as a social being, wrap bodies before death was declared, and not assign rooms to newly admitted patients near death, but rather leave them “throughout the night to die in the supply room”.12 Although these descriptions seem antiquated, they draw certain similarities to modern-day examples of dehumanising behaviour towards unconscious or noncommunicative patients in the ICU. Wilson et al. describe how in the ICU setting, “physicians may enter the patient's room without introduction, proceed to move parts of the patient's gown and touch the patient without notice, speak to the nurse about what is going on with the patient, and leave the patient's room all without uttering a single word to the patient”.14 The institutionalised nature of the ICU may enable such attitudes and behaviours by creating barriers to human connection, such as visitation restrictions, protective equipment and invasive interventions that make people unrecognisable.15 In addition, critically ill patients often have diminished capacity and limited ability to interact due to their underlying condition or the need for sedation that may reinforce perceptions of social death. However, even under such conditions, clinicians' perceptions of patients' ability to socially interact may vary. This may especially be true for patients in minimally conscious states who display subtle and intermittent signs of consciousness and responsiveness that leave room for interpretation of the degree of their intentionality.16 Studies have reported variations in clinicians' views on the mental and interactive abilities of patients in a minimally conscious state and the appropriateness of withdrawing life-sustaining interventions.17, 18 Social death perceptions in medicine have also influenced treatment decisions based on implicit appraisals of a patient's social importance; a practice known as social rationing. Past studies have shown that social death perceptions based on characteristics, such as a patient's age, social status or perceived moral standing, and how well clinicians know the patient impacted the likelihood of performing cardiopulmonary resuscitation.12, 19 On the other hand, clinicians with a greater social connection with patients, either because they can professionally or personally identify with the patient or have a longer clinical relationship with the patient or family, perceived a greater social importance of these patients and greater sense of social loss from these patients' deaths.12, 19 Contextualising the patient as a social being beyond their physiological status may, in turn, cause clinicians to evaluate the harms and benefits of interventions differently than those that lack this deeper insight. Conversely, having limited knowledge of patients' personal identities may cause clinicians to be biased towards poorer prognoses that impact end-of-life decisions.14 Team and family disagreement around goals-of-care based, a least in part, on differing interpretations of patients' ability to socially interact indicates that social death may play a role in what parties believe to be appropriate treatment for these patients. Since social death perceptions can influence decision-making and thus have consequences for end-of-life care, it is vital to understand the source of differing perceptions on social death. Team and family disagreement may in part be explained by differing philosophical paradigms that clinicians and families employ to assess and interpret important factors for decision-making at the end of life. An interview-based study of patients' families demonstrated that whereas clinicians tend to use a more clinical framework based on measurable physical parameters to interpret meaningful patient movement, families use a relational framework that can “encompass the uniqueness of the patient and their relationship to them, and thus a framework that constitutes the patient as a continuing member of a community—as a social being”.16 These different interpretations may be explained through the use of different ontological perspectives employed by clinicians and families. The current biomedical model of medicine is largely influenced by mind-body dualism, which distinguishes the mind from the body and views the human body as a biological organism that can be understood by its constituent parts or as a type of organic machine to be fixed.20 From this perspective, it makes sense that once the biological machine is irreparable and the mind is gone, the patient may seem “as good as dead.” Although families may also endorse a dualist perspective, they are more likely to think in less reductionist terms given that they have a more nuanced understanding of the patient as a person. They might embrace a more existentialist paradigm that sees the patient's identity as shaped by her unique experiences and relationships.15 The patient's present condition may be viewed as a critical experience in the context of the person's broader life, in which important relationships continue to play a vital part. Shared decision-making for patients near the end of life requires clinicians to bridge these perspectives by bringing their relevant biomedical expertise to the table, while also appreciating and treating the patient as a person. Person-centred care requires clinicians to see “the person behind the patient—as a human being with reason, will, feelings, and needs” and to define health in terms of the person's ability to function in relation to their vital goals.21, 22 Mitchell and Loughlin warn of “the danger of objectifying the patient and dehumanising care” for ICU patients who are near death and the importance of ensuring that “the personhood of the patient is protected and not forgotten”.15 This may be achieved through authentic solicitude, which requires those who care for others to recognise the humanity of those they care for.15 Authentic solicitude therefore supports patients or their families to make decisions according to what makes the person's life meaningful. Thus, person-centred care first requires clinicians to (1) see the humanity in their patients and (2) help patients or their families to uncover what is meaningful for the person in the context of their lives and incorporate this into decision-making. There seems to be something intuitively relevant about a patient's ability to socially interact, but should clinicians' perceptions of social death influence clinical recommendations during goals-of-care discussions for patients near the end of life? One argument is that social interaction is an essential element of quality of life and without it, there is no benefit to prolonging biological life. In this case one might feel that the harms of aggressive interventions would outweigh the benefits. However, the value of quality-of-life assessments may be questionable if they are reduced to mental functioning or influenced by clinicians' own values and experiences without considering the patient's personal understanding and experience of harm and benefit. Such quality-of-life assessments by clinicians are prone to the disability paradox, in which clinicians may underestimate the benefits of continued life for others with conditions perceived to be intuitively undesirable.23 Second, it could be argued that social interaction is an essential element of personhood, and therefore social death means that someone is no longer owed the moral consideration of a person. This argument is supported by relational concepts of personhood that define a person in terms of whether they can maintain and express specific mental states24 and how they are valued through important relationships.24, 25 If personhood is truly relational, then there may be different moral obligations towards patients who can socially interact versus those who cannot. Although an argument centred on relational personhood is compelling given the importance of human relationships and how interactions with others are constitutive of people's experience and identity, not all philosophers agree that personhood is relational. Some argue that personhood is an essential characteristic of the human species as a whole or even that personhood is a “problematic shortcut” for delineating our moral obligations to others.26, 27 It becomes apparent that the relational view of personhood may be problematic, particularly when considering critically ill patients. It seems clearly unjust to treat minimally interactive patients with strong family advocates with greater care and moral consideration than minimally interactive patients that are unrepresented. Furthermore, it is not always clear when patients retain the ability to interact in a relational way. There are clinical situations in which it is uncertain whether a patient is using their body movements to communicate specific mental states or whether such movements are simply unconscious, spontaneous reflexes. Social death perceptions can be dynamic due to shifts in patient condition or sedation levels, making it difficult to define a specific point at which a patient's lack of social attributes is permanent and would be a relevant criterion for withholding or withdrawing life-sustaining interventions.4 The potential to misinterpret a patient's body movements would therefore create a margin of error that risks treating a person as a nonperson. Given these uncertainties, clinicians should avoid making clinical judgements based on their perceptions and values related to social death. Such decisions may be prone to bias since perceptions of social death may be shaped by personal experience, beliefs, shared identities, valued social characteristics and the clinician's relationship with the patient.19, 28 A patient's ability to socially interact should only play a role in end-of-life decisions when there is compelling evidence that it supports fair allocation of resources, promotes benefit and reduces harm, and supports patient autonomy. Clinicians can guard against potential bias and dehumanising influences by challenging their own assumptions and using person-centred approaches to end-of-life decision-making and communication. Implementing standardised or consensus-based clinical decision-making may help challenge potential bias that can lead to social rationing and inaccurate prognostication based on social death perceptions. Although fair allocation of resources may play a role in decision-making, such decisions should be made using standardised criteria for rationing and triage made by committees separate from the treating team to guard against implicit social rationing. Standardised or consensus-based definitions of physiological harm and benefit may likewise reduce potential bias and medical uncertainty. It is typically considered ethically supportable to recommend withdrawal of life-sustaining interventions when broader medical consensus deems particular interventions to be clinically inappropriate because certain clinical goals can no longer be achieved. For example, the Society of Critical Care Medicine published a statement on inappropriate treatment at the end of life to create standards of practice and minimise bias and ad hoc decision-making. This statement asserts, “ICU interventions should generally be considered inappropriate when there is no reasonable expectation that the patient will improve sufficiently to survive outside the acute care setting, or when there is no reasonable expectation that the patient's neurologic function will improve sufficiently to allow the patient to perceive the benefits of treatment” (italics added).29 Even when appealing to such medical consensus statements, patient assessments should include team consensus or strong evidence of severe neurological injury to reduce potentially biased perceptions of a patient's current and future social viability. However, assessments based on physiological harm and benefits are insufficient for engaging in person-centred decision-making and communication that emphasise patient autonomy and individual experiences and interests. It is therefore essential to factor patients' ability to socially interact into decision-making when it respects a patient's autonomy based on known patient values and preferences. Evidence of such patient preferences may be identified through conversations with the patient's family, knowledge of how the patient values social interaction, or advance directives that specify a preference to withhold or withdraw life-sustaining interventions if the patient loses the ability to socially interact with others in a meaningful way. In the absence of irreversible severe neurologic injury and known patient preferences regarding socio-interactive attributes, clinicians should avoid making clinical judgements based on their perceptions and values related to social death. However, clinicians' awareness of the role that social death perceptions can play in decision-making can better equip them to make more reflective decisions, understand family perspectives, and engage in compassionate and effective goals-of-care conversations. Below, we offer recommendations to help clinicians avoid making decisions based on perceptions of social death and to promote fruitful and compassionate goals-of-care discussions with patients' families. We recommend clinicians strive to (1) reflect on and challenge their own assumptions or potential biases, (2) recognise patients' humanity and (3) support families to make decisions for the person near death that are meaningful in the context of the person's life. To do this, physicians can take part in self-reflection, speak to the patient's family to better understand the patient's values and preferences, and use philosophical counselling to uncover patterns of reasoning, expose presuppositions, and help create a common awareness of relevant issues for end-of-life decision-making.30, 31 For each recommendation, we provide accompanying questions for clinician self-reflection and suggested phrasing for goals-of-care discussions with patients' families that support patient-centred care (Table 1). Do I believe that the patient currently has and may in the future have meaningful social interaction with others? On what am I basing this assessment? How long and in what context have I known this patient and family? Are similarities or differences between myself and the patient playing a role in my medical recommendations? How is my own view on adequate quality-of-life shaping my clinical judgement? Am I consciously incorporating the patient's known values and preferences into my assessment? How and why might the family's perceptions of the patient's social viability differ from my own? “What are your biggest concerns right now regarding your loved one?” “What do you think your loved one is experiencing right now?” “How have your interactions with your loved one been since she has been sick? How have these interactions changed over time?” “How are you hoping to be able to connect with your loved one now and in the future?” Does the patient have irreversible severe neurologic injury affecting her social interactions? Are there any other treatable or reversible clinical conditions that may improve the patient's social interactivity? How confident am I in the patient's prognosis? If this patient was my family member or a patient with whom I had an established relationship, would my prognostication remain the same? Have I spoken with other clinicians about relevant criteria for recommending withdrawal? In doing so, were there any gaps in my assessment that may be attributed to perceptions about the patient's social viability or other nonclinical factors? What are the patient's known preferences regarding socio-interactive attributes? Have I fully assessed preferences the patient might have had by reviewing any advance care planning documents and discussing this with the patient's family? “I want to make sure we are all on the same page about your sister's condition and prognosis. What is your understanding of your loved one's health and what to expect moving forward?” “Different people want different things when they are in this condition and nearing the end of life. Based on what you know about your sister, what do you think would be important for her right now?” “Did your loved one ever speak to you about the importance of being able to interact with others if she ever become very sick?” Does the family perceive continued meaningful social interaction? How am I responding to a family's perception of continued social viability? Am I acknowledging the family's concerns? How can I describe and differentiate relevant clinical factors from perceptions of social viability without dismissing the family's perspective? “You have brought up that your mother squeezes your hand when you hold her hand. Tell me what that means for you.” “I am hearing that when your sister blinks, it feels like she knows you are with her. I can understand how that is important for you.” “From what you are saying, it sounds like when your uncle looks around, it feels like he knows you are with him. It is hard for us to know exactly what patients are experiencing, but I can see why this is important for you.” “Even if your loved one is aware, it is important that we make decisions that benefit and do not harm her in the long run and that align with her values given her current condition and prognosis.” Does the language I use communicate that I value this patient and see him as continuing to have worth? How can I affirm the patient's moral worth and acknowledge her importance to the family? Can any additional accommodations be made based on my knowledge of the patient and his family that may demonstrate that the care team continues to value the patient as a social and moral being? Am I still present for the patient and the family? If the option of comfort care has been discussed, have I made it clear to the family that we will continue to be present and care for their loved one? Use present tense when discussing the patient with family members, for example, “What would she want right now” instead of “What would she have wanted?” “Tell me about your loved one. From what I've heard so far, it sounds like he has led an amazing life.” “Even if we cannot heal your loved one, how can we honor her now?” “Does religion play a role in your loved one's life? How can we best incorporate this into her care now?” “‘Comfort measures’ mean I would continue to treat your loved one and continue to communicate with you as before. The goal has just changed to prioritize comfort.” Before making recommendations for care at the end of life, clinicians should reflect on their own perceptions of the patient's ability to socially interact and on other factors that might influence perceptions of social death. If clinicians discern that a patient is not exhibiting meaningful social interaction, they should consider their relationship with the patient and family and other factors like shared identities or valued characteristics that may influence social death perceptions. For example, having known the patient for years in the outpatient setting or identifying with the patient because of a shared race, ethnicity, profession, or sexual orientation may affect a clinician's judgement of a patient's social worth and viability. Conversely, lacking insight into a patient's social or relational context may cause clinicians to perceive the patient more as a biological body near death. Once clinicians recognise their own perceptions and potential biases, they should explore the family's perspective on the patient's socio-interactive state to ensure these considerations are taken into account during goals-of-care discussions. To differentiate perceptions of social death from relevant clinical factors, clinicians may consider discussing relevant clinical criteria for withdrawing life-sustaining interventions in a particular patient's case with colleagues to build consensus and challenge implicit personal biases. Clinicians can also speak with the patient's family to identify the patient's values, preferences, and patient-informed views of quality of life. Values elucidation and philosophical counselling may help families think through what their loved one finds important and meaningful. For example, if a family says that their loved one has always been social and valued the ability to talk with her grandchildren, then the patient's ability to socially interact would be ethically relevant. However, it may not be ethically supportable to base end-of-life decisions on perceived social death if a patient's family states that a patient never previously communicated the importance of social interaction, explicitly stated that social interaction was not a key factor for their quality of life, or clearly expressed other priorities, such as prolonging life. Families may still experience a social connection with their loved one even when they no longer possess the ability to interact with people the way they once did. To better understand the family's perception of their loved one and encourage shared decision-making, clinicians can identify how the family interprets the patient's continued social presence and acknowledge their concerns regarding shifting goals-of-care. For example, if family members point out that their loved one blinks when they talk to her, clinicians can ask the family what this blinking means for them and then validate the family's perspective and concerns. Once acknowledged, clinicians can incorporate the family's perceptions and concerns in the context of clinical and patient preference-based factors (e.g., “I acknowledge how important it is to you that your sister blinks when you are with her and it is important that we discuss options for your sister that benefit and do not harm her and that align with her values given her current condition and prognosis”). By doing this, clinicians can help families feel heard and understood, bridge communication gaps, and build trust. To affirm a patient's social worth and build trusting relationships when discussing options near the end of life, clinicians should offer authentic solicitude that requires that “the humanity (ontologically speaking) of the person receiving care is recognized and placed at centre stage”.15 Clinicians can humanise care through how they communicate that they value the patient and continue to see the patient as having moral worth. Even subtle language choices, such as referring to the patient in the present tense rather than the past tense can indicate to families that clinicians view the patient as a person who is alive and is owed full moral consideration. It may also be helpful to assure the patient's family that the health care team will continue to be present and provide excellent care for the patient even when goals-of-care may shift. In fact, a shift in the patient's goals-of-care may be a time for the care team to strengthen relationships with the family and learn more about the patient as a person. Sitting at bedside and speaking to the family about the patient's life, values and legacy allows the family not only to process the patient's impending death, but it may also help families feel supported in honouring their loved one and validating their loved one's social and moral worth. This can both build trust with the family and help facilitate what the family believes to be a good death for the patient. Even when critical care results in death, “person centred care does not end with the death of the patient, but continues into the process of the mourning and other rituals which surround death, and which include those connected to the deceased”.15 Honouring the patient near the end of life may come in the form of accommodations for a family who may believe the patient can feel their presence and is aware of the world around them. This may include family or pet visits, religious or secular ceremonies, or respecting families' preferences for specific dates for withdrawal of life-sustaining treatment (if within a reasonable timeframe) that are seen by the patient's family as beneficial or meaningful to the patient even if there is disagreement over whether the patient can appreciate these efforts. Recognising the humanity of their patients can help clinicians to support patient-centred death and meaning-making for the family and the health care team. Perceptions of social death may influence end-of-life decisions and can offer insight into a possible basis for families' and clinicians' disagreements regarding goals-of-care at the end of life. There are compelling reasons for clinicians to avoid applying their perceptions of social death to end-of-life decision-making since these perceptions may be prone to bias or misinterpretation and undermine person-centred approaches to care. However, awareness of social death perceptions may be useful for mitigating conflict with patients' family members and building mutual understanding. Clinicians can facilitate shared decision-making by recognising different perceptions of a patient's social viability, identifying relevant clinical and patient-centred factors for decision-making, and taking time to acknowledge the moral worth of the patient. This process can also empower clinicians to feel that they are doing something meaningful for the patient and the family when nothing curative can be done. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors. The authors declare no conflict of interest. This manuscript does not include primary source data.