The TRIBE Model: How Socioemotional Processes Fuel End-of-Life Treatment in the United States

Abstract

Prior interventions have repeatedly failed to decrease the prescription and receipt of treatments and procedures that confer more harm than benefit at the End-of-Life (EoL); new approaches to intervention are needed. Ideally, future interventions would be informed by a social-ecological conceptual model that explains EoL healthcare utilization patterns, but current models ignore two facts: (1) healthcare is an inherently social activity, involving clinical teams and patients’ social networks, and (2) emotions influence social activity. To address this omission, we scaffolded Terror Management Theory and Socioemotional Selectivity Theory to create the Transtheoretical Model of Irrational Biomedical Exuberance (TRIBE). Based on Terror Management Theory, TRIBE suggests that the prospect of patient death motivates healthcare teams to conform to a biomedical norm of care, even when clinicians believe that biomedical interventions will probably be unhelpful. Based on Socioemotional Selectivity Theory, TRIBE suggests that the prospect of dwindling time motivates families to prioritize emotional goals, and leads patients to consent to disease-directed treatments they know will likely be unhelpful, as moral emotions motivate deference to the perceived emotional needs of their loved ones. TRIBE is unique among models of healthcare utilization in its acknowledgement that moral emotions and processes (e.g., shame, compassion, regret-avoidance) influence healthcare delivery, patients’ interactions with family members, and patients’ outcomes. TRIBE is especially relevant to potentially harmful EoL care in the United States, and it also offers insights into the epidemics of overtreatment in healthcare settings worldwide. By outlining the role of socioemotional processes in the care of persons with serious conditions, TRIBE underscores the critical need for psychological innovation in interventions, health policy and research on healthcare utilization.

Introduction

More people than ever will need End-of-Life (EoL) care in coming decades, given the aging of the large baby-boom cohort in populations worldwide (Bartels and Naslund, 2013; Beck, 2009). Despite high rates of public preference for a peaceful death at home (Hamel et al., 2017), too many individuals with incurable disease receive disease-directed treatments, procedures, or services that confer more harm than benefit at the EoL (Duberstein et al., 2020; Luta et al., 2015). Conversely, too few receive interventions that would confer more benefit than harm, including pain control (Institute of Medicine, 2015; Sallnow et al., 2022). By some counts there are more than 300 potentially harm-conferring discretionary interventions at the end-of-life (DIALs), such as chemotherapy (Duberstein et al., 2020; Henson et al., 2020). Starting with the SUPPORT study (Anonymous, 1995), there have been numerous unsuccessful attempts to decrease the prescription and receipt of harm-conferring, disease-directed treatments (Abedini et al., 2019). To improve healthcare, there is a need for social-ecological conceptual frames that explain why disease-directed treatments are offered when there is no longer hope for sustained improvement, let alone a cure. In this paper, we offer a conceptual frame, the Transtheoretical model of Irrational Biomedical Exuberance (TRIBE), that is relevant to both EoL care in the United States, and to the broader problem of low-value care worldwide (Crowley et al., 2020; Elshaug et al., 2017; Fuchs, 1968; Le Fanu, 2012; Sallnow et al., 2022).

The language used to describe biomedical exuberance at the EoL is contestable. We chose a relatively value-neutral term (DIALs) instead of widely-used terms such as potentially avoidable (Daly et al., 2016), burdensome (Teno et al., 2013), inappropriate (Wolf et al., 2015), unduly intensive (Gidwani-Marszowsk et al., 2018), futile (Huynh et al., 2014), wasteful (Almoosa et al., 2016), and aggressive (Earle et al., 2008) because judgmental terms adversely affect policy (Tinetti, 2012) and practice (Maciasz et al., 2013). The acronym, DIALs, suggests that treatments could be calibrated -- dialed up or dialed down. Typically, they are dialed up (Aldridge and Bradley, 2017) and patient quality-of-life worsens (Zhang et al., 2012).

Hundreds of qualitative, quantitative and mixed-methods articles have highlighted other costs of DIALs to individuals, families, and society. For example, DIALs have been associated with worse bereavement outcomes in family caregivers (Wright et al., 2008), missed opportunities to divert resources elsewhere (Huynh et al., 2014), and clinician moral distress (Meier et al., 2001; Piers et al., 2011). Elevated heath care costs, financial strain on families, and catastrophic costs leading to poverty have also been documented (Osborn et al., 2014; Ramsey et al., 2016; Sallnow et al., 2022; Zhang et al., 2009).

In contrast to the burgeoning literature documenting the harms of DIALs, only two conceptual frameworks (Kelley et al., 2010; Prigerson and Maciejewski, 2012) have attempted to explain why dying patients receive potentially harm-conferring treatments (e.g., chemotherapy), procedures (e.g., surgeries), or services. Both frameworks can inform the development of individual-level interventions, but healthcare is an inherently social activity, involving two social groups: clinical teams (Nelson et al., 2002; Pask et al., 2018) and patients’ social networks (Pescosolido, 1992). Moreover, consistent with biopsychosocial (Engel, 1980) and social-ecological (Bronfenbrenner, 1979) approaches to human development, the provision and receipt of healthcare is shaped by the actions of individuals and entities operating at other levels of analysis in the healthcare ecosystem, including multiple government sectors (e.g., health financing, health policy, aging policy) and the private sector (e.g., for-profit and nonprofit entities). Although research has highlighted the role of emotions in team-based healthcare and family medical decision-making (Bluhm et al., 2016; Buiting et al., 2011; Djulbegovic et al., 2016; Finucane, 2013; Frosch et al., 2012; Good et al., 2004; Jackson et al., 2008; Steinhauser et al., 2000), we are aware of no conceptual frame that elevates the social dynamics and attendant moral emotions driving DIALs or healthcare utilization more broadly.

The Transtheoretical Model of Irrational Biomedical Exuberance (TRIBE) fills that gap. Just as a prior integration of two psychological theories led to novel hypotheses that would not have been formulated by either theory alone (Mikulincer et al., 2003), the TRIBE model scaffolds two theories, Socioemotional Selectivity Theory (Carstensen, 2006) and Terror Management Theory (Greenberg et al., 2014), to explain how conformity on clinical teams and superficial emotional deference within families motivate DIALs. Conformity on clinical teams and superficial emotional deference are behaviors that are themselves motivated by fears that, when making a morally salient decision, one might behave in a manner that fails to comport with someone else’s (imagined) expectations. Clinicians, patients, and family caregivers all thereby participate in a moral performance, signaling to each other via established scripts (Butler, 1988; Nentwich and Morison, 2018) that they are behaving in a manner that is morally acceptable and socially expected. For example, a patient may be prepared for death but, not wanting to disappoint his daughter, consents to dialysis; he ultimately regrets the decision given his need for repeated hospitalizations (Russ et al., 2007; Saeed et al., 2020). In a poignant account of what may be termed “performative chemotherapy,” palliative care physician Diane Meier (2014) described how an oncologist candidly confessed to her that he prescribed aggressive, futile treatments to a dying patient because he did not want the patient to think she was being abandoned. Nor did he want Meier to think he was abandoning the patient, as Meier revealed in a conference talk (2013). Thus, acts of prescribing and receiving DIALs, such as chemotherapy or dialysis, are occasionally performative, byproducts of moral performances that reflect patient and caregiver deference to each other and clinician conformity to professional norms. Before explicating how this occurs, we review two key concepts that are central to the TRIBE model: irrational biomedical exuberance (Fuchs, 1968) and the moral emotions (Haidt, 2003; Tangney et al., 2007).

Irrational Biomedical Exuberance

We derived the concept of irrational biomedical exuberance from former Federal Reserve Chairman Alan Greenspan’s (1996) famous question (“How do we know when irrational exuberance has unduly escalated asset values?”) and Victor Fuchs’ (1968) work on the technological imperative. Writing about soaring health care costs five decades ago, Fuchs observed that “medical tradition emphasizes giving the best care that is technically possible; the only legitimate and explicitly recognized constraint is the state of the art” (p. 192). Soberingly, it seems that there are no limits whatsoever. In one satirical account, a surgeon rushed to repair a fracture on a 97-year-old patient – who had died (Neuman, 2010). In contrast to “medical tradition,” Fuchs (p. 192) averred that manufacturers and those working in human service or in other health care professions, such as dentistry, “do not, and are not expected to, produce the best [product] that engineering skills permit.” Only those working in medicine or the military “in time of total war” are expected to use state-of-the-art technology, oblivious to the economic costs, environmental damage, or burden on future generations.

In many Western health care settings, irrational biomedical exuberance is emboldened by a norm, biomedicalism, that encourages norm-adherents to believe that even intractable, complex (Rittel and Webber, 1973; Glouberman and Zimmerman, 2004) problems can be solved or fixed by treatments, interventions or procedures. Anchored in a 17^th century worldview (Engel, 1992), biomedicalism is well-suited to the treatment of curable diseases that obey mechanistic laws of nature, but is ill-suited for incurable disease. The norm prioritizes disease-modifying biomedical treatment; not offering such treatment is conflated with abandonment (Quill and Cassell, 1995). The norm is accompanied by Western approaches to problem-definition and problem-solving, which tend to be mechanistic, piecemeal, focalist, and analytic, as opposed to holistic, relational, and interdependent (Henrich, 2020; Henrich et al., 2010). Biomedicalism spawns reductionism (Engel, 1992) and medicalization (Conrad, 2005; Illich, 1976), a psychocultural process by which powerful, high-status individuals, groups, and organizations, due largely to self-interest, authoritatively define ubiquitous experiences, such as aging (Clarke et al., 2009; Kaufman et al., 2006), as problems requiring costly medical intervention. Medicalization is poorly aligned with the need to craft innovative solutions for other pressing societal challenges that demand public and private resources, such as psychological suffering (Horwitz and Wakefield, 2007) and dying (Clark, 2002; Gawande, 2017; Sallnow et al., 2022). Contextualist social-ecological models can help resolve these and other complex, challenges, including gender-based violence (Heise, 1998), health disparities (Krieger, 2008), HIV prevention and AIDS care (Kaufman et al., 2014), homelessness (Fowler et al., 2019), and, as we argue here, irrational biomedical exuberance.

Motives to prescribe DIALs are prima facie innocuous - to avoid death, save a life, “do all we can,” “fight a battle,” and “never give up” (Good et al., 1990). In this frame, the risk of not using a DIAL is worse than using one, even when prescribers know that they may be engaging in futile biomedicalism or offering interventions that might confer indelible harm. Seemingly innocuous motives to prescribe DIALs are sacralized (McCleary and Barro, 2006; Weber, 2002) in medical culture and the healthcare workplace (Drought and Koenig, 2002; Kaufman et al., 2006). They are morally imbued, ritualistic defaults (e.g., cardio-pulmonary resuscitation [CPR]), even when performative, as in slow-code CPR (Lantos and Meadow, 2011).

As with any sacralized, group-based moral act, it is easy to turn a blind eye and not delve below the surface to understand the act’s origins or consequences (Haidt, 2012; Henrich, 2020; Kendi, 2017). In his explanation of the insatiable demand for medical care, Fuchs (1968) signaled that we must delve below the surface. By comparing medicine to the military in time of total war he implied that the prospect of death motivates decision-makers to make risk-benefit calculations that are inconsistent with widely-accepted rationalist principles. Our premise is that Fuchs’ “technological imperative” motivates ostensibly moral acts (e.g., prescribing, treating, military deployment, etc.) in a manner that blinds and binds participants to unintended consequences (Haidt, 2012) precisely because the moral emotions guide their decision-making.

Moral Emotions in Western End-of-Life Care

TRIBE is unique among conceptual models of healthcare utilization in its emphasis on group processes (e.g., conformity on clinical teams, superficial emotional deference in families) and its premise that emotions (Barrett, 2017), particularly the moral emotions (Haidt, 2003; Tangney et al., 2007, Table 1) motivate behavior in group settings. By definition, moral emotions “are linked to the interests or welfare either of society as a whole or at least of persons other than the judge or agent” (Haidt, 2003, p. 853). Emotions are recognized in models of patient/caregiver/clinician communication (White et al., 2018), but theorizing about the role of moral emotions in healthcare utilization is nonexistent. Andersen’s influential model of healthcare utilization (Andersen and Newman, 1973; Andersen, 1995; Bradley et al., 2002; Phillips et al., 1998) does not mention the role of emotion, let alone the moral emotions. Clinicians’ emotional competencies are ignored in impactful approaches to conceptualizing and assessing healthcare quality (Anhang Price et al., 2014; Donabedian, 1966). New theories of medical overuse (Nassery et al., 2015) pay little attention to emotion; the same is true of older theories of organizational change (DiMaggio and Powell, 1983). Influential models of health behavior, such as social cognitive theory (Bandura, 2001), the transtheoretical model (Prochaska, 2008), and the theory of planned behavior (Ajzen, 1991) acknowledge the role of norms (Birken et al., 2020) in behavior change, but not the critical role of the moral emotions in shaping or sustaining those norms. In the United States, where “Western” (Henrich, 2020; Markus and Kitayama, 2010) focalist (vs. holistic) approaches to problem-solving guide piecemeal fee-for-service (vs. capitated/bundled) medical payments that contribute to excess spending (Papanicolas et al., 2018) and the sacred cow of biomedicalism (Leff and Finucane, 2008) motivates overtreatment (Angell and Relman, 2002; Crowley et al., 2020; Relman, 1994; Rosenthal, 2017), there remains a dearth of theorizing that recognizes the foundational role of the moral emotions in healthcare.

Table 1.

Moral emotions, processes, and motives in healthcare

Shame and Guilt. Emotions that are primarily evoked by moral lapses; in shame, the primary concern is others’ evaluations of the self. In guilt, the primary concern is the negative effect of one’s behavior. Shame is more painful than guilt because one’s core self is impugned, not simply one’s behavior. People experiencing shame will often try to defend against, deny, conceal, hide, or escape from the shame-inducing situation. In contrast, guilt can generate remorse which can lead to proactive verbal actions (confessions, apologies) and enduring behavior change. Whereas guilt can increase sympathy and compassion, shame can have the opposite effect and lead to disgust, contempt or moral indignation.

Anticipatory Shame-Avoidance: Engaging in an action that is motivated by the desire to avoid others’ anticipated disgust, contempt, or moral indignation; contributes to clinicians’ emotion-driven conformity to the local norm. Anticipatory Guilt-Avoidance: Engaging in an action that is motivated by the desire to avoid impugning one’s behavior (to oneself) or giving the appearance to oneself of harming another. Also referred to as regret-avoidance; contributes to clinicians’ emotion-driven conformity to the local norm.

Sympathy, Compassion, Empathy: Interpersonal moral processes, not emotions, that contribute to superficial emotional deference in end-of-life care. These processes are evoked by the imagined or actual emotional state of another. Whereas sympathy merely involves feelings of concern for the emotional state of another, compassion motivates action that is propelled by the desire to alleviate the other’s imagined or actual suffering. Of these three interpersonal processes, only empathy requires an accurate interpretation of the other’s emotional state; empathy is thus less common than sympathy or compassion.

Disgust, Contempt, Moral Indignation: Strong emotional responses that are evoked by moral violations. Disgust is also evoked by physical objects (e.g., rotten food). No matter what the source, disgust motivates a desire to avoid, expel, or escape from the offensive object, often accompanied by a motivation to wash, purify, or remove any reminder of the object. Contempt is similar to disgust but is evoked primarily if not explicitly by moral or social violations, not physical objects, and is not accompanied by purification motives. Unlike expressions of contempt, expressions of moral indignation always involve an explicit reference to a moral violation. Clinicians’ emotion-driven conformity may be motivated by anticipatory avoidance of colleagues’ disgust, contempt, and indignation.

Note. The definitions provided here were synthesized from numerous sources, especially Haidt (2003, 2012) and Tangney et al. (2007).

No conceptual frame can attempt to explain all of the vagaries of EoL care that might lead to DIALs. Our focus is confined to clinician conformity to local norms on clinical teams and superficial emotional deference in patients and families. These tribes – care teams and social networks – shape the experience and expression of moral emotions that fuel EoL treatment.

Conformity on Clinical Teams and Superficial Emotional Deference in Families

Conformity to local norms on clinical teams.

Conformity is a basic value and universal (Schwartz et al., 2012) phenomenon. Conformity can explain the presence of local norms in healthcare settings (Barnato, 2017; Barnato et al., 2014; Cutler et al., 2019; Molitor, 2018) and why EoL treatment intensities vary from region-to-region and hospital-to-hospital (C-TAC, 2019; Dartmouth Atlas Project, 2019). When physicians re-locate from one part of the country to another, or one hospital to another, their practices often shift to reflect local norms (Molitor, 2018). Clinicians may deliberately conform to peer norms, but we posit that conformity behavior on clinical teams is partially accounted for by nonconscious, mimetic processes that have been shown to explain, for example, how spouses grow alike over time (Monin and Schulz, 2009; Zajonc et al., 1987). Procedural conformity to a norm can explain why advance directives are often ignored at the point-of-care and thus have had profoundly disappointing effects (Jimenez et al., 2018; Morrison, 2020) and how “clinical momentum” makes it impossible to slow down and deliberate in intensive care units (Kruser et al., 2017). In many healthcare settings, decisions are made without reflection, often with unfortunate consequences for patients (Kruser et al., 2017), clinical teams, families, and society. Although supply-side (e.g., clinician preference) and demand-side (e.g., patient preference) factors both explain variation in healthcare utilization, supply side factors are far more important (Cutler et al., 2019; Finkelstein et al., 2016), and are particularly vulnerable to conformity effects.

Superficial emotional deference in families.

Fear and uncertainty motivate patients to seek input from others after learning that they have a life-limiting diagnosis (Ferrer et al., 2021). High-stakes decisions around EoL care are, ideally, evidence-based and, inevitably, emotion-based (Ellis et al., 2019). Unfortunately, patients and families often misunderstand (Malhotra et al., 2019) the evidence and misconstrue others’ emotions. Moral emotions are paramount. Months before his death from cancer, a professor observed that “the experience of dying…may have been fundamentally about etiquette, about how one deals socially with people who don’t have cancer… a kind of wall seems to persist, a barrier of fear, shame, and perhaps most of all embarrassment” (Bell, 1996, p. 40). A son may know that his father is dying but, thinking that he would break his father’s spirit if he recommended hospice, he encourages third-line treatment. Moral processes (Rai and Fiske, 2011; Tangney et al., 2007), including sympathy and compassion, lead patients and family caregivers to subordinate their own judgement about the most appropriate course of action to someone else’s imagined emotions, life goals, and desires. Pretense (Seale et al., 1997) leads to collusive subversions of the truth in families. We call this phenomenon superficial emotional deference. Given the powerful role of culture in shaping human behavior (Markus and Kitayama, 2010; Stephens et al., 2012), specific family cultures (Kagawa-Singer et al., 2010) will influence how such deference is enacted. As Lown (1999, p. 73) put it, “Doctors, in the words of Reinhold Niebuhr, mean well, do ill, and justify their ill-doing by their well-meaning.” The same is true of patients and caregivers.

The TRIBE Model

Drawing from biopsychosocial (Engel, 1980) and social-ecological (Bronfenbrenner, 1979) approaches to human development, TRIBE offers a social-ecological, contextualist explanation of irrational biomedical exuberance. The model has two parts, depicted in Figures 1 and 2. Figure 1 is an adaptation of a standard social-ecological model based on Bronfenbrenner’s decades of theorizing. The center of the Figure shows the clinical microsystem, which includes clinical teams (physicians, nurses, social workers, pharmacists, medical assistants, office staff, etc.) that provide specialty care for life-threatening illness (e.g., oncology, nephrology, cardiology, emergency, intensive). The microsystem frames all interactions with patients and their families, who (ideally) are at the center of the microsystem. The outer rings are the meso-, exo-, and macrosystems. In TRIBE, the mesosystem refers to other systems that interact with the clinical microsystem; in theory, strengthening these other systems (e.g., primary care) could improve EoL care. The exosystem refers to institutional structures that interact indirectly with the micro- and mesosystems. The Figure depicts institutional arrangements that (unintentionally) undermine the structures and processes in the micro- and mesosystems that could mitigate irrational biomedical exuberance. The macrosystem refers to broader psychocultural beliefs (e.g., belief that complex problems are merely complicated; Rittel and Webber, 1975; Glouberman and Zimmerman, 2004) and structural inequities (e.g., racism, Kendi, 2017) that shape everyday life and affect biomedical exuberance indirectly and insidiously. The chronosystem reminds us that outcomes are historically-contingent, ever-evolving, and affected by chance (Taleb, 2007).

Figure 1. The Social Ecology of Biomedical Exuberance at the End-of-Life.

All microsystem activities aimed at mitigating the harmful effects of DIALs will be hindered or facilitated by activities in the mesosystem (depicted here as mitigating harms), the exosystem (depicted here as both conferring and mitigating harms), and the macrosystem (depicted here as conferring harms). See text and online Appendix for details.

1.5 column

Figure 2. Socioemotional Processes Fueling End-of-Life Care in the Clinical Microsystem.

Clinical Teams: Based on Terror Management Theory, TRIBE presumes that the prospect of death motivates healthcare teams to conform to a biomedical norm of care. Activation of the biomedical norm is accompanied by a defensive process, called worldview defense, which can lead directly to DIALs via procedural conformity to the local biomedical norm of care (Path A) as clinicians “go with the flow” (Renkema et al., 2008) and model themselves after their colleagues. Alternatively, the effect of worldview defense on DIALs may be indirect, through emotion-driven conformity and the activation of moral emotions (e.g., shame), processes (e.g., compassion) and motives (e.g., anticipatory shame-avoidance). In this scenario, clinicians initially resist going with the flow but ultimately conform and capitulate to the biomedical norm.

Families and Patients: Based on Socioemotional Selectivity Theory, TRIBE presumes that patients and members of their social network prioritize socioemotional goals, which precipitates positivity-driven beliefs about prognoses and leads patients to consent to interventions that confer more harm than benefit (Path D). Alternatively, the effect of socioemotional prioritization on DIALs may be indirect, via superficial emotional deference (Paths E and F).

Moral Performance (dotted oval): Both superficial emotional deference and emotion-driven conformity to norms are characterized by assumptions made about someone else’s beliefs or emotions about morally questionable acts. As such, acts of emotional deference and conformity in the EoL setting can be characterized as performative, leading to performative discretionary interventions, such as performative CPR or performative chemotherapy (Paths C and F).

Qualifiers: Terror Management Theory can explain patient and caregiver conformity to local social network norms, and Socioemotional Selectivity Theory can explain clinician deference to patients and caregivers. For simplicity, these processes are not depicted.

2 column

Figure 2 puts the clinical microsystem under a psychological microscope. A life-limiting diagnosis is a complex, intractable, emotion-generating (Ellis et al., 2019) problem. There is no optimal solution, rarely a single goal on which all parties can unambivalently agree, and the swirl of emotion makes it difficult to understand the gravity of the prognosis (Finkelstein et al, 2021; Glare et al., 2003; Malhotra et al., 2019, Loh et al., 2022).

A life-limiting diagnosis elicits two powerful, non-normative terrifying thoughts: thoughts of impending death and thoughts of limited time. Conceptual frames that elucidate how and why treatment decisions are communally “made” as all parties manage uncertainty while contemplating looming death and dwindling time are critical to improving healthcare delivery. Instead of creating a new conceptual frame, we scaffold robustly supported psychological theories of motivation, social behavior, and group processes that are particularly relevant to decision-making in the setting of Western healthcare.

Whereas Terror Management Theory (Greenberg et al., 1986; Greenberg et al., 2014) can illuminate behavior when the prospect of death looms large, Socioemotional Selectivity Theory (Carstensen, 2006; Carstensen et al., 1999) offers insights into behavior when individuals are faced with the prospect of imminent loss. Whereas Socioemotional Selectivity Theory is particularly well-suited to explaining superficial emotional deference within families, Terror Management Theory can explain the procedural conformity behavior that sustains practice norms (Barnato, 2017; Barnato et al., 2015; Barnato et al., 2014; Kaufman et al., 2004; Knutzen et al., 2020; Molitor, 2018).

Terror Management Theory: Conformity on Clinical Teams

Terror Management Theory (Greenberg et al., 1986) builds on Ernest Becker’s Pulitzer-prize winning book, The Denial of Death. Becker was a cultural anthropologist whose work was influenced by Sigmund Freud and other psychoanalytic theorists. Terror Management Theory has had a sustained influence in the fields of social, organizational, clinical and health psychology over the past three decades, but has rarely been applied to healthcare utilization (Arndt and Goldenberg, 2017; Goldenberg and Arndt, 2008; Solomon and Lawlor, 2011). The theory outlines what happens when people experience “mortality salience” – the sense of terror engendered by the prospect of death. People experience death anxiety consciously or nonconsciously upon exposure to a mortality cue. Such cues are episodically present in everyday experience (e.g., news) and in the built environment (e.g., cemeteries), but they abound in hospitals and are pervasive in specialty settings where dying patients are treated, such as cancer centers, dialysis clinics, or intensive care units. A key concept in Terror Management Theory is worldview defense (Figure 3), a mechanism by which people maintain their self-esteem while under threat of annihilation by adhering to a belief system that confers a sense of meaning, order, purpose, or solidarity. Worldview defense following exposure to a mortality cue was initially shown to explain derogatory attitudes toward sexual immorality and has subsequently been shown to be related to attitudes toward other divisive topics, including racism (Greenberg et al., 2001).

Figure 3. Irrational Exuberance on Clinical Teams: Conforming to the Biomedical Norm.

Worldview defense is a mechanism by which people maintain their self-esteem when exposed to reminders of death. Worldviews (i.e., ideologies, paradigms) that promote a sense of meaning and group connection are elevated. In healthcare settings, conformity on clinical teams is motivated by cognitive processes (e.g., decreased receptivity to threatening information, desire to identify with the in-group) that prioritize technical over emotional knowledge (Zussman, 1992). These cognitive processes are reinforced by mantras and metaphors (Reisfield and Wilson, 2004).

1 column

People tend to deny or diminish the prospect of death and suppress death-related thoughts, often with the encouragement of psychocultural ideologies, such as biomedicalism, religion, or racism. Physicians who are more comfortable with biomedicalism prescribe more DIALs to dying patients (Duberstein et al., 2019); dying patients who are more religious tend to receive more DIALs as they pray for a miracle (Shinall et al., 2014); Black patients receive more DIALs for reasons that remain unknown, but structural racism (Bailey et al., 2021) and racist stereotypes about Blacks’ treatment preferences might be implicated (Koenig and Gates-Williams, 1995).

The motivating effects of death anxiety on social behavior have been studied in social psychology laboratories. It would be unwise to ignore the implications for EoL care of these laboratory tests of Terror Management Theory, just as it would be imprudent to ignore the implications for human genetics of Drosophila research in genetics laboratories (Ellis et al., 2019; Ferrer et al., 2016). Still, it should be acknowledged that the theory is controversial, due to failed replications (Klein et al., 2019) and conceptual disagreements (Kirkpatrick and Navarrete, 2006; Tritt et al., 2012), particularly about the status of fear of death as a motivator of conformity behavior (Kirkpatrick and Navarrete, 2006; Navarrete et al., 2004). These concerns have been successfully rebutted (Chatard et al., 2020; Chen et al., 2022; Pyszczynski et al., 2015).

Research on Terror Management Theory typically involves exposing participants to a mortality cue and then observing participant behavior in the laboratory. These studies show that conformity behavior increases after exposure to a mortality cue. We theorize (Figure 2) that conformity to local norms can lead directly to DIALs by default via procedural conformity (Path A); alternatively, the effect may be indirect, via emotion-driven conformity (Paths B and C). The distinction between procedural vs. emotional subtypes of conformity is well-established (Schwartz et al., 2012).

Procedural conformity on clinical teams.

Without mechanisms to motivate conformity, organizations fail. Interpretation of data from decades of conformity experiments (Asch, 1955; Bond and Smith, 1996) typically presumes that people conform because they are motivated to behave in a manner that increases their psychological similarity to the in-group. Intuitively, it might seem as though conformity to norms is simply a benign, non-defensive, behavioral process. Terror Management Theory suggests otherwise.

In settings where death is a potent stimulus (e.g., ICU, cancer center), conformity to a norm serves as a defensive reaction to threat. In Figure 3, we show that worldview defense (Arndt et al., 2009; Martens et al., 2011; Vess et al., 2009) can lead to DIALs via decreased receptivity to both threatening information (Greenberg et al., 2014; Landau et al., 2004) (e.g., “the patient is dying”) and to alternative courses of action (“stop the current treatment plan”).

Well-controlled experiments have shown that exposure to mortality cues, even when outside of conscious awareness (Greenberg et al., 1994), activates neural mechanisms in a span of milliseconds (Dor-Ziderman et al., 2019) along with personally-relevant ideologies (e.g., biomedicalism) that motivate people to conform to worldviews associated with the ideology.

Terror Management Theory suggests that conformity to a psychoculturally-enabled biomedical norm (“more is better”, “doing less means giving up”) would mitigate death anxiety in clinicians. This idea, which has been discussed in the clinical literature for decades (Menzies, 1960), has experimental support. In one study (Solomon and Lawlor, 2011), researchers asked medical students to describe the emotions aroused by the thought of their own death (mortality cue) or a control cue. Participants then read about a hypothetical patient whose advance directive explicitly stated that he did not want any aggressive, disease-directed treatment. The students who had been asked to contemplate their own death expressed greater support for aggressive treatment. The authors concluded that clinicians “may strive to keep their patients alive to assuage their own death fears” (Solomon and Lawlor, 2011, p. 103).

The TRIBE model presumes that conformity to the biomedical norm, even in circumstances when such conformity is counterproductive, is enabled by unexamined death anxiety in clinicians (Menzies, 1960; Mount, 1986). There is support for this idea, too. (Eggerman and Dustin, 1986; Rodenbach et al., 2016; Schulz and Aderman, 1979). One study showed that when clinicians are more comfortable with the biomedical norm, their patients are more likely to receive DIALs (Duberstein et al., 2019). This might be especially true when clinicians practice a religion (Shinall et al., 2014). Other research has shown that clinicians with greater death anxiety are more likely to have negative attitudes toward disclosure of a poor prognosis (Cialkowska-Rysz and Dzierzanowski, 2013; Cochrane et al., 1991) and palliative care (Thiemann et al., 2015).

Emotion-driven conformity on clinical teams.

In EoL settings, emotion-driven conformity occurs when motivated by conscious fear of ridicule and ostracism or the conscious experience of moral emotions (e.g., shame, guilt, contempt). Meier’s (2014) description of how a physician candidly revealed that he prescribed a futile treatment because he was concerned about what others would think of him starkly illustrates that patients receive treatments for reasons other than their underlying illness (Bluhm et al., 2016; Jackson et al., 2008). It also exemplifies how clinicians conform to a norm after contemplating how their actions might be perceived by others. Nonconformity exacts a penalty, including ridicule and derogation (Arndt et al., 2009; Burke et al., 2010; Renkema et al., 2008). Such derogation can occur on care teams, where disagreements around a plan of action are common (Adamy and McGinty, 2012).

Socioemotional Selectivity Theory: Superficial Emotional Deference in Patients and Caregivers

Socioemotional Selectivity Theory (Carstensen, 2006; Carstensen et al., 1999) was developed in the 1980s in response to the then influential belief that age-related social withdrawal was a reaction to preconscious awareness of impending death and biological change. Carstensen and colleagues countered that this pattern of withdrawal was, instead, motivated by perceived changes in time horizon (Carstensen, 1995; Lang, 2001) and the concomitant prioritization of emotional (vs. instrumental) goals, the desire to spend time participating in enjoyable activities and avoiding unpleasant situations. With aging, social networks become smaller in part because older adults have learned to retain only the most emotionally meaningful relationships.

Whereas future-oriented instrumental goals (e.g., achieving the competencies needed for career aspirations) are prioritized in early adulthood, present-oriented socioemotional goals (e.g., spending meaningful time with grandchildren) become increasingly prioritized in middle age (Carmichael et al., 2015) when people experience a sense of time running out. Changes in time horizon are most obvious in older adulthood as people gradually appreciate that their lifespans are limited and earthly time is finite. However, beyond aging, there are other elicitors of changes in time horizon, such as life-limiting illness (Sullivan-Singh et al., 2015).

Laboratory studies suggest that time horizon changes lead to general motivational preferences for positive over negative stimuli (Mather and Carstensen, 2005). This positivity effect (or bias) is remarkably robust (Reed et al., 2014) and has been observed in studies of basic psychological processes, including attention (Samanez-Larkin et al., 2009) and memory (Mikels et al., 2005) as well as in studies of more complex processes, including healthcare decision-making (Löckenhoff and Carstensen, 2007, 2008). The positivity effect is presumed to be nonconsciously motivated by the desire for positive, hedonic experiences; it has been observed in laboratory tasks that require judgments made in approximately 500 milliseconds (Isaacowitz et al., 2009).

Positivity can explain the seemingly counterintuitive finding that emotional well-being increases (Blanchflower and Oswald, 2008; Carstensen et al., 2011) or remains stable (Scheibe and Carstensen, 2010) across the lifespan and is rarely compromised by disability (Albrecht and Devlieger, 1999). The positivity effect can explain why some individuals are vulnerable to deceit, including financial fraud (Scheibe and Carstensen, 2010), and why some may be unduly swayed (Tanco et al., 2015) by positively-valenced health communications, such as “Your scans look great!” (Good et al., 1990; Wolf and Wolf, 2013). Positivity can explain why prognostic understanding is poor (Finkelstein et al., 2021; Malhotra et al., 2019), as it motivates individuals to ignore, discount, or forget negatively-valenced prognostic communications. Positivity effects might be eliminated in high stakes situations, but there is evidence to the contrary (Löckenhoff and Carstensen, 2007).

Based on Socioemotional Selectivity Theory, we hypothesize that a constricted time horizon motivates a positivity effect, which leads people to believe that their prognoses are better than they actually are, and this in turn incentivizes DIALs (Figure 2, Path D). Alternatively, as discussed below, the effects of socioemotional prioritization on DIALs may be indirect, via superficial emotional deference, the subordination of one’s own judgment about the most appropriate course of action to someone else’s imagined, or anticipated, emotions, goals, or wishes (Figure 2, Paths E and F).

Superficial emotional deference within families.

Illness and aging motivate people to regulate their emotions by not fully engaging in emotionally taxing situations, like healthcare visits (Charles et al., 2009). Socioemotional Selectivity Theory suggests that, when family members accompany seriously ill patients to healthcare visits (Wolff and Roter, 2011), the family members are often providing emotional and cognitive support, in addition to instrumental support (e.g., transportation). When patients defer decisions to family caregivers, they do so both to regulate their own emotions and convey their moral concern about the caregiver’s emotional welfare (Maxfield et al., 2014; Maxfield et al., 2007). For example, dying patients may want to prepare for death but opt for experimental treatment out of a desire to please family members who implore them to keep fighting (Hoerger et al., in press). (They may consent to procedures to please their physicians as well (Saeed et al., 2020), but our focus here is family dynamics.)

If it is difficult to forecast one’s own ability to adjust to misfortune (Hoerger et al., 2012; in press; Smith et al., 2009), it is near impossible to accurately assess another party’s capacity to do so. Without satisfactory information about the other’s emotions and beliefs, one must leap to conclusions, “fill in the blanks” and lean on default assumptions and heuristics to fill the void. In most settings, the default is “more treatment” (Ellen et al., 2020; Le Fanu, 2012; Nassery et al., 2015; Rosenthal, 2017; Segal et al., 2015). As Kaufman, Shinn, and Russ (2004, p. 9) stated, “biomedical technique provides the most powerful logic, the most pervasive method” that patients and caregivers can show each other that they truly care about each other.

Superficial Emotional Deference and Emotion-Driven Conformity: Moral Performance in Healthcare

Both superficial emotional deference in families and emotion-driven conformity to local professional norms are characterized by assumptions one party makes about another’s attitudes, beliefs, or emotions about morally questionable acts. These assumptions are rarely tested, let alone verified. As such, some moral acts in the EoL setting can be characterized as performative (Butler, 1988) (Figure 2) and thus often not constitutive of core, essentialist, abiding preferences and values. Yet values and preferences are presumed to be foundational in EoL care (Institute of Medicine, 2015; Wolf et al., 2015).

The presence of performativity in EoL care raises troubling ethical questions about the ontological status of values and preferences in healthcare, clinician concealment of prognostic information (often nonconscious), clinician engagement in other performative acts (prescribing, intervening; often nonconscious), and the rationalist elicitation of so-called values and preferences to inform treatment decision-making (Drought and Koenig, 2002; Eyal and Sjöstrand, 2020; Fagerlin and Schneider, 2004; Kaufman et al., 2004; Mack and Smith, 2012; Panagopoulou et al., 2008; Wirtz et al., 2006). The activation and anticipation of painful emotions could explain why clinicians conceal prognostic information (Mack and Smith, 2012; Panagopoulou et al., 2008), and why, in contrast to idealized, rational models of EoL care, embodied in advance care planning, values-driven preferences must inexorably be constituted “on the spot” (Kaufman et al., 2004, 2006) or “in the moment” (Sudore and Fried, 2010). It is thus hardly surprising that rationalist approaches to EoL care often fail (Fagerlin and Schneider, 2004; Morrison, 2020), and that legal documents such as advance directives, while frequently helpful (Silveira et al., 2010), are ignored and become de facto works of fiction.

Compelling case reports and qualitative studies suggestive of moral performance in EoL care have been published (Fallowfield et al., 2002; Meier, 2014). Performative CPR was considered ethical until the 1980s (Lantos and Meadow, 2011). Likewise, the near absence of truth-telling in cancer care settings (Oken, 1961) until recent decades is partially a product of superficial emotional deference (Fallowfield et al., 2002). It was commonly assumed that white lies of omission, intended to not break the patient’s spirit, were relatively harmless; there is scant evidence for the validity of that assumption (Fallowfield et al., 2002; Levine et al., 2018) and patient satisfaction increases when physicians are forthcoming about prognosis (Fenton et al., 2018). In fact, white lies can be harmful to patients (Levine et al., 2018), patient satisfaction with care declines when patients perceive that physicians are emotionally inauthentic (Yagil and Shnapper-Cohen, 2016) and offering false hopes to patients could engender moral distress in clinicians (Saeed et al., 2021).

We are not suggesting that moral performance on the part of patients, caregivers, or clinicians is inherently bad or reflects an inauthentic, “false” self. Rather, moral performance in Western EoL settings is due, in part, to the terrifying socioemotional context of EoL care, the prospects of limited time and imminent death. Sometimes, patients receive treatments, interventions, and procedures because clinicians feel compelled to conform to a norm, and patients and caregivers feel emotionally moved to go with the flow in the service of the moral emotions, caring and love. These dilemmas cannot be addressed without first being identified, labelled, and conceptualized.

DISCUSSION

The TRIBE model synthesizes two established psychological theories to explain why dying patients receive DIALs despite repeated calls for alternative care models (Committee on Quality of Health Care in America and Institute of Medicine, 2001; Engel, 1992; Lown, 1999) and numerous interventions designed to improve EoL care over the past quarter century (Anonymous, 1995). After outlining future research directions, we call for empirically-informed psychological innovation in education and policy.

Future research directions:

TRIBE complements other conceptual models of EoL utilization (Kelley et al., 2010; Prigerson and Maciejewski, 2012) and overtreatment more broadly (Nassery et al., 2015) by encouraging researchers to consider conformity processes and moral emotions in healthcare settings. Whereas prior frameworks (Kelley et al., 2010; Prigerson and Maciejewski, 2012) can lead to the development of novel individual-level interventions, TRIBE has implications for the development of psychologically-informed dyadic, triadic, family, and organizational interventions to decrease the harms of overtreatment in EoL settings and other treatment settings where the prospect of death looms large. New interventions are needed to optimize EoL care, given that prior interventions have repeatedly failed (Abedini et al., 2019). Ideally, future intervention research would be informed by a theory of DIALs (e.g., TRIBE) as well as a theory of implementation (e.g., institutional theory; Birken et al., 2020; DiMaggio and Powell, 1983). Future intervention research should be conducted in diverse cultural settings, including diverse family sociocultural contexts and different treatment delivery settings (e.g., academic vs. community care settings). Researchers must acknowledge that healthcare is offered by groups of people (clinicians working on teams within organizations) to groups of people (patients in social and kinship networks) and that clinicians, teams, and organizations respond to emotional pressures related to maintaining their legitimacy. Birken et al. (2020) identify three such pressures: regulatory, mimetic, and, of greatest relevance to TRIBE, normative.

In Table 2, we identify two normative processes that, if appropriately modified, could lead to improvements in the use of DIALs and EoL care, clinician conformity to the biomedical norm and superficial emotional deference. For example, innovative dyadic (patient-caregiver), triadic (patient-caregiver-clinician) and family “emotional competency” interventions are needed that center the role of moral emotions and acknowledge the potential role of superficial emotional deference and performativity in EoL care (Hoerger et al., in press).

Table 2.

Norm-Modifying Interventions Suggested by the TRIBE Model

Modifiable normative process		Interventions
1) Clinician conformity	1a) Emotion-driven conformity	Training in emotional competencies, such as affective forecasting, emotion identification, and moral performance Interdisciplinary Schwartz rounds, Balint groups, and mindfulness interventions in healthcare settings
	1b) Automatic (Procedural) conformity	Norm-modifying interventions, including nudges, de-implementation and de-prescribing Interventions for death anxiety
2) Superficial emotional deference	2a) Patient and caregiver superficial emotional deference to each other	Innovative dyadic (patient-caregiver) and family interventions that center the role of moral emotions, motives, and processes in the context of serious illness
	2b) Superficial emotional deference in the patient/caregiver-clinician relationship	Same as above but adds a clinician training component

Beyond recommending intervention research, we offer three substantive directions for future scholarship on the implications of socioemotional processes for EoL healthcare delivery (Table 3). First, there is a pressing need to examine clinician conformity to potentially harmful norms. Mixed methods research and ethnographic research is needed to explore how and why some clinicians, care teams, specialties, or sub-specialties are more likely to conform to unhelpful local norms than others (Knutzen et al., 2020); research on the effects of clinician age, race, ethnicity and gender would be helpful. Theory (Menzies, 1960) and research (Solomon and Lawlor, 2011) suggest that clinicians embrace the biomedical norm because it quells clinician anxiety, including death anxiety; survey and mixed methods research on this question are needed. There is some evidence to suggest that race differences in DIALs are due to ICU culture (Barnato et al., 2006; Barnato et al., 2007; Quill et al., 2014); shared stereotypes about group preferences might also be relevant (Koenig and Gates-Williams, 1995). For example, based on the research literature on race differences in so-called preferences, clinicians may stereotype Black patients and think that all Black patients want aggressive interventions. Consequently, in some Western settings, White clinicians may be less likely to initiate a difficult conversation with Black patients or family members about the potential harms of the aggressive intervention because of this stereotype. Phenomenological, ethnographic, and quantitative research is needed on the contributions of shared stereotypes to patient care.

Table 3.

Research Questions Generated by the TRIBE Model

Research topic	Research questions
Clinician conformity	To what extent do clinicians believe they offer DIALs or are pressured to offer DIALs to conform with peer norms in their organization? To what extent do clinicians experience pressure to fake emotions and enact a moral performance for the benefit of their peers and patients? Does such moral performance increase risk of burnout? Is the relationship between conformity to the biomedical norm and DIALs partially mediated by moral emotions, processes, or motives (e.g., shame, pride)? Can brief psychological interventions for clinicians mitigate the risk of unhelpful procedures becoming appropriate by default?
Patient and caregiver superficial emotional deference	How do moral emotions (shame, guilt), interpersonal processes (e.g., sympathy) and motives (anticipatory shame-avoidance) shape communication and treatment decision-making within families, for better and for worse? To what extent do patients consider and accurately perceive and forecast the emotions of caregivers? To what extent do caregivers consider, accurately perceive and forecast the emotions of patients? How do these forecasts influence high-stakes treatment decisions and EoL care? Do patients and caregivers feel pressure to put on a moral performance in response to their loved one’s actual or anticipated emotions?
Moral emotions in treatment decision-making	To what extent do patients, caregivers, and clinicians express moral sentiments at the point-of-care, and how do these expressions affect EoL care? To what extent do patients agree to treatments out of guilt, shame, or compassion? To what extent do clinicians offer treatments out of a sense of sympathy or compassion, and how do power differentials, social distance, and cultural differences factor into these emotional responses? To what extent do the moral emotions, processes, or motives encourage or discourage technologically intensive interventions and referrals to palliative care or hospice?

Second, there is a need for programmatic research on patient and caregiver superficial emotional deference. Qualitative research would be particularly helpful. Although there have been numerous studies on negative emotions (e.g., anxiety, sadness) in EoL settings, there has been a dearth of quantitative and qualitative research on moral emotions (Ferrer et al., 2016). Third, a comparable program of research is needed on the role of the moral emotions in patient-clinician communication and treatment decision-making. TRIBE focuses on interactions within care teams and within patients’ networks that have historically been neglected in models of healthcare utilization. However, the TRIBE model also has implications for research on health communication and decision-making in clinical settings.

Psychological innovation in education and policy: Recognizing and responding to the biomedical bias.

A former editor of the New England Journal of Medicine once opined, tongue-in-cheek, that “only those who have been hospitalized…[should] be admitted to medical school” (Ingelfinger, 1980). He assumed that the first-hand experience of hospitalization would boost clinician empathy and mitigate the magnetic-pull of the biomedical norm. He was probably right, but it would be far more efficient to modify the processes by which clinicians are educated and professionally acculturated.

Curricular changes and the growth of palliative care programs bode well for decreasing the harmful effects of DIALs (Denney-Koelsch et al., 2018). As others have suggested (Berwick and Finkelstein, 2010; Eisenberg, 1988), immersion in the social, behavioral, and organizational sciences would prepare the healthcare workforce to manage and respond to intractable clinical problems like those that arise at the EoL.

Exposing trainees to coursework in these areas is necessary, but not sufficient. Fostering the development of emotional competencies (Barrett, 2017), particularly with respect to the moral emotions (Haidt, 2003; Tangney et al., 2007), is critical. The hidden curriculum (Branch Jr et al., 2017), and its attendant biomedicalist biases, encourages detached concern and denial of feeling while teaching students to view themselves as technically-competent fixers whose sense of identity (Cruess et al., 2015; Gawande, 2017) derives largely from taking action (Coulehan, 2009; Fox and Lief, 1963; Menzies, 1960; Mount, 1986). Licensure exams and the approach to healthcare financing reinforce the implicit connection between competence and action-oriented procedures, as opposed to cognitive and emotional labor (Crowley et al., 2020).

To foster cultural humility, healthcare professionals should be made aware of the implicit, automatic (vs. deliberate), nonconscious bias favoring biomedicalism in their training. Most patients and families receive care in clinical microsystems that are affected directly and indirectly by the culture of biomedicalism– for better and worse. Clinicians who are trained to gain insight into how and why their clinical decision-making might be influenced by macrosystem forces, such as structural racism (Figure 1), that perpetuate a culture of biomedicalism that fosters reductionism and essentialism (Engel, 1992; Dar-Nimrod and Heine, 2010) will be positioned to offer higher quality care – and perhaps advocate for system change. To further foster cultural humility, healthcare professionals should be trained to understand the limits of clinical prognostication (Glare et al., 2003), which is a product of the decades-long marginalization of prognostication in clinical science (Christakis, 1997) and a culture of optimism.

To help learners appreciate some of the biases and structures that sustain them, we recommend group interventions that focus on identity formation (Cruess et al., 2015), perhaps integrating communication principles (Epner and Baile, 2014); mindfulness exercises; and candid discussions of death and, when appropriate, interventions for death anxiety (Menzies et al., 2018). These interventions can help learners ranging from medical students in longitudinal, integrated, community-based clerkships (Hirsh et al., 2012) to seasoned clinicians participating in Balint groups (Epner and Baile, 2014) or Schwartz rounds (Lown and Manning, 2010) gain insight into how moral emotions and motives affect how they interact with patients and influence patient outcomes. They will come to view skilled communication as an active intervention that can be delivered with varying degrees of expertise. When combined with interventions to mitigate unhelpful norms at the point-of-care and decrease performativity in families and in the patient-caregiver-clinician relationship (Table 2), improvements in EoL care and outcomes could be expected.

Caveats and qualifiers.

First, TRIBE integrates two psychological theories that have been extensively tested in controlled laboratory settings, but health decisions cannot be readily simulated in laboratory settings. Second, EoL care involves multiple parties who occupy multiple roles (patient, wife, daughter, churchgoer) in multiple contexts (families, church, work). Membership in one or more groups, be it a family, a religion, a profession, a political party or an advocacy group poses challenges, as norms across roles and contexts may be incompatible. As such, healthcare professionals are not just clinicians but members of a profession or a religion. TRIBE, like most prior conceptual models, ignores role multiplicity and intersecting identities. Third, TRIBE ignores power hierarchies in conformity, but it may be psychologically easier for established attending physicians to question the biomedical norm than for nurses or residents to do so; indeed, more experienced physicians are less likely to offer DIALs (Frost et al., 2011). Fourth, TRIBE focuses on ways to avoid “bad deaths,” not how “good deaths” (Steinhauser et al., 2000) may be engendered. Fifth, TRIBE is a contextualist social-ecological conceptual model, not a formal theory. Like similar models (Heise, 1998; Kaufman et al., 2014; Nelson et al., 2002), it can inform implementation science, which, at its heart, is a science of sociocultural and systemic change (Hawe et al., 2009; Miller et al., 2019). As such, TRIBE can stimulate the development of novel clinical and health services interventions as well as curricular innovations in clinical education. Although causal elements of the model can and should be tested in isolation, TRIBE is not designed to be formally evaluated using methods of causal inference. Sixth, Terror Management Theory is well-suited to one of our goals, namely the explanation of conformity on clinical teams. It is a controversial theory, and may not explain findings that could emerge in future research, such as gender differences in clinician conformity to biomedical norms. Researchers interested in studying conformity in healthcare settings might also want to draw from evolutionary (Griskevicius et al., 2006) and coalitional psychology (Navarrete et al., 2004; Navarrete and Fessler, 2005).

Finally, this introduction of the TRIBE model focused primarily on the clinical microsystem, not broader sociocultural, political and economic forces and power structures operating in the meso-, exo-, or macro-levels of the healthcare ecosystem. The United States is the only country in the Organization of Economic Cooperation and Development without universal health insurance. It is also one of the few countries worldwide that allows the pharmaceutical industry to advertise directly to consumers. Forceful messages incentivizing the use of more treatments, interventions, procedures and services are propagated by other entities in the medical-industrial complex, including the healthcare (Vater et al., 2014), medical device, and private equity industries (Rosenthal, 2017). The United States has repeatedly resisted calls for federally-supported interventions to improve health, including universal healthcare financing (Quadagno, 2004). The resistance is not surprising; for decades, the American Medical Association has fiercely opposed national health insurance. Efforts to stop the runaway train of irrational biomedical exuberance must countenance and remove the perverse financial incentives that enrich entities in the exosystem (including professional associations) while bankrupting American families (Himmelstein et al., 2019). By elevating the role of socioculturally motivated moral emotions and processes in legitimizing and sustaining technological imperatives (Fuchs, 1968) that contribute to overtreatment worldwide (Crowley et al., 2020; Elshaug et al., 2017), social scientists could develop new interventions to subvert unhelpful biases favoring biomedicalism and thereby lessen the burden of DIALs on individuals, families, society, and future generations. A new evidence base would inform policy that leads to broader psychocultural change in healthcare delivery and financing.

CONCLUDING COMMENTS

Two foundational facts have been marginalized in the social science literature on healthcare utilization: 1) healthcare is offered by groups of people to groups of people, and 2) socioemotional processes, including the moral emotions, influence the behavior of people in groups. By specifying that both the prescription and receipt of discretionary interventions, treatments, and procedures reflect socioemotional processes, TRIBE underscores the critical need for psychological innovation in healthcare delivery and the education of health professionals. We hope that this contribution motivates creative, multifaceted improvements in research, education, clinical care, and policy to improve healthcare delivery, especially for individuals and families with advanced, life-limiting disease.

Online Appendix

Figure 1 suggests that activities in the clinical microsystem aimed at mitigating the harmful effects of DIALs will be hindered or facilitated by other entities in the social-ecological landscape of healthcare delivery. Some data support the idea that a more robust mesosystem could mitigate the harms of DIALs (Hanson et al., 2017; Seow et al., 2014), but there is a need for more research on this issue. The exosystem includes a mix of entities that either facilitate the harmful effects of DIALs (e.g., private equity; Braun et al., 2021) or both facilitate and mitigate the harmful effects of DIALs. University-owned healthcare systems and their affiliated academic health centers are prototypical “profitable nonprofits,” as some system components (e.g., palliative care) mitigate harms created by other components. Depicted in the macrosystem are psychocultural beliefs that facilitate harmful practices at the end-of-life. Our focus is the United States in the early 21^st century. Although some of these psychocultural beliefs will not be particularly relevant outside of that context, others (e.g., death denialism) are presumed to be universal.

• Beginning in the middle of the19th century in the United States, a culture of narcissism (Lasch, 1979) sprouted on the culture of individualism (Bellah,1985). Individualism itself is a relatively recent cultural invention; the word did not appear in the Oxford English Dictionary until 1835 (Bellah,1985). Narcissism has steadily worsened over the past few decades (Twenge et al., 2008). Psychocultural narcissism motivates clinicians to offer “heroic” interventions in the service of self-aggrandizement and could potentially motivate patients and families to seek interventions as well.

• A culture of optimism (Ehrenreich, 2010) is common in capitalist countries where risky gambles (e.g., venture capital, stock market “investments”) promise huge payoffs. Kahneman (2011) described optimism as the “engine of capitalism.” Psychocultural optimism contributes to poor prognostic understanding and motivates treatment risk-taking against the odds. Risk taking at the end-of-life is reinforced because previously incurable diseases occasionally become curable (Curti, 2018).

• A culture of death denialism (Becker, 1973) contributes to low levels of death literacy and death acceptance and high levels of death anxiety.

• A culture of religious fundamentalism (Hunter, 1991) sacralizes (Weber, 2002) healthcare debates and frames efforts to mitigate harmful DIALs as an abuse of power that undermines basic religious rights.

• A more is better culture contributes to the biased beliefs that spending more on healthcare leads to better outcomes, and that more intensive, invasive interventions are inherently better than less intensive interventions (Fagerlin et al., 2017; Verkerk et al., 2021).

• A newer is better culture motivates the biased belief that untested, unproven interventions will outperform established interventions (Carman et al., 2010).

• An action-oriented, problem-solving culture motivates the biased and misguided belief that complex, intractable problems can be solved (Rittel and Webber, 1975; Glouberman and Zimmerman, 2004).

• A conformity culture (Fromm, 1941/1969; 1970) or, less benignly, a stupidity culture (Alvesson and Spicer, 2016) makes it difficult to advocate for change in the healthcare workplace or actively work for systemic change in the exosystem.

• Adverse effects of psychocultural beliefs are amplified in groups that have historically been subjected to structural and legal discrimination (Kendi, 2017).

This Figure illustrates how systems interact across levels of analysis to produce adverse outcomes of end-of-life care. It is offered here for illustrative purposes and to stimulate research and educational innovations. For example, as health care professionals strive for cultural humility, the Figure suggests that humility requires not only self-reflection on one’s positionality and privilege but also an appreciation of the myriad influences at multiple levels of analysis both on the provision of clinical care and health outcomes.