You are here

Article Text

Article menu

Download PDFPDF

Extended essay
Practising what we preach: clinical ethicists’ professional perspectives and personal use of advance directives
  1. Jason Adam Wasserman1,
  2. Mark Christopher Navin2,
  3. Victoria Drzyzga3,
  4. Tyler S Gibb4
  1. 1 Foundational Medical Studies and Pediatrics, Oakland University William Beaumont School of Medicine, Rochester, MI, USA
  2. 2 Department of Philosophy, Oakland University, Rochester, MI, USA
  3. 3 Oakland University William Beaumont School of Medicine, Rochester, Michigan, USA
  4. 4 Program in Medical Ethics, Humanities and the Law, Western Michigan University Homer Stryker MD School of Medicine, Kalamazoo, Michigan, USA
  1. Correspondence to Dr Jason Adam Wasserman, Foundational Medical Studies and Pediatrics, Oakland University William Beaumont School of Medicine, Rochester, MI 48309, USA; wasserman{at}oakland.edu

Abstract

The field of clinical bioethics strongly advocates for the use of advance directives to promote patient autonomy, particularly at the end of life. This paper reports a study of clinical bioethicists’ perceptions of the professional consensus about advance directives, as well as their personal advance care planning practices. We find that clinical bioethicists are often sceptical about the value of advance directives, and their personal choices about advance directives often deviate from what clinical ethicists acknowledge to be their profession’s recommendations. Moreover, our respondents identified a pluralistic set of justifications for completing treatment directives and designating surrogates, even while the consensus view focuses on patient autonomy. Our results suggest important revisions to academic discussion and public-facing advocacy about advance care planning.

  • end-of-life
  • living wills/advance directives
  • clinical ethics

Data availability statement

Data are available upon request.

https://doi.org/10.1136/medethics-2020-106760

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    The profession of clinical ethics strongly advocates for advance care planning (ACP), that is, practices in which people make decisions about the medical care they would like to receive if they become unable to make decisions for themselves.1 ACP promotes a core value of clinical ethics—patient autonomy—so there is good reason for clinical ethicists to support it. But people do not always practice what they preach. Our findings suggest that clinical ethicists have more diverse and nuanced beliefs and practices surrounding ACP than might be expected given the nearly ubiquitous and univocal support that ACP receives from mainstream clinical ethics institutions and guidelines.

    Our study asked people working in clinical bioethics—in academic and/or clinical contexts—about their personal beliefs and practices about ACP. We solicited respondents’ perceptions of how the field of clinical bioethics views the importance of advance directives; whether they recommended ACP to others; whether they have a formally or informally designated surrogate decision maker for healthcare; whether they have, or intended to create, a treatment directive to articulate their end-of-life preferences; and their reasons for the ACP-related choices they have made. Our findings suggest that participants widely recognise a strong professional norm favouring ACP, but that their personal practices betray significant scepticism, particularly regarding the value of treatment directives and their usefulness for promoting patient autonomy.

    ACP has played a central role in the socio-historical development of bioethics and, especially, in efforts to promote the autonomy of patients. The familiar origin story of contemporary bioethics focuses on the rise of patients’ rights and the demise of medical paternalism, under pressure from larger social movements of the 1960s and 1970s.2 3 High-profile cases such as Karen Ann Quinlan and Dax Cowart shined a bright light on medical authority, while the push for advance directives from legal and public advocacy organisations collided with cases like Nancy Cruzan, culminating in the Patient Self-determination Act of 1990,4 and the need to protect the rights of patients and their families to refuse medical interventions. While both formally documented and informally stated preferences are now widely acknowledged as legitimate forms of advance directive, the latter are considered less stable, effective, and enforceable.5 As a result, completing formal documents is strongly encouraged.

    There is considerable variation in the language, structure and legal authority of ACP documents used across the USA and around the world.4 We can, however, distinguish broadly between two primary functions of ACP documents (ie, advance directives): surrogate designations and treatment directives. Surrogate designations name an agent who will possess durable power of attorney for healthcare, if the patient is deemed to lack decision-making capacity. While we will use the term ‘surrogate’ broadly here, in many jurisdictions it refers specifically to those who become decision makers by default (eg, next of kin), while the term ‘proxy’ is used for those specifically chosen to be third-party decision makers (eg, designated by a probate court). Treatment directives articulate specific kinds of treatments that a patient would or, more commonly, would not, want to receive under various presumptive states of debility. Examples of treatment directives include living wills and physician standing orders for treatment. The latter have different names and take diverse forms across different jurisdictions, but they include Provider Orders for Life-Sustaining Treatment (POLST) and Medical Orders for Life-Sustaining Treatment (MOLST).

    Importantly, surrogate designations and treatment directives are widely identified as being complementary.6 Treatment directives are often presented as the clearest and most direct expression of patient preferences, since they identify the detailed outcomes that patients would like to pursue or avoid. However, it is not possible for people to anticipate every possible medical decision that might have to be made about their care, and even the most expansive treatment directive documents are often underspecified with respect to the range of possible future situations. Therefore, surrogates can play a complementary role in ACP, by helping interpret a patient’s treatment directive and apply it in complex and dynamic clinical scenarios. Guidance documents often suggest that even if a patient is unwilling to complete a treatment directive, they should nonetheless discuss their preferences with their families.6 This common articulation of the relationship between the two kinds of advance directives supposes that treatment directives set a baseline, while surrogates draw on the values that treatment directives express to make judgements about possible interventions that treatment directives do not cover. Notably, studies show that surrogates struggle to accurately identify the patient’s preferences.7 8 The fallibility of surrogates may offer a further reason to think that treatment-preference documents are essential to ACP.

    Despite the political wins associated with patients’ rights—and the corresponding enshrinement of advance directives in the legal system—the public’s use of these documents has been modest: only one-third of US adults have completed an advance directive.9 10 Completion of advance directives is more common among those who are older, have higher levels of education and higher income, or who live with chronic disease.

    The socio-historical arc of advance directives—and their role in promoting patient autonomy—has generated a comparative consensus in the field of clinical ethics. Clinical ethics guidance documents routinely underscore the value of ACP for the promotion of patient autonomy.11 12 Consider the three goals that the Hastings Center Guidelines for Advance Care Planning and Advance Directives identify for ACP:

    1. To elicit a patient’s values and preferences concerning future treatment and record the patient’s treatment preferences.

    2. To use a patient’s values and preferences to establish goals of future care, often in light of a patient’s established diagnosis, prognosis and treatment options.

    3. To develop a future care plan that reflects a patient’s preferences and goals, and to modify the care plan so it remains consistent with changes in goals.6

    Each of these goals focuses clearly on the promotion of patients’ self-determination. Taken together, these goals suggest that the chief, if not sole, value of advance directives is that they articulate and record patient preferences to foster autonomous decision making by patients.

    Other important guidelines and textbooks similarly convey the centrality of autonomy in the process of ACP to various clinical ethics audiences ranging from professional clinical ethics consultants to physicians and other clinicians (and these groups overlap in notable ways).13 The American Medical Association (AMA) Code of Medical Ethics describes how:

    …respect for autonomy and fidelity to the patient are widely acknowledged as core values in the professional ethics of medicine. For patients who lack decision-making capacity, these values are fulfilled through third-party decision making and the use of advance directives (AMA, p77).14

    Bernard Lo’s provider-facing treatment of the subject offers a thorough discussion of the limitations of advance directives, but his ethics guidance for clinicians nonetheless comes to the conclusion that, ‘Physicians should tell patients about the advantages of written advance directives and encourage patients to complete them’ (Lo, p96).15

    In keeping with the accounts above, the American Society for Bioethics and Humanities (p36) suggests in their case-based study guide, that “[Advance directives] are grounded in the principle of respect for autonomy and are seen as enhancing autonomy by allowing a person to communicate in advance goals, values and preferences for future moments when they may lack decision-making capacity.”16 This guide also endorses the idea that surrogate decision makers are helpful complements for treatment-preference documents, but that treatment-preference documents are (or should be) the primary mechanism for ACP. It argues that treatment-preference documents should be expanded to include not only preferences about particular interventions, but to identify the patient’s broader values, goals and concerns. However, it notes that, even with the development of better treatment-preference documents, it is still necessary to have well-informed surrogates who can help interpret and adapt the treatment directive to the contours of the actual clinical scenario. Similarly, Jonsen et al (p83) call treatment-preference documents the “most important element of advance care planning,” and they argue for the development and proliferation of more elaborate treatment directives, for example, Five Wishes and POLST.17 ,18

    Despite what we might see as a relative consensus in clinical and public-facing guidance on ACP, the academic literature has featured critical engagement with the shortcomings of advance directives. Some have argued that efforts to promote ACP among the public have failed.18 19 Others have suggested that the concept of precedent autonomy is flawed or incoherent, particularly in the ways it prioritises the preferences of a rational former-self over the preferences of a decisionally incapacitated current-self.20–22

    Even in the academic literature, which sometimes criticises advance directives, there remains a strong presumption that ACP is motivated primarily, if not exclusively, by concern about protecting autonomy. Few people mention any other ethical justifications for ACP. There are, however, some notable exceptions. For example, treatment directives may facilitate conversation about death and dying, which can be positive in its own right.23 Rie and Engelhardt note that there are other goods entailed by advance directives—such as cost containment and the expression of altruism in the form of self-sacrifice by those who opt to limit care—though their discussion clearly centres autonomy as a primary value.24 Moreover, some discuss the value of ACP for decreasing burdens on family, though discussions of this value are often intermingled with discussions about protecting patient autonomy.25 Hackler et al (p4) briefly mention that advance directives can offer ‘physicians and families confidence in a questionably competent patient’s refusal of treatment.’26 Indeed, studies show that advance directives can offset psychological burdens for surrogates.27

    Although there are exceptions, the overall shape of the ACP literature, and the clear message of core guidance documents and public-facing positions, reflects three core themes: (1) advance directives are centrally important; (2) treatment directives are at the primary form of advance directives, supplemented by surrogate designations and (3) autonomy is the dominant (if not sole) value served by advance directives.

    Research often unveils disconnects between a profession’s public-facing consensus and the personal beliefs and behaviours of members of that profession. In particular, previous research has shown that physicians often make choices about their personal ACP that are inconsistent with the treatment recommendations they make for their patients, especially surrounding end-of-life care.28 This paper reports a similar study about the relationship between the beliefs and choices of clinical ethicists and the recommendations that they and their profession make about ACP.

    Methods

    Data were collected using Qualtrics online survey software to allow participants to respond anonymously. Participants were recruited via an email list that was constructed by the researchers from publicly available institutional directories that listed ethics faculty at medical schools, faculty in graduate biomedical ethics programmes and ethics committee members at health systems.

    The survey employed a branching structure to route participants through relevant questions based on their answers to previous questions. Survey items queried:

    Whether the participant

    1. Had a formally designated surrogate (yes or no; then why or why not).

    2. Had an informally designated surrogate (yes or no).

    3. Had a formal document specifying treatment preferences (yes or no; then why or why not).

      1. If no, then whether they intended to complete a treatment directive document (5-point Likert Scale ranging from ‘definitely will not’ to ‘definitely will’, with an ‘undecided’ middle option).

    Basic descriptive information including

    1. Age.

    2. Professional positions (allowing select all that apply, including academic faculty, clinical ethicist, physician, nurse, social workers, chaplain/pastoral care and a residual option with text entry). In the analyses, responses dichotomised into ‘non-clinical’ (ie, academic only) and ‘clinical,’ (at least one clinical position).

    Perceptions about the field and their deployment of ACP for others

    1. The participants’ perception of the general view of clinical bioethics toward the completion of a treatment-preference document (with a 4-point Likert Scale ranging from unimportant to very important).

    2. Whether they recommended formal designation of a surrogate to others (including patients, family, friends and students; evaluated on a 5-point Likert Scale ranging from ‘strongly discourage’ to ‘strongly encourage’ with a middle option of ‘neither’).

    3. Whether they recommended completion of treatment-preference documents to others (including patients, family, friends and students; evaluated on a 5-point Likert Scale ranging from ‘strongly discourage’ to ‘strongly encourage’ with a middle option of ‘neither’).

    Coding of narrative responses

    Responses to open-ended questions regarding reasons for having or not having designated a formal surrogate and/or completing a treatment directive were coded in a two-step process by identifying the basic justification and then combining like justifications. These resulted in a variety of kinds of justifications viable for analysis: ‘autonomy’ (related to realising one’s own preferences), ‘duties to others’ (related to avoiding conflict among family members or, in a few cases, among healthcare teams), ‘not necessary’ (relating to suggestions that one was too young or too healthy to warrant designation of a formal surrogate or completion of a treatment directive), ‘default okay’ (relating specifically to comfort with next-of-kin statutes as a justification for not formally designating a surrogate), ‘surrogacy is better’ (relating to justifications that suggested a preference for their surrogate as a reason for not having a treatment directive) and ‘not beneficial’ (relating to justifications which offered explicit critique of treatment directives as too inflexible). Other justifications were not offered frequently enough to be analytically viable, including references to professional responsibilities either in the clinic or the classroom, not having enough knowledge about how to fill out treatment directives or that it was too hard to do so, and the notion that the strictures of their religious community superseded their own preferences.

    Analyses

    Analyses consisted of basic descriptive, as well as parametric difference of means and difference of proportions tests to identify or rule out underlying group differences in these descriptive data (eg, differences by age or by professional position type). A multivariate logistic regression was employed to assess potential differences in likelihood of having a treatment directive by professional position type or age while controlling for the other. Representative quotations are provided for some analyses, to characterise the kinds of reasons participants offered for having or not having completed elements of ACP.

    Findings

    Emails were sent to 1093 potential participants, and 46 bounced back as undeliverable. Of the remaining 1047, 17 individuals initiated but did not complete the survey and 2 declined to participate, leaving a total analytical sample of 224 and a response rate of 21.39%, which is comparable to other studies using email recruitment among professional populations.29 The mean age of the sample was 55.19 years (SD=13.47), with a range of 28–84 years. In terms of professional position, respondents were allowed to select all categories that they felt applied to them (see table 1). Among the ‘other’ category, only 10 respondents selected this and no other category. Of those 10, nearly all described themselves as bioethics researchers or as retired academic bioethicists. Categorisation of professional position resulted in the following subsamples: academic only (n=111, 49.55%), clinical only (n=13, 5.80%), and academic and clinical (n=84, 37.50%). In keeping with the idea that clinical exposure might affect ACP, the latter two categories were combined to compare those with actual clinical involvement to those without. This resulted in professional categories of clinical versus non-clinical (academic only).

    View this table:
    Table 1

    Self-reported professional position

    As we turn to the descriptive data, it is useful to remember that many analyses report only on subsamples of the total set of respondents because of ways in which respondents may have answered questions (eg, reasons for not having a formally designated are reported only for the subgroup that reported that they did not have a designated surrogate). Numbers also vary slightly if questions were skipped by respondents who still submitted answers to other questions that were able to be retained in other analyses. Finally, the percentages reported for the reasons given for different ACP decisions also is modified by the fact that some respondents did not provide any written reasons for their selection. These adjustments are identified for each analysis.

    Clinical bioethicists’ use of advance directives

    Participants were nearly uniform in reporting that the field of clinical bioethics supports treatment directives. Specifically, 216 of 222 (97.3%) believed that the professional view of clinical bioethics is that treatment directives are important to some degree, with over half (120, 55.6%) suggesting that the profession considers these documents to be ‘very important.’ High numbers of respondents reported encouraging or strongly encouraging others (patients, families or friends) to specify a formal surrogate (201 of 223, 90.1%). However, fewer respondents report encouraging or strongly encouraging others to complete a treatment directive (160 of 222, 72.1%).

    In terms of their personal practices and reasons for their choices, additional discrepancies emerge between consensus guidance and respondents’ beliefs and practices. In terms of having designated a surrogate for themselves, 168 of 224 (75.0%) respondents reported having a formally designated surrogate. Of the 56 who did not have a formally designated surrogate, 45 (80.36%) indicated that they have informally designated a surrogate. In total, then, 213 of 224 (95.1%) have formally or informally designated a surrogate. Older participants were more likely to have a formally designated surrogate (F=14.802; p<0.001). However, whether or not someone was in a non-clinical versus clinical position had no effect on their likelihood of having formally designated a surrogate, even while controlling for the effect of age (OR=0.861; p=0.640).

    Of the 168 respondents who had a formal surrogate, 140 gave reasons. Of those 140 respondents, 73 (52.1%) suggested the appointment of a surrogate was a way to exercise their own autonomy. For example, one participant wrote, “I know the importance of advance care planning so as to avoid unwanted treatment.” Having a formal surrogate because of a duty to others was cited as a reason by 66 (47.1%) respondents. For example, one respondent noted, “I want my loved ones to avoid conflict and headaches after I die.” There were no statistically significant differences in reasons given for having a formal surrogate by job role or by age.

    Among the 58 respondents who said they did not have a formally designated surrogate, 46 articulated a reason. Of these 46, 12 (26.1%) suggested there was questionable benefit to doing so, often because they reported being young and/or healthy. One noted, for example, “I am still relatively young and [it] has not been as pertinent to me at this stage.” Another 14 (30.4%) respondents, who reported not having formally designated a surrogate, suggested that they were comfortable with their legally designated next of kin. One wrote, “I am content with the state mandated order of surrogate decision making: my spouse and then my adult children.” However, 24 (52.2%) noted that they simply had not done so yet, suggesting they intended to do so at some point. There were no statistically significant differences in reasons given for not having a formal surrogate by job role or differences by age (though given the preponderance of surrogate designation, in some cases the subsamples were simply too small to analyse).

    In contrast to the near ubiquity of having formally or informally designated a surrogate, only 110 of 223 (49.3%) respondents indicated that they had a treatment directive. Of the 113 respondents that did not have treatment directive, only 43 (38.1%) responded that they probably or definitely intended to complete such a document, with 35 (31.0%) being undecided and 45 (39.82%) saying they probably or definitely would not. In total, then, 153 of 223 (68.6%) respondents either have or intend to complete a treatment directive. There were no statistically significant differences in having a treatment directive by age (F=0.844; p=0.359), however, those in non-clinical professional positions were nearly two times as likely to have a treatment directive when controlling for age (OR=1.92; p=0.027).

    Of the 110 (49.3%) of 223 respondents who had a treatment directive, 89 gave reasons for doing so. Of those, 58 (65.2%) indicated that the document expressed their autonomous preferences in important ways. One wrote, for example, “I have strong convictions pertaining to various possible future situations, about not using medical treatment or oral feeding to extend my life.” Another 33 (37.1%) respondents listed duty to others as a reason for having a treatment directive. One put it, “I believe based on experience with ethics consults, that specifying wishes makes it easier on families to decide what to do and alleviates guilt when life sustaining treatments are withdrawn.” There were no statistically significant differences in reasons given for having a treatment directive by job role or by age.

    Of the 113 (50.7%) respondents who did not have a formal document clarifying their treatment preferences, 93 articulated a reason and 74 (79.6%) of those suggested that a surrogate designation is a better form of advance directive in light of the limitations of treatment documents to address complex real-world situations. As one respondent put it, “Because a well-informed surrogate will be able to make decisions regarding my wishes. Living wills like Five Wishes will never be able to get it right in all future circumstances.” There were no statistically significant differences in reasons given for not having a treatment directive by job role or differences in age.

    Discussion

    Respondents widely recognised that treatment directives are seen as important within the field of bioethics. However, clinical ethicists’ behaviour, in terms of recommending these ACP options to others, reflects a clear priority placed on designating a surrogate relative to completing a treatment directive. This divide becomes even more pronounced in terms of respondents’ personal ACP choices, where nearly all respondents had a formally or informally designated surrogate (95%), while only 65% of respondents had or said they intended to complete a treatment directive, with just under half of respondents having actually completed such a document (49.33%). While the prevalence of having a treatment directive among our sample still exceeds the prevalence in the general population, it is strikingly low for members of a profession whose public-facing statements about ACP are so positive about treatment directives.

    Notably, there were relatively few significant differences in having designated a surrogate or completed a treatment directive by age or professional position. While one might suppose older individuals would be more likely to have engaged in all aspects of ACP, this was true only for having designated a formal surrogate. Older respondents were no more likely to have completed a treatment directive than younger ones. Additionally, one might suppose that respondents with professional clinical exposure might be more prone to engage in ACP, but no differences in terms of formally designating a surrogate were detected. In contrast, those in strictly academic positions were nearly two times as likely to have a treatment directive, tentatively suggesting that clinical exposure may serve to emphasise the limitations of treatment directives.

    How might we explain the apparent mismatch between the personal choices of bioethicists and the acknowledged recommendations of their profession when it comes to choices about treatment directives? One explanation is that professional bioethicists may have seen the many ways that treatment directives can fail to be well implemented (or even acknowledged) in clinical contexts, and their practical knowledge of ACP may have led them to rely more on surrogate designations, seeing them as more effective in practice. If this is the case, then one wonders why public-facing ACP advocacy still places so much emphasis on treatment directives.

    Another reason for the comparative dearth of having completed treatment directives, in contrast to having designated a surrogate and recommending treatment directives to others, may be scepticism about ACP as a means for promoting autonomous judgement. Consider that the reasons that our participants offered for engaging in either aspect of ACP often did not hinge on the kinds of autonomy-based justifications that are dominant in the clinical ethics guidance literature. In fact, in terms of designating a formal surrogate, only slightly more than half (52.1%) cited having their autonomy respected as their reason for that choice, while nearly as many (47.1%) cited duties to others (ie, making things easier on their families). Similarly, of those who had completed a treatment directive, which was just under half of the sample (49.3%), only 65.2% of those cited protecting their autonomy as the reason for doing so. This means that only 32.14% of respondents had a treatment directive and cited autonomy as the reason for doing so. This stands in stark contrast to the academic literature which suggests, with few exceptions, that treatment directives are centrally important and that autonomy is the dominant reason for preparing such documents.

    Our results tell against justifications for treatment directives that are rooted exclusively in patient autonomy. Many people who do not have treatment directives identify autonomy as their reason for choosing not to have those documents. They claim that their surrogates’ exercise of substituted judgement will better reflect their own preferences than will the preferences they choose on treatment directives. Additionally, some people who do have treatment directives do not identify autonomy as their (only) reason for choosing such documents. Instead, other regarding considerations, such as reducing conflict and psychological burdens on family, were cited nearly as often among those who had completed treatment directives.

    Perhaps pragmatic or even political imperatives explain the predominance of autonomy justifications for ACP in clinical ethics guidance and public-facing advocacy work. Even if many individuals want to do ACP for other-regarding reasons (such as reducing burdens on family members), it may be unseemly, or even politically contentious, for bioethics and healthcare institutions to suggest that other-regarding considerations are good reasons for ACP. We suspect that one reason public advocacy of ACP focuses so narrowly on the value of autonomy—often ignoring other-regarding considerations—is to avoid accusations that ACP is a ‘death panel’ programme that aims at releasing families from the burdens of caring for infirm relatives.

    Among clinical bioethicists who did not have a treatment directive, the overwhelming majority reasoned that their surrogate would be more effective and cited the limited applicability of treatment directives to complex clinical scenarios. This response echoes some of the criticism of treatment directives that have been in the academic literature. While critiques of treatment directives appear in a relatively small corner of the literature, they clearly reflect the beliefs and inform the choices of many clinical bioethicists. This finding stands in stark contrast to the consensus view of the field about the value of those types of directives and their important role in promoting patient autonomy.

    Limitations

    There are several limitations to this study. There is risk of sampling bias where those with advance directives may have been more likely to opt to take a survey titled ‘Advance Directive Utilisation and Recommendations’ than those without them. Importantly, however, this would predictably inflate the number of clinical ethicists with advance directives and therefore underscores the surprising variability we found on this account. Additionally, the self-described reasons articulated by participants for having or not having designated surrogates or treatment directives were coded by the researchers. This sometimes required us to parse multiple justifications contained within single statements. For example, someone might have suggested they have a treatment directive because of “an obligation to my family to let them know my preferences.” In this case, the participant would have been coded as articulating both an autonomy justification and a duty to others justification. While these are clearly complementary, they nonetheless are oriented differently and were coded as such. This strikes us as appropriate given the potential independence of such justifications (where someone might suggest the document will alleviate stress on their families without having any hope that it will sufficiently protect their own preferences). Additionally, our subgroups, particularly with respect to professional position, may have less separation than our analysis would suggest. For example, someone in a non-clinical professional position may nonetheless do remote clinical consultation, but have not described themselves as such.

    Conclusions

    Several important considerations emerge from our findings. First, bioethicists, in both academic and clinical positions, often take seriously the limitations of ACP, and especially the limits of treatment directives, when they make decisions about their own care. Importantly, surrogates appear to be more important than treatment directives for many respondents. Moreover, autonomy is not the primary motivating factor for ACP for many clinical ethicists, considering that less than one-third of our respondents reported having a treatment directive and also cited an autonomy-related justification as the reason for that choice. This finding contrasts with the impression one might get from the guidance and public advocacy literature about ACP, which suggests that completing a treatment directive is primary way to ensure one’s preferences are respected. In contrast, among clinical bioethicists, easing the burden on loved ones was nearly as common a reason for having completed a treatment directive.

    In light of these findings, several recommendations emerge. First, given the divergence in completion rate and justification for having a designated surrogate and a treatment directive, use of the global term ‘advance directive’ should be reconsidered. The function and value of surrogates, relative to treatment directives, and the reasons clinical ethicists provide for having or not having these documents, suggest these two kinds of ACP are distinct constructs, even if they are both part of planning for the end of life. A decomposition of the broader construct of ‘advance directives’ into its two component parts may be warranted, both in theory and practice. This also affirms an existing practice where those resistant to completing a treatment directive are nonetheless advised to select surrogates and converse with them about their values and preferences regarding end of life (see for example, guidance in Berlinger et al 6).

    Additionally, it may be wise for the field to identify additional reasons, beyond precedent autonomy, for completing ACP. Our data suggest that a common motivating factor for ACP is to decrease burdens on one’s family members. Pragmatic or political considerations may continue to limit the public advocacy of these other regarding reasons for ACP, but advocates of ACP should work to identify ways to overcome these limits.

    Placing even greater emphasis on the value of a strong surrogate may also represent a positive shift in discussions about ACP.18 Further emphasising the primacy of surrogate designation, and a more pluralistic range of justifications for treatment directives in both public-facing and provider-facing documents and guidance, would align with what appear to be the real perspectives of clinical bioethicists.

    Our results may also support efforts to improve treatment directives, such as the use of POLST documents and other more extensive forms of ACP.30 Data suggest, however, that the complexity of end-of-life scenarios has continued to outstrip these efforts.8 Dwight D Eisenhower famously said, “Plans are worthless, but planning is everything.” This appears to ring true for bioethicists as they think about their own end-of-life planning.31

    Data availability statement

    Data are available upon request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This research was approved by the Oakland University Institutional Review Board.

    References

    1. ↵
      2. Pecanac KE ,
      3. Repenshek MF ,
      4. Tennenbaum D , et al
      . Respecting Choices® and advance directives in a diverse community. J Palliat Med 2014;17(3):282–7.doi:10.1089/jpm.2013.0047 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24325558
    2. ↵
      2. Khuse H ,
      3. Singer P
      . What is bioethics: A historical introduction. In: Singer P , Kuhse H , eds. A companion to bioethics. Oxford: Wiley-Blackwell, 2009: 1–10.
    3. ↵
      2. Garrett JR ,
      3. Jotterand F ,
      4. Ralston C
      . The development of bioethics in the United States. Dordrecht: Springer, 2013.
    4. ↵
      2. Simon A
      . Historical review of advance directives. In: Lack P , Biller-Andorno N , Brauer S , eds. Advance directives. Dordrecht: Springer, 2014: 3–16.
    5. ↵
      2. Carr D ,
      3. Luth EA
      . Advance care planning: contemporary issues and future directions. Innov Aging 2017;1:igx012. doi:10.1093/geroni/igx012 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30480109
    6. ↵
      2. Berlinger N ,
      3. Jennings B ,
      4. Wolf SM
      . The Hastings center guidelines for decisions on life-sustaining treatment and care near the end of life. 2nd ed. New York: Oxford University Press, 2013.
    7. ↵
      2. Chen EE ,
      3. Pu CT ,
      4. Bernacki RE , et al
      . Surrogate preferences on the physician orders for life-sustaining treatment form. Gerontologist 2019;59(5):811–82.doi:10.1093/geront/gny042 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29788197
    8. ↵
      2. Shalowitz DI ,
      3. Garrett-Mayer E ,
      4. Wendler D
      . The accuracy of surrogate decision makers: a systematic review. Arch Intern Med 2006;166(5):493–7.doi:10.1001/archinte.166.5.493 pmid:http://www.ncbi.nlm.nih.gov/pubmed/16534034
    9. ↵
      2. Rao JK ,
      3. Anderson LA ,
      4. Lin F-C , et al
      . Completion of advance directives among U.S. consumers. Am J Prev Med 2014;46(1):65–70.doi:10.1016/j.amepre.2013.09.008 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24355673
    10. ↵
      2. Yadav KN ,
      3. Gabler NB ,
      4. Cooney E , et al
      . Approximately one in three us adults completes any type of advance directive for end-of-life care. Health Aff 2017;36(7):1244–51.doi:10.1377/hlthaff.2017.0175 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28679811
    11. ↵
      2. Silveira MJ ,
      3. Wiitala W ,
      4. Piette J
      . Advance Directive completion by elderly Americans: a decade of change. J Am Geriatr Soc 2014;62(4):706–10.doi:10.1111/jgs.12736
    12. ↵
      2. Spranzi M ,
      3. Fournier V
      . The near-failure of advance directives: why they should not be abandoned altogether, but their role radically reconsidered. Med Health Care Philos 2016;19(4):563–8.doi:10.1007/s11019-016-9704-7
    13. ↵
      2. Hauschildt K ,
      3. De Vries R
      . Reinforcing medical authority: clinical ethics consultation and the resolution of conflicts in treatment decisions. Sociol Health Illn 2020;42(2):307–26.doi:10.1111/1467-9566.13003 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31565808
    14. ↵
      1. American Medical Association
      . Code of medical ethics of the American Medical association. American Medical Association, 2016: 77.
    15. ↵
      2. Lo B
      . Resolving ethical dilemmas: a guide for clinicians, 5th Ed. Philadelphia: Wolters Kluwer, 2013.
    16. ↵
      1. American Society for Bioethics and Humanities
      . Addressing patient-centered ethical issues in health care: a case-based study guide, 2017.
    17. ↵
      2. Jonsen AR ,
      3. Siegler M ,
      4. Winslade WJ
      . Clinical ethics: a practical approach to ethical decision-making in clinical medicine. 2nd edn. New York: McGraw Hill, 2010.
    18. ↵
      2. Fagerlin A ,
      3. Schneider CE
      . Enough. The failure of the living will. Hastings Cent Rep 2004;34(2):30–42.doi:10.2307/3527683 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15156835
    19. ↵
      2. Jennings B ,
      3. Burt RA
      . Preface: improving end of life care: why has it been so difficult? Hastings Cent Rep 2005;Spec No:S9–13.pmid:http://www.ncbi.nlm.nih.gov/pubmed/16541535
    20. ↵
      2. Davis JK
      . The concept of precedent autonomy. Bioethics 2002;16(2):114–33.doi:10.1111/1467-8519.00274 pmid:http://www.ncbi.nlm.nih.gov/pubmed/12083154
    21. ↵
      2. Stahl D ,
      3. Banja J
      . The persisting problem of precedent autonomy among persons in a minimally conscious state: the limitations of philosophical analysis and clinical assessment. AJOB Neurosci 2018;9(2):120–7.doi:10.1080/21507740.2018.1459932
    22. ↵
      2. Wasserman JA ,
      3. Navin MC
      . Capacity for preferences: respecting patients with compromised decision-making. Hastings Cent Rep 2018;48(3):31–9.doi:10.1002/hast.853 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29806903
    23. ↵
      2. Emanuel L
      . Living wills can help doctors and patients talk about dying. West J Med 2000;173(6):368–9.doi:10.1136/ewjm.173.6.368 pmid:http://www.ncbi.nlm.nih.gov/pubmed/11112736
    24. ↵
      2. Rie MA ,
      3. Engelhardt T
      . The financial enforcement of living wills: Putting teeth into natural death statutes. In: Hackler C , Moseley R , Vawter DE , eds. Advance directives in medicine. New York: Praeger, 1989: 85–92.
    25. ↵
      2. Green MJ ,
      3. Schubart JR ,
      4. Whitehead MM , et al
      . Advance care planning does not adversely affect hope or anxiety among patients with advanced cancer. J Pain Symptom Manage 2015;49(6):1088–96.doi:10.1016/j.jpainsymman.2014.11.293 pmid:http://www.ncbi.nlm.nih.gov/pubmed/25542552
    26. ↵
      2. Hackler C ,
      3. Moseley R ,
      4. Vawter DE
      . Introduction: Advance directives – An overview. In: Hackler C , Moseley R , Vawter DE , eds. Advance directives in medicine. New York: Praeger, 1989: 1–8.
    27. ↵
      2. Hickman RL ,
      3. Pinto MD
      . Advance directives lessen the decisional burden of surrogate decision-making for the chronically critically ill. J Clin Nurs 2014;23(5-6):756–65.doi:10.1111/jocn.12427 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24330417
    28. ↵
      2. Periyakoil VS ,
      3. Neri E ,
      4. Fong A , et al
      . Do unto others: doctors' personal end-of-life resuscitation preferences and their attitudes toward advance directives. PLoS One 2014;9(5):e98246–9.doi:10.1371/journal.pone.0098246 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24869673
    29. ↵
      2. Shih T-H ,
      3. Fan X
      . Comparing response rates in e-mail and paper surveys: a meta-analysis. Educational Research Review 2009;4(1):26–40.doi:10.1016/j.edurev.2008.01.003
    30. ↵
      2. Hickman SE ,
      3. Hammes BJ ,
      4. Moss AH , et al
      . Hope for the future: achieving the original intent of advance directives. Hastings Cent Rep 2005;Spec No:S26–30.doi:10.1353/hcr.2005.0093 pmid:http://www.ncbi.nlm.nih.gov/pubmed/16468252
    31. ↵
      2. Eisenhower DD
      . Public papers of the presidents of the United States, containing the public messages, speeches, and statements of the President, remarks at the National defense executive reserve Conference, date: November 14. Washington D.C: Federal Register Division, National Archives and Records Service, General Services Administration, 1957: 818.

    Footnotes

    • Contributors JAW conceptualised the study, conducted the analysis, drafted the initial manuscript, and reviewed and approved the final draft. MN conceptualised the study, conducted the analysis, revised the manuscript, and reviewed and approved the final draft. VD conducted portions of the analysis, revised the manuscript, and reviewed and approved the final draft. TSG conceptualised the study, revised the manuscript, and reviewed and approved the final draft.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.