Article Text

Download PDFPDF

We don’t need unilateral DNRs: taking informed non-dissent one step further
  1. Diego Real de Asúa1,2,
  2. Katarina Lee3,4,
  3. Peter Koch5,
  4. Inmaculada de Melo-Martín6,
  5. Trevor Bibler7,8
  1. 1 Department of Internal Medicine, Hospital Universitario de la Princesa, Madrid, Spain
  2. 2 Division of Medical Ethics, Cornell University Joan and Sanford I Weill Medical College, New York, New York, USA
  3. 3 Department of Family Medicine, University of Manitoba, Winnipeg, Manitoba, Canada
  4. 4 Health Care Ethics Service, St. Boniface Hospital, Winnipeg, Manitoba, Canada
  5. 5 Department of Philosophy, Villanova University, Villanova, Pennsylvania, USA
  6. 6 Department of Public Health, Weill Cornell Medical College, New York, New York, USA
  7. 7 Center for Medical Ethics and Health Policy, Baylor College of Medicine, Houston, Texas, USA
  8. 8 Ethics Consultation Service, Houston Methodist Hospital, Houston, Texas, USA
  1. Correspondence to Dr Diego Real de Asúa, Department of Internal Medicine, Hospital Universitario de la Princesa, Diego de León 62, 28006 Madrid, Spain ; diego.realdeasua{at}


Although shared decision-making is a standard in medical care, unilateral decisions through process-based conflict resolution policies have been defended in certain cases. In patients who do not stand to receive proportional clinical benefits, the harms involved in interventions such as cardiopulmonary resuscitation seem to run contrary to the principle of non-maleficence, and provision of such interventions may cause clinicians significant moral distress. However, because the application of these policies involves taking choices out of the domain of shared decision-making, they face important ethical and legal problems, including a recent challenge to their constitutionality. In light of these concerns, we suggest a re-conceptualization of informed non-dissent as an alternative approach in cases where the application of process-based policies is being considered. This clinician-directed communication model still preserves what is valuable in such policies and salvages professional integrity, while minimising ethical and legal challenges.

  • clinical ethics
  • end-of-life
  • ethics committees/consultation
View Full Text

Statistics from


End-of-life decisions are among the most challenging ones patients, surrogates, and clinicians face and discussions surrounding them abound.1 2 Disputes between the clinical team and the patient’s family account for 26%–57% of the conflicts in the intensive care setting.3 4 Many of them involve communication breakdowns and disagreements over the provision of life-sustaining treatments or cardiopulmonary resuscitation (CPR).2 5 6 As a last resort measure to address entrenched end-of-life conflicts, some healthcare organisations and professional societies have advocated for unilateral decision-making through process-based conflict resolution policies (unilateral decision-making policies).5 7 8 Such policies are currently legal in most US states and rest primarily on the judgement that the potential harms of a given intervention, such as CPR, are likely to outweigh the possible benefits to the patient.9

There are good reasons to defend the ethical soundness of unilateral decision-making policies.5 10 First, arguably, the provision of CPR can violate clinicians’ professional integrity when there is no reasonable expectation that patients’ neurological function will improve enough to allow them to perceive the benefits of treatment or that they will improve sufficiently to survive outside the acute care setting.11 Second, in patients who do not stand to receive proportional clinical benefits, the physical harms involved in CPR seem to run contrary to the principle of non-maleficence.12 Third, in contexts where the possible harms of providing CPR outweigh its potential benefits, performing CPR can cause significant moral distress to the healthcare team.13 Fourth, because healthcare resources are scarce, their use creates opportunity costs and physicians also have the duty to steward expensive resources, such as intensive care unit (ICU) beds or advanced life support therapies.

However, these conflict-resolution policies are less neutral than they appear. Despite their process-based approach, these strategies are nonetheless ‘unilateral’ as the decision is implemented regardless of whether and how much surrogates disagree. Because the application of these policies involves a move away from shared decision-making, they face important ethical and legal problems, including a recent challenge to their constitutionality.14

In this paper, we defend informed non-dissent as an alternative approach that can preserve what is valuable in unilateral decision-making policies while avoiding some of their most pressing shortcomings. This strategy was introduced over a decade ago with the purpose of unburdening surrogates of the guilt of weighty end-of-life decisions.15 However, its usefulness remains relatively unexplored. Arguments for the use of informed non-dissent are applicable to a broad array of clinical scenarios. Nonetheless, given that the most recent challenges in the US legal arena surrounding unilateral decision-making policies have been in the context of do-not-resuscitate (DNR) orders, we will focus our attention on conversations over resuscitation status. We begin by reviewing the most relevant ethical and legal challenges unilateral decision-making policies confront. We then provide an ethical justification for the use of informed non-dissent as an alternative.

Unilateral decision-making policies: some ethical and legal concerns

Unilateral decision-making policies are grounded on an evaluation of the potential harms and benefits of a specific intervention, such as CPR, to a particular patient.5 8 Unfortunately, a coherent, established understanding of harms and benefits shared by relevant stakeholders has proven elusive.16 Evidence shows that individuals’ socio–economic, cultural, ethnic and religious backgrounds inform their particular accounts of welfare and harm.17 18 These variations pose significant challenges to clinicians’ determinations regarding which harms matter to patients and their surrogates, how much they matter, and what potential benefits would offset what possible harms. Given their multifaceted nature, these normative evaluations clearly fall outside the realm of clinicians’ expertise.19 Moreover, they are particularly contentious in the inpatient context, where healthcare professionals need to make rapid decisions, often based on incomplete information, while establishing trust in a short-term relationship with patients about whom they know little.20 Thus, patients’ surrogates are arguably better positioned than healthcare providers to determine the patient’s preferences and values despite the biases and charged emotions that are known to influence their decisions.21 22

These policies, however, set up a conflict-focused relationship between the clinical care team and the patient’s surrogates. In shifting the locus of decision-making from the physician–patient–surrogate triad to an institutional committee- or a judicial system- unilateral decision-making policies can minimise the normative weight of the surrogates’ voices. Indeed, this occurs regardless of how much surrogates are invited to participate in the process.5 As a recent report on the effects of one such policy shows, in almost half of the instances where an ethics consultation was requested for conflicts over resuscitation status, a DNR order was written despite the patient’s or surrogates’ refusal.7 Furthermore, an account of the experience after the implementation of Chapter 166 of the Texas Health and Safety Code or Texas Advance Directive Act (TADA), observed that ethics committees agreed with the physicians’ determination of futility in 90% of the cases presented to them over 2 years, despite the fact that one in every six families rejected the committees’ decisions.23 Equally troublesome, patients affected by these unilateral decisions, and whose surrogates persist in requesting CPR, were disproportionately non-white, foreign-born or non-English speaking, raising the question of whether these policies might have the unintended outcome of further marginalising already vulnerable populations.24,25

Unilateral decision-making policies are not only problematic on ethical grounds, but possibly on legal grounds as well. Although a number of states permit institutional process measures that allow unilateral decision-making with respect to requests for so-called ‘medically inappropriate’ treatments,26–28 their constitutional validity has recently been called into question in Texas. The lawsuit involves a case where a medical decision to withdraw life-sustaining treatments was upheld by the institutional ethics committee through a dispute resolution process, consistent with TADA. The plaintiffs sought, however, a declaratory judgement that this legal mechanism violated procedural and substantive due process protections under the Texan and US constitutions.29 The plaintiffs argued that the process did not provide sufficient notice for surrogates to participate in any meeting and lacked standards by which to evaluate the physician’s or the committee’s decisions, denying patients an impartial arbiter to decide their case.

Although its final ruling is still pending, the case exemplifies some of the major legal challenges unilateral decision-making policies face. Institutional imbalances of power shape these policies in practice, preventing the meaningful participation of those who love the patients and are likely to know them best.30 Is it then still possible to give back clinicians a more active role in discussions about resuscitation status while respecting incapacitated patients’ preferences as voiced through their surrogates?

Revisiting informed non-dissent

Although these ethical and legal concerns make unilateral decision-making policies difficult to defend, this does not mean that physicians must limit themselves to presenting options about performing or withholding CPR as if all options were equally appropriate in all cases.10 Indeed, a clinician-directed communication strategy can be appropriate and might even be the surrogates’ preferred communication strategy in end-of-life situations. Surrogates who struggle with end-of-life choices on behalf of incapacitated patients have shared that they would welcome guidance from clinicians, and that only a minority of them would prefer to make these decisions completely on their own.31,32 Similarly, a recent a study evaluating how parents made difficult decisions in the paediatric intensive care setting identified physicians and nurses as parents’ main sources of support and guidance.33

One such model of clinician-directed communication is informed non-dissent. In this communication strategy, physicians state that an intervention, such as placing a DNR order in a patient’s chart, will happen, while also permitting patients or surrogates to object.15

Informed non-dissent must meet informed consent requirements if it is to be ethically sound. It is the clinicians’ duty to ensure that a patient’s or surrogate’s desire to cede decision-making authority is not based on a lack of information, support or understanding.34 Thus, physicians must attempt to understand their patients’ and surrogates’ values before discussing the patient’s diagnosis and prognosis. Clinicians should convey their specific recommendation to the surrogates (eg, that a DNR order will be placed in the chart), as well as the rationale for such recommendation, including discussion of risks, potential benefits and alternatives. Clinicians should make special efforts to clarify what will and will not happen to the patient after the decision is implemented, that is, how the decision about CPR might influence the provision of other life-sustaining interventions (such as intubation) but is independent of any other treatments or symptom management. As with any consent process to an intervention, surrogates need to be aware of their right to object and that their silence will be interpreted as tacit agreement. Although implementing informed non-dissent in daily conversations in the wards may not be an easy task, there are a number of concrete practices that physicians may find helpful when putting this strategy into practice (table 1).

Table 1

Sample language to use informed non-dissent in clinical practice

One reason for the relative lack of attention to what appears an ethically sound alternative to both business as usual and unilateral decision-making is that the original proposals advocating for the use of clinician-directed communication models came with a caveat. Some of their proponents, such as Curtis and Burt, argued against the use of these strategies when ‘clinicians are uncertain about the possibility of success or when the clinicians’ convictions about withholding or withdrawing treatment are based on their value judgments about the patient’s resulting quality of life’.35 Uncertainties and value judgements, however, pervade every clinical decision, in particular end-of-life ones. Thus, the caveat severely limited the value of directive communication strategies, rendering the model inapplicable in most end-of-life settings.

This caveat need not be accepted. Clinicians should use informed non-dissent despite the presence of uncertainties and biases around the potential success of an intervention. The uncertainties around CPR involve considerations about what success means in the mid- and long-term, which are determinations inevitably influenced by personal context and, often unconscious, biases. But this is no different than most other clinical decisions. Implicit biases and clinicians’ personal value judgements influence clinical decision-making, including the decision to pursue a conflict-resolution process. We acknowledge that clinicians’ social, economic, cultural and religious backgrounds can significantly impact management strategies and treatment choices.36,37 Clinical practice guidelines and consensus documents appropriately emphasise the relevance of fostering awareness of those biases, and recommend remaining ‘open to the patient’s and surrogates’ values, goals and preferences’.34 Our conceptualization of informed non-dissent does not negate this concern. Rather it encourages physicians to make clear recommendations in very delicate situations while recognising these biases. Oftentimes, it is the clinicians’ fear of those same biases, and of their potential to contaminate, rather than inform, momentous decisions what precludes many physicians from speaking up. A change in initial decision-maker from patient representative to physician can lead to an early clarification of why a certain course of action is preferable under the proposed circumstances, helping both clinicians and families ponder with appropriate weight all treatment options (CPR vs DNR), and thus decreasing conflict downstream.

To further enhance the usefulness of informed non-dissent, this model should not be simply a physician’s empathic response to struggling surrogates, who would rather defer their responsibilities to the treating team.15 Instead, it should be more broadly offered in all those cases where clinicians judge that patients will die imminently or that they have no chance of surviving CPR to the point of leaving the hospital.11 To be sure, physicians should base such decisions not only on the information derived from the clinical context but also on what they know about the patient’s values, and because knowledge of patients’ values is often inadequate, clinicians need to ensure that surrogates have opportunities to disagree.

The advantages of this version of non-dissent over business as usual and unilateral decision-making are clear. It protects professional integrity while still allowing bidirectional communication. Professionalism is exercised by respecting the patients’ and surrogates’ autonomy. But it also involves an appropriate assertion of expertise and professional authority.38 Informed non-dissent clearly promotes the latter. It also fosters the former by reframing and maintaining the conversation within the clinical relationship. This model preserves a modicum of bidirectional communication by presenting surrogates the option of objecting to the DNR order and accepting the consequences of CPR. Surrogates who clearly object to the clinicians’ proposal will be able to provide arguments in favour of CPR, and clinicians must commit to hearing their objections. Surrogates who want to be unburdened from the decision also have an opportunity to express their autonomy by not dissenting to the recommendation. Contrary to unilateral decision-making, informed non-dissent preserves the autonomous expression of preferences of all stakeholders involved in these charged situations.34

We recognise, however, that our proposal faces several limitations. Changing the initial locus of decision-making from the patients and their families to the physician may benefit some surrogate populations, namely: those overburdened by the decision and those who, wanting to let go, are unwilling or unable to actively request a DNR order. It could, however, have a negative impact on others. Surrogates who have a preference for CPR would have needed no justification for their position in a context where CPR is the default option. Informed non-dissent would now impose the burden of justifying their preferences and rejecting the physician’s recommendation. In a context where minority populations happen to be more frequently involved in disputes over resuscitation status,7,24 the imposition of this burden raises justice concerns. Minority populations are more likely to lack sufficient resources—in terms of education, socio–economic resources or communication skill—to object to physicians, and are therefore more likely to be negatively affected by power differentials.39

Nonetheless, this concern also affects unilateral decision-making policies and thus informed non-dissent does not fare worse in this respect. Moreover, although informed non-dissent cannot address it entirely, it can mitigate its effects more readily. Once recognised the implications that informed non-dissent might have for minority populations, institutions could provide resources to ensure that patients and families are aware of their right to object, that they have access to advocates who could defend their interests, and that physicians are sensitive to the burdens they are imposing on surrogates. Informed non-dissent will not solve all cases of disagreement. Discrepancies will still arise. By giving due voice to surrogates and respecting their choices we hope to mitigate the impact of an uninformed acceptance of a questionable default practice—with its continuation causing harm to the patient and impinging on the professional integrity of the care team.


Physicians should strive for shared decision-making in clinical encounters. However, in contexts such as the provision of CPR disparate views over goals and treatments can lead to intractable disagreements and make this ideal difficult to achieve. As clinical ethicists, we are often faced with clinicians grappling with the question “what can we do in these circumstances?” Although unilateral decision making on the side of clinicians might seem appealing in these cases, an enhanced conceptualisation of informed non-dissent can be a more ethically sound alternative. Such an alternative can protect professional integrity while maintaining bidirectional communication and thus also protect patients’ and surrogates’ interests. It also has the potential to address, at least partially, some of the most relevant justice concerns that stem from the application of unilateral decision-making policies.


The authors would like to thank Drs Augustine I Choi, Michael G Stewart, Anthony Hollenberg and Laura L Forese for their support of the Fellowship in Medical Ethics at Weill Cornell Medicine. The authors acknowledge the support of the New York-Houston Clinical Ethics Consortium of New York Presbyterian Hospital, Weill Cornell Medicine, Houston Methodist, and Baylor College of Medicine


View Abstract


  • Contributors DRdA, KL, PK, IMdM and TB developed the original idea and drafted the initial version of the manuscript. DRA, IMM and TB revised the manuscript and made additional contributions. IMM and TB supervised the process.

  • 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.

  • Patient consent for publication Not required.

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.