Show Summary Details
Page of

Ethical decision making in withdrawing and withholding treatment 

Ethical decision making in withdrawing and withholding treatment
Ethical decision making in withdrawing and withholding treatment

Margaret Isaac

and Jared Randall Curtis

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 04 July 2022

Key points

  • The process of intensive care resource allocation and practices around withholding and withdrawing life support vary by physician, hospital, region, and country.

  • Shared decision-making is the recommended default approach to medical decision-making, though there is cultural geographic and individual variability in physician attitudes and patient/family preferences toward shared decision-making.

  • Surrogate decision-making is difficult, with decisions being shaped by many factors including family members’ preferences and patients’ values.

  • Communication with patients and families regarding complex medical decisions and shared decision-making can be enhanced with specific approaches and techniques.

  • Previously completed advance directives can help inform decision-making when patients lose decisional capacity in the ICU.


Intensive care units (ICUs) are the centre of some of the highest technology, and most advanced care that medicine has to offer and have the potential to offer life-saving treatment to critically-ill patients. Death preceded by ICU care occurs in approximately 20% of patients in the US [1]‌, creating an imperative for critical care physicians and staff to thoughtfully and ethically address end-of-life issues. Particularly in settings in which resources are limited, an ethical approach to resource allocation must be advanced, both in considering if patients are sick enough to warrant critical care services, and, at the other end of the spectrum, determining when patients are too sick to benefit from this type of care. A variety of models for decision-making exist, and many professional medical societies advocate for a shared decision-making approach. However, patients and their surrogate decision-makers are widely variable in their preferences for decision control, so their role and the role of the treating physician is best approached on an individualized basis. Similarly, the literature on communication in the ICU setting describes key features associated with increased family satisfaction, though physicians are best served to use these as general guidelines, tailoring their specific approach to the individual patient and family. Other challenges in physician-patient/surrogate communication include cross-cultural communication and working with interpreters, implementation of wishes expressed in advance directives and discussing resuscitation preferences.

Resource allocation

Critical care provision requires utilization and mobilization of many resourcesresources not universally available across all regions and countries. Even within urban centres in the developed world, there is significant variability in the availability of ICU beds. Although the number of ICU beds does not correlate with societal health expenditures within the developed world, it does correlate significantly with mortality [2]‌. The manner in which finite resources are allocated varies significantly within the US and across other countries. The process by which these resources are allocated raises specific ethical considerationsboth in terms of who is deemed sick enough to warrant critical care, and who is too sick to benefit. Principles of resource allocation outlined by the American Thoracic Society highlight the following [3]:

  • Each individual’s life is valuable and equally so.

  • Respect for patient autonomy, as represented by informed consent, is a central tenet for providing health care, including ICU care.

  • Enhancement of the patient’s welfare, by providing resources that meet an individual’s medical needs and that the patient regards as beneficial, is the primary duty of health care providers.

  • ICU care, when medically appropriate, is an essential component of a basic package of health care services that should be available for all.

  • The duty of health care providers to benefit an individual patient has limits when doing so unfairly compromises the availability of resources needed by others.

Withholding and withdrawing of life support

Most patients who die in the ICU do so after having life-sustaining treatments withheld or withdrawn [4]‌. Though many medical ethicists and medical societies, including the American Thoracic Society and the American Medical Association view withholding and withdrawal of life-sustaining treatments to be ethically equivalent, this view is not universally shared. The degree to which these treatments are withdrawn and the process by which these decisions are made varies significantly by country, shaped in part by cultural and religious differences. In a large study that included seven different geographic regions throughout the world, 36% of in-hospital deaths followed withdrawal or withholding of life-sustaining treatments [5]. In a study comparing regional practice variability in Europe, rates of withdrawal of life-sustaining treatments were over two times as high in northern Europe (47%) as in southern Europe (18%) [4].

Physicians’ actions are shaped by their own cultural and religious beliefs and individual clinician characteristics affect decisions to limit life-sustaining treatments. Specifically, religious affiliation impacts physicians’ decisions to either withdraw or withhold these treatmentswith decisions to withhold treatment being more common among Jewish, Greek Orthodox, and Muslim physicians in Europe, and decisions to withdraw treatment being more common among Protestant, Catholic, and unaffiliated physicians [6]‌. Time from ICU admission to limitation of these treatments also varied with physicians’ religious affiliations. Overall religiosity among Jewish physicians is inversely related to willingness to withdraw life-sustaining treatments. In a survey of US physicians, black physicians were far more likely than white physicians to personally want treatments such as CPR, mechanical ventilation, and artificial feeding when given hypothetical scenarios of living with severe brain damage or in a persistent vegetative state, though the majority of all physicians surveyed did not desire such treatments.

Use of intensive care services during terminal hospitalizations varies in different regions and countries, but in most parts of the developed world, intensive care is increasingly being used at the end-of-life among older individuals [7]‌. Therefore, given the aging of populations in the developed world, increasing attention must be paid towards thoughtful advance care planning, health care utilization, and optimal end-of-life care in the ICU setting.

Decision making

Models of decision making

Many models and frameworks of medical decision-making have been developed, encompassing a spectrum from maximal patient autonomy to a more traditional parentalist approach. At the patient autonomy end of the spectrum, informed decision-making (or the ‘informed choice’ model) describes a process in which patients are informed about health care treatments and their alternatives and then independently choose from these options. On the other hand, parentalism describes a relationship in which physician is guardian, providing information to steer a patient towards a decision that the physician considers preferable. This model, the traditional one used in the intensive care setting, assumes that ‘physician knows best’ by virtue of their training and experience and will act in the best interest of the patient. This is still commonly used in certain countries as the default approach, and is used universally in the setting of emergency care, in which consent cannot be explicitly obtained and is often considered implied. In between these extremes is shared decision-making. Shared decision-making involves:

  • Involvement of both patient and physician in the medical decision-making process.

  • Sharing of information between patient (about personal values and goals) and physician (about prognosis and specifics of medical treatment options).

  • Discussion of treatment preferences and evaluation of understanding.

  • Consensus between both parties on the treatment to be implemented.

Geographic and cultural variability

Although shared decision-making was endorsed as the preferred default model for medical decision-making in the ICU by an international group of critical care societies in 2003, significant variability in decision-making processes exists across practice sites, countries, and cultures. Patients and families themselves express varying preferences regarding the locus of decision control, with many expressing preference for a shared decision-making model [8]‌, but significant percentages desiring either more autonomy, or a more directive approach on the part of physicians [9]. In short, there is tremendous variability amongst family members regarding their desired level of participation in medical decision-making, so clinicians are best served by assessing this on an individualized basis and adapting their approach to the preferences of the family member as well as the circumstances of the patient. In Europe and in Israel, for example, a parentalist approach is more common than in the US. The ETHICUS study, evaluating differences in end-of-life care in ICUs across European countries and Israel, found that a minority of end-of-life decisions were discussed with family. The authors speculated that this mainly reflected situations in which further medical intervention was thought by physicians to be unwarranted. Geographic differences in sharing of prognostic data with patients exist as wellwith only a minority of Japanese physicians and patients agreeing that a patient should be informed of an incurable cancer diagnosis before their family, and a majority of US physicians and patients agreeing with such a statement. Other studies as well have supported the finding that particular ethnic groups (Asians and Latinos, specifically) support a more family-centred approach to medical decision-making than other groups. These studies again suggest the need for an individualized, tailored approach specific to each clinical situation, taking into account geographic location, family dynamics, prognosis, and cultural factors.

Surrogate decision-making

Because most patients receiving end-of-life care in the ICU either lack decisional capacity or the ability to communicate, clinicians often turn to surrogate decision-makers for assistance with medical decision-making. Laws regarding designation of a proxy decision-maker, in the absence of a patient-selected surrogate, vary widely by country [10]. Many physicians expect that surrogates will utilize substituted judgment when making decisions on behalf of the patient (i.e. ‘What would the patient say about their preferences if they were able to talk to us right now?’). However, some patients express a preference for surrogates to abide by best interest standard when participating in decision-making. Data suggest that, in practice, proxies actually rely on many factors including substituted judgement and the bests interests of the patient, as well their own values and best interests. Multiple studies from a variety of countries have shown that surrogate accuracy in describing the wishes of patients is often poor. However, the accuracy of substituted judgments has been positively associated with patients having spoken with their surrogate in advance about care preferences.

The job of a surrogate decision-maker can come at a cost: in one study, over half of surrogate decision-makers suffered symptoms of learned helplessness, characterized by reduced motivation, difficulty in determining causality, and depression. Multiple studies have found an increased risk of psychological symptoms for family members and surrogate decision-makers of patients who died in an ICUincluding depression, anxiety, panic disorder, post-traumatic stress symptoms, and complicated grief [11,12,13].

Because effective surrogate decision-making can be very difficult for many family members and many patients have not expressed clear treatment preferences prior to critical illness, ICU clinicians are frequently in the position of helping family members identify patients’ values, exploring how those values might lead to specific treatment preferences, and interpreting how the preferences inform specific treatment decisions that need to be made. ICU clinicians can help family members with this process by asking directed questions about the patient’s values and goals and the aspects of life most important to the patient.


Communication with an interdisciplinary team

Critical care provision relies on the skills and coordination of various staff and specialists. Interdisciplinary communication around end-of-life care in the ICU significantly augments the quality of care that is provided, though is variable across practice sites and countries. Interdisciplinary communication is valued by terminally ill patients and their families [14]. Caring for dying patients, withdrawing life support, and conflict among clinicians are all independently associated with burnout among nursing staff as well [15]. Burnout among critical care physicians is lower among those who maintain positive working relationships with nursing staff, and higher in individuals who experience conflict with colleagues and staff [16], highlighting the need to enhance interdisciplinary collaboration to provide a sustainable work environment for all ICU staff.

Communication with patients and families

Skilled communication is one of the core tools of any physician addressing end-of-life care. While every medical case, patient, and family, are different, several principles can help guide clinicians in leading family conferences:

  • Families of critically-ill patients are more satisfied when physicians spend more time listening and less time talking [17].

  • Families benefit from explicit reassurance that patients will not suffer, from statements of support for the medical decisions they have made, and from statements that they and the patient will not be abandoned [18].

  • Family satisfaction increases when physicians express empathy explicitly.

A randomized controlled trial from France found that use of a specific communication strategy (see Box 387.1) and a bereavement brochure led to lower rates of post-traumatic stress symptoms, anxiety, and depression among surviving family members [19]. Notably, the conferences implementing this standardized approach also were longer in duration, had a higher proportion of family speech, and a higher proportion of nurse speech, though it is not clear which of these factors was causative.

Reprinted from Journal of Critical Care, 17(3), Curtis JR et al., 'Studying communication about end-of-life care during the ICU family conference: Development of a framework', pp. 147–60. Copyright © 2002, with permission from Elsevier and World Federation of Societies of Intensive and Critical Care Medicine (WFSICCM).

The multi-disciplinary ICU family conference is a key component of high quality care for critically-ill patients and their families.

Table 387.1 Approach for effective family conferences in the ICU


Components to consider


  • Conduct conference early within the ICU stay

  • Find a quiet, private space for the conference [12], with chairs for all participants

  • Hold a ‘pre-conference’ with participating clinicians to ensure common goals and messages

Conducting the conference

  • Speak less, listen more [18]

  • Use empathic language [19,20] to acknowledge family emotions

  • Explicitly acknowledge the difficulty of having an ill loved one, and of surrogate decision-making

  • Discuss patient values, hopes, fears, and treatment preferences

  • Remind surrogate decision-makers of their role in invoking substituted judgment

  • Assure families that all attempts will be made to prevent patient suffering

  • Make explicit statements about non-abandonment of the patient and family [19]

  • Provide support for decisions made by family [19]

Closing the conference

  • Ask the family to summarize what they understand from the discussion and what further questions they have

  • Summarize next steps and a follow-up plan

Data from various sources (see references).

Addressing spirituality

Spiritual concerns on the part of patients and/or their surrogates are prominent at the end of life, and physicians often feel ill-equipped to address these issues. Family satisfaction and feelings of support increase when spiritual needs are discussed with clinicians and addressed by spiritual care providers.

Cross-cultural communication and working with interpreters

Communication across cultures and languages can create additional barriers in communication and can lead to miscommunication and misunderstandings. In an observational study of family conferences utilizing a professional medical interpreter, alterations (including additions, omissions, substitutions, and editorializations) occurred over 50% of the time, with three-quarters of the alterations thought to have clinically significant impact, the vast majority of those being negative [20]. Interpreted conferences have been found to be shorter in length, and contain fewer physician statements of family support than non-interpreted conferences. Best practices include:

  • Briefing and debriefing with interpreter before and after family conferences.

  • Clarifying the interpreter’s role as strict interpreter or cultural broker.

  • Introduction of interpreter and their role to the patient/surrogate(s).

  • Triangular seating arrangement for patient/surrogate(s), interpreter, and clinician.

  • Eye contact with patient/surrogate(s), addressing patient/surrogate(s) in the first-person.

  • Speaking slowly and ensuring only one person is speaking at at time.

  • Request for direct and complete translation from the interpreter.

  • Explicit identification of misunderstandings by the interpreter.

Specific communication challenges—health care directives and advance care planning

Advance care planning and completion of health care directives/living wills have been promoted by many as a mechanism to allow better understanding of patients’ care preferences and values. These are rarely completed in the critical care setting, but occasionally, a pre-existing advance directive helps to shape and frame discussions with surrogate decision-makers. An extensive literature demonstrates the poor performance of health care directivespointing to their inadequacy in realistically identifying the types of medical decisions that are most salient at the end of life, and failing to capture the evolution in preferences that often occur over the course of a complex illness. However, advance directives are associated with higher surrogate ratings of quality of dying and with lower burden on surrogates.

Recent studies suggest that communication about end-of-life care among patients with cancer is associated with increased quality of life and reduction in the use of intensive life-sustaining treatments and reduced health care costs at the end of life. A randomized trial targeting hospitalized patients over age 80 showed that advance care planning was associated with improved quality of life and reduced ICU use at the end of life as well as reduced psychological symptoms among family members. These studies suggest promise for advance care planning to improve quality of life and quality of care while simultaneously reducing intensive care use at the end of life.

Specific communication challenges—discussing resuscitation preferences

Resuscitation preferences are commonly addressed in discussions with patients and families in the critical care setting and these discussions present unique challenges to clinicians. Patient preferences, followed by short-term prognosis, are the biggest factor in determining Do Not Attempt Resuscitation (DNAR) status in the ICU setting. Studies have shown that patients routinely overestimate their survival post-resuscitation and these overestimates may be associated with consumption of television medical dramas, which themselves have historically portrayed unrealistically high in-hospital CPR survival rates. Knowledge of actual CPR survival rates has been shown to affect patient preferences. Patients who choose to be full code are more likely to think of resuscitation abstractly as a means to restore life, and DNAR status as a sign of inferior care while patients who opt for DNAR status are more likely to think of resuscitation in terms of suffering and futility, and see DNAR orders as a means towards a natural death. With this in mind, several core messages can help clinicians communicate with patients and surrogates about resuscitation:

  • Emphasizing that resuscitation occurs in the setting of patient death, rather than resorting to euphemistic language, such as ‘restarting the heart’.

  • Describing the alternative to resuscitation, namely, natural death with comfort-focused care.

  • When appropriate, using CPR survival data to help patients understand their probability of surviving resuscitation.


In summary, critical care services, while potentially life-saving, are increasingly utilized at the end of life, particularly for older patients. Utilization of these intensive resources can cause stress and distress for critically-ill patients and their families and requires careful consideration on the part of clinicians. Though shared decision-making has been endorsed as a preferred approach, patients and their surrogate decision-makers have variable preferences with regard to their desired level of involvement in medical decision-making, and this is best evaluated on an individual basis. Effective communication, both outside of the ICU through, for example, advance care planning, and inside the ICU, such as during family conferences, can benefit patients, their families, clinicians, and health care systems as a whole.


1. Angus DC, Barnato AE, Linde-Zwirble WT, et al. (2004). Use of intensive care at the end of life in the United States: an epidemiologic study. Critical Care Medicine, 32(3), 638–43.Find this resource:

2. Wunsch H, Angus DC, Harrison DA, et al. (2008). Variation in critical care services across North America and Western Europe. Critical Care Medicine, 36(10), p. 2787–93, e1–9.Find this resource:

3. Fair allocation of intensive care unit resources. (1997). American Thoracic Society. American Journal of Respiratory Critical Care Medicine, 156(4 Pt 1), 1282–301.Find this resource:

4. Sprung CL, Cohen SL, Sjokvist P, et al. (2003). End-of-life practices in European intensive care units: the Ethicus Study. Journal of the American Medical Association, 290(6), 790–7.Find this resource:

5. Azoulay E, Metnitz B, Sprung CL, et al. (2009). End-of-life practices in 282 intensive care units: data from the SAPS 3 database. Intensive Care Medicine, 35(4), 623–30.Find this resource:

6. Sprung CL, Maia P, Bulow HH, et al. (2007). The importance of religious affiliation and culture on end-of-life decisions in European intensive care units. Intensive Care Medicine, 33(10), 1732–9.Find this resource:

7. Wunsch H, Linde-Zwirble WT, Harrison DA, et al. (2009). Use of intensive care services during terminal hospitalizations in England and the United States. American Journal of Respiratory Critical Care Medicine, 180(9), 875–80.Find this resource:

8. Azoulay E, Pochard F, Chevret S, et al. (2004). Half the family members of intensive care unit patients do not want to share in the decision-making process: a study in 78 French intensive care units. Critical Care Medicine, 32(9), 1832–8.Find this resource:

9. White DB, Evans LR, Bautista CA, et al. (2009). Are physicians’ recommendations to limit life support beneficial or burdensome? Bringing empirical data to the debate. American Journal of Respiratory Critical Care Medicine, 180(4), 320–5.Find this resource:

10. Lautrette A, Peigne V, Watts J, et al. (2008). Surrogate decision makers for incompetent ICU patients: a European perspective. Current Opinions on Critical Care, 14(6), 714–19.Find this resource:

11. Siegel MD, Hayes E, Vanderwerker LC, et al. (2008). Psychiatric illness in the next of kin of patients who die in the intensive care unit. Critical Care Medicine, 36(6), 1722–8.Find this resource:

12. Pochard F, Azoulay E, Chevret S, et al. (2001). Symptoms of anxiety and depression in family members of intensive care unit patients: ethical hypothesis regarding decision-making capacity. Critical Care Medicine, 29(10), 1893–7.Find this resource:

13. Azoulay E, Pochard F, Kentish-Barnes N, et al. (2005). Risk of post-traumatic stress symptoms in family members of intensive care unit patients. American Journal of Respiratory Critical Care Medicine, 171(9), 987–94.Find this resource:

14. Carline JD, Curtis JR, Wenrich MD, et al. (2003). Physicians’ interactions with health care teams and systems in the care of dying patients: perspectives of dying patients, family members, and health care professionals. Journal of Pain Symptom Management, 25(1), 19–28.Find this resource:

15. Poncet MC, Toullic P, Papazian L, et al. (2007). Burnout syndrome in critical care nursing staff. American Journal of Respiratory Critical Care Medicine, 175(7), 698–704.Find this resource:

16. Embriaco N, Azoulay E, Barrau K, et al. (2007). High level of burnout in intensivists: prevalence and associated factors. American Journal of Respiratory Critical Care Medicine, 175(7), 686–92.Find this resource:

17. McDonagh JR, Elliott TB, Engelberg RA, et al. (2004). Family satisfaction with family conferences about end-of-life care in the intensive care unit: increased proportion of family speech is associated with increased satisfaction. Critical Care Medicine, 32(7), 1484–8.Find this resource:

18. Stapleton RD, Engelberg RA, Wenrich MD, et al. (2006). Clinician statements and family satisfaction with family conferences in the intensive care unit. Critical Care Medicine, 34(6), 1679–85.Find this resource:

19. Curtis JR Engelberg RA, Wenrich MD, et al. (2002). Studying communication about end-of-life care during the ICU family conference: development of a framework. Journal of Critical Care, 17(3), 147–60.Find this resource:

20. Pham K, Thornton JD, Engelberg RA, et al. (2008). Alterations during medical interpretation of ICU family conferences that interfere with or enhance communication. Chest, 134(1), 109–16.Find this resource: