◆ Limitation of life-supporting therapies in anticipation of death has become common in Intensive Care Units (ICUs) around the world.
◆ Like other clinical procedures performed by intensive care clinicians, life support limitation should be approached with careful attention to relevant evidence, standards of quality, and involvement of the interdisciplinary ICU team.
◆ ICU clinicians should avoid use of phrases like ‘withdrawal of care’ and ‘nothing more we can do’, but instead emphasize their ability to achieve the patient’s values, control patient distress, and provide family support.
◆ A decision to withhold any life-sustaining therapy should prompt a critical review of the rationale for continuing other such therapies.
◆ Knowledge of underlying principles and best practices for limitation of life support, together with skills for clear and empathic communication, are core competencies for intensive care clinicians.
The withdrawal or withholding of life-supporting therapies in anticipation of death has become common in ICUs around the world . Most ICU deaths in the US involve life support limitation . In other parts of the world, the proportion of deaths following withdrawal or withholding of life support ranges from about one-quarter to two-thirds, depending on the region . Thus, knowledge and skill related to limitation of life support, including processes of communication, decision-making, and implementation, are essential for delivery of high-quality intensive care. In this chapter, we synthesize existing relevant evidence and provide recommendations based on the best available data for approaching this important area of critical care practice.
Legal and ethical frameworks
In the US and many other countries, withholding (refraining from starting a treatment) and withdrawal (stopping a treatment that is already underway) of life supporting therapy are both acceptable, legally and ethically . Thus, when prognosis is uncertain, a time-limited trial of life support may be initiated without an irrevocable commitment to continued treatment. In practice, clinicians as well as patients and families often find withdrawing a treatment more emotionally difficult than withholding. Intensive care treatment is not withdrawn in some countries . Key ethical and legal principles related to US practice are summarized in Box 388.1.
Advance directives allow the patient to specify preferences regarding specific life-sustaining therapies (a living will), and/or to authorize a surrogate to make medical decisions if the patient becomes incapacitated (a health care proxy or durable power of attorney for health care). When the patient lacks decision-making capacity, the right to refuse life-sustaining treatment is extended to a surrogate appointed by the patient or designated by applicable law. Standards for life-support limitation by surrogates vary from state to state in the US, as does the legal hierarchy of individuals who serve as surrogate decision-makers in the absence of a designation by the patient. Wherever they practice, intensive care clinicians must be familiar with the applicable local provisions.
In the US, a growing number of states have established templates that allow written directives to be translated into medical orders that are accessible across multiple venues in the health care system; The ‘POLST’ (Physician Orders for Life Sustaining Treatment)  is one example. A pre-printed, checklist order form addressing specific life-sustaining treatments provides the basis for discussion by the clinician with the patient or the surrogate. The completed form operationalizes the patient’s preferences as medical orders, which can be changed or updated at any time.
Communication and decision-making
Since some patients and families may choose to continue intensive care therapies primarily out of fear that limitation will inevitably be accompanied by patient suffering, it is important for clinicians to provide assurance that the patient’s comfort can and will be maintained. Families are more satisfied when assured that the patient undergoing life support limitation will not suffer or be abandoned prior to death. ICU clinicians should avoid use of phrases like ‘withdrawal of care’ and ‘nothing more we can do’, but instead emphasize their ability to achieve the patient’s values, control patient distress, and provide continuing patient and family support.
Physicians frequently miss opportunities to acknowledge and address emotions experienced by patients and families facing life-threatening illness . Incorporation of empathic statements in clinician communication, which helps to modulate emotions and to enhance trust in clinicians, is supported by expert opinion and empirical data . The intensivist needs skills for listening as well as delivering information about the patient’s condition and prognosis. The family’s proportion of speaking time in an ICU meeting with clinicians is related to family satisfaction . Recommendations have been provided for approaching patients and family members who prefer continuation of life-supporting treatments based on the belief that a miracle will occur , and those who insist on ‘everything’ despite a clinician’s view that the treatments will not have meaningful benefit.
Critical care guidelines endorse a shared decision-making model in which both the patient or family and the physician play an active role , but there is a wide continuum of preferred roles among patients and families, and varying roles for clinicians across different countries, regions, and cultures . In general, it is appropriate for clinicians, after investigating the patient’s values and preferences, to offer a recommendation reflecting best medical judgment. Besides fulfilling a professional responsibility, the clinician’s recommendation may help to address the stress, guilt, doubt, and fear that plague many surrogates making major treatment decisions for others. Some ICUs distribute printed informational materials such as a brochure to families to help them prepare for surrogate decision-making.
Decisions about the limitation of life support often give rise to conflict . In a large, multinational survey, conflicts (between nurses and physicians, among nurses, and between staff and relatives) were reported by over 70% of responding clinicians, with end-of-life care representing a major source of discord and attendant job strain . Clinicians’ decisions to limit life sustaining treatments are influenced by diverse factors apart from patients’ wishes, including such provider-level factors such as number of years since completing training, specialty, experience working in ICUs, gender, cultural background, and religious beliefs . For a variety of reasons that are thought to include an exaggerated sense of personal responsibility for the patient’s outcome and a perceived ‘covenantal bond’ to continue life-sustaining treatment , surgeons may be less willing to accept withdrawal of intensive care therapies after surgery. Recognition of diverse influences including nonclinical and idiosyncratic factors is valuable for prioritizing attention to patient-focused concerns and for promoting consensus among the patient, family, and multiple members of the health care team.
When conflicts arise over the use of life support, a process oriented approach informed by the most common reasons for conflicts may facilitate consensus . The intensive care clinician should always first explore the family’s current understanding of what has happened and what is expected. After describing the medical situation as perceived by the clinical team, the intensivist can then address other potential sources of conflict. These include discordant perceptions of appropriate roles (e.g. the family feels obligated to pursue every treatment to extend life, while the physician feels responsible to limit treatments thought to be futile and the nurse to protect the patient’s dignity and comfort during the dying process); distrust of the team based on past experiences or cultural factors; and strong emotions such as grief or fear. Exploration of spiritual beliefs may also be important. Experts have recommended a ‘differential diagnosis’ approach to conflict over decisions to limit treatment . Strategies for promoting consensus are summarized in Box 388.2.
Mechanical ventilation is the form of life support that is most frequently withheld or withdrawn in anticipation of death in the ICU. Life-sustaining therapy may also include vasopressor medications, renal replacement, transfusions, antibiotics, tube feeding, intravenous hydration, and other modalities without which the patient’s death would be expected . Although ethically equivalent, limitation of individual treatments tends to be sequential rather than simultaneous, and withdrawal of mechanical ventilation usually occurs as a late event . Although withdrawal of artificial nutrition rarely causes discomfort for seriously ill patients, nutrition is often limited only as a last step, likely because of the symbolic importance of food for many families.
An observational study of 14 hospitals in one metropolitan area in the Northwest US found that withdrawal of all life-sustaining interventions occurred over more than 1 day for nearly half of the patients . Among long stay patients, a longer duration of the process of withdrawal was associated in this study with greater family satisfaction with care . ‘Stuttering withdrawal’, however, is not considered optimal by most expert guidelines. Guidelines suggest that a decision to withhold one life-sustaining therapy should prompt a critical review of the rationale for continuing all other therapies . Ongoing phlebotomy and imaging should be reconsidered and all medications should be re-evaluated. Deactivation of a pacemaker or automatic implantable cardiac defibrillator should also be addressed. A Task Force of the Society of Critical Care Medicine has provided a conceptual framework and recommendations for the use of non-invasive ventilation in patients who have foregone endotracheal intubation and invasive ventilation, emphasizing clarity about care goals and weighing of therapeutic benefits and burdens .
Life support limitation: a clinical procedure
Like other clinical procedures, limitation of life support requires expertize, careful preparation, and consistent implementation in accordance with the best available evidence, appropriate documentation, interdisciplinary collaboration, and ongoing evaluation. Inexperienced clinicians should not perform this procedure independently until they have demonstrated their competence in a supervised setting. Ideally, an attending physician and an experienced nurse with specific relevant training will both play an active role. Input from all team members should be invited in formulating and implementing the final plan.
Preparing the patient and family
Most patients and families do not know what to expect, fear the unknown, and take comfort from a straightforward, sensitive, description of common scenarios after limitation of life support. A key point for clinicians to communicate, however, is that a degree of uncertainty is inherent with respect to the patient’s responses—a small minority of patients will survive to hospital discharge after life support limitation. For non-survivors, the duration of the dying process is variable. Broad estimates based on the patient’s clinical condition (for example, minutes to hours or hours to days) can be cautiously offered.
The family should be reassured that the clinical team will respond immediately and effectively to any patient distress. Since families (and even some clinicians) may not understand acceptable approaches to symptom control at the end of life, it is helpful to review the basic principles. The doctrine of double effect provides justification for the use of opiates and sedatives at any dose needed to maintain the comfort of the patient, even if the clinician sees a potential for drug-related respiratory depression or hypotension . Clinicians may wish to address timing and setting along with other aspects of the process of forgoing life support. A family may have reasons to delay implementation—for example, to give an out-of-town relative time to make a final visit, or to allow arrangements for a funeral that by religious law must occur quickly after death. Conversely, rapid withdrawal of life support may be requested by a family that finds it more painful to postpone an inevitable outcome. The clinical team should endeavour to accommodate the family’s preferred schedule. Consideration may be given to transfer of the patient to a more private and peaceful setting, such as a Palliative Care Unit or a private room on a regular hospital ward with arrangements for close monitoring of symptoms. When a patient and family have developed important relationships with the critical care team, symptom needs are expected to require intensive nursing, or continuity of care would be compromised by transfer, the ICU may be the optimal place for withdrawal of life sustaining therapies. Efforts to optimize the environment might then include closing curtains, liberalizing family visits, and removing all tubes, lines, and drains that are not necessary for the patient’s comfort. Electronic monitoring should be discontinued so that alarms do not disturb the patient or family and attention remains focused on the patient rather than on irrelevant technology. Expert recommendations suggest a specific conversation with family and staff about the rationale for removing monitors in order to alleviate anxiety.
Families should feel they have the option, but not an obligation to be with the patient during or after discontinuation of treatments or at the time of the patient’s death. A non-judgmental, open-ended approach by the clinician, with reassurance that the patient will be continuously attended by staff, will allow the family to express true preferences and help relieve guilt or other concerns for those who decide to leave. For families who plan to stay, a description of the procedure and expected patient responses should be offered. The family can be prepared for the possibility of noise from respiratory secretions, a range of breathing patterns (fast, slow, deep, shallow), snoring, muscle twitching, and progressive slowing of heart rate, and lowering of blood pressure. Most families welcome an inquiry about religious or spiritual traditions they may wish to observe, and an offer of pastoral care. The family may also value the opportunity to speak with a social worker and/or an organ donation coordinator.
Methods of withdrawing mechanical ventilation from an endotracheally-intubated patient
No comparative evidence identifies the optimal strategy for separating a patient from the ventilator. Some clinicians withdraw ventilator and endotracheal tube in one step. Others maintain the artificial airway to protect against obstruction and permit access for suctioning. Another approach involves incremental reduction of oxygen and ventilatory support over a relatively short period (e.g. 10–30 minutes), followed by complete withdrawal of the ventilator. The last strategy allows for titration of medications to address symptoms during the period of final ‘weaning’.
Approaches to symptom control can be broadly categorized as ‘pre-emptive’ (i.e. the patient is medicated with opiates and/or sedatives before withdrawing life support) and ‘reactive’ (i.e. medications are given if patient distress is actually observed). Pre-emptive treatment might be given when the patient’s death is expected imminently after withdrawal and the patient is either already in distress or such distress is anticipated. Although the use of general anaesthesia (‘pre-emptive high doses of opioids and sedatives for anaesthesia or at least deep sedation to assure comfort’) has been proposed for the ‘potentially conscious and imminently dying patient’, this strategy may be appropriate only when consistent with the patient’s values and when the risk of distress is sufficiently high to justify the risk of hastening the patient’s death. For the patient with a reasonable prospect of surviving after separation from the ventilator, treatment would be deferred unless and until the patient manifested signs of (or self-reported) distress. There is no reason to medicate a patient who is brain dead. All other patients undergoing withdrawal of life support must be carefully assessed for distressing symptoms throughout the process . Tools are available for behavioural assessment of pain and dyspnoea . Ample doses of opiates and sedatives (as infusions and boluses), along with sufficient staff, should be immediately available at the bedside. Dosing is limited only by the principle of proportionality (i.e. treatment should be proportional to the patient’s symptoms). Documentation in the medical record by both physicians and nurses should clearly reflect this relationship. Expert recommendations are available to guide management of symptoms including pain, dyspnoea, and delirium for critically-ill patients at the end of life .
Noisy breathing, with unpleasant, gurgling sounds, often occurs during the dying process, causing distress for some families and ICU staff. Pharmacologic approaches include glycopyrronium bromide, hyoscine (scopolamine), atropine, and octreotide. Since glycopyrronium bromide does not cross the blood brain barrier, there is reason to believe it can reduce pharyngeal and tracheobronchial secretions with less risk of central effects than hyoscine or atropine. Non-pharmacologic interventions, such as repositioning and gentle suctioning, may also be helpful for some patients.
Neuromuscular blocking agents should never be introduced when the ventilator is withdrawn . For patients who have already received such a medication, withdrawal of mechanical ventilation should be deferred until neuromuscular function is restored. If neuromuscular blockade cannot be reversed without an unacceptable delay, it may be appropriate to proceed, with close attention to physiologic signs of distress and liberal use of comfort medications.
The ICU nurse and withdrawal of life support
The ICU nurse is involved more directly and frequently with the patient undergoing limitation of life support, and with the patient’s family, than any other clinician. In a U.S. national survey, two-thirds of nurse respondents experienced emotional difficulty with the process of limiting life support, but more than 60% reported that physicians did not usually provide emotional support, for which they looked instead to their colleagues and even to patients’ families . The presence of a physician during this process is an important form of support both for nurses and families. Ideally, this presence continues during the post-mortem care, which is often a lonely and psychologically difficult part of the nurse’s work. Reducing the nurse-patient ratio to 1-to-1 during withdrawal of life support, and providing the primary nurse at least a brief period of respite after a patient’s death, are among strategies to improve the practice environment and patient care. The End-of-Life Nursing Education Consortium (ELNEC) project  provides specialized training and extensive curricular materials in palliative and end-of-life care including limitation of life support for nurses.
Increasing sophistication and application of life-supporting technologies have been accompanied by a growing need for expertise in limiting their use when burdens exceed potential benefits for critically-ill patients. Like other clinical procedures performed by intensive care clinicians, withholding and withdrawing life support should be approached with careful attention to relevant evidence, standards of quality, and involvement of the interdisciplinary ICU team. Knowledge of underlying principles and best practices for limitation of life support, together with skills for clear and empathic communication, are core competencies for intensive care clinicians.
1. 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, 790–7.Find this resource:
2. Prendergast TJ and Luce JM. (1997). Increasing incidence of withholding and withdrawal of life support from the critically ill. American Journal of Respiratory Critical Care Medicine, 155, 15–20.Find this resource:
3. 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, 623–30.Find this resource:
4. Luce JM and Alpers A. (2000). Legal aspects of withholding and withdrawing life support from critically ill patients in the United States and providing palliative care to them. American Journal of Respiratory Critical Care Medicine, 162, 2029–32.Find this resource:
5. Physician Orders for Life Sustaining Treating Program Paradigm. Available at: http://www.ohsu.edu/polst/ (accessed 20 February, 2014).
6. Selph RB, Shiang J, Engelberg R, Curtis JR, and White DB. (2008). Empathy and life support decisions in intensive care units. Journal of General Internal Medicine, 23, 1311–17.Find this resource:
7. 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, 1484–8.Find this resource:
8. Widera EW, Rosenfeld KE, Fromme EK, Sulmasy DP, and Arnold RM. (2011). Approaching patients and family members who hope for a miracle. Journal of Pain Symptom Management, 42, 119–25.Find this resource:
9. Davidson JE, Powers K, Hedayat KM, et al. (2007).Clinical practice guidelines for support of the family in the patient-centered intensive care unit: American College of Critical Care Medicine Task Force 2004–2005. Critical Care Medicine, 35, 605–22.Find this resource:
10. Azoulay E, Timsit JF, Sprung CL, et al. (2009). Prevalence and factors of intensive care unit conflicts: the conflicus study. American Journal of Respiratory Critical Care Medicine, 180, 853–60.Find this resource:
11. Cook DJ, Guyatt GH, Jaeschke R, et al. (1995). Determinants in Canadian health care workers of the decision to withdraw life support from the critically ill. Canadian Critical Care Trials Group. Journal of the American Medical Association, 273, 703–8.Find this resource:
12. Schwarze ML, Bradley CT, and Brasel KJ. (2010). Surgical ‘buy-in’: the contractual relationship between surgeons and patients that influences decisions regarding life-supporting therapy. Critical Care Medicine, 38, 843–8.Find this resource:
13. Goold SD, Williams B, Arnold RM. (2000). Conflicts regarding decisions to limit treatment: a differential diagnosis. Journal of the American Medical Association, 283, 909–14.Find this resource:
14. Truog RD, Campbell ML, Curtis JR, et al. (2008). Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College of Critical Care Medicine. Critical Care Medicine, 36, 953–63.Find this resource:
15. Gerstel E, Engelberg RA, Koepsell T, and Curtis JR. (2008). Duration of withdrawal of life support in the intensive care unit and association with family satisfaction. American Journal of Respiratory Critical Care Medicine, 178, 798–804.Find this resource:
16. Curtis JR, Cook DJ, Sinuff T, et al. (2007). Noninvasive positive pressure ventilation in critical and palliative care settings: understanding the goals of therapy. Critical Care Medicine, 35, 932–9.Find this resource:
17. Billings JA. (2012). Humane terminal extubation reconsidered: the role for preemptive analgesia and sedation. Critical Care Medicine, 40, 625–30Find this resource:
18. Campbell ML, Templin T, and Walch J. (2010). A Respiratory Distress Observation Scale for patients unable to self-report dyspnea. Journal of Palliative Medicine, 13, 285–90.Find this resource:
19. Kirchhoff KT and Kowalkowski JA. (2010). Current practices for withdrawal of life support in intensive care units. American Journal of Critical Care, 19, 532–41.Find this resource:
20. American Association of Colleges of Nurses (2014). End-of-life nursing education consortium (ELNEC). Available at: http://www.aacn.nche.edu/ELNEC/ (accessed 4 March, 2014).