3.3 Palliative care in the emergency department

Training and practice in emergency medicine prioritizes acute illness and life-prolonging interventions. As the elderly population continues to grow, more patients with crises related to illness and debility are expected to rely on emergency departments for help. Severely ill and functionally impaired elderly patients seek emergency care not only for physical concerns, but also for concerns related to psychological or financial distress, and access to care (Grudzen et al., 2010). Clinicians may feel ill-equipped to provide the necessary care for these patients, but they recognize the importance in providing this care (Smith et al., 2009; Stone et al., 2011). There is increasing evidence of the benefit of early palliative care intervention in acute care settings (Campbell and Guzman, 2003; Lustbader et al., 2011).

As the entry point to the hospital and often the health-care system, the emergency department is uniquely positioned for early identification of patients with unmet palliative care needs (Mahony et al., 2008). In addition, the emergency department is often a staging area for negotiating the difficult decisions surrounding life-sustaining interventions and an environment where sudden life-altering events are managed. In recognition of the need to evolve expertise in managing these critical concerns in patient care, there is wide support for improving palliative care practices in the emergency department (American College of Emergency Physicians, 2014; National Consensus Project for Quality Palliative Care, 2013; Emergency Nurses Association, 2013).

Models have been proposed for encouraging the integration of palliative care practices into routine care in the emergency department (Grudzen et al., 2011). Also, resources are available and continue to develop for improving education, clinical care, and research (Education in Palliative and End of Life Care— Emergency Medicine, 2010) Center to Advance Palliative Care, 2011). This chapter will introduce these principles and provide structured approaches to the care of patients with palliative care needs in the emergency department.

The initial assessment of all patients presenting to the emergency department is a triage process based on the level of acuity to direct care to the most critical patients first. Patients with significant unmet palliative care needs may present to the emergency department at any stage of illness and at any level of acuity. The emergency clinician must recognize these patients as early as possible, and assure that their care is not only efficient and effective, but also concordant with the goals of the patient and the family. Two tools have been developed to rapidly guide care in the emergency department. The first developed as an initial screen for palliative care needs in stable patients, the Screen for Palliative and End of Life Care Needs in the Emergency Department (SPEED) assessment tool. The second is the mnemonic ‘ABCD’, which was developed to assist in the approach to patients in critical condition.

The SPEED tool is designed for use as a brief multidimensional symptom assessment for quick administration in the emergency department. It is a 13-question assessment tool adapted from the previously developed NEST assessment tool (Emanuel et al., 2001). It can be performed rapidly by anyone on the care team (physician, nurse, chaplain, social worker, or other health-care worker), but may have particular utility at the point of triage. This is the first tool designed for use in the emergency department and tested favourably for both reliability and validity in screening for palliative care needs of cancer patients (Richards et al., 2011). SPEED addresses five major domains of palliative care needs: physical, therapeutic, psychological, social, and spiritual. A preliminary study applied the SPEED instrument to 1842 patients in routine emergency department triage and found nearly half of the patients had at least moderate to severe pain, and approximately one-quarter of the patients had at least moderate difficulty with care needs at home, difficulty with medications, were feeling overwhelmed, and had difficulty getting medical care that fitted their goals (Quest et al., 2011). These findings support other emergency department-based efforts to identify patients with unmet palliative care needs (Mahony et al., 2008; Waugh, 2010; Glajchen et al., 2011; Grudzen et al., 2011) with the advantage of brevity and simplicity of application.

Patients arriving at the emergency department in critical condition at the final stages of chronic progressive illnesses are a complex challenge. Despite the inherent chaos of the environment, the emergency clinician must immediately intervene within the context of the patient’s illness and goals of care. Therefore, a simultaneous rapid palliative care assessment must occur concurrently with temporizing medical interventions until either goals are clear or life-sustaining interventions are necessary, despite a lack of clarity.

The mnemonic, ABCD, was developed by the Education in Palliative and End of Life Care for Emergency Medicine (EPEC-EM) project (EPEC, 2010) to assist in these situations and is summarized in Table 3.3.1 . This approach provides a framework to assess and address the immediate symptom needs of the patient while simultaneously acquiring information to guide medical interventions that can reasonably achieve the goals of the patient and family (B in the mnemonic). There are three remaining elements of the mnemonic: existence of an advance directive (A); the presence of a knowledgeable caregiver (C); and the capacity of the patient to participate in medical decision-making (D). These factors allow the medical provider to place the patient’s current condition in the context of the underlying illness, recent functional decline, and goals for medical care. This information can be used to provide recommendations for reasonable medical interventions within the value system of the patient and family.

Patients with malignancy often present to the emergency department with uncontrolled pain (Vandyk et al., 2012). Overall, pain is also the most common reason patients seek emergency care, yet analgesics are underutilized, and delays to treatment are common (Todd et al., 2007). Historically, there have been barriers to optimal pain management in the emergency department, including lack of training (Lamba et al., 2012), inadequate assessment (Stalnikowicz et al., 2005), misinterpretation of reasons for seeking care (Wilsey et al., 2008), and disparities in treatment practices among clinicians (Cone et al., 2003; Pletcher et al., 2008). Therefore, patients coming to the emergency department with severe breakthrough cancer pain may be justifiably fearful that their pain may not be addressed in a timely or adequate manner.

Assessing severity of malignant pain in the emergency department using a standardized pain scale such as the numerical rating scale, is an important first step (Todd, 2005). Severe pain in malignancy has been referred to as a ‘cancer pain emergency’ and rapid titration of parenteral opioids has been demonstrated to provide effective and safe control of pain while concurrently addressing other diagnostic concerns (Hagen et al., 1997; Mercadante et al., 2002; Soares et al., 2003). In addition to severity, initial assessment of cancer pain must also quickly identify underlying patient characteristics, such as prior exposure to opioids or underlying organ dysfunction, which would alter initial dose calculations. Also, new or progressive pain with the potential for emergent interventions, such as spinal cord compression with motor dysfunction, must be pursued concurrently.

In an effort to encourage emergency clinicians to provide timely and effective intervention, a five-step approach was originally introduced through the EPEC-EM curriculum and is based, in part, on recommendations from the American Pain Society (Miaskowski et al., 2005, pp. 39–48; EPEC, 2010; DeSandre and Quest, 2009) (see Table 3.3.2 ).

Navigating patients and families through the chaotic experience of the emergency department and negotiating reasonable plans for medical care is a skill required of all emergency clinicians. When patients and families are faced with situations related to mortality in the emergency department, communication requires particular sensitivity and focus. Structured approaches can improve communication by reducing clinician distress in these difficult situations (Baile et al., 2000). A structured approach can be used to address three important scenarios in the emergency department: (1) rapidly establishing goals of care, (2) death disclosure, and (3) family presence during cardiopulmonary resuscitation (CPR) (if it is assessed as appropriate to initiate CPR).

Goal setting is a routine practice of emergency care. Patients present to the emergency department at all phases of illness, and the emergency department is often the staging area for complex medical decisions that set a trajectory for future interventions. Patients are placed on ventilators, sent to surgery, admitted to intensive care units, and sometimes supported symptomatically as they die. Patient and clinician goals must ultimately align in order to ensure medical interventions are likely to be beneficial. Prior to any diagnostic test or medical intervention in the emergency department, the clinician anticipates an outcome. If a test or intervention would not change the outcome, then it would not be medically reasonable to proceed. Other interventions may be more reasonable, in order to limit suffering, or ensure a person can die in a peaceful environment.

‘Goals of care’ have been defined as ‘physical, social, spiritual, or other patient-centered goals that arise following an informed discussion of the current disease(s), prognosis, and treatment options’ (Weissman and Meier, 2011). There are several perceived barriers facing emergency clinicians needing to address goals of care prior to intervening with diagnostic evaluations or therapies. Unlike more time-intensive approaches, the purpose of discussing a patient’s goals of care in the emergency department often seeks to answer questions about interventions, such as whether or not a ventilator or the intensive care unit will achieve the most appropriate clinical outcome. While it is clear that family meetings are an effective communication strategy for many complex medical decisions (Hudson et al., 2009), emergency clinicians may feel ill-prepared for such meetings in the emergency department or the clinical situation is changing too rapidly for a family meeting to occur.

Similar to the guidelines for conducting a family meeting (Hudson et al., 2008), one strategy for effective goals of care discussions in the emergency department uses a seven-step approach:

Like other communication strategies, start by acquiring sufficient information and limiting environmental distractions in order to conduct the meeting effectively. Have the questions clear in your mind, have the necessary family and support staff present or on the phone, and introduce everyone involved in the meeting along with their relationship to the patient or role in clinical care. Start with reassurance, such as, ‘In an effort to provide the best possible care for. . . , we need to determine which investigators and interventions would be most beneficial’.

In order to proceed from a common perspective, it is important that the patient or family have a similar understanding of the disease and the prognosis. If there are questions or a lack of clarity, it may be necessary to provide or obtain additional information before proceeding further.

Whenever possible, the patient’s goals should be communicated directly from the patient either verbally or in the form of an advance directive. Eliciting the patient’s perspectives may be challenging if the patient has lost the capacity to interact, has never communicated these concerns to their family previously, or both. Often, the patient has spoken to family about conditions they would find acceptable or unacceptable, such as ‘I want to be at home with my family when I die’, or ‘I never want to be left on a machine if I won’t recover’. In order to preserve autonomy and objectivity, in such situations, it is helpful to ask the family questions from the perspective of the patient, and to imagine what the patient would want or not want in particular scenarios. If the family is unable to imagine or report the patient’s direct perspectives, it may be more useful to ask the family if the patient has ever spoken about others who have died or been critically ill in ways the patient found acceptable or unacceptable. Was hospice involved? Were they held on life-sustaining interventions beyond what appeared reasonable? Was there evidence of suffering? Were there things the patient wished were different? Importantly the question ‘What do you want for. . . ?’ should never be asked. Instead, ‘What would be acceptable for. . . under these circumstances?’ assuming that there are viable clinical options.

A common fear among emergency clinicians is that questions about hopes and expectations may lead to responses such as ‘do everything’. A productive approach to understanding this general hope is to redirect the focus onto specific burdens and benefits of particular interventions. For example, some patients may want to sustain biological existence regardless of the likelihood of ever achieving awareness. This is very uncommon. For most patients and families, ‘everything’ often refers to any interventions whose benefits would outweigh adverse outcomes and is most often focused in the minds of families on physical comfort. For persistent requests for high-burden interventions that are unlikely to achieve patient goals, a ‘harm reduction strategy’ is recommended. This approach attempts to support reasonable medical interventions within the context of patient-centred goals (Quill et al., 2009).

In order to offer reasonable suggestions for goals, the emergency clinician should frame the suggestions in the context of prognosis of hours to days, days to weeks, weeks to months, or months to years. An example might be, ‘People close to death often are most concerned with assuring comfort, peace, and dignity in their dying process’ or ‘People with only weeks to months to live often focus on avoiding the risk of dying in an intensive care unit, and may consider treatments that are not too high of a burden, such as a trial of intravenous antibiotics’.

If the expectations or hopes communicated are medically unachievable, reframing hope towards medically achievable goals can be an effective strategy. Examples include controlling physical symptoms, providing access to emotional or practical support, and assuring dignity (Clayton et al., 2005).

Clear recommendations from the emergency clinician can be a powerful source of comfort for patients and families. The recommendations should be based on medically achievable goals from the patient’s perspective whenever possible. For example, a patient may present to the emergency department with advanced cancer and severe functional impairment who is demonstrating ventilatory insufficiency, but had previously communicated they never wanted to ‘die on machines’. If the patient is unlikely to ever achieve ventilator independence or survival off a ventilator, then a recommendation to focus on controlling dyspnoea and avoiding intubation will be more concordant than a recommendation to proceed with intubation. Likewise, if the same patient were at risk for imminent cardiac arrest and death, an understanding of likely outcomes of CPR would inform your recommendations. Patients and families often have unrealistic understanding of CPR (Heyland et al., 2006). Cardiac arrest itself portends an extremely poor prognosis, and patients with advanced illness with poor function at baseline are unlikely to survive CPR and almost certainly will not regain pre-arrest levels of function (Ebell et al., 1998; Ewer et al., 2001; Peberdy et al., 2003; Reisfield et al., 2006; Young 2009; Sasson et al., 2010). For patients who survive attempts at CPR with such severe underlying disease, dependence on life-sustaining interventions through end of life is the expectation. Therefore, a recommendation in this setting is to avoid CPR.

Once goals and recommendations are established, consensus dictates any immediate interventions. Summarize the available medical options to support the patient’s goals, and assure understanding and agreement. If no consensus can be achieved reasonably, the emergency clinician may need to initiate specific life-sustaining interventions to allow the family time to process what is happening to the patient. In such complex situations, further communication with the family will be necessary after the initial attempts to medically stabilize the patient.

A clear plan of care sets the course for appropriate disposition and follow-up. The emergency clinician may focus initially on symptom management while other needs, such as psychosocial distress, are being addressed by other interdisciplinary professionals. If life-sustaining interventions are initiated in the emergency department, then a timeframe and conditions for future discussions should be established.

Informing a family of the devastating news that a loved one has died is particularly challenging in the emergency department, where the death is often relatively unexpected or sudden. Emergency clinicians may feel uncertain of the best approach, yet they are required to have these extraordinarily sensitive conversations routinely (Stone et al., 2011). Death disclosure is a form of breaking bad news, and educational models exist to teach these skills in the emergency department (Quest et al., 2002; Benenson and Pollack, 2003; Hobgood et al., 2005). Each of these models share four common elements:

A nine-step approach incorporates these elements from the educational models and is recommended for managing death disclosure in the emergency department (EPEC, 2010).

When a family arrives in the emergency department, they may not be aware that the patient has died. In order to limit additional emotional trauma to the family, staff should be prepared to receive them on arrival while the clinician is informed of their presence. The room should be quiet and private with adequate seating, away from other emergency department traffic, and equipped with facial tissues. The emergency clinician should be prepared to immediately respond having all available information (including the patient’s full name), and assure that the body is being cleaned and prepared for viewing by the family. Relevant staff such as the physician, nurse, social worker, and chaplain should enter the room together. Mobile phones or pagers should be handed over to colleagues until after the death disclosure.

While universal introductions are essential, when entering the room, the appropriate sense of gravitas must be evident. The initial introductions must state the full name of the patient and all involved staff, while unequivocally identifying the primary survivor and the relationship to the patient. If there are children present, depending on their age and maturity, it might be best to communicate first to the parent separately from the children, then collaborate to determine how best to communicate with them.

Sit down, preferable closest to the primary survivor (whether or not they choose to sit down). Maintain awareness of body language to communicate respect, cultural sensitivity, and engagement.

The preferred opening communication to a sudden death in the emergency department is to provide information about the immediate circumstances prior to the patient’s death. One approach may be, ‘I’m not sure what you know about what has happened, but I’m afraid I have some bad news about [identify deceased patient by name and relationship to the survivor]. . . ’ or ‘I’m not sure if you are aware that [deceased patient’s name] was involved in a serious accident . . . ’.

It is imperative that any euphemisms be avoided at this stage. Survivors are predictably in a state of emotional shock from the transition statements, so unequivocal language must follow to avoid misperception. An example of clear communication would be, ‘I’m sorry to tell you that [patient’s full name] has died’.

Emotional response to devastating news is an unpredictable and individual experience. It may be highly expressive as agony or anger, or it may be profoundly withdrawn or stunned. Allowing time, presence, and empathic but clear communication should quickly stabilize the situation and direct further explorations.

Understanding and engaging empathic responses is a skill that can be both taught and practised. Many clinicians may be uncomfortable with the risk of creating an emotionally uncomfortable situation while in the emergency department. With preparation and practice, emergency clinicians can engage in these necessary interactions with both efficiency and compassion. One useful technique for communicating empathy can be performed in under 40 seconds while improving satisfaction and decreasing the anxiety of the emotional experience for the patient (Fogarty et al., 1999):

The surviving family hopes for a lack of suffering. Although it may not be possible to empirically report on the patient’s experiences, it is reasonable to assume that the patient was not experiencing physical suffering once unconscious. If the death was violent or otherwise a difficult death, the clinician can speak to the importance of the interventions used to control pain and suffering prior to death.

Information about the patient’s underlying medical conditions is often needed to understand and report on the likely cause of death, and may help provide additional closure if related to a known illness. An approach might be, ‘To better understand what happened to [deceased patient’s name], may I ask you some questions about his/her medical history? I can also tell you what we know and answer any questions you may have’.

Family typically will choose to view the deceased person, and it should be offered. It may work best to have support staff guide the family into the room, and assure ongoing availability nearby.

Prior to leaving the family, the clinician should offer condolences and provide contact information for any additional questions.

If the deceased is a child, support staff can offer to provide the bereaved something tangible from the child, such as a lock of the child’s hair to take with them.

Death experiences in the emergency department can be exhausting for everyone involved, particularly if a child has died, or if the death was violent, or if the death triggers transference of personal memories for clinicians or other staff. It can be helpful to ask involved staff how they are managing and to verbalize feelings or distress. Staff should consider a routine debriefing just after the death disclosure and possibly after work, or with other friends and family or with a trusted third party.

There is increasing acceptance for allowing family presence during resuscitation in the emergency department in both adults and paediatrics (Kleinman et al., 2010; Morrison et al., 2010). Parents in particular want to be given the choice to be present at the bedside of their child during CPR and those who have lived through the experience were grateful and would repeat the experience if faced with it again. Although family members generally do not create interference, there remains significant apprehension among clinicians and variability in practice. Clinicians vary in their level of comfort or even their perceptions of the appropriateness of including family members in a resuscitation event. Policies encouraging the use of family presence during resuscitation are recommended to promote more uniformity in practice (Dingerman et al., 2007; Tinsley et al., 2008; Dudley et al., 2009).

Although no consensus guidelines are available to provide universal recommendations, certain common themes have emerged from the literature to guide practice (Boudreaux et al., 2002; Agard, 2008; Doolin et al. 2011). First, although family presence should certainly be encouraged, the situation may dictate whether or not to proceed. Certain conditions should be met to assure comfort of the family members and the medical team. The number of family members may need to be limited during the initial phases of resuscitation, while ‘final goodbyes’ may be encouraged by a controlled process of bringing in other family members as the patient ceases to respond to reasonable resuscitative efforts. A ‘family support person’ should be identified from the staff, freed of other clinical responsibilities to remain with the family members throughout the resuscitation. The family support person can be a social worker, chaplain, or nurse. They should be trained sufficiently to guide the family members through the medical process by interpreting events and provide psychosocial support, while monitoring behaviours and assuring a controlled situation (Farah et al., 2007).

Given the common discomfort expressed by clinicians regarding family presence, a stepwise structure may help to guide clinicians while also meeting the needs of the staff and the family.

As the essential family members are brought into the room by the family support person, the family support person should introduce the family members. The resuscitation team leader should then introduce themselves as the ‘doctor responsible for the care of your [state relationship to the patient, such as “husband”]’.

Either the resuscitation team leader or a delegate reviews the situation and current status in clear and concise terms. For example, ‘Your husband’s heart has stopped beating and we are trying to restart the heart’.

Given the high likelihood of death in cardiac arrest, it is prudent to warn the family members of the seriousness of the situation. A statement such as, ‘As you can see, this is an extremely serious situation, and we are worried that your husband may not survive’.

State the plan of further care, such as, ‘We are giving him powerful medications and electricity to try to restart his heart’.

Provide standard resuscitation as indicated by the situation.

It is essential to acknowledge with gratitude the difficulty of the event, the skill of the staff, and effort put forth to attempt to save the life of the patient. For example, ‘Thank you team. This is never easy. I appreciate your skill and hard work’.

Either offer to inform the remaining family independently, or offer to bring the family members in the room to the others to inform them of the patient’s death. As with other death disclosures, this would be an appropriate time to gather additional medical information to better understand the cause of death.

The same principle applies as with death disclosure (see earlier subsection). Debriefing after attempts at resuscitation is an excellent method of assuring that all concerns about the resuscitation are addressed. It is also an important time for self-reflection and staff assessment prior to returning to patient care.

When the usual systems of care outside the hospital cannot meet the needs of patients and families, the emergency department offers the hope of relief or resolution. The highest priority for care in the emergency department is necessarily the threat to life or limb. Unfortunately, this may leave an increasing number of patients and families suffering in a state of uncertainty in a chaotic environment. For emergency clinicians, the extent to which any consultation can expedite or improve care determines its success.

Emergency clinicians are invested in improving the experience of the patient and family by attending to their needs efficiently. Timeliness of care, empathy, technical competence, information dispensation, and pain management have been identified as particular elements associated with patient satisfaction in the emergency department (Welch, 2010). These five elements of patient satisfaction in the emergency department align easily with palliative care goals for patients presenting with life-limiting or life-threatening illness, and may be an important focus for efforts to improve palliative care specialty involvement in the emergency department.

Acute concerns related to chronic illness is an increasingly large proportion of the emergency department patient population, and while palliative care consultation in the emergency department may be available, large variability exists in the access to palliative care specialty services. In order to assist the general clinician to identify and assess patients with palliative care needs, a checklist approach has been proposed for use as early as the time of admission. The hope of this approach is to provide structure to limit variability in practice and to encourage the development of generalist skills in identifying and intervening appropriately for patients with palliative care needs (Weissman and Meier, 2011). For emergency clinicians, recommendations for essential palliative care skills include the ability to assess illness and trajectory of decline with a basic formulation of prognosis; expertly manage difficult communications including breaking bad news and death disclosure, and basic spiritual and cultural competency; assist in urgent advance care planning; manage family presence during resuscitation, manage pain and other symptoms, manage withdrawal or withholding of non-beneficial treatments, manage the care of the imminently dying, manage hospice patients and palliative care systems referrals, manage ethical and legal issues, and manage the care of the dying child (Quest et al., 2009). The following sections address several available structured approaches to efficiently address these essential skills.

Bad news can be viewed as information that is expected to have a profound and negative influence on a person’s future. When confronted with situations such as unexpected or devastating diagnostic findings in the emergency department, clinicians may be uncomfortable delivering the bad news. A few guiding principles can help to assure that the communications are performed with skill, compassion, and efficiency. Optimally, the goal is for the communications to be perceived as ‘understanding and positive’ while demonstrating empathy in a patient-centred approach. The clinician should also first consider the patient’s willingness or interest in receiving the information, and conclude by assuring that the patient understands the information given (Schmid et al., 2005). A useful approach that incorporates these principles well is the six-step S-P-I-K-E-S strategy (Baile et al., 2000). Although likely more familiar in practice to non-emergency clinicians, this strategy can be easily adapted for emergency department use (EPEC, 2010).

Review the available information and mentally rehearse what needs to be communicated, while anticipating critical questions. Establish a private setting whenever possible, and allow the patient to determine whether to bring family or significant others in the room. Sit down and make eye contact with those in the room. Silence or hand over your mobile phone or pager to a colleague. Consider including support staff such as a social worker, who can answer additional questions and offer further support to the patient and significant others.

The clinician should use open-ended questions to determine how much the patient understands or ‘perceives’ their medical condition. This important step allows the clinician to recognize if there is an informational gap between the patient’s perception of illness and the clinician’s knowledge of the patient’s illness. An example might be, ‘Can you tell me your understanding of why we did the CT scan of your head?’.

Most patients would prefer to have information given to them directly when available. Some patients may choose to avoid hearing the information altogether or defer to a significant other as the proxy to receive information. Neither is more correct or appropriate, and both are valid coping strategies for individuals. The important element for the emergency clinician is to ask either ‘how’ they would like the information given, or ‘if’ they prefer to receive the information directly, through a proxy significant other, or later by their primary physician.

Before delivering the difficult information, it may lessen the shock and create more receptivity to information by preparing the patient with a ‘forecasting’ or ‘warning shot’ communication such as, ‘I’m afraid I have some bad news’. Communicate at the level of the patient’s ability to understand. Provide small pieces of information at time, avoiding technical jargon. Stop frequently to assess understanding and comfort with continuing. Avoid being overly blunt or overly emotional, with the goal of being perceived as competent and forward thinking.

Understanding and engaging empathic responses is a skill that can be both taught and practised. Many clinicians may be uncomfortable with the risk of creating an emotionally uncomfortable situation while in the emergency department. With preparation and practice, emergency clinicians can engage in these necessary interactions with both efficiency and compassion. One useful technique for communicating empathy can be performed in under 40 seconds while improving satisfaction and decreasing the anxiety of the emotional experience for the patient (Fogarty et al., 1999):

Review what was covered in the interaction, and assure that a clear plan is in place for follow-up and symptom management that is acceptable to the patient. Also provide contact information, either directly or indirectly, to assure continuity of care.

When patients present to the emergency department in extremis without advance directive documents, life-sustaining interventions are often underway prior to family arrival. Once additional information is available to guide medical decision-making in the context of the patient’s underlying illness and goals, it may be appropriate to consider withdrawal of life-sustaining interventions in the emergency department to relieve the patient from unintended or non-beneficial interventions at the end of life. The withdrawal of a ventilator warrants particular attention by the emergency clinician. Prior to consideration of terminal ventilator discontinuation (Curtis, 2005), an interdisciplinary group of available clinicians (physician, nurse, social worker, chaplain, or others) should meet with family members to discuss the current situation, and establish clear overall goals of medical care to determine whether or not ventilator discontinuation is concordant with the goals of the medical team and the patient and family. Documentation must unequivocally define the purpose, concordance among family, and the benefit to the patient of proceeding. In addition, an order must also be placed in the record to not attempt resuscitation along with appropriate documentation.

Before proceeding, the interdisciplinary team should focus on controlling the environment of care and preparing the family for the process of ventilator discontinuation. The patient may be moved to a room with less external noise or traffic, and monitors and any unnecessary medical devices may be removed from the patient to avoid distraction from a focus on the patient and family comfort. In addition, the family must be meticulously prepared for what to expect during and after the procedure. The medical team should offer recommendations for either removing (most situations) or maintaining (if airway obstruction is anticipated) the endotracheal tube. The family should be reassured that while abnormal breathing reflexes are expected and may be disturbing to witness, any symptoms of distress will be aggressively managed. Focus the family to observe for signs of comfort on the face of the patient rather than expected breathing reflexes such as a gasping appearance during agonal respirations. Encourage the family to provide calm verbal and tactile reassurances for the patient throughout the procedure. Meanwhile, have the medical team prepared to respond to observable signs of physical distress such as severe tachypnoea, accessory muscle use, restlessness, nasal flaring, grunting at end-expiration, and facial grimace. The family must also be prepared for the unpredictability of continued spontaneous breathing, from as little as minutes to as long as several days depending on the patient’s condition. If the patient is expected to continue spontaneous breathing and is maintaining vital signs, appropriate disposition plans should be in place, such as inpatient hospice transfer or hospital admission for further symptom management through the end of life (Campbell, 2007).

Once a decision to proceed is clear and the effect of any paralytic medications have reliably subsided, the patient may be either ‘terminally weaned’ by reducing oxygen dose and ventilation rate incrementally, or placed on ventilator ‘standby’ while adjusting medications to limit distress prior to ventilator discontinuation. With proper titration of opioids for dyspnoea and benzodiazepines for anxiety or agitation, control of distressing symptoms can be achieved with little concern for influencing the time to death (Campbell et al., 1999; Chan et al., 2004). Recommendations for dosing are similar to the management of other symptom distress and is relative to the patient’s condition. Observable signs of distress should be minimized with the lowest effective dose. Once efficacy has been established, then maintenance dosing should be continued to avoid emergence of additional distress (Campbell et al., 1999; Chan et al., 2004; Curtis 2005; Campbell 2007).