Cultivating Conscience: Learning to Make End-of-Life Decisions in the Emergency Department

Abstract

Resuscitation is the essence of our training as emergency medicine residents. The essentials of training include becoming comfortable with when to initiate resuscitation and learning proficiency with lifesaving techniques and procedures. Equally important as learning the mechanics of resuscitation, and perhaps even more difficult to become comfortable with, is knowing when and how to terminate resuscitation. Making end-of-life decisions in the emergency department is very different than in other practice settings. In the ED we often confront dying patients without any knowledge of their medical histories or advance directives. We face the added difficulties of all with diverse problems that demand time and focus. Deciding what care constitutes “the right thing” in such a high-pressure environment can be challenging.Residency is a time for us to become comfortable with decisions of when to terminate resuscitation. However, most of us receive little formal training in ethical decisionmaking. Much of the training focuses on the inpatient setting, where there is plenty of time to gather information, deliberate, and reach consensus. The following real case highlights some of the dilemmas we will face practicing on our own and provides an introduction to the issues involved in end-of-life decisionmaking in the ED.CASE SUMMARYSettingThe ED in this case is in a small community hospital surrounded by a large rural catchment area. There is single-physician coverage with the emergency physician providing EMS medical direction for the extensive rural area. There is also a military hospital in the area, which serves primarily dependents and retirees.The radio callDuring a busy night shift, paramedics contact the base physician requesting permission to terminate a field resuscitation. The patient is a 71-year-old man who was found unconscious and unresponsive by his family. He did not have bystander CPR. The fire department first-responders applied an automated external defibrillator and delivered 2 shocks. When the paramedics arrived, the patient had pulseless electrical activity (PEA). They intubated the patient, established intravenous access, administered 3 rounds of medication according to standard Advanced Cardiac Life Support (ACLS) protocols, and noted that the patient was still in PEA. Forty-five minutes have elapsed since the initial 911 call, and the estimated transport time to the closest ED is an additional 40 minutes. Paramedics further state that the patient may have a do-not-resuscitate (DNR) order, but no documentation is available on scene.ISSUE: OUT-OF-HOSPITAL TERMINATION OF RESUSCITATIONAs base physicians, we are legally and morally responsible for the care provided by paramedics under our direction. Typically EMS guidelines require paramedics to initiate resuscitation of a patient unless the patient or family can provide a valid, written, system-approved DNR order. Once CPR is initiated, EMS providers may only terminate resuscitative efforts if the patient meets the system’s requirements for discontinuation of CPR or the criteria for death in the field. In many EMS systems, death may be pronounced in cases of asystole with other signs of lifelessness (rigor mortis, dependent lividity), or in the case of asystole that does not respond to appropriate ACLS interventions. PEA is not a specific criterion for field determination of death, although some EMS systems give discretion to the base hospital physician in recognition of unusual circumstances. ACEP offers guidelines to base physicians in its policy statement on discontinuing resuscitation in the out-of-hospital setting. This policy supports terminating resuscitative efforts if “the patient is in asystole or a wide-complex bradycardic rhythm with a rate less than 60, normothermic, and fails an adequate trial of resuscitation therapy.”1American College of Emergency Physicians: Nonbeneficial (“futile”) emergency medical interventions [policy statement]. Approved March 1998. To obtain a copy, call 800-798-1822, touch 6.Google ScholarIf these guidelines are applied to this case, there is little question that discontinuation of CPR should be authorized. Few physicians would argue that an elderly patient in PEA unresponsive to ACLS, when 45 minutes have elapsed since the initial call and transport time of an additional 40 minutes, would likely benefit from ongoing resuscitation.DecisionThe physician in this case granted the paramedics’ request to terminate resuscitation in the field.Return of spontaneous circulationShortly after the initial call, another radio call is made by the same paramedics. Moments after the previous call, the patient had return of spontaneous circulation. Paramedics request permission to transport the man to the military hospital where he receives his usual care. Because that hospital receives ambulances and is about the same distance from the scene as the base hospital, permission is granted.Approximately 40 minutes later, the patient arrives unexpectedly at the base hospital. The medic who was driving misunderstood the direction. While en route, the patient maintained a pulse the entire time. His respirations were provided by a bag-valve device through the endotracheal tube, and dopamine infusion was started for low blood pressure according to system protocols.On arrival, the patient’s airway is secure and breath sounds are equal. The patient has a strong carotid pulse and a weak femoral pulse, with an irregular heart rate of 133 beats/min. Blood pressure is measured at 82/48 mm Hg on dopamine at 10 μg/kg/min. The cardiac monitor shows a wide, almost sine-wave rhythm.The paramedics state the patient might have a DNR advance directive but no documentation of such was available. They state the patient’s wife is also elderly and homebound and that she is not coming to the hospital. They learned that the patient had a history of recent kidney problems, and had surgery at the military hospital. The paramedics have temporarily misplaced the patient’s home phone number, but they offer to call their dispatcher to locate the number.ISSUE: ADVANCE DIRECTIVES NOT AVAILABLEUnavailability of advance directives is unfortunately a common scenario for emergency physicians. We often encounter the patient for the first time in a critical moment, and are working with limited information about the patient’s medical history and prior wishes. In a recent study, 88% of high-risk patients presenting to the ED did not have advance directives and of those who did, fewer than a fourth actually brought the DNR orders with them to the ED.2Ishihara KK Wrenn K Wright SW et al.Acad Emerg Med. 1996; 3: 50-53Crossref PubMed Scopus (20) Google Scholar What options does the clinician have in this case? The information provided to the emergency physician is very incomplete. The patient’s wishes are not entirely clear, and the family is not immediately available. The patient’s prior medical problems and health are mostly unknown. The patient is critically ill and immediate intervention seems indicated. Failure to act would probably result in the immediate death of the patient and would foreclose any opportunity to further investigate the patient’s wishes.The limited available evidence suggests that this patient would not want further resuscitation. Certainly, the patient has a right to refuse medical care. However, there is also a legal presumption that an emergency patient would want treatment. Emergency physicians have an ethical obligation to honor the patient’s rights but also an obligation to preserve life. Withholding further treatment has a greater moral significance for clinicians in the ED than withdrawing care later when more information might be available.3Iserson KV Ann Emerg Med. 1996; 28: 51-54Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar Furthermore, emergency physicians have the added pressures of working in a highly public arena where certain expectations for providing emergency medical treatment exist.3Iserson KV Ann Emerg Med. 1996; 28: 51-54Abstract Full Text Full Text PDF PubMed Scopus (38) Google ScholarDecisionBecause of the extreme time pressure, the emergency physician in this case chose to err on the side of maintaining life. The patient was treated with intravenous calcium chloride for presumed hyperkalemia. Immediately ECG complexes narrowed and blood pressure increased. Further treatment with sodium bicarbonate and glucose with insulin led to further improvement in the blood pressure and the cardiac rhythm.Availability of more informationDuring these efforts, the clerk contacted the military hospital. They faxed a copy of a generic DNR advance directive, signed by the patient 14 years previously, which stated that “no heroic measures” be undertaken but contained no directions for what specific interventions this might include. The form did not contain any information about the patient’s medical condition, and did not list an alternate decision-maker.ISSUE: VALIDITY OF AN ADVANCE DIRECTIVEAlthough there is no expiration date on an advance directive, the physician must ensure that it reflects the patient’s current wishes. Individual preferences, beliefs, and circumstances may change over time. Without other supporting evidence, this 14-year-old directive has very limited utility. Laws vary about what type of directive is valid and the “portability” of such directives from setting to setting.4Adams J Wolfson AB Emerg Med Clin North Am. 1990; 8: 183-192PubMed Google Scholar The form in this case also does not provide enough guidance to the physician. Most emergency physicians might agree that an ED thoracot-omy is “heroic,” but there is much less consensus about mechanical ventilation or intravenous medication. At one time even defibrillation was considered highly experimental.Furthermore, the patient recently had surgery, so it is clear that the patient does not reject all medical care, or even some invasive care. More information about the patient’s beliefs and desires is urgently needed. The only available sources are the family and any primary care physician. Unfortunately, neither is available.DecisionThe emergency physician continues to provide advanced life support measures for the patient.StabilizationAfter treatment for hyperkalemia, the patient remained critically ill but his condition was less unstable. The emergency physician had a brief opportunity to investigate further. The physician telephoned the patient’s home and a woman who identified herself only as “a friend of the family” answered. Although she would not give her name, she claimed to know the patient and his wife very well, and stated she was “like a daughter” to them. She emphatically stated that the patient would not want artificial life support of any kind, but she refused to bring the wife to the telephone. She claimed that the wife was too upset to come to the phone, but that the wife also did not want the patient to have artificial life support.ISSUE: WHO CAN ACT AS SURROGATE?In some states, the law clearly defines who may act for the patient when the patient is unable to express his or her wishes. In other jurisdictions, only the person designated on a “durable power of attorney” or a court-appointed conservator has authority to make decisions for the patient. The individual who answered the telephone is neither a relative nor a legally designated substitute. Her statements at best provide only minimal support for the patient’s apparent wish to avoid resuscitation.Subsequent to the telephone contact, paramedics identified the woman on the telephone as the wife of their supervisor. They stated that she works as a counselor for the local home hospice agency. Does this increase the persuasive force of her statements, or restore her credibility? Likely she is expressing the patient’s wishes as she knows them. However, because she is very familiar with the long-term care of the terminally ill, she may be predisposed against resuscitation, even in cases where the patient has a significant chance of recovery.DecisionBecause the “friend of the family” refused to identify herself, her credibility was in doubt, and no change of treatment plan was made.The ED grows busyDuring this time the patient’s blood pressure remains stable with infusion of moderate-dose dopamine, but he has no spontaneous respiratory efforts and his pupils remain fixed and dilated. Although an ICU bed is available, the on-call physician is busy at another hospital at least an hour away. During this time, several other patients have arrived in the ED and are awaiting evaluation. Here, as in many community hospitals, the emergency physician is the only physician who is continuously in- house.ISSUE: TREAT OR TRANSFER?The emergency physician has a busy ED and must remain available. If the patient is moved to the ICU, it is highly likely that the emergency physician will be called repeatedly to the ICU. As ACEP’s policy on “Emergency Physicians’ Patient Care Responsibilities Outside of the Emergency Department” states, “an emergency physician must be available at all times to respond to emergency department patients in a timely and safe manner.”5American College of Emergency Physicians: Emergency physicians’ patient care responsibilities outside of the emergency department [policy statement]. Approved August 1992. Reaffirmed March 1997. To obtain a copy, call 800-798-1822, touch 6.Google Scholar The emergency physician cannot abandon his responsibilities to the other ED patients. Unfortunately, this patient requires immediate ongoing attention, which the admitting doctor would be unable to provide at this time. Furthermore, ICU beds can be a scarce resource. The emergency physician is already aware that this patient may die soon despite care, and there is a strong possibility that therapy could be withdrawn if the patient’s DNR status is confirmed. In addition to pulling the emergency physician away from ED patients, moving this patient to the last ICU bed may not be the best allocation of resources.DecisionThe patient remained in the ED to allow the emergency physician to continue providing safe and timely care for every patient.ClarificationThe emergency physician again called the patient’s home and asked to speak with the patient’s wife. The emergency physician was gentle but adamant that life support would not be discontinued until the patient’s wishes could be discussed. Although the “friend of the family” was still reluctant, she conceded the importance of verifying that the patient’s wishes were current, and she brought the wife to the telephone. The wife stated a long-standing desire that neither she nor her husband be kept alive by mechanical means. In support of this, she related that the patient had recent prostate surgery. She said he consented very reluctantly to the general anesthesia, because of his concern that he might end up as “a vegetable.” When the current situation was described to her, she stated that her husband “wouldn’t want all that.” She verified that the 14-year-old DNR order was still consistent with her husband’s wishes. After discussion of the prognosis, she was offered the opportunity to come to the hospital, but she declined. She had thought he was probably dead when the paramedics left the house. She asked that her husband be kept comfortable.ISSUE: FAMILY INVOLVEMENT AND DNR ORDERSStudies indicate that families approve of contact by an emergency physician to request DNR orders from the ED and that the majority allow the emergency physician to write such orders.6Balentine J Gaita T Rao N et al.Acad Emerg Med. 1996; 3: 54-57Crossref PubMed Scopus (4) Google Scholar The role of the family in assisting with end-of-life decisions and ascertaining the patient’s wishes is emphasized repeatedly by ethicists.4Adams J Wolfson AB Emerg Med Clin North Am. 1990; 8: 183-192PubMed Google Scholar However, family wishes should not override valid advance directives.7American College of Emergency Physicians: Ethical issues for resuscitation [policy statement]. Approved March 1997. To obtain a copy, call 800-798-1822, touch 6.Google Scholar In this case, both were in accordance and supported the physician’s belief that further treatment was unlikely to benefit the patient. A short discussion with a family member can be invaluable in validating beliefs and actions when terminating a resuscitation. Addressing the needs and concerns of the family can be the most rewarding part of an otherwise “nonbeneficial” resuscitation.DecisionThe emergency physician entered a “no CPR” order into the patient’s record at this time.Final courseAfter speaking to the patient’s wife, the emergency physician obtained consensus from the ED staff and the dopamine infusion was discontinued. However, the patient remained in apparently stable condition with normal blood pressure and mild tachycardia. The patient still had no spontaneous respirations and no motor activity. Pupils remained fixed and dilated, and there were no corneal or gag reflexes. Approximately 50 minutes had elapsed since the patient’s arrival, or about 2 ¼ hours since the initial collapse.ISSUE: WITHDRAWING CARE IN THE EDAt this point, the patient and his wife clearly do not desire mechanical ventilation. All medical evidence suggests that death is the likely outcome in this case, regardless of any further intervention. Because of the patient’s severe neurologic impairment, we cannot know what, if any, discomfort he may experience either from treatment or its discontinuation. From experience with other patients during training, we know that sudden removal of airway support may result in agonal movements of the body or sounds that give the appearance of distress. The ED staff is not accustomed to the withdrawal of ventilator support; most patients in critical condition are quickly moved to the ICU.Although commentators claim that there is no legal or ethical difference between withdrawing a treatment and withholding it, both lay people and medical professionals do perceive a difference.8Beauchamp TL Childress JF Principles of Biomedical Ethics.in: ed 4. : Oxford University Press, New York1994: 149-150Google Scholar If a treatment is not started, we may be “allowing nature to take its course.” Removal of life support is a more active process, and is more easily perceived as hastening the death of the patient. Removal of life support should be done with care and deliberation, and with the consensus of all involved parties. In particular, nursing staff may have a different and more immediate perspective on the patient’s care, and they may be more likely than the physician to experience discomfort as life support is removed. In several legal cases in which the physician was charged with hastening the death of a patient, a nurse was the first to report the case to authorities.DecisionIn keeping with the wife’s clear request for comfort care but no mechanical life support, the emergency physician decided not to remove the endotracheal tube at this time. Instead, the ventilator was discontinued (the “machines” were turned off) and the patient received oxygen through the endotracheal tube. Oxygen and fluids were still given, maintaining the appearance of comfort even though the family was not expected to come to the hospital. Importantly, each of the emergency nurses and the respiratory therapist assented to this course of action. Each was given an opportunity to express any misgivings before any irreversible acts were taken.The patient’s condition slowly deteriorated and his heart stopped approximately 55 minutes after the ventilator was turned off. He never had spontaneous respiratory efforts, supporting the emergency physician’s impression of severe anoxic encephalopathy. The wife and “family friend” were notified by phone, and they thanked the ED staff for the care of the patient.This case highlights several unique aspects of end-of-life decisionmaking faced by emergency physicians. As physicians we have a duty to act in the patient’s best interest. In the case of incomplete or contradictory information, usually acting to preserve life will be in the patient’s best interest, at least initially. Treatment can and should be withdrawn later, if further information makes this appropriate. Sometimes these choices are difficult. In such scenarios we may know in our hearts that further resuscitation will not alter the patient’s outcome. However, the nagging questions that we have not done everything we should have, or that the family, patient, or other staff members wanted more, can persist.Even excellent, experienced clinicians may feel uncertain about their decisions. The physician involved here has been practicing for many years and sits on the ethics board of his hospital. Yet the following morning, still disturbed by the case, he sought validation for his actions from his colleagues. At some time in our career we will all struggle with such decisions.As residents we need to begin building our personal foundations for solving ethical dilemmas. Understanding what constitutes beneficial or nonbeneficial care and knowing when to stop is an important aspect of our emergency medicine training. Providing some patients a dignified death may be just as critical as saving the lives of others. Learning which to choose in a particular situation is one our greatest challenges as developing physicians. Resuscitation is the essence of our training as emergency medicine residents. The essentials of training include becoming comfortable with when to initiate resuscitation and learning proficiency with lifesaving techniques and procedures. Equally important as learning the mechanics of resuscitation, and perhaps even more difficult to become comfortable with, is knowing when and how to terminate resuscitation. Making end-of-life decisions in the emergency department is very different than in other practice settings. In the ED we often confront dying patients without any knowledge of their medical histories or advance directives. We face the added difficulties of all with diverse problems that demand time and focus. Deciding what care constitutes “the right thing” in such a high-pressure environment can be challenging. Residency is a time for us to become comfortable with decisions of when to terminate resuscitation. However, most of us receive little formal training in ethical decisionmaking. Much of the training focuses on the inpatient setting, where there is plenty of time to gather information, deliberate, and reach consensus. The following real case highlights some of the dilemmas we will face practicing on our own and provides an introduction to the issues involved in end-of-life decisionmaking in the ED. CASE SUMMARYSettingThe ED in this case is in a small community hospital surrounded by a large rural catchment area. There is single-physician coverage with the emergency physician providing EMS medical direction for the extensive rural area. There is also a military hospital in the area, which serves primarily dependents and retirees.The radio callDuring a busy night shift, paramedics contact the base physician requesting permission to terminate a field resuscitation. The patient is a 71-year-old man who was found unconscious and unresponsive by his family. He did not have bystander CPR. The fire department first-responders applied an automated external defibrillator and delivered 2 shocks. When the paramedics arrived, the patient had pulseless electrical activity (PEA). They intubated the patient, established intravenous access, administered 3 rounds of medication according to standard Advanced Cardiac Life Support (ACLS) protocols, and noted that the patient was still in PEA. Forty-five minutes have elapsed since the initial 911 call, and the estimated transport time to the closest ED is an additional 40 minutes. Paramedics further state that the patient may have a do-not-resuscitate (DNR) order, but no documentation is available on scene. SettingThe ED in this case is in a small community hospital surrounded by a large rural catchment area. There is single-physician coverage with the emergency physician providing EMS medical direction for the extensive rural area. There is also a military hospital in the area, which serves primarily dependents and retirees. The ED in this case is in a small community hospital surrounded by a large rural catchment area. There is single-physician coverage with the emergency physician providing EMS medical direction for the extensive rural area. There is also a military hospital in the area, which serves primarily dependents and retirees. The radio callDuring a busy night shift, paramedics contact the base physician requesting permission to terminate a field resuscitation. The patient is a 71-year-old man who was found unconscious and unresponsive by his family. He did not have bystander CPR. The fire department first-responders applied an automated external defibrillator and delivered 2 shocks. When the paramedics arrived, the patient had pulseless electrical activity (PEA). They intubated the patient, established intravenous access, administered 3 rounds of medication according to standard Advanced Cardiac Life Support (ACLS) protocols, and noted that the patient was still in PEA. Forty-five minutes have elapsed since the initial 911 call, and the estimated transport time to the closest ED is an additional 40 minutes. Paramedics further state that the patient may have a do-not-resuscitate (DNR) order, but no documentation is available on scene. During a busy night shift, paramedics contact the base physician requesting permission to terminate a field resuscitation. The patient is a 71-year-old man who was found unconscious and unresponsive by his family. He did not have bystander CPR. The fire department first-responders applied an automated external defibrillator and delivered 2 shocks. When the paramedics arrived, the patient had pulseless electrical activity (PEA). They intubated the patient, established intravenous access, administered 3 rounds of medication according to standard Advanced Cardiac Life Support (ACLS) protocols, and noted that the patient was still in PEA. Forty-five minutes have elapsed since the initial 911 call, and the estimated transport time to the closest ED is an additional 40 minutes. Paramedics further state that the patient may have a do-not-resuscitate (DNR) order, but no documentation is available on scene. ISSUE: OUT-OF-HOSPITAL TERMINATION OF RESUSCITATIONAs base physicians, we are legally and morally responsible for the care provided by paramedics under our direction. Typically EMS guidelines require paramedics to initiate resuscitation of a patient unless the patient or family can provide a valid, written, system-approved DNR order. Once CPR is initiated, EMS providers may only terminate resuscitative efforts if the patient meets the system’s requirements for discontinuation of CPR or the criteria for death in the field. In many EMS systems, death may be pronounced in cases of asystole with other signs of lifelessness (rigor mortis, dependent lividity), or in the case of asystole that does not respond to appropriate ACLS interventions. PEA is not a specific criterion for field determination of death, although some EMS systems give discretion to the base hospital physician in recognition of unusual circumstances. ACEP offers guidelines to base physicians in its policy statement on discontinuing resuscitation in the out-of-hospital setting. This policy supports terminating resuscitative efforts if “the patient is in asystole or a wide-complex bradycardic rhythm with a rate less than 60, normothermic, and fails an adequate trial of resuscitation therapy.”1American College of Emergency Physicians: Nonbeneficial (“futile”) emergency medical interventions [policy statement]. Approved March 1998. To obtain a copy, call 800-798-1822, touch 6.Google ScholarIf these guidelines are applied to this case, there is little question that discontinuation of CPR should be authorized. Few physicians would argue that an elderly patient in PEA unresponsive to ACLS, when 45 minutes have elapsed since the initial call and transport time of an additional 40 minutes, would likely benefit from ongoing resuscitation.DecisionThe physician in this case granted the paramedics’ request to terminate resuscitation in the field.Return of spontaneous circulationShortly after the initial call, another radio call is made by the same paramedics. Moments after the previous call, the patient had return of spontaneous circulation. Paramedics request permission to transport the man to the military hospital where he receives his usual care. Because that hospital receives ambulances and is about the same distance from the scene as the base hospital, permission is granted.Approximately 40 minutes later, the patient arrives unexpectedly at the base hospital. The medic who was driving misunderstood the direction. While en route, the patient maintained a pulse the entire time. His respirations were provided by a bag-valve device through the endotracheal tube, and dopamine infusion was started for low blood pressure according to system protocols.On arrival, the patient’s airway is secure and breath sounds are equal. The patient has a strong carotid pulse and a weak femoral pulse, with an irregular heart rate of 133 beats/min. Blood pressure is measured at 82/48 mm Hg on dopamine at 10 μg/kg/min. The cardiac monitor shows a wide, almost sine-wave rhythm.The paramedics state the patient might have a DNR advance directive but no documentation of such was available. They state the patient’s wife is also elderly and homebound and that she is not coming to the hospital. They learned that the patient had a history of recent kidney problems, and had surgery at the military hospital. The paramedics have temporarily misplaced the pat