The emergency department is where the patient and potential ethical challenges are first encountered. Patients with acute neurologic illness introduce a unique set of dilemmas related to the pressure for ultra-early prognosis in the wake of rapidly advancing treatments. Many with neurologic injury are unable to provide autonomous consent, further complicating the picture, potentially asking uncertain surrogates to make quick decisions that may result in significant disability. The emergency department physician must take these ethical quandaries into account to provide standard of care treatment.
Neuroethics in the emergency department is part of daily practice and requires and understanding of the pillars of medical ethics and their application to the neurologically injured emergency department patient.
Self-fulfilling prophesies inherently conflict with the traditional pillars of medical ethics and ultra-early prognostication in the neurologically injured is not morally sound.
Resource limitations in neurologic therapeutics markedly alter the ethical framework of care, especially in the context of neurointerventional strategies for acute stroke management.
True autonomy is rare in the neurologically injured patient; substituted consent and proxy autonomy are critical and susceptible to extreme bias.
Ethics is an integral part of standard of care. In tort law, the standard of care is the degree of prudence and caution required of an individual providing the care who thereby has a legal obligation to adhere to certain standards while performing care that could cause foreseeable harm others, that is, maleficence. The requirements of this standard are closely dependent on circumstances; whether a standard of care has been breached is usually framed in terms of a “reasonable person” and if the intended beneficence outweighs any maleficence.
The emergency department (ED) is often the first point of hospital entry for patients with neurologic diseases and the location in which many ethical challenges related to care are initially identified. The emergency care ecosystem is a unique environment and one in which complex decisions related to neurologic care and the attendant ethical issues must be made with limited information and under marked time constraints. Neuroethics as a discipline has recently been identified as its own entity, with subspecialty training now available as well. As a subdiscipline of biomedical ethics, neuroethics focuses on the unique disease states, prognostic challenges, and end of life care that often accompany neurologic illnesses. Both chronic and acute neurologic illness and injury share common features that are specific to this subdiscipline and the interface of neuroethics and emergency medicine has become more complex as treatments have become more advanced and the push for ultra-early prognostics has followed. Faced with decision making in real time that can have a profound impact on the likelihood of survival, and expected to make these decisions as a function of outcome prediction, the emergency care practitioner is placed in a very challenging position with these patients. As a result, a discussion of neuroethics in the emergency medicine environment is warranted and timely.
Definitions and Discussion
Generally speaking, the 4 most commonly accepted pillars of medical ethics are autonomy, beneficence, nonmaleficence, and justice. The application of these foundational underpinnings of bioethics to the practice of emergency medicine is not unfamiliar to most emergency physicians owing to the challenging ethical dilemmas faced by emergency physicians in everyday practice. However, there are nuanced differences inherent in the application of neuroethics to certain ED patients that warrants exploration. Specifically, there are 2 overarching realities of the care of the neurologically injured patient in the ED that challenge some of the more common understandings of the application of the pillars of medical ethics.
First, ultra-early prognostication in the neurologically injured is fraught with uncertainty, bias, and error.
Second, the true risks and benefits of many therapeutic interventions is not known with definitive certainty
Prognostication in the ultra-early phase is so challenging that the Neurocritical Care Society guidelines for the management of devastating brain injury specifically caution against limiting lifesaving care within the first 72 hours of injury in most neurologically injured patients. The supposition is that self-fulfilling prophesies regarding devastating brain injury lead to early limitations in resuscitative care, yet data continue to demonstrate the inability of clinicians to accurately prognosticate in severely injured neurologic patients, even with the best of intentions and the best technology available. In 2009, a landmark investigation was published describing the outcomes of patients with blunt head trauma and Glasgow Coma Score of 3 at presentation. In this review of 189 patients, 13.2% of the cohort achieved good functional outcomes after 6 months. In a similar style review of patients with intracerebral hemorrhage (ICH), the validity of the ICH score as a criterion standard for risk stratification and prediction of 30-day mortality was questioned. The authors found that the ICH score does not accurately predict the mortality rate regardless of surgical intervention and that survival exceeded the index assessment of outcome from the baseline ICH score. These examples clearly underscore the challenges faced by emergency physicians when attempting to align clinical decision making with time-sensitive prognostic models that are likely to induce mortality bias, the end result of which is loss of therapeutic opportunity for the patient.
When the challenges noted are applied to the pillars of medical ethics, certain critical assessments must be addressed. With respect to autonomy, it is required that the patient be able to make an informed decision, with the necessary facts, risks, and benefits clearly delineated and the likelihood of success reasonably estimated. In the case of acute, severe neurologic injury, the information regarding treatment and prognosis is highly technical and it is nearly impossible for those patients to operate with fully informed consent when making decisions about their health, even in the rare circumstances when they can participate. The very nature of the injuries themselves renders these patients’ capacity for expression of autonomy incomplete. When unable to participate in those discussions, proxy expressions of consent and autonomy are sought from those with close relationships to the patients under the pretext that those identified as such will have particular insights into the goals of care wishes of the patient and express them accordingly. However, it is again important to note that the expression of proxy decision making as a substitute for autonomy is held to the same technical standards regarding the possession of a reasonable understanding of risks and benefits and prognosis. In the acutely neurologically injured patient, the veracity of autonomous decision making, even when expressed via proxy, is subject to much disagreement and warrants careful scrutiny to ensure internal consistency and the absence of bias.
Beneficence (the requirement that care be provided with the intent of doing good for the patient) is a more straightforward principle, even in the care of the neurologically injured patient. Neurotherapeutics are developed and deployed for the benefit of the patient. Although all therapeutics carry a risk of side effects, the overall intent is the good of the patient and the risk of the side effects is balanced against the absence of therapeutic intervention. Nonmaleficence (requiring that the care rendered does not harm the patient or society at large) is less straightforward in neuroethics discussions in the emergency setting. Despite the intent to do good for the patient, many neurotherapeutics carry high risks of rare but devastating complications. The treatment of acute ischemic stroke with intravenous tissue plasminogen activator carries a risk of symptomatic ICH of 6.4% and the use of endovascular therapy for mechanical thrombectomy in stroke carries an aggregate risk of symptomatic ICH of 4.4% in the largest 5 trials of endovascular therapy. , When attempting to provide these life-saving therapies to acute stroke patients, a consideration of the risk of intervention is critical in assessing the ethical underpinnings of the treatment. Only patients with acute, disabling strokes are eligible for treatment, but many of those patients would survive without the interventions. And some, although mildly disabled at time of treatment, are infrequently worsened by attempts at therapeutic intervention. The weighing of neurologic disability against the risk of intervention-associated morbidity and mortality requires a nuanced understanding of the belief system of the patient and an expression of societal norms that would support the view that, in fact, some disabilities are worse than death.
The fourth pillar of medical ethics, justice, has particular applicability to the emergency medicine ecosystem, and has been applied during times of scarcity and widespread disease. Most recently, the global severe acute respiratory disease coronavirus-2 (coronavirus disease-19) pandemic has brought the discussion of justice to the modern practitioner in a way that many have not seen before. Specifically, justice requires the fair and equitable distribution of care and resources, and attention to competing needs, rights and obligations. It is important to make the distinction between equal distribution and equitable distribution in emergency medicine. Fairness, as applied to need, defines equitable care, but not necessarily equal care. These competing claims have been illuminated in times of scarcity; as an example, when neurointensive care unit beds are used for surge capacity during pandemics, the specialized neurologic care that would otherwise be available to the patients being admitted through the ED may not be available. Neurointerventional procedures, including endovascular therapy, may also not be available as a result of resource scarcity including a lack of appropriately trained proceduralists as a result of staff illness and quarantine requirements. However, assigning resources to those with the most need respects the principle of distributive justice, even in times of scarcity. Broadly speaking, justice in emergency medicine requires the fair and impartial treatment of patients in the ED; however, issues related to treatment and prognostic bias in neurologically injured patients are potentially in conflict with justice and must be recognized and mitigated thoughtfully.
The ethics of endovascular stroke therapy for patients in the emergency department
In acute stroke care, tissue plasminogen activator has been an accepted standard of care treatment for decades with endovascular treatment (EVT) via mechanical thrombectomy more recently introduced. Per the American Heart Sacculation/American Stroke Association stroke guidelines, mechanical thrombectomy is the standard of care for patients with low premorbid modified Rankin scale (mRS), significant disability, causative proximal occlusion, and early treatment initiation; class of recommendations I, level of evidence A. The latest guidelines have even expanded the time window for EVT up to 24 hours (and beyond in clinical trials) in cases of mismatch between clinical deficit and volume of infarct, effectively evaluating salvageable tissue, based on imaging criteria.
A standard of care, however, is not without its ethical dilemmas, as has been highlighted by the coronavirus disease-19 pandemic. In times of scarcity, whether owing to limited providers or facilities, the standard of care may informally shift as increased emphasis is placed on justice, both in terms of distribution and rationing. For EVT, scarcity may be in the form of limited expert physicians to perform the procedure, limited availability of transport to capable facilities, or limited intensive care beds in which to take care of postprocedural patients.
As a standard of care treatment, there is already a degree of explicit rationing of EVT based on clearly defined and well-accepted criteria for patient appropriateness for therapy. But for those who do not clearly meet class of recommendations I, level of evidence A criteria, the guidelines are less defined and at higher risk of implicit rationing based on subjective usefulness. Per the American Heart Sacculation/American Stroke Association Guidelines, patients with a premorbid mRS of greater than 1 who do not meet other criteria have a class of recommendations IIB, level of evidence B-R for thrombectomy (ie, more randomized data is needed), but may be reasonable candidates. It can be reasonably assessed that there is decreasing marginal benefit relative to increasing baseline mRS, but where is the threshold of futility?
Patient autonomy, insofar as definition of quality of life, should be a primary consideration when considering interventions, even in the acute setting. Intensive care unit-level interventions should generally be considered inappropriate when there is no reasonable expectation that the patient will improve significantly enough to allow him or her to perceive the benefits of the treatment. But if marginal improvement can be obtained, even in the face of severe remaining deficits, that may be acceptable to some. For patients not capable of autonomous consent, a proxy decision maker is asked to express what that patient would have wanted, which can be fraught with bias and difficult decision making. Although these conversations around quality of life and goals of care are challenging owing to time and resource limitations, it remains very reasonable to expect that they be done in the ED when time allows.
Brain death in the emergency department
Unlike prognostication in early catastrophic brain injury where various prediction models and shared decision making factor significantly, a declaration of death by neurologic criteria is decided by clear-cut legally and morally binding criteria.
The first criteria for brain death determination were created in 1968 with multiple similar, although not identical, guidelines created by various societies since that time, all based fundamentally on the Uniformed Determination of Death Act of 1981. The American Academy of Neurology published brain death criteria for adults in 1995, revised in 2010, and clarified in 2019, which are considered the primary guidelines throughout the United States for death by neurologic criteria, although practices vary widely. Using these criteria for diagnosis, no case of recovery of neurologic function has been observed after determination of brain death.
The diagnosis of brain death requires multiple components including exclusion of confounders, clearance of sedating and paralytic agents, clear causative and irreversible etiology, correlative neuroimaging, and an examination consistent with cessation of whole brain function. Although the mechanics of the examination itself may be accomplished in the acute period, the other components are not as easily achieved. To ensure clearance, waiting at least 5 half-lives for any central nervous system depressants that are present may require days based on pharmacokinetics and pharmacodynamics in critically ill patients. The addition of therapeutic hypothermia after cardiac arrest specifically alters both the pharmacokinetics and pharmacodynamics of sedatives and requires cautious analysis to exclude residual effects that can obfuscate the neurologic examination. Although this factor is less likely to be a confounder in the ED, hypothermia on arrival may have similar effects on drug clearance. Equally important, early neuroimaging can be falsely reassuring, leaving prerequisites for brain death unfulfilled.
Although it may be tempting to declare brain death in the acute phase to spare the patient’s family the agony of waiting, to limit potentially futile interventions, or to preserve intensive care unit capacity during times of scarcity, an early diagnosis of brain death should rarely, if ever, occur in the clinical course of devastating neurologic injury.
The Neurocritical Care Society published recommendations that specifically caution against early prognostication in devastating brain injury, recommending a 72-hour window immediately after the injury during which withdrawal from life-sustaining treatment should be delayed and maximal resuscitative efforts should be pursued to avoid a self-fulfilling prophecy. Although there is no minimally acceptable period of observation before a diagnosis of brain death after a devastating brain injury, it is reasonable to follow similar guidelines for a minimum of a 72-hour window, effectively excluding the diagnosis of brain death in the ED setting.
If brain death were to be inappropriately declared too early, at worst it becomes a self-fulfilling prophecy and at best that patient disproves the diagnosis and the trust between the physician and family is damaged. At the base of the diagnosis is nonmaleficence; give a patient time to declare themselves and ensure criteria are met to ensure you inflict no harm. It is not ethically justified, therefore, to declare brain death early in the acute phase when harm outweighs beneficence.
If death by neurologic criteria is not to be declared in the acute period, is there then a limit to interventions, specifically resuscitation? With advancement of resuscitative technology such as extracorporeal membrane oxygenation, the continuum between life and death has become increasingly complicated. Mired in this controversy is organ donation and the potential for competing interests between resuscitative efforts and potential donation. Ethically, the physician’s responsibility lies in the patient in front of them rather than a potential transplant candidate elsewhere; therefore, all resuscitative efforts should continue when futility is not yet established, as is often the case in the acute phase before death by neurologic criteria. If the patient becomes hemodynamically unstable, Dalle Ave and colleagues concluded in their work that organ preserving cardiopulmonary resuscitation ought to be performed only when specific consent is obtained from the patient’s family.
Airway management in devastating neurologic conditions
The astute physician knows the airways, breathing, and circulation take precedence over diagnosis and prognosis, but in devastating neurologic conditions, prognosis may be key to medical decision making before even the basic airways, breathing, and circulations are considered.
If an evidence-based prognosis offers reasonable hope for improvement, many patients and their surrogates opt for aggressive therapies for maximal benefit. When the prognosis for neurologic recovery is poor; however, it is reasonable to opt for more palliative goals. In the case of airway management, the case is fairly straightforward for a compromised airway in time-limited acute conditions because, once a treatment is instituted, there is reasonable hope for not only improvement, but likely return to baseline in the near future. The situation becomes complex when the timeline is extended or the prognosis is less clear; when offering therapy may equate to prolonged suffering the decision to proceed bears significant moral consequence.
Stroke is a common diagnosis for ethics consults in neurocritical care patients, with a majority of the consults for withdrawal of life support or futility. In large territory infarcts or hemorrhages, ventilatory support may be required to sustain life while the stroke evolves, and potentially longer if the patient survives the acute phase with significant continued deficits. Interventions may be possible during this phase that will prolong quantity with an unknown effect on quality of life, potentially prolonging suffering. Does the beneficence of potential therapy outweigh the potential maleficence of associated complications? The patient’s wishes and prognosis will often tip the scale, although research has shown significant variability in physician prognosis, particularly in the early phases. It is key to strike a balance between careful optimism and nihilism because at one end of the spectrum is the self-fulfilling prophecy of a dismal prognosis and at the other is prolonged suffering without additional gain.
The situation is similar for neuromuscular respiratory failure. A patient with Guillain-Barre syndrome requiring early intubation may eventually wean from mechanical ventilation after weeks or months with unknown residual deficits or he or she may require mechanical ventilation indefinitely. The same is true of acute high thoracic spinal cord injuries; a ventilator may be needed initially and possibly indefinitely, although it is often too early to know in the acute phase of injury. The acute phase will be difficult, but is the suffering worth the unknown end result? Only the patient can truly make that decision with input from the critical care team. Again, prognosis and autonomy are cornerstones of this dilemma with careful avoidance of clinical nihilism in self-fulfilling prophecies.
Consent in critical neurologic patients
All the tenets of medical ethics stand alone, but beneficence and autonomy are linked in that it is reasonable for individuals to want the best for themselves and others. This tenet comes under scrutiny in the form of consent in incapacitated critical neurologic patients. Consent is rooted in respect, dignity, and autonomy with 2 prerequisites: disclosure, capacity, and voluntariness. Ideally, when a patient is incapacitated, the legal proxy would be present and well-informed of the wishes of the patient but in time-sensitive cases, the proxy may be unknown or unreachable in the acute phase. Is it ethically justifiable to make educated guesses based on mRS and prognosis?
A common example is a patient presenting with a stroke—perhaps with aphasia or otherwise altered—who is unable to provide consent for tPA or EVT. Without therapy, the deficits will likely remain, whereas with therapy, there is a potential for significant recovery. Is it ethically justifiable to assume what a patient would want? Per American Heart Sacculation/American Stroke Association Stroke Guidelines (2018), when a patient cannot provide consent and a legally authorized representative is not immediately available, it is justified to proceed with IV thrombosis in an otherwise eligible adult patient with a disabling acute ischemic stroke. Although not explicitly stated, it can be extrapolated that the same holds true of EVT as treatment for the same etiology so long as eligibility criteria are met. To withhold these treatments owing to lack of consent would be to deny the patient their right to therapeutic opportunity.
A more complicated case exists in the encephalopathic patient. There are no clear guidelines for a standard of care in undiagnosed encephalopathy, particularly for diagnostic testing.
When there is no clear prognosis, it is uncertain if any risk is worth the reward, and the “reasonable person” standard becomes nebulous. This situation brings into question whether assumed consent can be decoupled from prognosis. Failure to obtain informed consent is a basis for medical malpractice but, in practice, few medical malpractice suits are predicated on the basis of absent informed consent; rather it serves to undermine cases where it is absent.
Retrospectively, aggressive treatment will be looked upon more favorably in those with a good outcome than those with a poor outcome. For those physicians not fortunate enough to have a crystal ball, the “reasonable person” must remain the guiding standard—would most people, given the known facts, accept the known risks to gain the possible rewards—with any nonurgent procedures delayed until clear informed consent can be properly obtained.
Organ donation in emergency medicine
The respect for autonomy may, at times, require that care be rendered for those who would otherwise be limited from receiving aggressive care and resuscitation. Specifically, when it is known that a patient with a neurologically devastating injury has a first-person consent for organ donation, either through his or her own communication or that of a proxy, it is reasonable to consider that the preservation of their opportunity for donation respects their autonomy, beneficence and nonmaleficence, as well as simultaneously paying deference to justice. In general, discussions related to organ donation are not had with patients and their families in the ED by emergency medicine providers. Instead, those discussions are left for the so-called effective requestors, specially trained staff of the organ procurement organizations or nonphysician staff at the hospitals with unique skills to facilitate the necessary discussions about organ donation, donor status, and consent for donation. Even though discussions regarding organ donation generally do not occur in the ED, the care rendered in the ED may dramatically affect the opportunity for organ donation after admission. Resuscitative efforts aimed at hemodynamic stabilization preserve organ function and the potential for delayed decision making regarding donor suitability. These opportunities are often seen as beneficial to the surviving families and should be preserved whenever possible. Often described in terms of duty, efforts to preserve organ donation possibilities are generally consistent with justice and should be the default action unless specifically understood, a priori, to be contrary to the wishes of the patient expressed via legal documentation or proxy.
The application of traditional pillars of medical ethics to the care of the neurologically injured patient in the emergency setting requires attention to the risks of bias, errors in prognostication, and a lack of evidence-based therapeutics. The unique environmental conditions of emergency medicine require rapid assessment and application of nuanced ethical principles, often under the shroud of stress and emotional shock. It is precisely these challenges that necessitate an a priori understanding of the limitations of both prognosis and autonomy in many of these situations. Generally speaking, when moral conflict arises in the emergency setting, erring on the side of preserving life and allowing time to observe the trajectory of prognosis is ethically sound and may eliminate moral distress in caregivers, patients, and surrogates.
Clinics care points
Ultra-early prognosis in neurologic injury is fraught with bias, error and a lack of compelling data. Even patient with an initial Glasgow Coma Score of 3 after blunt trauma may have the potential to recover.
Neurologic injury renders traditional assessments of autonomy impossible in many situations. Care must be taken when using proxy expressions of patient’s wishes after neurologic injury.
Many interventions in neurologic injury carry a risk of rare but devastating consequences; understanding the belief system of the patient is crucial to know if these risks are acceptable.
Scarcity may shift the standard of care; be aware of resource limitations that may affect patient care.
The authors attest that they have no relevant disclosures for this work. Dr J. Bonomo has served as a speaker and consultant for the Portola and Genentech companies.