Who Should Make Medical Decisions When a Patient Lacks an Advance Directive?

Case 1

In the spring of 2019, Jennifer Friedlin’s father had a stroke, leaving him incapable of making goals-of-care decisions. ² Over the next two months, Jennifer battled with a rehab center in New York over whether to deactivate her father’s pacemaker. Jennifer believed that her father would want the pacemaker turned off and that it was merely prolonging his death. The medical team refused to turn off the pacemaker, as they felt this would cause Jennifer’s father to die sooner.

Case 2

In the winter of 2019, Shelina Begum and Mohammed Raqeeb’s five-year-old daughter suffered a hemorrhagic stroke in her brainstem after rupture of an arteriovenous malformation. ³ During a brain death evaluation, she demonstrated the ability to breathe spontaneously. This categorized her as being alive and in a vegetative state. The Royal London Hospital, feeling that further medical treatment would not improve her condition or be in her best interest, planned to withdraw life-sustaining treatment. Her parents, however, believed they saw signs of improvement and objected to this plan based on their religious beliefs. Her parents sought and found a hospital in Italy that was willing to continue treatment. This case ultimately went to court where it was determined that she could be transferred to the Italian hospital. A key consideration in the court’s decision was her parents’ description that she followed Islamic practices, as evidenced by a video of her encouraging her brother to pray with her, and that, based on her beliefs, she would want to have her life prolonged despite her current condition.

Discussion

In both of these cases, there is no objectively correct answer about what is best for the patient, so the key question is, “Who should make medical decisions when a patient lacks an advance directive?”

Historically, clinicians were responsible for making all medical decisions for their patients. Doctors created a plan and prescribed treatments, and their authority was not questioned. A central assumption of this framework was that clinicians, because of their medical knowledge, knew what was in the best interest of their patients. The American Medical Association endorsed this paradigm in the Code of Medical Ethics in 1847:

The obedience of a patient to the prescriptions of his physician should be prompt and implicit. He should never permit his own crude opinions as to their fitness, to influence his attention to them. A failure in one particular may render an otherwise judicious treatment dangerous, and even fatal. ⁴

Achieving a balance between paternalism and autonomy can be challenging when making decisions with a surrogate on behalf of a patient. ⁶ In some cases, clinicians may feel that a surrogate wants something inappropriate and inconsistent with what they believe is good and right for the patient. This can create competing ethical obligations between respecting patient autonomy, promoting patient wellbeing (beneficence), and avoiding harm. These principles must all be followed, but prioritization may vary based on the situation, as well as social and cultural context, recognizing that the balance of power between clinicians and surrogates/patients differs across countries; even in the Western World this varies, with surrogate/patient views often weighing more in the United States as compared to Europe.

In the case of Jennifer Friedlin’s father, Jennifer was in the best position to know what her father would have wanted. This principle is termed “substituted judgment,” and has been proposed as a method of promoting autonomy. However, one study found that while patients believe their family and doctors were knowledgeable about their wishes, neither could routinely predict patient wishes on scenario testing, with agreement ranging from 59-88%. ⁷ Although the clinicians in this case best understood the potential side effects of turning off a pacemaker, including dizziness, syncope, and shortness of breath due to inadequate cardiac output, guidelines on withdrawal of treatment in this scenario provide straightforward advice. The Heart Rhythm Society published a consensus statement which indicates that it is both ethically and legally permissible to honor a patient or surrogate request to turn off a cardiac device. ⁸ The American Medical Association similarly stated, “A patient’s surrogate may decline an intervention or ask that an intervention be stopped even when that decision is expected to lead to death, and regardless of whether or not the individual is terminally ill.” ⁹ The first step to resolve conflicts regarding withdrawal of treatment is to hold an open conversation seeking consensual agreement between the surrogate/patient and clinicians. If disagreement persists, consultation with an ethicist and/or palliative care specialists can be of great benefit. Ultimately, however, surrogates should be empowered to refuse or withdraw treatment altogether and thus define “beneficence” in light of the preferences of their family member.

In contrast to withdrawal of treatment, there is little consensus among ethicists about continuation of treatment for patients in a vegetative state. In Shelina and Mohammed’s case, they had the best understanding of their daughter’s social and cultural values. In Islam, a life with a severe disability is held, in both domestic and international law, to be of equal value to all other lives. ³ On the contrary, the perspective of the clinicians in this case was that life-sustaining treatment should be targeted towards the sanctity of life, and that in this case, continuation of support was futile. Some regions have developed policies to address disagreements about treatment futility. For example, in Texas, an ethics board is consulted if there is a conflict between the medical team and family about potentially futile interventions. If the ethics consultant agrees with the medical team that a treatment is futile, but the family insists on continuation of treatment, the hospital is required to facilitate transfer to another institution willing to provide treatment. ¹⁰ In addition to the consideration of whether treatment is futile, is the importance of avoiding harm. The clinicians in this case best understood the potential consequences of prolonging treatment including the development of a range of complications. Furthermore, there was no way to know with certainty if Shelina and Mohammed’s daughter was in pain or distress. Another important consideration in this situation is resource availability and ICU costs involved in sustaining the life of a patient in a vegetative state. Ultimately, culture-specific values need to be incorporated into surrogate decision-making, but clinicians need to recognize their obligation to beneficence when considering life-prolonging treatment.

Of additional importance in these cases is the potential for uncertainty in neuroprognostication. While brain death is a finite determination, prognostication in neurological disease is challenging and laden with uncertainty. There is no dispute that Shelina and Mohammed’s daughter had irreversible brain damage. However, she did not meet the criteria for brain death, and the natural history and prognosis of her condition is not well defined. Predictors of stroke outcome include degree of initial deficits, age, infarct size and location, ischemic stroke mechanism, and medical comorbidities. ¹¹ A few models are available for predicting global outcome from ischemic stroke, including the ASTRAL score, iScore, and PLAN score. However, none are established as valid nor widely used in clinical practice. The ICH score, which is used for intracerebral hemorrhage, is more accepted and used clinically to predict mortality and neurological outcome. The AHA/ASA guidelines, however, warn against using cohort data alone to predict individual patient outcomes. ¹² Improved neuroprognostication is hopefully on the horizon. Of course, even if neuroprognostication were perfect, it would continue to be important for clinicians to include a patient’s surrogates in goals-of-care discussions and the decision-making process.

These cases highlight the complexity and nuances of decision-making for, and management of, patients with neurological diseases. When considering withdrawal of care, autonomy should be valued over other ethical principles. However, when prolonging care, the focus should be on beneficence to minimize pain and distress. Neurologists must educate surrogates and correct any misconceptions in conjunction with medical ethicists and/or palliative care specialists to ensure decision-making is always patient-centered. Relatedly, these cases also highlight the importance of establishing formal advance directives, particularly for patients at risk for acute neurological conditions, to ensure wishes are clear, mitigate the risk for disagreements and improve quality of care.