Advance directives and substituted judgment

The modern concept of an advance directive for healthcare dates from Luis Kutler’s seminal 1969 law view article, ‘Due Process of Euthanasia: The Living Will, A Proposal’ in which he advocated for the right an individual to specify in advance, ‘while fully in control of his faculties and his ability to express himself … to what extent he would consent to treatment’. Kutler proposed a ‘living will’ through which a patient could ‘provide that if the individual’s bodily state becomes completely vegetative and it is certain that he cannot regain his mental and physical capacities, medical treatment shall cease’.3 This approach sought to further patient autonomy, effectuating what the patient’s actual preferences were, over the then dominant approach in which ‘constructive consent [was] presumed and the doctor [was] required to exercise reasonable care in applying ordinary means to preserve the patient’s life’.3 This type of advance directive creates a ‘Ulysses pact’—a term derived from the Homeric tale of Odysseus and the Sirens—in which an individual irrevocably binds himself to a decision in a future context.4 The concept of the living will has since been expanded widely to cover nearly all medical decisions unrelated to psychiatric illnesses. In addition, patients in the USA now have the right to appoint proxies, third-party decision makers who can steer care for incapacitated persons based on the earlier expressed wishes of those individuals. Every US state has now established provisions for patients to execute binding advance directives, and clearly expressed wishes are largely enforced across jurisdictional lines.5 The importance of advance directives was enshrined in American federal law with the passage of the Patient Self-Determination Act of 1990, which requires healthcare facilities to provide patients with information on these directives on admission.6 In addition, nearly every US state has established some form of surrogacy process through which, in the absence of a formal advance directive, a close third-party can steer care in accordance with an incapacitated patient’s previously expressed wishes.7

The doctrine of ‘substituted judgment’ has informed decisions related to estate administration since the 19th century and became widely accepted in American medical care following legal decisions in Strunk v. Strunk (1969), Hart v. Brown (1972) and In re Quinlan (1976).8 9 In the Quinlan case, addressing whether the parents of Karen Ann Quinlan, a woman in a persistent vegetative state, had authority to remove their daughter from a ventilator, Chief Justice Hughes of the New Jersey Supreme Court explained that while ‘Karen’s independent right of choice … would ordinarily be based on her competency to assert it’, that right ‘should not be discarded solely on the basis that her condition prevents her conscious exercise of the choice’. Rather, the court permitted ‘the guardian and family of Karen to render their best judgment … as to whether she would exercise it in these circumstances’.9 That approach has been widely, although not uniformly, adopted by states in end-of-life cases. While some jurisdictions require clear and convincing evidence of a patient’s wishes, others accept a preponderance of the evidence, including general past statement and actions indicative of a patient’s broader views on healthcare and life prolongation. More generally, ‘substituted judgment’ has become the standard by which third-party decision makers render guidance for the medical care of incapacitated adults.

Substituted judgment is not without itself critics.10 Frequent objections include the evolution of patient attitudes over time, the non-concordance of patient attitudes with both physicians and chosen third-party decision-makers, and some evidence that patients do not ‘want their prior wishes to be the sole basis for decisions made on their behalf’.10 A particular concern is the phenomenon of ‘bargaining down’, in which patients finding themselves in a compromised state may prove willing to accept life at a level of impairment that they previously had not thought they would wish to tolerate.11 These concerns notwithstanding, substituted judgment remains the approach to medical decision-making believed to best vindicate the right to autonomy of incapacitated patients. That is why it has been nearly uniformly embraced in a wide range of clinical circumstances. Decision-making for patients with mental illness remains an outlying exception.

Mental illness

Care of the mentally ill—historically the subject of criminal jurisdiction, and until recently, shrouded in secrecy and stigma—has increasingly come to be viewed by physicians, policy makers and the broader public as a subset of general medical care.12 The adoption of the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III) (1980), which based classification on symptomatic presentation rather than psychodynamic causes, and the acceptance of the biopsychosocial model of psychiatric disease, has lowered historic barriers between mental healthcare and general medicine. A series of American court cases including Wyatt v. Stickney (1970) and O’Connor v. Donaldson (1975) guaranteed psychiatric patient’s rights similar in nature to those of the medically ill. More recently, psychiatric advance directives (PADs) have gained tractions as a way for those with mental illness to specify their care wishes for future episodes in which they lose decisional capacity. Many jurisdictions do not recognised the authority of PADs and none in the USA do so in the cases of patients seeking to end their own lives either via care withdrawal or active suicide. However, the Joint Commission on the Accreditation of Healthcare Organizations now requires psychiatric hospitals to ask admitted patients if they have PADs and at least one federal court has enforced a PAD for a patient who refused medication.13

A glaring exception to the increasing uniformity between medical and psychiatric practice has been the resistance to adopting the standards of non-psychiatric medicine for PMP/LQL patients with depressive disorders who seek to terminate medical care or end their own lives. Such cases arise frequently for consult-liaison psychiatrists.14 Depression is extremely common among patients with advanced cancer, for instance, despite the widespread adoption of palliative care.15 Reports suggest that the incidence of depression in patients with terminal illness may run as high as 77%.16 The desire for death is less common, but still significant, and often tied to negative affective state.17 Whether such a reaction is pathological, if the condition is so frequent, or merely a normal response to these specific adverse conditions, requires a subjective, values-based determination. What matters from a policy perspective is that refusal of care in these cases is treated in a fundamentally different manner from similar patients refusing care who are not psychiatrically ill.

The primary justification for such different approaches is that depressed patients may be restorable and that psychiatrists can relieve their depression so that they can ultimately make informed decisions regarding care. Constrained by current law and ethical cannons, physicians do not investigate whether, if these patients were not depressed, they would have wanted to be restored or whether they would have wanted care. That differs greatly from the approach used with medical patients. A patient who has informed family members that he does not wish to be dialysed if his kidneys fail, and presents without decisional capacity secondary to uraemia, will not be forcibly dialysed. It does not matter that he can be restored to decisional capacity through dialysis. What is dispositive is the substituted judgment of the patient’s third-party decision-maker based on the patient’s prior known or inferred wishes. In contrast, if a third-party decision-maker states that a patient would not have wanted to be treated for depression, but would have preferred withdrawal of care to mental health treatment, providers over-ride that preference. No American jurisdiction permits a third-party decision-maker to refuse hospitalisation for a patient who is suicidal under such circumstances, or to authorise aid in dying, even though that might have been the rational choice of the patient if he possessed capacity.

The palliative care movement has succeeded in mitigating the most extreme excesses of the paternalist approach. For example, patients with C-3/C-4 spinal fractures are no longer forced to remain on ventilators indefinitely while psychiatrists adjust their antidepressants. However, these incremental improvements result from psychiatrists reassessing the patient’s capacity in light of ongoing circumstances, or some collective consensus between psychiatry and palliative care on a reasonable course of action that meets the patient’s best interests, not on the substituted judgment of third-party decision-makers.

Right to die

The right of the adult patient with decisional capacity to refuse life-saving medical care has become an established principle of American law and western medical ethics. Many jurisdictions in both the USA and Europe have also sanctioned aid-in-dying for the terminally ill. Philosophers and ethicists increasingly distinguish between active suicides that are rational and those that are induced by mental illness.18 19 Yet maybe because one of the core tenets of allopathic psychiatry is the prevention of self-harm, psychiatrists and ethicists are generally unwilling to acknowledge that cases exist in which depression-induced desires for death and rational desires for death may overlap. In other words, a PMP/LQL depressed patient may voice a desire for death under circumstances in which they would prefer death even if not depressed. Such a wish could reflect a longstanding attitude toward a proposed medical intervention, a desire to control one’s means and time of dying, or a preference for death over an LQL. Complicating matters, such a desire could also reflect a preference for a shorter life over one constrained by psychiatric hospitalisation or the side effects of psychiatric medication. A healthy person might well instruct a proxy, ‘If I ever end up critically ill and depressed, please let me die rather than impose unwanted psychiatric care upon me, even if I have some prognosis for recovery’. Unfortunately, psychiatrists do not ask whether such a statement has occurred because that answer currently receives no weight in medical decision-making.

A substituted judgment approach for PMP/LQL depressed patients with diminished capacity may also result in cases, although more rarely, in which the patient demands care over objections from his proxy or surrogate. What is to be done when such a patient does want treatment for either his underlying medical condition and/or his psychiatric illness, but whose third-party decision maker reports that, based on previous statements, he would not have wanted such care. Should substituted judgment still prevail? Under a model that values autonomy and honours Ulysses pacts, the logical answer is yes. At the same time, such an approach in reverse cases like this example will likely prove less palatable to clinicians, at least under some circumstances. The superficial difference is that in the original case, the PMP/LQL depressed patient with diminished capacity is expressing goals consistent with those previously indicated during periods of full capacity, while the later case the PMP/LQL depressed patient with diminished capacity is expressing goals inconsistent with those previously indicated during periods of full capacity. Maybe the better explanation for the viscerally different reaction that many providers will display is that diminished capacity is not synonymous with the complete absence of capacity, although the practical nature of hospital decision-making often requires the imposition of this false dichotomy. In the second case, psychiatrists are acknowledging what they know as human beings: namely, that even in the absence of formal decisional capacity, ethics demand some leeway for the patient with diminished capacity to reverse course. This flexibility is no different than that afforded to other medical patients. Why should a psychiatric diagnoses require rigid thinking on this matter when the absence of one does not?