Background on Refusal of Treatment

The “right to die” litigation that dominated American healthcare jurisprudence in the last three decades of the twentieth century, culminating in the Supreme Court decisions in Cruzan,¹

Glucksberg,² and Quill,³ confirmed the almost unqualified right of competent patients to refuse any and all medical interventions, for any reason or no reason, even when those interventions may be absolutely essential to preserve life. Although the courts acknowledged certain “countervailing” interests of the states that must be taken into account when patients directly, or indirectly through designated surrogates, refuse “medically indicated” treatment, only rarely have those interests been deemed of sufficient weight to override the clearly articulated wishes of patients.⁴

The hospital policy considered in this installment of “Ethics Committees at Work” constitutes an effort to institutionalize this well-recognized constitutional, common law and, in some states, statutory right to refuse treatment. Treatment characterized as “life-sustaining” usually refers to such medical interventions as mechanical ventilation and artificial nutrition and hydration. However, depending on the patient's diagnosis and the extent of disease progression, treatment might also include medications to manage such conditions as congestive heart failure and chronic obstructive pulmonary disease, or renal dialysis for patients with renal failure. The institutional policy addresses situations in which treatment is “forgone,” which might conceivably encompass both withdrawing interventions currently provided and withholding measures that might otherwise be indicated. An example of the latter would be the entry of a “do not resuscitate” order and hence the withholding of cardiopulmonary resuscitation in the event of a cardiac arrest. In the case of patients with advanced malignancy, such procedures as chemotherapy, radiation therapy, and in some instances surgery may be considered life-prolonging rather than curative.

There is a general presumption that forgoing life-sustaining treatment relates only to disease-directed (curative) or life-extending interventions and not those intended to make the patient comfortable and relieve physical or mental distress (palliative). This presumption is based on a reasonable belief that, whereas many patients may no longer wish to prolong what has come to be recognized as their inevitable death, one rarely encounters a patient who does not seek the relief of suffering. Hence the policy under consideration establishes a basic premise that “supportive care will be provided.” The policy goes on to provide, however, that this presumption may be rebutted by a statement to the contrary from “the patient or a surrogate.” For purposes of discussion and analysis, it may be helpful to consider separately patients with and without decisional capacity.⁵

Patients with Decisional Capacity

When a patient possesses decisional capacity, the clinical and nonclinical implications of forgoing treatment can be fully explored in discussions between physician and patient. Curative and palliative measures can be described and distinguished as to their nature and purpose. If a patient who clearly has decisional capacity declines measures that would be provided solely to maintain comfort, it is very important to be clear about the basis for that refusal. Initially, the physician should seek to verify that the patient understands both the nature and the purpose of the measures, as well as the consequences of not providing them. The patient's refusal may be based on the mistaken assumption that they are life-prolonging or that their side effects may impose burdens that exceed whatever benefits might be anticipated. Such discussions should not focus excessively on particular comfort measures (means) but rather on the patient's goals (ends). Quality care at the end of life is care that is consistent with and promotes the patient's goals and values.

Because comfort measures for dying patients often include the administration of opioid analgesics to control pain and relieve other distressing symptoms, it is critically important that patients, and family members who are close to and may influence the patient, be educated about these medications. Opiophobia is not restricted to healthcare professionals.⁶

Quite understandably, many lay persons do not comprehend the distinction between addiction and physiological dependence on opioid analgesics to relieve severe, persistent pain. Patients and family members may also have exaggerated concerns about severe, unmanageable side effects from opioids. Pain management remains one of the vast wastelands of informed consent because a lack of physician knowledge quite naturally yields the perpetuation of myths, misinformation, and groundless fears among patients and families.⁷

If the patient's refusal does not appear to be based on ignorance or unfounded fears (his own or those of close family members) concerning opioid analgesics, then psychological, social, cultural, or religious factors should be taken into account. Good end-of-life care often requires an interdisciplinary team effort, and members of that team, or consultants readily available to it, should ideally include social workers, psychologists and/or psychiatrists, and chaplains or pastoral counselors. The patient may believe, erroneously, that particular cultural norms or religious doctrines demand or promote the experience of pain or suffering as an essential element of the dying process. Given the general disinclination of people of many different social and cultural backgrounds to discuss death and dying, one should not assume that discussions of this nature have previously taken place and hence inform the patient's perspective. If the patient has any association with a faith community, it is important to encourage that a leader of that community (e.g., minister, priest, rabbi, imam) provide spiritual guidance and counsel if the patient agrees.

In the unlikely event that, after exploring all of these factors, the patient continues to decline comfort measures, then the patient's wishes must be respected, just as they must when curative measures are declined. In some instances, healthcare professionals may seek to be relieved of the responsibility to provide further care to the patient because they do not wish to participate in what they consider to be the infliction or the tolerance of unnecessary suffering.

Patients without Decisional Capacity

The ethical standards for surrogate decisionmaking reflect a primary objective of respecting the wishes, goals, and values of the patient who has lost decisional capacity. If a patient has provided clear written directives that indicate the circumstances under which she would or would not wish certain types of treatment to be provided, the responsibility of the surrogate is to make decisions consistent with such directives. Regrettably, such comprehensive and unambiguous directives rarely exist.⁸

At best, there may be one of two common types of advance directive. The first, a living will, provides that once a patient has lost decisional capacity and has been determined by two physicians to be in a terminal, irreversible condition, life-sustaining measures should be withdrawn. The second, a durable power of attorney for healthcare, designates a specific individual as the patient's proxy and confers on that person the authority to make medical decisions on behalf of the patient. Often a durable power of attorney contains no specific guidance on how such decisions, in the face of a terminal or serious, irreversible condition, are to be made. Sometimes the term “substituted judgment” has been used to describe the basis on which such a proxy is to make decisions that most closely approximate how the patient would make them if he had not lost decisional capacity.⁹ Substituted judgment presupposes some minimal level of knowledge about the patient as a unique individual, his values, and how he might make decisions in similar circumstances. It is a subjective standard, but the subjective view is that of the patient, not the proxy.

When the patient has provided no written directives, the ethical standard for surrogate decisionmaking is that of the patient's best interests.¹⁰

At least in theory, the best-interests standard is what a proxy must resort to for guidance when no information is available on which to base a substituted judgment. Traditionally, it has been considered an objective standard, based on what a reasonable person in the patient's circumstances might be expected to prefer. However, some recent state legislation has sought to impart a quasi-subjective tone to best-interests determinations by proxies. Consider, for example, the following language from the California statute regulating healthcare decisions by a conservator:

One particularly salutary feature of such a provision is the recognition that best interests cannot, in any intrinsic sense, be a generic concept. Two patients with very similar diagnoses and prognoses might still have radically different perspectives on what, clinically, is in their best interests. One might view a high-risk surgical procedure or experimental treatment as in her best interests, whereas another would not. Those acting as surrogates for patients who lack decisional capacity have an ethical obligation to make a good-faith effort to apply the “best interests” criterion from that particular patient's perspective.

We come, once again, to the situation addressed by the policy in question. The decision to discontinue or not provide life-sustaining treatment has been made by the appropriate proxy. Presumably that decision is consistent with the patient's directive, values and priorities, or best interests and is not inconsistent with the diagnosis, prognosis, and current clinical status of the patient. The critical issue is whether it would ever be appropriate for a surrogate to refuse comfort measures on behalf of a patient. The policy rightly posits that comfort measures are the standard of care when life support is withdrawn. There is a developing literature on how to manage patients as they are taken off of a ventilator or artificial nutrition and hydration are discontinued.¹²

Carefully managed, the suffering of all involved, but particularly the patient, can be prevented or alleviated. However, there are also data suggesting that often “best practices” elude us and dying in the intensive care unit becomes an ordeal, especially for patients and their families.¹³

Just as I earlier conceded that there might well be circumstances, however rare, when a patient with decisional capacity declines comfort care when life support is discontinued, so too there might be circumstances in which a surrogate is authentically respecting the values of a patient by declining comfort measures. Nevertheless, an institutional policy that simply gives a surrogate carte blanche to decline measures that are intended only to prevent patient suffering seems deeply flawed. Patients who prefer to receive no comfort measures whatsoever are extremely rare. More common are those who wish to accept a certain level of discomfort rather than be completely deprived of an opportunity to be present with those closest to them in the final days or hours of life. However, patients whose conditions afford them no real opportunity to experience the presence of others may still have the capacity to feel pain and suffering. In such instances, there is no meaningful trade-off available between comfort and engagement.

The healthcare team can responsibly challenge surrogates who are unable to provide some independent, credible basis for a judgment that a patient would not wish to receive appropriate comfort measures. In the absence of reasonably persuasive evidence to support the surrogate's decision to decline comfort measures, the question can legitimately be raised whether the surrogate is acting in the patient's best interests. Such situations warrant consideration by the institutional ethics committee. Ultimately, judicial review of the surrogate's position may well be required, and the institutional policy should recognize and provide for it.