Case Report

Top Case Report Discussion Summary Author & Article Info References

A.B. is a 70-year-old widow with end-stage renal failure due to analgesic nephropathy who has been receiving hemodialysis for 7 years. She lives alone in an apartment and receives dialysis at a limited care facility where she is responsible for many of the procedures associated with her dialysis. Her closest relative is a sister, to whom she is very attached.

Throughout 1993, A.B. had a series of complications related to her fistula that necessitated several revisions and reconstructions. Her illness was also complicated by recurrent falls and chest infections, and she became very wasted and frail. Altogether in 1993, she spent 188 days as a hospital in-patient. During this period, she requested withdrawal of dialysis. However, because she was believed to be clinically depressed at the time, antidepressant therapy was instituted. Her mood improved, and she decided to continue with dialysis.

Early in 1994, a large transitional cell carcinoma was found that arose from her bladder and infiltrated much of her pelvis. Sinuses formed between her bladder and her vagina and through to a large sacral pressure sore. Because of her pelvic pain and vaginal bleeding, she required radiotherapy, recurrent blood transfusions, and large doses of oral morphine. She also developed a severe, purulent corneal abscess. A.B. was eventually discharged from the hospital with maximum community support but soon required readmission because of a series of falls. She was so wasted that she appeared cadaveric and was virtually bedridden. She was covered with bruises and lacerations and became intermittently confused. A.B. was nonetheless assessed by the liaison psychiatrist who considered her to be legally competent.

Throughout 1994, the option of discontinuing dialysis was repeatedly raised with her by the medical team because of concerns that continued dialysis was prolonging her suffering unnecessarily. She repeatedly stated that she wanted to continue dialysis because she did not want to die. She refused all offers of counseling and pastoral care and refused to discuss this issue further with her sister or with staff.

Caring for A.B. often resulted in confrontations between staff members about the appropriateness of her care, and at times these issues even threatened the overall morale of the unit. Her management raised difficult ethical issues, because respect for her autonomy appeared to clash with the duties of beneficence (providing care) and nonmaleficence (doing no harm). To many of those involved in her care, respecting her wishes by continuing dialysis appeared to be both expensive and futile.

Discussion

Top Case Report Discussion Summary Author & Article Info References

Death and Renal Dialysis

Cardiac disease is the most common cause of death among patients who receive dialysis. [1]. This reflects a marked association between long-term dialysis and accelerated atherosclerosis [2].

The second most common cause of death is discontinuation of dialysis. This is responsible for 8.9% to 22% of dialysis-related deaths [3, 4]. Most of the patients who discontinue dialysis include those older than 65 years, those with diabetes, those living in nursing homes, and those who have been receiving dialysis for long periods [5]. The average length of time from discontinuation of dialysis to death is 8 days. In most instances, withdrawal of dialysis occurs at the request of patients or their families [4]. It is a striking aspect of dialysis that by its very nature, it enables patients to schedule their own death.

Approximately 4% of patients receiving dialysis die of cancer [6].

Guidelines for the Withdrawal of Treatment

Some of the most difficult decisions in medicine are concerned with initiating and withdrawing treatment. Several studies have examined the ways in which physicians make these decisions with regard to dialysis. The two main criteria that have been proposed are the patient's wishes and the likely medical benefits to the patient [7].

The wishes of patients are determined by the patients themselves if they are mentally competent or, if they are not mentally competent, by advance directives and the use of surrogate decision makers.

Unfortunately, estimates of the medical benefits of dialysis are difficult to quantify. They are for the most part determined by the patient's eligibility for dialysis, which in turn is determined by the patient's meeting certain minimum standards.

In a paper prepared by the Institute of Medicine (United States), it was recommended "... that patients, nephrology clinicians and bioethicists develop guidelines for evaluating patients for whom the burdens of treatment may outweigh the benefits, that such guidelines should be flexible and be followed with discretion by physicians in individual cases, and that clinicians discuss with all patients with ESR their wishes regarding dialysis, cardiopulmonary resuscitation, and other life-sustaining treatments and encourage the formulation of advance directives" [8].

Such guidelines have been proposed by Lowance [9], who suggests that patients should be counseled to consider not accepting or not continuing dialysis when any of the following circumstances exist:

1. The patient is physiologically or chronologically old, with an estimated life expectancy of less than 2 years.

2. The patient has dementia or impending dementia.

3. The patient's life expectancy is less than 2 years because of the existence of coexisting disease such as advanced diabetes, vascular disease, heart disease, the acquired immunodeficiency syndrome, cancer, or other systemic illness.

4. The patient has a history of substance abuse.

5. The patient has a coexisting illness that will produce intractable pain or suffering [9].

Such guidelines are usually justified on utilitarian grounds of maximizing efficiency. For example, Kilner [10] states: "Anyone who will benefit medically from treatment, in the medical judgment of the health care giver, is given treatment.... For the majority of directors, this criteria provides one of the most satisfactory and effective bases for resisting unreasonable demands for treatments that will provide no medical benefit whatsoever but will waste valuable resources." Unfortunately, these guidelines tend to be "quasi-medical" rather than based on strictly medical grounds. Concepts such as physiologic age, impending dementia, and life expectancy are not well defined clinically. Also, judgments about the significance of such concepts tend to be based on the values, beliefs, and biases of those who make them. For example, in some societies, groups such as the elderly, substance abusers, and those with dementia are considered to be of low social value and utility. The moral and clinical validity of guidelines that stress the purely "scientific" basis of clinical decision making is questionable also because these guidelines deny that the patient's own judgments regarding his or her life and death have a role in the decision. Such guidelines may in fact provide a convenient means of excluding patients who are judged to have poor compliance or to be psychologically unstable.

At best, guidelines such as these provide a starting point for discussion between physicians and patients regarding the aims and goals of therapy. At worst, they "medicalize" decisions that are in essence deeply personal moral choices. Either way, if "medical benefits" criteria are judged to be of greater importance than "patient choice" criteria, it is clear that a patient such as A.B. would not be offered continued dialysis.

Overriding Patient Autonomy

Patient autonomy is a cornerstone of decision making in modern medical practice. It is now clearly established in medicine, ethics, and law that a competent patient has the right to choose or refuse medical treatment. What is not clearly established, however, is the extent to which society and health care providers are obligated to accede to these choices. There is no difficulty if a patient's request coincides with clinical judgment, but the issue becomes problematic when a patient's request conflicts with medical opinion. What then are the justifiable limits of patient autonomy?

When resources are scarce, some form of rationing may override the patient's wishes. However, although concerns about resource allocation are recognized at the administrative level, they are not always apparent at the patient's bedside. This may be because of doctors' unwillingness to apply economic principles to patients under their personal care, or it may simply reflect an unwillingness on the part of physicians to recognize the reality of scarce resources.

Autonomy may also be restricted if the physician believes that a patient would be harmed by the requested treatment. In such instances, it is the physician's estimate of harm rather than the patient's that forms the basis for treatment decisions. For example, a doctor, believing that he or she is acting in a patient's best interests, need not administer morphine to an addicted patient, even if the patient believes that it would be beneficial.

It has also been argued that one need not accede to a patient's demands for treatment if the suggested treatment is considered futile. Unfortunately, providing an adequate definition of a "futile" medical intervention has proved to be extraordinarily difficult. Futile therapy has been variously defined as therapy that is useless or ineffective, that cannot achieve the patient's wishes or goals, that does not offer a reasonable chance of survival, that fails to achieve a physiologic effect, or that fails to offer a minimum quality of life or medical benefit [11-14]. Thus, a treatment that does not reduce distress or disability, shorten the course of the illness, or reduce the threat to (meaningful) life might arguably be considered futile.

Finally, a patient's wishes may be discounted if the patient is mentally incompetent. In such instances, decisions about dialysis are made by the physician in consultation with the patient's surrogate. In the case of A.B., this approach had been used when she was profoundly depressed in order to continue dialysis against her wishes.

It could be argued that the very fact that A.B. wished to live in such terrible circumstances proved her incompetence—that such a decision is inherently an incompetent one. The basis for such an argument is that the ability to deliberate rationally is regarded as one of the components of competence [15]; if a decision reached by the patient is judged to be irrational, then it is believed that the patient himself or herself must be irrational. Perhaps the best analogy is the treatment of patients who attempt suicide: In the emergency room, all such patients are regarded as incompetent and are therefore resuscitated and given psychiatric care.

However, there are significant problems with this line of reasoning. To assess competence, one needs to determine whether patients are able to understand their condition, to reflect on it in accordance with their own values and beliefs, and to communicate their decisions to others [15, 16]. Because competent patients need only be sufficiently rational to understand and reflect on their conditions in accordance with their own belief systems, their decisions may not necessarily appear rational to others. Unfortunately, no clear consensus exists about how free and voluntary a decision must be or how much irrationality may be displayed before a patient is considered incompetent. It seems that rationality in medical decision making is not a quality that is present or absent but rather exists along a continuum. Decisions made by competent patients in different circumstances and at different times may vary in the degree of rationality. Under the stress of severe illness, for example, patients may experience a reversible loss or decline in cognitive function [17]. Furthermore, unrealistic expectations, inadequate understanding, misconceptions, exaggerated fears, and magical hopes may distort the choices made by doctors as well as by patients [18].

Summary

Top Case Report Discussion Summary Author & Article Info References

Several criteria may be used by renal physicians to make decisions about dialysis acceptance and withdrawal, and there are arguments for overriding patient's wishes when they conflict with medical opinion. However, it is clear that in the case of A.B., none provides a satisfactory directive for clinical and ethical decision making.

For A.B., continuing dialysis could not be considered futile by any of the accepted definitions. Dialysis was consistent with her wishes, was physiologically effective, and provided her with a quality of life that she obviously found sufficient in that she wanted to continue to live. Continuing dialysis was not "useless" or "ineffective," because it provided her with her only chance of survival. We should, however, emphasize that although the care of A.B. could not clearly be considered futile according to any of the current formal definitions of futility, caring for her did invoke a deep sense of futility in some of her caregivers.

Many patients who receive dialysis ask that treatment be discontinued if therapy becomes excessively burdensome [19]. The fact that A.B. had been receiving dialysis for 7 years suggests that she could make a reasonable assessment of the benefits and burdens of continuing dialysis. Indeed, despite her terrible suffering, A.B. appeared to want life at any cost. Finally, although A.B. was known to have a history of depression, the liaison psychiatrist determined that she no longer had significant depression and was now mentally competent. Under these circumstances, we could not convince ourselves that withdrawing dialysis against her wishes was in her best interest.

Conclusions

The management of A.B. illustrates an area of medical decision making that is fraught with difficulties. In the care of such a patient, there is an inherent conflict between the need to respect the patient's autonomy and freedom of choice and the role of physicians as providers of expert and compassionate care. In the final analysis, decisions to provide or withdraw treatment are not merely philosophical, technical, or economic but are extremely complex and intense personal decisions. In these instances, decision making is more than merely a summation of ethical principles: Although the concepts of autonomy and beneficence can be considered individually, they cannot be easily measured against each other.

We do not believe that it is possible to be dogmatic about these decisions, but, where all other considerations appear to be in balance, the benefit of the doubt should in most instances go to the patient. The physician's role in these circumstances remains to give support, to be compassionate, and to continue to try to understand patients' perception of their own needs [20].

Postscript

As her pelvic pain increased, A.B. requested stronger pain relief, and a subcutaneous morphine infusion was instituted. Her pain became well controlled, but she became increasingly obtunded. Given the deterioration in her condition, a decision was made by the health care team to discontinue dialysis. She died 20 days later.