Chapter 102. Nutrition and Medical Ethics: The Interplay of Medical Decisions, Patients' Rights, and the Judicial System

“Principalism”

The traditional Hippocratic system of professional ethics was increasingly questioned in the United States in the mid-1960s in conjunction with many professional and

societal changes that included depersonalization and fragmentation in many physician-patient relationships and challenges to governmental authority with the rise in

the civil rights, antiwar, feminist, and consumer rights movements in a multiethnic pluralistic democratic society. As physician-patient relationships came under

increasing scrutiny with claims for more patient rights, a series of principles of medical ethics were advanced (designated by some as “principalism”). These were

effectively set out by Beauchamp and Childress in the first edition of Principles of Biomedical Ethics (7) and its successors. Its first two principles (nonmaleficence and

beneficence) were similar to those in the Hippocratic ethic; the third concept, patient autonomy, contravened paternalism; while the fourth, justice (an area of diverse

views often involving elements of fairness and entitlement for individuals), was completely at odds with the old tradition.

The term autonomy signifies the right of the competent individual to make choices freely about medical care and denotes the obligation of the healthcare provider to

communicate effectively with the patient and solicit those decisions. As detailed by Beauchamp and Childress, autonomous actions require that definitions be

available including criteria so that someone could give or refuse informed consent and make other decisions. Furthermore, the principle of respect for autonomy must

be evaluated in given situations in competition with the other principles ( 7). Widespread acceptance of principalism in the United States was circumscribed to a period

of about 20 years from the mid-1960s, as other ethical concepts became more prominent ( 4).

Alternative Ethical Approaches

There has been increased advocacy in the past decade for alternative ethical approaches that place more emphasis on the value of a more personal decision-making

interaction of physician and patient ( 8, 9, 10, 11 and 12). Alternative ethical approaches include casuistry (analysis of a specific case in relation to a system of

principles) and cross-cultural, narrative, and virtue ethics. These gained advocacy as changes continued in society and in medical practice and care. Key examples of

the latter are institution of diagnosis-related groups (DRGs) in 1983 and, more recently, the widespread movement toward managed care systems owned or directed

by for-profit corporations and nonprofit organizations vying competitively, which raised serious concerns, divided physician loyalties, and imposed limitations on

medical care (10).

In addition to secular ethical positions, some bioethicists have formulated their positions on the basis of particular religious traditions ( 13, 14) and some attempt to

construct a basis for secular ethics for those “who do not share a contentful moral vision” ( 15).



ISSUES IN NUTRITIONAL SUPPORT OF THE COMPETENT PATIENT

Many of the ethical and legal issues concerning medical practice and patient-physician relations relate to provision of nutrition and hydration. Prior to the practical

application of forced-feeding techniques, the inability to maintain adequate nutrition by mouth meant progressive body wasting until death. Except for deaths occurring

very rapidly from violence, trauma, or acute overwhelming infection, the more usual direct causes of death were dehydration and starvation associated with more

chronic illness. Under these circumstances, there was relatively little that a physician could do in restoring health while maintaining a good nutritional state. This

situation began to change significantly in the late 19th century as effective therapies (initially surgical) and other advances in diagnosis and treatment enhanced the

societal importance of the physician.

Concerning oversight of medical care in general, the prevailing judicial position in the past was that of “compelling state interest” such as preservation of life,

prevention of suicide, and certification of physician education. This legal position in conjunction with the paternalistic role of the physician eventually led to

contentious legal issues.

The patient's right to refuse medical treatment was not established until 1891, in a case in which the United States Supreme Court upheld the right of a personal injury

plaintiff to refuse a medical examination ( 16). This decision established the general rights of individuals to make choices regarding bodily examination and treatment

under the principle that common law guards the right of every individual to the possession and control of his or her own person on the basis of the doctrine of informed

consent. However, this only began to establish judicial decisions that more specifically defined the laws of informed consent and the rights of the incompetent patient.

In the past, the issue of provision of nutrition and fluids has caused the most controversy among adult patients, their surrogates, physicians, hospital administrators,

and state courts. The right of the competent individual to decide whether or not to receive medical therapy (autonomy) was not automatically granted. Withholding or

withdrawal of artificial nutrition and hydration became a contentious issue because some physicians, ethicists, theologians, and “right-to-life” groups considered such

acts assisted suicide or euthanasia (and hence likely to involve criminal penalties) or were cessation of “ordinary” medical procedure.

The Bouvia Case

The Bouvia case posed the issue of the right of a competent individual to starve to death. Elizabeth Bouvia, a quadriplegic in her mid-20s, was hospitalized in

California with severe cerebral palsy and arthritic pain requiring morphine injection. While able to ingest food orally with assistance, she asked that such feeding be

stopped. The hospital refused to accede to her wishes, and a lower court ruling in 1984 supported the hospital ( 17). When her condition deteriorated to the point that

feeding by nasogastric tube was necessary, the appellate court ruled in 1986 that her refusal of treatment was not a form of suicide, thus rejecting the arguments of

hospital officials that removing the tube would make them a party to suicide ( 18). With the additional support given later by the United States Supreme Court decision

in the Cruzan case (see below), the right of an irreversibly ill but competent patient to refuse artificial feeding is unlikely to be seriously challenged again.

Bouvia's condition was classified as a state of severe and permanent paralysis, which may acutely or progressively result in a “permanent locked-in state.” This was

first described as a medical condition in 1966 by Cranford, as irreversible loss of motor function with preservation of normal consciousness, possible long-term

survival of years or even decades, and physical and psychologic suffering of a degree that may become extreme because of the patient's awareness of the condition

(19). Eleven years later Bouvia was still alive, totally dependent on others for her care and, despite several half-hearted attempts, unable to bring herself to insist on

actual discontinuation of artificial feeding ( 19a).

Patient-Physician Interaction and Decision Making

In an incisive essay, Pellegrino stated “that a more sensitive and compelling guide to the care of the sick is to be found in the fact of illness as a human experience

rather than in the assigned role of the profession. Without supplanting traditional professional ethics, the intrinsic dehumanizing nature of illness imposes additional

obligations of greater sensitivity” ( 20). Individuals with “illness as an acute event or as a chronic accompaniment of life are deprived in varying degrees of those things

which distinguish humanity from other forms of existence.” These include loss of freedom of action, freedom to make choices, and freedom from the power of others

as well as threats to personal self-image. These disabilities “must be the infrangible base for the obligations of physicians and all others who profess to heal ..., these

obligations constitute the substance of professional medical ethics, ... Its rooting in the existential situation is more authentic and more human ... than the traditional

one in the self declared duties of the profession ... The professional can make a valid claim for technical authority but this no longer extends to moral authority ... The

patient has the human right to his own moral agency if he or she wishes to exercise it. The physician has the moral obligation to ascertain the degree to which the

patient wishes to exercise his moral prerogatives and to provide the fullest exposition which will enable the privilege to be exercised” ( 20).

The principle of respect for autonomy of the competent patient may result in tension between a patient or the parent of a minor and the physician when the patient's

decision seems inappropriate. What appears legally to be a clear-cut situation, i.e., the rights of a competent patient, may be a complex and difficult situation for all

concerned with the welfare of the individual. The old Spanish proverb quoted elsewhere in this context is apt: “The appearance of the bull changes as one leaves the

grandstand and enters the ring” ( 21). The patient's viewpoint may reflect a lucid and rational decision or an undisclosed or undiagnosed problem(s) such as

depression, other mental difficulty, side effect of medication ( 22, 23), an unspoken complicating social problem, strong opposing family influence, misunderstanding

the physician's intentions, or the patient's failure to know, comprehend, or accept the severity or irreversibility of the disease.

Disagreement may occur, for example, when the patient wishes parenteral or enteral feeding stopped and the physician believes that continued feeding is in the

patient's best interest. An even more difficult situation pertains to stopping parenteral feeding in an intestinally obstructed cancer patient who has failed all therapy but

is still competent and ambulatory, and for whom such continued support at home could conceivably extend life for months.



Recommendations of the President's Commission

Several recommendations on this issue were considered by the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral

Research (24). It held that “health care professionals serve patients best by maintaining a presumption in favor of sustaining life while recognizing that competent

patients are entitled to choose to forgo any treatments, including those that sustain life” and “the voluntary choice of a competent and informed patient should

determine whether or not life-sustaining therapy will be undertaken, while healthcare institutions and professionals should try to enhance patients' abilities to make

decisions on their own behalf and to promote understanding of the available options.” In such situations, members of the hospital nutritional support team, who are

often and intimately involved in a major aspect of the patient's care, may be helpful in affording insight into the patient's status and expressed position. The

physician's discussion of the medical situation with patient and family, the basis for the medical recommendations, the therapeutic alternatives with their probabilities

of success, and the offer of a second opinion are all essential to the decision-making process.

This commission also examined the role of traditional moral distinctions as they relate to decisions about medical care and whether they are acceptable or

unacceptable. It noted that from the viewpoint of most competent patients, decisions about alternative available courses of treatment are made on the basis of factors

that include treatment benefits in terms of extension of life, the nature and quality of that life, the degree of suffering involved, and the various costs to themselves and

to others (24). It noted that other criteria have been used in judging the acceptability or unacceptability of life-and-death decisions and stated ( 24):

these bases are traditionally presented in the form of opposing categories. Athough the categories causing death—by acting versus by omitting the act;

withholding versus withdrawing treatment; the intended versus the unintended but foreseeable consequences of a choice; and ordinary versus

extraordinary treatment—do reflect factors that can be important in assessing the moral and legal acceptability of decisions to forego life sustaining

treatment, they are inherently unclear. Worse, their invocation is often so mechanical that it neither illuminates an actual case nor provides an ethically

persuasive argument.

Several of the commission's conclusions about such distinctions are relevant to the issue of nutrition and hydration and have, in fact, had a significant influence on the

attitudes of physicians and the courts. For example ( 24):

The distinction between acting and omitting to act provides a useful rule-of-thumb by separating cases that probably deserve more scrutiny from those that

are likely not to need it. Nonetheless, the mere difference between acts and omissions—which is often hard to draw in any case—never by itself

determines what is morally acceptable. Rather, the acceptability of particular actions or omissions turns on other morally significant considerations, such

as the balance of harms and benefits likely to be achieved, the duties owed by others to a dying person, the risks imposed on others in acting or refraining,

and the certainty of outcome ... A justification that is adequate for not commencing a treatment is also sufficient for ceasing it. Moreover, erecting a higher

requirement for cessation might unjustifiably discourage vigorous initial attempts to treat seriously ill patients that sometimes succeed.

The guidelines of the Hastings Center also expressed concern that the terms ordinary and extraordinary “obscure ethically important questions rather than helping to

resolve them. Prevalence of a treatment or its degree of technological complexity is sometimes used to make the distinction between ‘ordinary' and ‘extraordinary'. We

reject the distinction. No treatment is intrinsically ‘ordinary' or ‘extraordinary'. All treatments that impose undue burdens on the patient without overriding benefits or

that simply provide no benefits may justifiably be withheld or withdrawn. While traditional definitions of `extraordinary' hinged on this comparison of benefits and

burdens, the term has become so confusing that it is no longer useful” ( 25).

The Law Reform Commission of Canada, an official agency for evaluating and recommending reform of Canadian federal law, issued in 1983 a series of principles

and recommendations for amendments to the criminal code. These amendments suggested changing existing laws so that they would not be interpreted as requiring

a physician to initiate medical treatment against the wishes of the patient or to continue such treatments when they have become therapeutically useless and not in

the best interest of the patient ( 26). These and related amendments have not been adopted as of early 1996 ( 27).

Resolving Differences between Patient and Physician

When all the stated precautions and efforts have been honored in decision making but the choice of the competent and informed patient is contrary to the judgment of

the physician, where does this leave the physician? “The physician too has a set of values to which he owes allegiance. He has a double obligation, to protect those

of the patient and to be faithful to his own” ( 19). To deal humanely with the conflicts that must occasionally arise, “the physician must know enough about his own

beliefs to decide when he can compromise, when he cannot, and when he must give the patient the opportunity to transfer his care to another physician whose values

more closely coincide” (20). In this situation, the President's Commission took the position that “health care professionals or institutions may decline to provide a

particular option because that choice would violate their conscience or professional judgment, though in doing so they may not abandon a patient” ( 24). In the case of

the attending physician, responsibility for the care of the patient must be transferred to another physician who accepts the patient's decision and acts accordingly.

Abandonment means leaving the patient without care, a serious and punishable infraction of both the legal and ethical obligations of the physician. There has been a

recent emphasis on nonabandonment “as an open-ended commitment over time” (28) of the physician to the patient ( 28, 29); examples include accepting competent

patient decisions on demanding or stopping artificial feeding. Objection has been raised to the apparent elevation of nonabandonment to a principle rather than

considering it one of the essential derivatives of medical ethics ( 30).

When the patient's request is opposed by the hospital administration, the solutions open to the hospital are to yield to staff pressure on behalf of the patient, to

transfer the patient to another hospital willing to accept the patient, or to yield to a court order when it supports the patient. Often in the past and continuing to some

degree today, it has been the reluctance of the hospital administration to discontinue tube feeding or some other life support—usually because of fear of civil or

criminal penalties or on religious grounds—that has led to legal actions by patients or family members.

Distinguishing Patient Refusal from Requests

Physicians may differentiate between the competent patient's refusal of a recommended treatment and the request for a nonrecommended treatment ( 31). Refusals

must be honored even when the physician knows that death will result. In contrast, a physician may feel that there is no moral or legal obligation to honor a request

that the physician firmly believes will have an undesirable effect or result in death. This is a matter of judgment focusing on the issue of treatments deemed useless or

futile by the physicians or hospital.

Denial of Treatment as Medically Futile

Statements such as those made by the President's Commission (24) and the chief counsel for the American Medical Association (AMA) ( 32) supporting the

physician's right to deny futile treatments have been under increasing scrutiny and criticism as being too simplistic. This area has considerable relevance to artificial

nutrition, especially by the intravenous route, because of its expense, increased risk of infection, and evidence of little or no benefit in advanced disease, particularly

cancer unresponsive to disease-specific therapy.

Justifications for physician refusal of futile interventions for the competent patient include (a) the specific intervention proposed has no physiologic rationale; (b) the

intervention has already been tried and failed to achieve any apparent benefit; (c) maximal treatment for the underlying advanced disease is a failure; and (d) the

intervention(s) will not help achieve the goal of care, is burdensome, and may worsen the quality of life ( 3).

With respect to nutrition, how valid are these justifications? Item (a) above is obviously questionable or incorrect in the case of intravenous feeding, because it does

provide essential nutrition and hydration; (b) would not hold if nutrition has not previously been given; arguments (c) and (d) may hold true. However the challenge to

statement (d) rests on the answer to the question: whose goal or value is most valid—the physician's or the patient's? If the patient wishes to continue the requested

treatment—even for a limited period—to achieve his or her goal (e.g., survival and additional months to see an only child married or the first grandchild born), what is

the justification for saying “nay”?

Even if the likelihood of success, in the opinion of the physician, is very small, how can one know if a particular patient will or will not respond? Furthermore, can the

patient be certain that the physician is fully conversant with the medical literature on this point, or has not misinterpreted the situation, or that other physicians in the

same hospital or elsewhere will be in agreement? Uncertainty and changes of therapeutic options in medicine are hardly uncommon ( 33, 34).



The basis of physician opinion may be that the prospective benefit for the patient does not justify the required resources. However, futility is not to be confused with

resource allocation (which requires well-tested clinical data plus administrative and societal decisions for acceptability), even though they do share a common

purpose (35, 36). A continuum of conditions and situations extends from those that are obviously futile, to those that produce an effect but are deemed by the

physician to be of no net benefit, to those in which the physician is believed to be in error by other physicians or by the patient. In the case of the incompetent patient,

the same issues about futile treatments arise in the physician-surrogate relationship.

This issue has resulted in a large number of papers, letters, editorials, and conferences in the medical and bioethics literature (e.g., 32, 33, 34, 35, 36, 37, 38, 38a, 39,

40, 41, 42, 43 and 40), to the point where the question has been posed: “Is futility a futile concept?” This is based on the argument that its proponents have not paid

sufficient attention to the problematic nature of the data supporting the use of their definitions ( 40).

It is not surprising that strongly held differences of opinion may arise between and among physicians, patients, families, and surrogates about treatment that arguably

is futile. Collaborative procedures have been advocated or instituted to help solve such problems and settle such differences constructively, short of the courts ( 41,

42).

Caring for the Hunger Striker

The hunger striker “is a mentally competent person who has indicated that he/she has decided to embark on a hunger strike and has refused to take food and/or fluids

for a significant interval” ( 43). Once the physician agrees to attend a hunger striker, that person becomes his or her patient. Guidelines of the World Medical

Association (WMA) include thorough examination of the patient, daily visits, information on the clinical consequences of a hunger strike, treatment of trauma and

infection, and advice on fluid intake; any treatment must have the patient's approval. “Treatment or care of the hunger striker must not be conditional upon his

suspending his hunger-strike” ( 43). Autonomy must be respected. In Israel, however, a district judge authorized force-feeding of a political prisoner on a hunger strike

(43a).

Serious moral dilemmas arise when the patient is comatose when first seen and the person's wishes are not known. In that situation, the WMA, the British Medical

Association, and other medical societies hold that treatment should be initiated until resuscitation has occurred and the patient's wishes are known. In the instance of

a hunger striker becoming severely confused or comatose during medical observation, “the physician shall be free to make the decisions for his patient which he

considers to be in the best interest of the patient taking into account the decision he has arrived at during his preceding care of the patient” ( 43). The problems related

to management of asylum seekers on hunger strike has been reviewed (43b, 43c).

The Patient Self-Determination Act

As a result of the recommendation of the President's Commission and other similar suggestions, federal legislation (the Patient Self-Determination Act of 1990) was

enacted and became law on December 1, 1991 (44). This law applies to hospitals, nursing homes, hospices, healthcare maintenance organizations, and healthcare

companies receiving Medicare and Medicaid funds. At the time of admission, enrollment, or on initiating home care, they are required to inform patients about their

legal rights in that specific state and assist them in making decisions concerning their medical care and formulating advance directives. Such a directive must be in

the patient's medical record, indicating whether life support has been rejected. This federal law, as well as most state laws and court decisions on this subject, also

allows hospitals and nursing homes to express their administrative positions on these subjects to the patient before or at admission. One goal of the statute is to

encourage (but not require) adults to complete advanced directives in the form of treatment directives, a proxy appointment, or both while they are competent. Another

goal is to influence both healthcare givers and institutions to honor advance directives.

As a result of the majority decision of the United States Supreme Court on the Cruzan case (see later) and the Patient Self-Determination Act now in force, advance

directives must be based on the legal requirement of individual states. Thus, standards in the United States have been established in varying degrees for truth in

prognosis (45), informed consent, and advanced care directives.

The existence of advanced care directives has stimulated a number of empirical studies. It has been noted that most patients do not execute directives in part

because physicians frequently fail to discuss this issue ( 32) or because patients fail to see the directive when it is presumably distributed on admission ( 46). Even

when executed, only about 39% in one study wished to have the directives observed strictly by the physician ( 47). In addition, a significant number of directives are

either ignored or overridden by the physician; sometimes more treatment is given than requested and sometimes less ( 32, 48).

Attitudes toward executing directives, especially by older individuals, are significantly influenced by a number of factors including dependence on family decisions,

educational and economic status, religious views, ethnic backgrounds, and national origins. For example, Korean Americans and Mexican Americans were more likely

to hold a family-centered model of medical decision making, rather than the patient autonomy model favored by most African American and European American

respondents (49). The great majority of Navajo informants considered advance care planning a dangerous violation of their traditional values ( 50). To develop the

understanding of and respect for cultural differences that will help avoid inappropriately paternalistic judgments, physicians do well to seek assistance from family

members and others who are knowledgeable in this area ( 51).

Recently, a series of survey reports introduced another view of the purpose of advanced directive and planning, namely, its intent to limit life-sustaining treatments

and reduce healthcare costs. Such an approach creates the potential conflict between the original purpose of advanced directives (stimulating informed consent and

patient autonomy) to that of reducing medical care ( 51a).

Patient Choices for Life-Sustaining Treatments

Studies have been done on the process of decision making concerning medical intervention and life support of the seriously ill and the issue of quality-of-life and its

expected duration. Earlier studies used measures of physiologic competence (e.g., organ function or metabolic indices), economic or environmental factors, or overall

individual performance to ascertain whether any of these factors are improved by specific therapies. More recently, efforts have been spent in obtaining information on

the relative importance patients assign to the value of their lives and how it may be influenced by their therapies ( 52, 53). The subjective health values and ratings of

1438 seriously ill but competent patients with at least one of nine diseases with a projected overall mortality rate of 50% were compared with similar ratings obtained

from the patients' surrogates and their physicians. The health values of the patients varied greatly from one to another, but the individual patient showed excellent

test-retest reliability. Their preferences for living at current health levels compared with living in excellent health correlated strongly with their preferences for care that

extended life even though it was at the expense of pain and discomfort. Furthermore, their individual health values and ratings were generally higher than those

postulated by their surrogates and physicians. The study concluded that when it comes to the issue of resource allocation for care, the patient is the true “expert on

the values of living,” whereas the value of society is that of taxpayers or insurance premium payers who often foot the healthcare bill ( 54).

There are important differences in the responses of elderly individuals about the use of living wills and desire for life-sustaining treatments between those who have

chosen to forgo life-sustaining treatments and those who have not indicated such a preference. Those with living wills were less likely to change their expressed

wishes (14%) over a 2-year period than those without (41%) ( 55). During the study period, patients who had been hospitalized, had an accident, become more

immobile or depressed, or had less social support showed appreciable change in the stability of choice and now wanted increased treatment. Such data indicate the

need for periodic review of preferences for life-sustaining treatments.



NUTRITIONAL SUPPORT OF THE INCOMPETENT PATIENT

The major medical, ethical, and legal issues on cessation of involuntary nutritional support have involved incompetent patients. More than 30 separate state court

decisions on this issue between 1983 and 1995 have been summarized ( 56).

There are a number of causes of incompetence, ranging from serious memory loss without other physical disability, through progressive loss of cognitive ability

requiring hand feeding, to one of several forms of persistent unconsciousness. The patient in the last category requires artificial nutrition and hydration to avert

starvation or dehydration. Until physicians have had sufficient time to develop a firm prognosis of the individual who is unable to take sufficient food and fluid by

mouth, nutrition and hydration are indicated.

Traditionally, the right of the state to preserve the life of the incompetent patient has been manifest in the requirement for a court-appointed guardian or surrogate to



act on behalf of the adult patient. In the case of minors, courts have almost always affirmed the authority of the parents or (if needed) a surrogate to make decisions.

The Persistent Vegetative State (PVS)

The Multi-Society Task Force on PVS has summarized current knowledge of this state, which it defined as a “clinical condition of complete unawareness of the self

and the environment accompanied by sleep wake cycles with partial preservation of hypothalamic and brain stem autonomic functions” ( 57). Seven diagnostic criteria

were given. Facial and reflex movements, groans, and cries may occur, and small amounts of food placed in the mouth may be swallowed ( 19, 57). Clinical

observation, the results of positron-emission tomographic (PET) studies, and neuropathologic examination support the belief that these patients are unaware and

insensate and lack the cerebral cortical capacity to be conscious of pain ( 58).

When this “wakeful unconscious” state persists for a month in a patient with an acute traumatic or nontraumatic brain injury (e.g., from stroke or cardiorespiratory

arrest) or with one of a number of degenerative or metabolic disorders or developmental malformations, the term persistent vegetative state is in order. Of 257

patients, the cumulative mortality rate at 3 years was 70%, and at 5 years, 84% ( 58a); survival beyond 10 years was unusual, but some lived for 30 years ( 19, 58).

Progressive spasticity and bed rest lead to muscle atrophy and limb, hand, and foot contraction ( 13). The term permanent vegetative state is prognostic, because the

outcome cannot be known with absolute accuracy. It may be used after the time lapses noted, which strongly indicate that recovery is very unlikely.

The task force noted that few patients in PVS had verified recovery of consciousness after 12 months following traumatic injury, or more than 3 months after a

nontraumatic injury; the few who recovered were left with severe disability ( 58). A British unit specializing in neurodisability reported that 40 (41%) of 97 patients with

profound brain damage admitted from 1992 to 1995 had been diagnosed by the referring physician as being in a vegetative state; of these 40, 17 (43%) were

considered misdiagnosed because all but one responded in some manner to a command within 16 days of admission ( 59). An accompanying editorial questioned the

diagnosis of vegetative state in 9 of the patients and emphasized the severity of their physical, neurologic, and cognitive states despite some responsiveness ( 60).

The need for competent and persistent serial evaluation of all vegetative patients and use of additional evaluation techniques is emphasized ( 60, 60a).

The prevalence of PVS is not known because it was not a codable diagnosis until recently in ID-9-CM and elsewhere. The task force referred to estimates of PVS in

the United States of 10,000 to 25,000 adults and 4,000 to 10,000 children ( 57).

Guidelines were published recently by a multidisciplinary group convened by the Royal College of Physicians, working on the diagnostic criteria and terminology of

PVS (61). This group preferred the general term “vegetative state: to describe the condition which may be transient with recovery/or which might persist to death.” The

term continuing vegetative state (CVS) was applied when the condition persisted for more than 4 weeks and recovery from coma became unlikely. A “permanent” state

exists “when the diagnosis of irreversibility can be established with a high degree of clinical certainty.” The working group accepted the U.S. Multi-Society Task Force

on PVS (52) suggestion for use of the term permanent for adults and children with traumatic injury of 12 months duration, and 3 months for those with nontraumatic

injuries.

In contrast to incompetent adults with PVS, forgoing medically provided nutrition and hydration in children has received less discussion and legal review. The current

status of this aspect has been reviewed recently ( 62). Nelson et al. noted that pediatricians were more likely than internists to continue artificial nutrition and hydration

in PVS against parental desires, even though other life-support systems had been discontinued. They listed three major medical categories of children for whom

consider it is ethically permissible to forgo such nutrition and hydration: neurologic devastation, such as PVS from severe brain injury or anencephaly; irreversible

total intestinal failure, such as total bowel necrosis or neural dysplasia; and proximate death (days or weeks) in conditions such as advanced unresponsive cancer.

The issue of total intestinal failure is discussed in light of the possibility of a successful bowel transplant and the presence of other serious disease in recipient and

donor. Various judicial decisions concerned with cessation of artificial feeding (in the great majority of which the courts have upheld the desire by the parents for

termination) are reviewed as well as federal statutory law and state laws ( 62).

Forgoing nutritional support in children with severe intestinal dysfunction requires careful consideration because most of these children are not in the terminal state or

in PVS. Total parenteral nutrition, despite the cost and probability of complications, has proven effective in improving and maintaining adequate nutrition and a

reasonably good quality of life. Despite major intestinal loss, in a number of patients the remaining bowel can often adapt to the point that parenteral nutrition can

eventually be discontinued (see Chapter 101). For those who may develop severe hepatic insufficiency, bowel transplant is an increasingly successful option ( Chapter

101).

The impact of PVS on the patient's family and friends is one of distress and helplessness. Psychologic consequences of seeing the loved one remain unresponsive

with increasing atrophy and contractures as weeks and months pass are exacerbated by developing monetary and social problems. As noted above, a large number

of such cases have led to court decisions that have focused the attention of the public, other courts, and medical societies on biomedical and ethical issues related to

the requests for withdrawal of life-sustaining treatments including withdrawal of artificial nutrition and hydration. Some instructive cases and decisions affecting adults

are summarized below.

The Quinlan Case

The case of Karen Ann Quinlan was a landmark effort to discontinue life support. She was a young woman in New Jersey who was in PVS on a respirator. The state's

supreme court, on a reversal of a lower court decision, made medical legal history in 1976 when, at the request of the parents and despite the opposition of the

patient's physician and hospital, it ruled that the respirator could be discontinued ( 63). When this was done, the patient was found to be able to breathe

spontaneously, and she survived for 9 years on tube feeding until she died from overwhelming infection. Withdrawal of tube feeding was not considered an option by

the parents. This position reflected a view then widely held, namely, that food and hydration were not in the category of a special life-support system but rather a

humanitarian action necessary for the comfort of the patient.

The Barber-Nejde Case. Another troubling judicial issue was the legal liability of physicians when nutrition was withheld from an incompetent patient who was not

terminally ill and who, while competent, had not clearly indicated in writing or verbally to a reliable witness a desire not to be force fed when terminally or incurably ill.

This concern is illustrated by the legal developments resulting from the medical management of Clarence Herbert, who had suffered respiratory arrest in association

with routine intestinal surgery ( 64, 65). Following resuscitation, he remained comatose over the following several days. On the advice of physicians that prognosis for

recovery was poor and with consent of family, use of the respirator was stopped. The patient then breathed spontaneously but without change in his comatose state.

Two days later, again after consultation with the family, the attending physicians ordered removal of the nasogastric tube, intravenous nutrition, and air mist. The

patient died 6 days later. He had not previously executed a formal directive under the California Natural Death Act, nor had he written anything concerning his wishes

in such circumstances; however, he had stated to his wife that he did not want to be kept alive by machine or “become another Karen Ann Quinlan.”

A nurse reported the actions of the physicians to local authorities, who then filed criminal charges for murder against both attending physicians, Drs. Barber and

Nejde. Despite a municipal court magistrate ruling that death had resulted from brain damage secondary to anoxia and that the conduct of the physicians was not

“unlawful,” the district attorney appealed the case to Los Angeles Superior Court. The murder charge was reinstated by a judge who decided that there was no legal

justification for the physicians' action. The physicians then appealed to the California Court of Appeals, which dismissed the criminal charges.

The Conroy Case

Another case, occurring at the same time, further delineated the complex medicolegal issues that arise when the status of an incurably ill patient is presented to

different jurists. Claire Conroy was an 84-year-old woman severely ill with advanced atherosclerosis, diabetes, and organic brain syndrome ( 66, 67). Although the

case description stated that she was not in PVS, her described behavior approximated closely the clinical description of this state. A nephew filed a petition to

authorize removal of the nasogastric feeding tube from the patient, although there was no clear evidence of what the patient would have desired. A trial judge

authorized removal of the feeding tube but not cessation of any voluntary or assisted oral feeding. The decision was appealed by the institution, and pending the

hearing, tube feeding was continued to the patient's death.

Despite Conroy's death, the New Jersey Appellate Court heard the appeal and reversed the earlier trial judge's decision. It held that (a) she was neither comatose nor

terminally ill, (b) the feeding tube was not a particularly invasive treatment as was the respirator in the Quinlan case, and (c) if the nasogastric tube had been removed

in accordance with the trial judge's decision, Mrs. Conroy would have died not as a result of her condition but from a “new and independent condition: dehydration and

starvation and that this would constitute murder (euthanasia).” The case was appealed to the New Jersey Supreme Court, which sanctioned withdrawal of artificial



nutrition and hydration in this type of case, but with a time constraint of expected survival of 1 year or less ( 66).

This final decision in the Conroy case is of special interest in the context of criteria established by judges as guidelines for decisions in terminating medical care of

incompetent patients. Emanuel noted that the court laid down three standards for terminating such care and stipulated that one of them be satisfied before treatment

could be stopped (67). The court designated the first a “subjective standard,” which permits termination of treatment when an incompetent patient left clear indication,

such as a living will, that he or she would have refused that treatment. It called the second a “pure objective standard,” which permits termination of life-sustaining

treatment for patients if the burden of the care outweighs the benefits, although they have not left indication of their preferences about life-sustaining care. For

example, administering life-sustaining care would be inhumane because it would perpetuate severe pain. Others label this the “best interests standard,” in which the

patient's surrogate objectively evaluates the benefits and risks of a treatment and decides what most benefits the patient. In the Conroy Case, the court, acting as

surrogate, accepted this standard. The third standard, called by the court a “limited objective standard,” is a combination of the first two and permits termination of

care if there is some evidence (such as remarks made during a conversation) or opinion of the surrogate that the patient would not want the treatment and if the

burden and pain of continued life outweigh the benefits; this is similar to the substituted judgment standard.

The Jobes Case

A subsequent ruling by the same court involved the case of Nancy Jobes, who was in PVS following hemorrhage ( 68). Her husband and parents requested that a

feeding tube be removed and that she be allowed to die. The nursing home refused to honor this request. When the case was brought before the New Jersey

Supreme Court, the court sanctioned removal of the feeding tube but rejected as “remote,” “general,” and “casual conversation” statements by the patient's friends

and relatives that she had said in conversation that she did not wish to be dependent on a respirator. The court eliminated its previous “Conroy” criterion and instead,

based its ruling on the Quinlan case, ruled that the “substituted judgment standard” exercised by the family should apply to patients in a PVS, and therefore, the

feeding tube could be removed. In reviewing this case, Emanuel noted that the substituted judgment standard means, in effect, that the surrogate for an incompetent

patient is put into position of attempting to make a decision for the patient as if that individual were competent; this standard thus differs from the best interest

standard, which is based on an objective standard of benefits and burdens ( 68).

The judicial and legislative history and endorsement of decision making for incompetent patients by proxy or surrogate and their ethical justification have been briefly

reviewed by Emanuel and Emanuel (69). They summarize objections to proxy making on the grounds that “proxy decision makers cannot divine or implement the

incompetent patient's wishes regarding the termination of life sustaining care.” Alternative solutions are suggested, with recognition that each has limitations. While

agreeing with the array of concerns about proxy decision making, Lynn, in an accompanying editorial, takes the position that “a morally justifiably and pragmatic policy

for decision making for incompetent adults, at this point, will have to rely heavily on appointed and family proxies, with a morally defensible and practical plan” ( 70)

which provides options similar to those suggested by the Coordinating Council on Life Sustaining Medical Treatment Decision-Making by the Courts ( 71).

AMA Statement of 1986

In the midst of these and other pertinent judicial decisions, the AMA, through its Council on Ethical and Judicial, issued in 1984 (and in revised form in 1986)

statements on withholding or withdrawing life-prolonging medical treatment. The 1986 statement consists of four short paragraphs that include the following, which

also relate to food and fluids ( 72):

In the absence of the patient's choice or an authorized proxy, the physician must act in the best interest of the patient ... For humane reasons, with

informed consent, a physician may do what is medically necessary to alleviate severe pain, or cease or omit treatment to permit a terminally ill patient

whose death is imminent to die. Even if death is not imminent but a patient's coma is beyond doubt irreversible and there are adequate safeguards to

confirm the accuracy of the diagnosis and with the concurrence of those with responsibility for the care of the patient, it is not unethical to discontinue all

means of life prolonging medical treatment. Life prolonging medical treatment includes medication and artificially or technologically supplied respiration,

nutrition or hydration.

The changed physician attitudes resulting from the widespread interest and discussion of these issues is evidenced by significant differences in the 1984 and 1986

statements. For example, the first version refers only to terminally ill patients, while the later one includes those in irreversible coma; the 1984 version did not define

“life prolonging medical treatment,” whereas in 1986, these are specifically designated and include nutrition or hydration.

The Cruzan Case

This case was the first to involve the United States Supreme Court in the issue of discontinuance of tube feeding an incompetent patient. Like the Quinlan case, it

aroused widespread public and professional interest and resulted in a decision with far-reaching implications. In January 1983, at the age of 25 years, Nancy Cruzan

suffered irreversible brain damage secondary to prolonged hypoxia following an automobile accident. She was then supported in the hospital by food and fluid fed

through a tube. In 1986 her parents requested discontinuance of the feedings but had to resort to legal action at the insistence of the Missouri state hospital

administration. In July 1988, a trial court ruled that tube feeding could be withheld; however on appeal, the Missouri Supreme Court, by a 4 to 3 decision, reversed the

trial court on the grounds that no reliable evidence was presented indicating that Cruzan would have refused artificial feeding ( 73).

On June 25, 1990, the United States Supreme Court affirmed the reversal by a vote of 5 to 4 ( 74). Five justices stated that the Constitution did not prohibit Missouri

from choosing to rule as it did. Three of the four dissenting justices stated that incompetent as well as competent patients had the constitutional right to be free of

unwanted medical treatment and the fourth stated that the Constitution required that the best interest of the patient be followed. Six of the nine justices explicitly found

no distinction between fluids and nutrition delivered artificially and other medical treatments; none of the other three found a constitutionally relevant distinction ( 74,

75). Six months later, nearly 8 years after the parents' first request, a judge in Missouri Circuit Court, in a brief order, authorized the parents as coguardians “to cause

the removal of nutrition and hydration from our ward, Nancy Beth Cruzan” ( 76). This was done, and Cruzan died 12 days later.

While some bioethicists felt that the Supreme Court's opinion would make a constitutional right to refuse artificial nutrition more difficult to achieve in the future ( 77,

78), others believed that the decision had important positive aspects. These were summarized in a statement by physicians, nurses, lawyers, and bioethicists ( 79).

The best evidence of the impact of the Supreme Court's decision was the rapid collapse of the medical, institutional, surrogate, and legal opposition in Missouri to

discontinuing Cruzan's feedings. Another effect of the Cruzan decision was the speed with which many states enacted health proxy legislation ( 80). Problems still

remained, however, because of the variability in important requirements among state laws. As a result of publicity about the Cruzan case and variability of relevant

state laws, some families in similar situations have requested that physicians not start artificial feedings because they and the patient may become prisoners of

technology and state law (19). This type of request presents a professional problem for physicians who are hesitant to start or to stop such treatment before the

prognosis seems certain. To help resolve this dilemma, Cranford proposed that the family be assured that once a diagnosis and prognosis were established with great

certainty, physicians would be willing to discontinue artificial nutrition and hydration, thus freeing the family of the fear that the patient and family would be held

hostage by medical technology (19).

The problem is further complicated by family members or surrogates who, despite physicians' recommendations, insist that life-support systems be continued in

patients in a PVS (e.g., cases of Wanglie [ 81] and of Baby K [82]). While various recommendations have been offered in an effort to increase legal support for medical

decisions, in the United States, the legal outcome in a suit depends on state court decisions as they arise.

In some other countries, the highest national courts have upheld the right of physicians to withdraw lifesaving treatment in specific instances of patients in PVS. In the

United Kingdom, the House of Lords made that decison on February 4, 1993, in the case of Anthony Bland, who had suffered severe brain damage in 1989 ( 83); the

wording of this decision led the Royal College of Physicians to provide standards for decision making by physicians that will have High Court or Parliamentary

approval (58).

The Irish Supreme Court by a 4 to 1 majority ruled that a gastrostomy feeding system could be withdrawn from a 45-year-old woman in “near permanent” vegetative

state for 23 years and that she be allowed to die ( 84). In recent years, courts in New Zealand and South Africa have ruled it lawful to discontinue treating PVS patients

(85).

Patients with Progressive Neurologic Deterioration

In addition to the state of severe and permanent paralysis (“locked-in syndrome” described above in the Bouvia case) and PVS, a third syndrome, dementia, is in the



category of disabling neurologic conditions with the potential for prolonged survival and the resulting issue of life-sustaining medical treatment. Dementia includes

Alzheimer's disease, the most common cause in the elderly, with its variable destruction of the neocortex; multiinfarct and other vascular dementias with variable

cognitive deficits ( 19, 23); and the variants of Creutzfeldt-Jakob disease and dementias associated with parkinsonism and progressive supranuclear palsy ( 85a).

These disorders are characterized by gradual onset of progressive neurologic deterioration occurring over years to decades, with time to prognostic certainty usually

months to years, and with suffering decreasing with increasing impairment of cortical function.

The PVS is the ultimate and terminal form of dementia as the result of complete loss of neocortical function ( 19, 57). An estimated 4 million patients in the United

States have various stages of dementia, of whom 1.3 to 1.9 million have Alzheimer's disease ( 23).

The President's Commission (24), the American College of Physicians ( 86), the AMA (87), and other relevant organizations draw a distinction between the

incompetent patient in PVS and one in a “conscious” state but with severe, irreversible, and deteriorating mental impairment. The critical difference between these

states is retention until late in the disease of self-awareness, motor function, and feelings of pain and suffering ( 57). At what stage and on what grounds is the family

or surrogate to make the decision to accept or forgo further life-sustaining treatments?

There are major differences in the judgmental criteria of various medical ethicists. One opinion is that the decision should be that which the patient when fully rational

would have made, based on what Dworkin calls “critical interests” ( 88, 89 and 90), which is equivalent to the “substituted judgment” standard mentioned above. An

opposing opinion is based on the proposition of an existing awareness on the part of the patient and the possibility that his or her interests and wishes might have

changed. In this situation, the family, surrogate, and physician should attempt to understand the experiential world of the patient by observation and testing so as to

better weigh the benefits and burdens of current treatment ( 88, 89 and 90). This is essentially the “best interest,” “objective,” or “benefit burden” judgment standard.

The situation changes when the patient becomes unable to eat voluntarily and swallow adequately when hand fed and has minimal consciousness. In such

circumstances, some bioethicists have favored withholding nutrition and hydration ( 19, 68, 90). Undoubtedly, this issue will receive increasing medicolegal scrutiny as

long as such conditions remain irreversible and family members or surrogates continue efforts to erase the distinction between PVS and severe dementia.

Attitudes and Beliefs Influencing Decisions

The major professional societies, most bioethicists, and many state legislatures and courts support withholding or terminating artificial feeding for the competent

patient who requests it. Some states still limit parental authority or advance directives for the incompetent terminally ill patient or for children and adults in a PVS

(reviewed in 62 and 93). The attitudes of physicians about artificial feeding vary. As mentioned above, physicians may be unaware of advance directives or do not

make serious efforts to obtain them or may undertreat or overtreat despite directives, and pediatricians are believed to be more reluctant than internists to discontinue

feeding with PVS (62).

Long-term comparative data are not available on attitudes and behavior of large numbers of physicians concerning the issue of artificial feeding. However, some data

on physician attitudes are at hand. A national survey of 169 neurologists and 150 nursing home medical directors indicated general concurrence (89%) that it was

ethical to withdraw artificial nutrition and hydration from patients with PVS in an irreversible stage ( 94). In a survey of 32 internists familiar with tube feeding, 80%

favored withdrawing tube feeding and the remaining 20% opposed it when presented with a case of a woman in PVS for 1 year who was being fed via gastrostomy

(95).

A more complex study was conducted with 456 university internists presented with the issue of withdrawing life support; preferences were obtained on withdrawal of

eight different life-support systems including TPN, enteral feeding, and intravenous fluids in relation to 13 attributes including cost, invasiveness, patient discomfort,

rapidity of death. The order of preferences of withdrawal of specific supports was marked. Physicians were much more likely to withdraw TPN before tube feeding

before intravenous fluids—the latter two being classed with artificial ventilation. Older physicians were less likely than younger ones to prefer withdrawing a number of

such supports. The authors concluded that “even when physicians may have agreed that life support should be withdrawn, the choices that they make about the

manner of withdrawing life support reflect other moral, social, and clinical goals” ( 96). A study of actual patterns in withdrawing life-support systems for dying patients

indicated a stepwise retreat in forgoing different types of support until death occurred; enteral and parenteral nutrition and, finally, ventilatory support were the last to

be withdrawn (97).

Patients, surrogates, and physicians may hold religious beliefs that influence their decisions on terminating artificial nutrition. Leading church figures and other

theologic scholars have stated their opinions. The Archbishop of Canterbury concluded in 1977 that extending life by artificial means was distinct from euthanasia and

that removal of life support was acceptable if it would be better for the patient to be allowed to die ( 98). The National Conference of United States Catholic Bishops in

1992 restated the Pope's earlier view that Catholics are not obliged to use extraordinary or disproportionate means when there is no hope for recovery and only the

burden of care remains (99).

For Orthodox Jews, the legal ethical system known as halacha governs most aspects of daily life, including the medical ethic that the “Sanctify of life, the halachic, the

imperative to preserve life, supersedes, with a few exceptions, quality-of-life-considerations” ( 100). Freedman states that the concept of duty and obligation rather

than rights is important and that Halachic Judaism has a concept of informed consent and guidelines for individual medical decision making ( 14). In his review of

relevant rabbinical decisions, an Orthodox Jewish theologian-ethicist has noted differing positions on artificial feeding and concludes that “the question of withholding

and withdrawing artificial nutrition and hydration—tube feeding—is not clear cut in halacha” ( 100). His own opinion is that in geriatric patients, advance directives

should be limited to withholding nutrition and hydration as opposed to withdrawing it. In Israel, the preservation of life in the sick has been embodied in parliamentary

legislation that empowers hospital ethics committees that can overrule a patient's expressed will ( 43a). At the same time, its Supreme Court has recognized the right

of a terminally ill suffering patient to reject intrusive and uncomfortable therapy that cannot cure the basic illness ( 43a).

The physician and hospital must each make their positions clear to patients and surrogates concerning such end-of-life issues so that, if necessary, alternatives can

be found without rancor and undue distress. It is equally important that physicians, nurses, and other members of the healthcare team “be skilled in the two essential

aspects of good clinical care of dying patients—technical issues in the compassionate withdrawal of life-prolonging therapy and counseling and emotional support for

patients, families, and staff during this process” ( 101).

Physiologic Responses to Restriction

Young, healthy, active individuals in negative water balance develop thirst, dry mouth, and headache, followed by fatigue; cognitive impairment occurs as dehydration

progresses and becomes severe with abnormal electrolytes, rising blood urea, and hemoconcentration. Renal failure ensues unless water is available. In healthy

elderly people, in contrast, reduced thirst may occur with water deprivation ( 102).

Those involved in palliative care of terminally ill patients have reported frequently that the conscious competent patient with an advancing severe illness unresponsive

to therapy and in little or no pain becomes progressively weaker, with decreasing communication, increasing anorexia, decreased desire for fluids, and progressive

apathy. In such patients with little intake of fluids, signs and symptoms of dehydration appear: dryness of skin and mouth, decreased urinary output, and occasional

thirst. Nausea, vomiting, or cramps are reportedly rare in this situation, and the dehydrated patient rarely needs oral pharyngeal suction; this is in contrast to the

hydrated patient (19, 103, 104 105 and 106). Obtundation usually progresses to a peaceful death.

There are a significant number of reports (summarized in ref. 92) that many terminally ill patients have relatively few of the changes in blood and urine chemistry that

indicate clinical dehydration and that thirst is usually absent. In a brief report about terminally ill patients, no differences were noted in biochemical parameters and

state of consciousness in those given intravenous fluids and those given only oral fluids ( 107). In their detailed study of 32 competent terminally ill cancer patients,

McCann et al. concluded that “food and fluid administered beyond the specific requests of patients may play a minimal role in providing comfort to terminally ill

patients”; 97% either did not experience hunger at all or only initially, 66% either did not experience thirst or only initially, and the rest were comfortable with mouth

care and sips of water (108). The relationships among symptoms, laboratory evidence of dehydration, and medications were studied in 82 individuals dying of cancer

in a hospice; none were given artificial fluid therapy, and the time from entry into the study to death ranged from 1 to 5 days (median, 2 days). Fifty percent had normal

serum osmolalities below the upper limit of normal and a urea concentration below 11.0 mmol/L, which was not considered evidence of dehydration. The others had

evidence of dehydration. There was no statistically significant relationship between hydration status and the symptoms of dry mouth and thirst noted by about 85% of

those able to respond clearly to questions. Almost all of the latter group were on medications known to cause dry mouth and/or had other causes for this condition

such as mouth breathing, candidiasis, and past treatment with chemotherapeutic agents and radiotherapy ( 109). Such reports put into question the use of tube and



orally administered fluids to the dying patient.

The common use of intravenous isotonic glucose—with or without electrolytes or vitamin additives—in this situation also seems less defensible because this prevents

ketosis and inhibits some dehydration, thereby overcoming their contribution to comfort and prolonging the dying process.

In the management of such patients with the stated goal of comfort, one must also consider some of the undesirable effects of tube feeding, whether via nasal or

gastrostomy tube. These include increased agitation, accidental or self-extubation often requiring restraints, and aspiration pneumonia ( 110, 111); in addition,

leakage, diarrhea, or impaction may occur. As Cranford has stated “to a large extent limiting nutrition and hydration is more a medical than a moral issue” ( 19).

Management of symptoms that may arise when artificial nutrition and hydration are withheld or withdrawn is outlined in Table 102.1.



Table 102.1 Management of Symptoms That Occasionally Complicate Withdrawal of Artificial Nutrition and Hydration



CHAPTER REFERENCES

Note: The references to judicial decisions follow the form generally used in legal writing. Following the case name, the first number refers to the volume of the reporter series for the decisional court;

the reported series follows as an abbreviation, then the number of the first page of the printed judicial decision, and finally the year of the decision. For example, Smith v Jones, 261 F2d448 (1990)

can be found in vol. 261 of the second series of the Federal Reporter beginning on page 448.

Published court decisions are generally available to the public through state, county, or city bar association libraries or law school libraries. West Publishing Company and Mead Data Services also

provide case text databases on a fee-for-service basis to libraries and law firms.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.



Pellegrino ED. JAMA 1974;227:1288–94.

Pennington CR. Nutrition 1996;12:56–7.

Lo B. Resolving ethical dilemmas. A guide for clinicians. Baltimore: Williams & Wilkins 1995.

Conrad LI, Neve M, Nutton V, et al. The western medical tradition: 800 BC–1800 AD. New York: Cambridge University Press, 1995, chapter 1.

Siegler M. Mayo Clinic Proc 1993;68:461–7.

Pellegrino ED. JAMA 1993;269:1158–62.

Beauchamp TL, Childress JF. Principles of biomedical ethics. 1st ed. New York: Oxford University Press, 1979.

Mahowald MD. Clin Geriatr Med 1994;10:403–18.

Wolf SM. Am J Law Med 1994;20:395–487.

Pellegrino ED. JAMA 1994;271:1668–70.

Beauchamp TL, Childress JF. Principles of biomedical ethics. 4th ed. New York: Oxford University Press, 1994.

Jones AH. Lancet 1997;349:1243–6.

Pellegrino ED, Thomasma DC. The Christian virtues in medical practice. Washington, DC: Georgetown University Press, 1996.

Freedman F. Duty and healing foundation of a Jewish bioethics. Available only on the World Wide Web at http://www.mcgill.ca/CTRG/bfreed/ .

Engelhardt HT Jr. The foundations of bioethics. 2nd ed. New York: Oxford University Press, 1996.

Union Pacific Ry v Botsford, 141 US 250, 251 (1891).

Bouvia v county of Riverside, 159780 Riverside Co. CA. Sup. Ct. 1984.

Bouvia v Superior Court (Glenchur), 179 cal. App. 3d 1127 1986;225 cal. Rpts. 297.

Cranford RD. Law Med Health Care 1991;19:13–22.



19a. Columbia Broadcasting Co. Sixty Minutes. Interview with Mike Wallace broadcast Sept 8, 1997.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.



Pellegrino ED. NY State J Med 1977;77:1456–2.

Nevins M. Am Coll Physicians Observer 1986;Mar:13–16.

Applebaum PS, Grisso T. N Engl J Med 1988;319:1635–8.

Howe EG, Gordon DS, Valentin M. Law Med Health Care 1991;19:27–33.

President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. Deciding to forego life-sustaining treatment. A report on the ethical, medical,

and legal issues in treatment decisions. Washington, DC: United States Government Printing Office, 1983;3, 61–62.

Hastings Center. Guidelines on the termination of life sustaining treatment and the care of the dying. Briarcliff Manor, NY: Hastings Center 1987;50.

Law Reform Commission of Canada. Report on euthanasia, aiding suicide and cessation of treatment. Report 20 to the Minister of Justice and Attorney General. Ottawa 1983;11–12, 32, 35.

Personal communication from Y. Roy, Senior General Counsel, Criminal Law Policy Section, Department of Justice, Canada, March 28, 1996.

Quill TE, Cassel CK. Ann Intern Med 1995;122:368–74.

Cimino JE. Topics Clin Nutr 1993;9:29–34.

Pellegrino ED. Ann Intern Med 1995;122:377–8.

Gert B, Bernat JL, Mogulnicki RP. Hastings Cent Rep 1994;24(4):13–5.

Orentlicher D. JAMA 1992;267:2101–4.

Tiemstra JD. J Clin Ethics 1995;6:163–5.

Logan RL, Scott PJ. Lancet 1996;347:595–8.

Veach RM, Spicer CM. Am J Law Med 1992;18:15–56.

Gatter RA Jr, Moskop JC. J Med Philos 1995;20:191–205.

Lantos JD, Singer PA, Walker RM, et al. Am J Med 1989;87:81–4.

J Med Philos 1995;20:109–21, 145–63.



38a. J Clin Ethics 1995;6:112–26, 128–32, 138–48.

39.

40.

41.

42.

43.



Schneiderman LJ, Jecker NS. Doctors, patients, and futile treatment. Baltimore: Johns Hopkins University Press, 1995.

Brody AB, Halevy A. J Med Philos 1995;20:123–44.

Spielman BJ. Law Med Ethics 1995;23:136–42.

Halevy A, Brody B. JAMA 1996;276:571–74.

British Medical Association. World Medical Association declaration on hunger strikers in medicine betrayed. The participation of doctors in human rights abuse. London: Zed Books, 1992:212–4,

119–31.



43a. Glick SM. N Engl J Med 1997;336:954–6.

43b. Silove D, Curtis J, Mason C, et al. JAMA 1996;276:410–5.

43c. Annas GJ. (Editorial) Br Med J 1995;311:1114–5.

44.

45.

46.

47.

48.

49.

50.

51.



The Patient Self-Determination Act, sections 4206 and 4751 of the Omnibus Budget Reconciliation Act of 1960; P.L. 101-508 Nov 5, 1990.

Annas GJ. N Engl J Med 1994;330:223–5.

Cugliari AM, Miller T, Sokal J. Arch Intern Med 1995;155:1893–8.

Sehyal A, Gallbraith A, Chesney M, et al. JAMA 1992;267:59–63.

Danis M, Southerland LI, Barrett JM, et al. N Engl J Med 1991;324:882–8.

Blackhall LJ, Murphy ST, Frank G, et al. JAMA 1955;274:820–5.

Carrese JA, Rhodes LA. JAMA 1995;274:826–9.

Gostin LO. JAMA 1995;274:844–5.



51a. Levinsky NG. N Engl J Med 1996;335:741–3.



52.

53.

54.

55.

56.



Gill TM, Feinstein AR. JAMA 1994;272:619–26.

Guyatt GH, Feeny DH, Patrick DL. Ann Intern Med 1993;118:622–9.

Tsevat J, Cook EF, Green ML, et al. Ann Intern Med 1995;122:514–20.

Danis M, Garrett J, Harris R, et al. Ann Intern Med 1994;120:567–73.

American Medical Association Council on Ethical and Judicial Affairs. Code of medical ethics. Current opinions with annotations, 1996–1997 ed. Chicago: American Medical Association,

1996;41–51.

57. The Multi-Society Task Force on PVS. N Engl J Med 1994;330:1499–508.

58. The Multi-Society Task Force on PVS. N Engl J Med 1994;330;1572–9.

58a. Ashwell S, Cranford R. N Engl J Med 1995;333:130 (correction).

59. Andrews K, Murphy L, Munday R, et al. Br Med J 1996;313:13–6.

60. Cranford R. Br Med J (Editorial) 1996;313:5–6.

60a. Zeman A. Lancet 1997;350:795–9.

61.

62.

63.

64.



Royal College of Physicians Working Group. J R Coll Physicians (London) 1996;30:119–21.

Nelson LJ, Rushton CH, Cranford RE, et al. J Law Med Ethics 1995;23:33–46.

In re Quinlan 70 NJ 10, 355 A2d; 647, 1976.

People v Barber A025586 Los Angeles Sup. Ct. 1983.



64a. Barber v Superior Ct. 147 CA 3d 1006 Ca, 1983.

65. Myers DW. Arch Intern Med 1985;145:125–7.

66. Matter of Claire Conroy 453, 464 A 2d 303 (NJ App. 1983).

66a. Matter of Claire Conroy rev'd cf also 98 NJ 321, 486A 2d 1209 (1985).

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.



Emanuel EJ. Lancet 1988;1:170–71.

Emanuel EJ. Lancet 1988;1:106–07.

Emanuel EJ, Emanuel LL. JAMA 1992;267:2067–71.

Lynn J. JAMA 1992;267:2082–84.

Coordinating Council on Life-Sustaining Medical Treatment Decision-Making by the Courts. Guidelines for state court decision-making in authorizing or withholding life sustaining medical

treatment. Williamsburg, VA: National Center for State Courts, 1991.

Council on Ethical and Judicial Affairs of the American Medical Association. Withholding or withdrawing life prolonging medical treatment. Chicago: American Medical Association, 1986.

Cruzan v Harmon 760 SW 2d 408 (MO 1988).

Cruzan v Missouri Department of Health 760 sw 2d 408 aff 110 S. Ct. 2841 (1990).

Annas GJ. Law Med Health Care 1991;19:52–9.

Cruzan v Mouton Estate No CV 384-9P (Circ. Ct Jasper Co, filed Dec. 14, 1990 (Teel).

Lo B, Steinbrook R. Ann Intern Med 1991;114:895–901.

Annas GJ. N Engl J Med 1990;323:670–3.

Bioethicists' statement on the US Supreme Court's Cruzan decision. N Engl J Med 1990;323:686.

King PA. Law Med Health Care 1991;325:511–2.

Angell M. N Engl J Med 1991;325:511–2.

Annas CJ. N Engl J Med 1994;330:1542–5.

Lancet (Editorial) 1993;341:41.

Wall M. Lancet 1995;346:368.

Brahams D. Lancet 1992;340:1534–5.



85a. Case Records of the Massachusetts General Hospital—Case 26-1997. N Engl J Med 1997;337:549–56.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.



American College of Physicians ethics manual. Ann Intern Med 1992;117:947–60.

Council on Scientific Affairs American Medical Association. JAMA 1990;263:426–30.

Dworkin R. Life's dominion: an argument about abortion, euthanasia, and individual freedom. New York: Alfred A Knopf, 1993:195.

Nelson JL. J Law Med Ethics 1995;23:143–8.

Callahan D. Hastings Center Rep 1995;25(6):25–31.

Dresser R, Whitehouse PJ. Hastings Center Rep 1994;24:6–12.

Fletcher JC, Spencer EM. Lancet 1995;345:271–2.

Kapp MB. J Am Geriatr Soc 1992;40:722–6.

Payne K, Taylor RM, Stocking C, et al. Ann Intern Med 1996;125:104–10.

Hodges MO, Tolle SW, Stocking C, et al. Arch Intern Med 1994;154:1013–20.

Asch DA, Cristakis NA. Med Care 1996;34:103–11.

Faber-Langendoen K. Arch Intern Med 1996;156:2130–6.

Coggan F. Proc R Soc Med 1977;70:75–81.

Maillet JO. J Am Diet Assoc 1995;95:231–4.

Schostak Z. J Med Ethics 1994;20:93–100.

Brody H, Campbell ML, Faber-Langendoen K, et al. N Engl J Med 1997;336:652–7.

Phillips P, Rolls BJ, Ledingham GG Jr, et al. N Engl J Med 1984;311:753–9.

Zerwekh JV. Nursing 1983;83:47–51.

Printz LA. Geriatrics 1988;43:84–8.

Schmitz P. Law Med Health Care 1991;19;19:23–6.

Dunlop RJ, Ellershaw JE, Baines MJ. J Med Ethics 1995;21:141–3.

Waller A, Adunski A, Hershkowitz J. Lancet 1991;337:745.

McCann RM, Hall WJ, Groth-Juncker A. JAMA 1994;272:1263–6.

Ellershaw JE, Sutcliffe JM, Saunders CM. J Pain Symptom Manage 1995;10:192–7.

Giocon JO, Silverstone FA, Graner LM, et al. Arch Intern Med 1988;148:429–3.

Quill TE. Arch Intern Med 1989;149:1937–41.



SELECTED READINGS

Beauchamp TL, Childress JF. Principles of biomedical ethics. 4th ed. New York: Oxford University Press, 1994.

Lo B. Resolving ethical dilemmas. A guide for clinicians. Baltimore: Williams & Wilkins, 1995.

Pellegrino E, Mazzarella P, Corsi P, eds. Transcultural dimensions in medical ethics. Frederick, MD: University Publishing, 1992.

Pellegrino ED, Thomasma DC. The virtues in medical practice. New York: Oxford University Press, 1993.



Chapter 103. Recommended Dietary Intakes: Individuals and Populations

Modern Nutrition in Health and Disease



Chapter 103. Recommended Dietary Intakes: Individuals and Populations

GEORGE H. BEATON

Evolution of Thinking and Understanding

Perceived Purposes of Standards

Scientific Basis of Estimated Human Nutrient Requirements and RDAs

Terminology and Concepts

Estimating and Describing Group Needs

Homogeneous Groups

Heterogeneous Populations

Addressing Upper Limits to Nutrient Intake

Criteria of Requirement and Adaptation

Dietary Reference Intakes

Epilogue: Possibilities and Limitations for Application

Dedication

Chapter References

Selected Readings



We know accurately only when we know little; with knowledge doubt increases.

(Goethe)

That which is sure is not sure; as things are they shall not remain.

(Bertolt Brecht)



EVOLUTION OF THINKING AND UNDERSTANDING

Perceived Purposes of Standards

From ancient times man has known that health and physical well-being depend on diet and that particular foods have special virtue for curing diseases ( 1). More than

2000 years ago, liver was known to cure night blindness. There was, and still is, an abundance of information about diet and disease in the folklore of almost every

culture. Unfortunately, in modern times in North America, that folklore continues to expand. The work of Lind (1753) was a milestone for human nutrition for it is

perhaps one of the earliest recognized applications of the methods of science to the claims of folklore. Lind, with his now-famous treatise ( 2), established the

relationship between diet and scurvy, validating what had been folklore. The first “dietary standard” given any sort of official status was created in 1862 at the request

of the British Privy Council ( 3). That standard was intended “to prevent starvation diseases” and was to serve the purpose of evaluating estimated intakes in the first

dietary surveys conducted in Britain shortly thereafter. Table 103.1 provides a brief summary of some markers of the changes in perceived purposes of dietary

standards.



Table 103.1 Changing Purposes of Dietary Standards—Some Markers



Since these early works, much has changed, slowly at first and then with ever-increasing speed. Today that speed is feverish. Even as this chapter is being written, it

is being superseded by events. The changes have been in many areas. First, our knowledge concerning the diverse functions of nutrients has increased

phenomenally. However, this rapid progress has not necessarily improved our ability to estimate human nutrient needs; rather, it has fostered an evolving concept of

what we hope to achieve (the markers of “inadequacy” are changing). Change is also evident in shifts in the type of evidence used in attempts to estimate human

needs. The earliest requirement estimates were based on epidemiologic observation—the intakes of healthy versus sick people. Gradually this gave way to

experimental approaches seen as more respectable—balance studies, planned deprivation and repletion studies, and theoretical modeling based on experimentally

established functions of nutrients. It was soon recognized that estimates based on experiments conducted under controlled conditions had to be confirmed under field

conditions; thus, epidemiologic evidence came back into its own.

We have now gone full circle and are starting again. Epidemiologic evidence currently suggests relationships between consumption of certain types of foods and

some forms of cancer, cardiovascular disease, and other noncommunicable chronic diseases. Unfortunately, although the dietary associations seem clear, identifying

the active factors in the foods is not clear. For that reason, attempts to develop estimates of human needs for fat, specific fatty acids, carotenoids, tocopherols,

ascorbic acid, and “antioxidant substances,” and to incorporate this knowledge into public health policy has faced many difficulties and much controversy. It is not

necessarily that we know too little; perhaps we know too much. Yet we cannot seem to put all the pieces together into a coherent whole (for discussion of these

issues, see Garza et al. [17]).

Another phenomenal and highly problematic evolution, particularly since World War II, has been in the scope and sheer divergence of applications of the

recommended dietary allowances (RDAs). A few years ago, the Food and Nutrition Board (FNB) established a committee to identify and catalogue the then-current

uses of the RDAs. The 1983 report of that committee was never published, though a summary of the findings was included in the FNB 1994 report ( 18) “How Should

the Recommended Dietary Allowances be Revised?” That summary (reproduced verbatim as Table 103.2) did not imply endorsement by the FNB of these uses and,

as stated, they certainly are not all endorsed by the present author. What is not apparent from the table is that many of these applications have been given a legal

dimension by being incorporated into regulations issued by federal and state bodies. This has further complicated a tangled situation. For example, the Food and

Drug Administration (FDA) established the U.S. recommended daily allowances, derived from the 1968 recommended dietary allowances, as a reference point for food

labeling. Numerically and conceptually these two sets of numbers are very different, although they were, at one time, related. The most confusing aspect is that the

abbreviation RDA applies to both. In fact the name used by the FDA is identical to the name used in the first FNB report on RDAs ( 19).



Table 103.2 Reported Uses of the RDAs



With the growth of scientific knowledge about the function of nutrients as well as about factors likely to affect requirements, inevitable pressure developed to alter the

criteria of adequacy, the specification of “requirement for what?” The nutrition community adopted the vague and unmeasurable goal of establishing “optimal intake.”

The North American public has monumental expectations of the relationship between diet and all aspects of their health and well being. Although these expectations

may be unrealistic, they are now a part of the climate (and folklore?) in which the nutritional scientist must work. This phenomenal belief in the power of nutrients has

led to massive growth in the use of nutritional supplements and pressure to allow health claims for dietary supplements ( 20). As criteria of adequacy shifted,

requirement estimates tended to increase. In some cases, as the base of knowledge increased, requirement estimates actually decreased, either because previously

accepted evidence was found inappropriate or incorrectly interpreted or because committees felt comfortable in lowering the “margin of safety,” sometimes called the

“margin of doubt.”

Another rapidly evolving feature of the estimation, description, and application of human nutrient requirement estimates is the use of statistical concepts. The need to

address the distribution of both requirements and intakes has been progressively realized. When assembling evidence to estimate human requirements, one must be

aware of exactly what each type of evidence actually estimates. It could be the average of individual requirements for a class of individuals; it could be the required

concentration of nutrient in a diet consumed ad libitum by subjects in an experimental study; it could be a necessary group mean intake in an epidemiologic study.

These estimates are not the same conceptually or numerically. Nevertheless, they can all be related by proper application of statistical concepts concerning univariate

and bivariate (joint) distributions of intakes and requirements. In parallel, given the appropriate construct and the relevant information, one can move from estimating

human nutrient requirements to specifying the necessary concentration of nutrients in the diet or to estimating necessary group mean intakes.

Only relatively recently (in the last 20–25 years) has there been a real effort to merge the evolving biologic and statistical knowledge bases concerning estimation and

application of human nutrient requirements. Most experimental scientists and even many epidemiologists, consider statistics a tool used to test hypotheses about

effects or associations, which it is. Statistics can also be used to understand and describe relationships; in the past this has received too little recognition. The Food

and Agriculture (FAO)/World Health Organization (WHO) series of reports on nutrient requirements, and the report of the eighth Joint Expert Committee on Nutrition

(21) chronicle the progressive integration of statistical concepts into the estimation, description, and application of nutrient requirement estimates. Specifically, one

can follow the attempts to describe and work with distributions of nutrient requirements and gradually see conceptual development of approaches to concurrently

dealing with distributions of requirements and of intakes. At first this involved assessing the apparent adequacy of diets in population groups. Gradually, merging of

nutrition and statistics led to estimation of necessary nutrient:energy ratios and, in 1996, to estimation of necessary group mean intakes ( 16).

In 1945, when Canadian and U.S. committees charged with developing a Dietary Standard and Recommended Dietary Allowances, respectively, disagreed and went

their own ways, a major issue was the stated goal and suggested interpretation of the 1945 revision of the U.S. RDAs ( 22):

The allowances are intended to serve as a guide for planning an adequate diet every normal personof the population and not for the average member of the group categories . . . In order to meet the needs of the whole population it is

for

necessary to satisfy the requirements of those with less efficient usage. Because the allowances take into consideration the requirements of those at the upper level of the normal range of requirements, they allow a factor of safety for persons

who have an average or less than average requirement. In most categories this factor of safety is approximately 30%, but for many persons the allowances cover only the amounts needed for maintenance . . . When the tables are used to

estimate the per capita needs of population groups and the agricultural production required to supply such groups, the allowances should be weighted by the composition of the population according to the categories and ages. When the

allowances are weighted in this manner the average allowance for each individual [the per capita allowance] is lower than that for a man for all factors except calcium. . . . Even when weighted in this manner, the allowances are greater than the

average or the mode of the need for many factors because the allowances are planned to be sufficiently high to meet the needs of normal persons who are less efficient in utilization.



The apparent disagreement between Canadian and U.S. committees at that time appeared to relate more to a difference in philosophy and approach than to any real

difference in scientific judgment. In the United States, strong emphasis was placed on the individual and ensuring adequacy for every individual. In Canada, greater

emphasis was and, until relatively recently, continued to be, placed on the group or population. Parallel differences between U.S. and Canadian perspectives and

approaches can be seen in aspects of their respective healthcare systems and, for example, approaches to asymptomatic hypercholesterolemia screening (favored in

the U.S. and opposed, except for high-risk individuals, in Canada). The open debate appears to have centered on how the U.S. RDAs were to be applied to

population groups. Both Canada and the U.S. accepted the reality of variation in need among individuals, but the Canadian group felt that a more realistic estimate of

needs could be developed if factors believed to affect requirement (and hence explaining part of the variation among individuals) were built into the dietary standard

(13). Unfortunately, that Canadian report ( 13) described the figures as representing a “nutritional floor beneath which maintenance of health in people cannot be

assumed.” That phraseology generated major reaction and for years split the two communities. Had the report, instead, said that the numbers “indicate a nutritional

floor above which maintenance of health [with reference to the tabulated nutrients] could be assumed,” the debate might have been much less acrimonious, and

harmonization of standards might have followed many years ago instead of still being discussed today.

What is interesting and important is that, by 1945, the major issues that have plagued the development, interpretation, and application of the RDAs were already on

the table. A paper written in that era captures the crux of one of the major issues—the need to marry statistical and biologic concepts. In 1945 Pett et al. ( 23) noted

four basic uses of dietary and nutritional standards:

i. for use in calculating the nutrient requirements of a population in connection with a national food and nutrition policy for health.

ii. for use in evaluating the dietary status of a group of people from the total and average quantities of foods eaten, or providing foods for such a group.

iii. for use in establishing regulations under the Foods and Drugs Act [a Canadian Act] governing the contents of foods, dietary supplements or drugs and of

allowable claims for them.

iv. for use in evaluating the dietary status of an individual person from the foods eaten, or from the foods purchased.

The authors commented (13) (absolutely correctly), “it is not likely that a single set of figures in a table could be used properly for all these purposes . . . Nevertheless

it is possible to marshal information about requirements in a manner that forms a uniform basis for various calculations suited to different purposes . . . It is thus

apparent that the problem of arranging and interpreting figures for human requirements is basically statistical.” The paper went on to speak of “the probability of the

individual's suffering from a serious deficiency or from a suboptimal state of nutrition” and “the likelihood of finding specific medical signs in any particular individual.”

They pointed out that “Medical evidence is necessary before the story is known about any individual. However, if the [intake and requirement] curves are properly

determined it would indeed be possible and correct to estimate the proportions of the population receiving an insufficient amount of any nutrient.” Those authors had

the elements of the puzzle in hand and mapped out a solution. However, at the time, the story fell on “stony ground” (L.B. Pett, personal communication, 1989), and

there was little or no follow-up.

It has taken the rest of us 50 years to reachieve the insights expressed in that paper. We are finally there, this time equipped with more complete scientific knowledge

and statistical concepts and tools that are more widely accepted and applied. We still have a long road to follow before the nutrition community and other users of the

RDA reports are ready to use the new insights and new techniques they bring. The FNB ( 18) has indicated that in the current revision of the RDAs, under way as this

chapter is being written, the board will direct major effort to developing a guidelines report relating to interpretation and application of whatever Dietary Reference

Intakes (DRIs) or RDAs it chooses to publish. Perhaps the dream of Pett and his colleagues will finally be realized. Perhaps also we will see a much greater

willingness to harmonize nutrient requirement reports around the world. It is amazing that in an international community of knowledgeable scientists,

misunderstandings rather than true scientific disagreements have persisted for so long.

This chapter attempts to document and illustrate some of the major evolutions mentioned above. In so doing, the author hopes that the reader will recognize that while

new concepts and approaches at first seem complicated and nonintuitive, the benefits of understanding and applying them correctly are great, and before long they