Print This Article

Ethical Conflicts in the Care of Anorexia Nervosa Patients

By Arnold E. Andersen, M.D.
University of Iowa School of Medicine, Iowa City
Reprinted from Eating Disorders Review
March/April 2008 Volume 19, Number 2
©2008 Gürze Books

Four principles form the basis of modern biomedical ethics.1 These guiding principles balance Emmanuel Kant's Categorical Imperative ("You gotta do the right thing, no matter what it takes") versus John Stuart Mill and Jeremy Bentham's Utilitarianism ("make the best decision for everyone all around"). As a group, the resulting principles are often called Principalism:

  1. Respect for autonomy (the patient's informed choices take priority).
  2. Nonmalfeasance (do no harm, Primum non nocere).
  3. Beneficence (doing what is best for patients, not always with their consent).
  4. Justice (balancing individual and social costs, benefits, and risks).

Implied in the above group is the assumption that practicing these ethical principles takes precedence over self-interest.

Nonmalfeasance and beneficence have historically been honored. In contrast, only recently has respect for autonomy and (social) justice become prominently emphasized in medical practice. This in great part resulted from growing distrust of all institutions in general, manifested by the social conflict during the Vietnam War.

Four guiding principles are generally applicable in most situations, but none are without exceptions in specific cases. There are conflicts within and between these principles that are relatively specific for anorexia nervosa (AN) because of the unique features of this eating disorder, fo. For example:

  1. The intermix of medical starvation signs and symptoms with the mental and behavioral consequences of the overvalued beliefs underlying the disorder;
  2. The common perception that AN is a voluntary disorder responsive to self-determination, both in development of the illness and in the process of regaining health;
  3. Moralizing disorders of eating behavior and weight as disorders of personal failure;
  4. The common stereotype of AN that limits perception of illness to teenage Caucasian girls;
  5. The multi-factorial concept of etiology that generates competing approaches to treatment; and
  6. The relatively robust appearance and behavior of many patients, even those with serious AN.

Respect for Autonomy vs. Beneficence

Severely ill AN patients often look far more competent than they actually are. Deciding when a patient with AN is too ill to make an informed decision about the need for treatment requires placing the patient into a category ("too sick to be autonomous, make the decision for her") on a linear dimension (gradually decreasing weight and mental function). Carefully delineated respect for both autonomy and beneficence, combined with experienced assessment of mental and behavioral functioning in AN, will determine when beneficence trumps autonomy. The situation is analogous to determining that a deluded, depressed patient with suicidal plans or a patient with schizophrenia experiencing command hallucinations has to be involuntarily committed for their safety and treatment despite their protests. Patients with severe AN give the illusion of sanity even when they are driven by deadly irrationality, failing to appreciate that AN has the highest death rate in psychiatry, approximately 19%.

Guarda et al2,3 have shown that shortly after patients feel coerced to receive inpatient treatment, they change their minds and come to appreciate the need for treatment. At discharge, they say, in so many words, "Thank you," rather than "I'll sue you."

Nonmalfeasance vs. (Unintended) Malfeasance

The issue is not about intended malfeasance, which is beyond the pale, but more subtle forms of malfeasance in practice, either active, albeit unintended and even good-willed, or passive. Both forms may occur in a number of ways, illustrated by the following specifics:

Active harm. Inappropriate refeeding may be deadly. Enough empirical literature exists documenting ways to minimize refeeding problems that should alert clinicians to avoid fluid overload, excess caloric content, or premature use of central lines. Studies have documented that severe medical complications as a result of refeeding include death, cerebrovascular accidents, pneumothorax, and severe metabolic shifts. Some refeeding peripheral edema and nonspecific gastrointestinal complaints are common and expected, but repeated severe medical consequences from refeeding suggests a review of methods used.

Another example of active harm involves using endless work-ups to look for a medical cause for symptoms, using a "rule-out" approach of discovering medical etiology rather than a "rule-in" approach based on a competent mental status examination. Until relatively recently, the diagnosis of AN was often delayed for years4 because of endless medical examinations, each with their own morbidity, such as gastrointestinal scoping, organ biopsies, and excessive laboratory tests. At times, the medical consequences of starvation are mistaken for causation, resulting, for example, in the prescription of thyroid replacement for starvation-related adaptive low lab values.

A third example of active harm is abandonment of the AN patient because of failure to recognize the countertransference that is stirring up feelings and behaviors of rejection, especially in narcissistic clinicians and lack of capacity for empathy.

Finally, of limited current implementation (thankfully), but nonetheless bruited about in European circles primarily, has been the devastatingly unethical suggestion of consent to euthanasia or passively agreeing to withhold treatment in some cases of AN.5

Passive harm. Some examples of passive harm include failure to refer a patient with AN, either for diagnostic confirmation or a more experienced level of care when personal skills are not matched to the seriousness or complexity of the illness. Another example is failure to develop a biopsychosocial formulation with integrative treatment, and instead relying on single modalities determined to be insufficient by themselves, such as medications alone, or a psychodynamic approach without attention to weight restoration.

Failure to ask administrative colleagues to apply for use of third-party medical insurance benefits for the medical starvation component of anorexia, relying only on the generally much more limited psychiatric benefits is one more example of passive harm. Federal Courts have determined that starvation is a medical diagnosis appropriate for medical benefits, and AN is a medical disorder up to return to 85% of healthy weight.6 Treatment costs for AN may be huge, but they are not larger than those for transplant surgery or treatment of leukemia, and every effort should be made to actively secure full coverage of treatment costs. Not to act on the patient's benefit when such action would be in their best interest and is a reasonable part of the professional role can be seen as passive neglect.

Authentic Beneficence vs. Authoritarian or Pseudo-beneficence

True beneficence is at the heart of patient care. It does not imply a patronizing or authoritarian approach, but can veer in those directions. On the one hand, an out-of-date beneficence fails to discuss with the patient the diagnosis, treatment options, prognosis, benefits and risks from treatment options or from no treatment at all. Into this category would fall the "doctor knows best" attitude and behavior (authoritarian beneficence). It would also include limiting the treatment options presented to an overly narrow category consisting only of the skill-set of the clinician, even if not best suited to the needs of the patient. On the other hand, a false view of updated beneficence presents a "cafeteria approach" to treatment options, with no guidance or recommendations, but an unguided list of all possible approaches. "Whatever the patient wants" is pseudo-beneficence and not in the best interest of the patient.

An ethical clinician will present a clear understanding of the diagnosis, treatment, and prognosis of AN, and will offer recommendations based on his/her experience and evidence from sound studies. True informed consent does not fail to offer guidance and an opinion of the best course of treatment most likely to lead to lasting improvement. True informed consent will also acknowledge limitations in the clinician's training, and will consent to a request for a second opinion or suggest a referral when appropriate. Not every clinician who can competently diagnose eating disorders can treat all forms of eating disorders at every level of severity. Any significant conflicts of interest should be disclosed, such as financial interests in the program to which the patient is referred.

Authentic beneficence grows out of a warm, non-possessive empathy that appreciates the patient's struggles with giving up the perceived benefits of illness (anorexia is often ego-syntonic) for what seems to be perceived as a loss of control resulting from becoming healthy. It is crucial to listen with the "third ear," and to try to reach out to the healthy part of the patient's mind and heart. The healthy side always constitutes a larger part of the patient's being than does the ill side. Beneficence will appreciate that the patient's concerns (i.e., hair falling out, dry skin) may seem to be clinically less important than the threat of death or chronic illness, but these patient concerns need to be incorporated into a treatment plan. The view of illness from within the illness overlaps only partially with a textbook description of the disease. Authentic beneficence always includes a biopsychosocial formulation that integrates palpable concern for patient suffering with the best available interventions treating the medical, intra-psychic, and interpersonal aspects of the disorder.

Social Justice vs. Business as Usual.

Social justice views the patient's disorder within a systems approach, as a tapestry into which individual and societal costs, needs, and benefits are interwoven. The patient is the de facto part of a broad social network of family members, caregivers, communities, governmental regulators, third-party payers, legislators, and judges. While in theory each component has the patient's welfare at its center, in practice, each "planet" in the solar system of care has its own orbit, its own self-serving goals, and, at times, its own counter-therapeutic practices.

On the part of legislatures, lip service has been given to parity, but seldom has it resulted in a truly impartial approach to all illnesses as worthy of comparable levels of care based on intensity of illness. So-called "mental disorders" have been carved out for more limited health care benefits than "medical disorders," and eating disorders specifically have been the most ruthlessly reduced in coverage limitations. Many psychiatric disorders have 30 days at best of inpatient care coverage, a small number of outpatient visits, and large co-payments. In contrast "medical disorders" may have almost unlimited benefits. The result has been shorter and shorter inpatient treatments for patients with AN, despite evidence-based studies documenting that inpatient treatment short of an attempt at acute full remission leads to more relapses and more frequent re-hospitalizations.7 Repeat hospitalizations are considered "business as usual," while an adequate length of stay is considered a money-losing proposition, although over the course of illness the latter costs less.

True social justice does not tolerate discrepancies in service based on race, gender, or sexual orientation. Excluding males from many inpatient or residential care programs represents an injustice, as does failure to reach out to diverse ethnic and racial communities. Data increasingly demonstrate that these groups have a high and increasing prevalence of eating disorders.

The healthcare training system may teach the importance of evidence-based treatment, but fail to provide training for psychiatrists in the most evidence-based psychotherapy for eating disorders, cognitive-behavioral therapy (CBT). Training in CBT has been largely abandoned in many hospital training programs. Unfortunately, too many programs neglect to provide residents with supervised experiences in the multiple therapeutic skills necessary for comprehensive treatment that include evidence-based psychotherapeutics.

Advocacy for social justice includes, besides a call for comprehensive treatment programs and adequate reimbursement of care, a call for funding of both clinical and transitional research studies for increased knowledge of AN leading to clinical advances, an end to discrimination in programs regarding gender, community education programs, screening for early diagnosis, and preventive intervention in demonstrated areas of effectiveness. It also calls for an end to stigma regarding eating disorders in particular, interaction with legislative bodies for actualizing parity, and willingness to appear before administrative justices and other personnel within the justice system to plead the patient's case for funding. States that currently do not admit severely ill AN patients for involuntary treatment even when clearly necessary in life-threatening cases need to remediate this deficit in view of evidence that selective involuntary admission and treatment can be life-saving.

Guiding one's practice within the boundaries of these four principles not only does not stifle vigorous treatment or compassionate clinician interaction with patients, it broadens historic guidelines to accentuate an appropriate increased appreciation of patient autonomy as a therapeutic partner and widens the horizon of concern to a comprehensive advocacy for social justice within and between all the components of the care system.

(Note: The ethics of care of AN patients are not synonymous with legal requirements or with "guild" guidelines, but rather spring from a profound appreciation of the "I-thou" nature of the clinician-patient relationship. The word "patient" rather than client is used because patient comes from the word for suffering, whereas "client" implies a primarily economic interaction. AN patients truly suffer even when they doth protest. Ethical principles will continue to evolve as biomedical science presents new and currently unforeseeable challenges.

About the Author

Dr. Andersen is Professor of Psychiatry at the Unviersity of Iowa College of Medicine, Iowa City, IA. He is the author of numerous books and articles, and is a charter member of the EDR Editorial Advisory Board.


1. Beauchamp TL, Childress JF. Principles of Biomedical Ethics, 5th Ed. Oxford University Press: Oxford, England, 2001.

2. Guarda AS, Pinto AM, Coughlin JW, Hussain S, Haug NA, Heinberg LJ. Perceived coercion and change in perceived need for admission in patients hospitalized for eating disorders. Am J Psychiatry. 2007. 164(1):108.

3. Andersen, AE. Eating disorders and coercion (Editorial). Am J Psychiatry. 2007. 164(1):9.

4. Andersen AE. Unmasking a current medical pretender: anorexia nervosa in gastrointestinal practice. Eur J Gastroenterol Hepatol. 2004. 16; 1123.

5. Draper H. Anorexia nervosa and respecting a refusal of life-prolonging therapy: a limited justification. Bioethics. 2000 14; 120.

6. U.S. District Court, Eastern District of New York. Manhein vs. Travelers Insurance. 92-CV-5466 (JG). September 15, 1995.

7. Baran SA, Weltzin TE, Kaye WH. Low discharge weight and outcome in anorexia nervosa. Am J Psychiatry. 1995. 152; 1070.