AMA Journal of Ethics®

Illuminating the art of medicine

Journal of Ethics Header

AMA Journal of Ethics®

Illuminating the art of medicine

Virtual Mentor. October 2003, Volume 5, Number 10.

Clinical Cases

  • Print

Psychiatrist's Role in Involuntary Hospitalization

Psychiatrists face difficulties when deciding on treatment plans for patients who are not interested in receiving help or are lacking the capacity to make responsible decisions.

Commentary by Jennifer Bremer, MD, Roy Lubit, MD, and Robert Orr, MD, CM

Psychiatrist Lisa Feinberg had been working with Suzanne Martin for 2 years. Miss Martin was referred to Dr. Feinberg by her primary care physician who suspected that Suzanne's extreme low weight was indicative of anorexia nervosa (AN). Dr. Feinberg agreed with the diagnosis of AN and began meeting with Suzanne weekly. Suzanne Martin, a 19-year-old sophomore at the state university, was an excellent student and fine musician. She managed course work, a 3-hour per day practice schedule, and a regular exercise routine with little sleep and little food. Suzanne Martin made light of what others called her "illness." She met with Dr. Feinberg mostly to keep her parents "off her back." She chatted easily with Dr. Feinberg, but the psychiatrist found it difficult to get Suzanne beyond superficial chatter, on the one hand, and deep theoretical discussions of her studies and her music, on the other hand. Suzanne avoided talking about her illness and the behaviors that must be necessary to maintain her dangerously low weight. She managed to remain just above a level of physical exhaustion and weakness that would have necessitated hospitalization.

One night Suzanne collapsed and was brought to the ER by friends over her protestations. She had received glucose and was gaining enough strength to demand to go home when her parents arrived. Her physician had been called, and he was present also. Suzanne's parents appealed to the physician to say that Suzanne was endangering her life—for all practical purposes, she was suicidal, they said—and hence should be declared incompetent to make medical decisions. Suzanne's physician had been reluctant make the declaration and had summoned to the hospital to confer about involuntary admission and artificial nutrition.

By the normally applied standards, Suzanne Martin was not incompetent to make medical decisions. She could understand the information she was given; she could analyze and measure the consequences of her refusal of treatment against an internal set of values and goals; and she could give back her decision in a coherent and consistent way. Dr. Feinberg figured that Suzanne's finely calibrated system had slipped out of control that day—a bit too much exercise or too little food. She was like a diabetic who takes too much sugar or too little insulin on a given day. One wouldn't hospitalize the diabetic against her will once physiologic balance had been restored. Dr. Feinberg feared that if Suzanne were hospitalized against her wishes and refused to eat all the food that was given her, she would be fed through a nasogastric tube. Lisa Feinberg knew Suzanne well enough to know that Suzanne would consider this a grave and obscene violation. She thought that hospitalization and the treatment Suzanne would receive if declared incompetent would set her work with Suzanne back seriously. Suzanne might even consider Dr. Feinberg's role in the commitment so serious a betrayal of trust that she would discontinue coming for therapy.

Commentary 1

Mandating Refeeding
by Jennifer Bremer, MD

From the information provided, this patient should go to an inpatient specialty eating disorders unit. Outpatient treatment has failed her; she is approaching an age where no treatment is especially effective for anorexia. Suzanne's recent medical events point towards acute danger. This patient will likely gain weight and benefit immensely from a stay on a specialty eating disorders unit.

More specifics about her medical condition would help elucidate the status of her medical condition. The more unstable the patient's medical condition is, the more justification there is for hospitalization—even if a patient refuses and must be hospitalized against her will [1]. Her response to glucose makes it appear that the cause was hypoglycemia which can be deadly [2-3]. Indicators of medical instability may include: syncope, rapid weight loss, seizures, organic brain syndrome, bradycardia, exercise-induced chest pain, arrhythmias, renal dysfunction, dehydration, tetany, and decreasing exercise tolerance [4]. Abnormal electrolyte levels can also be of acute concern. The assessment of medical urgency in Suzanne's case remains a clinical judgment though, and detailed medical information is lacking here, making this case somewhat difficult to assess.

Fortunately, involuntary hospitalization probably can be avoided with this patient. The parents' request for their daughter's hospitalization suggests that they may be willing to use their influence to effect her hospitalization. Such vigorous persuasion is sometimes viewed as controversial. Using parental influence to help a child's nourishment and survival is reasonable and effective and can be effectual in cases of anorexia nervosa. Data supports the efficacy of the Maudsley family therapy approach [5]. The first phase of such therapy guides parents to use whatever measures they must—within reason—to mandate regular meals large enough to cause weight gain.

It is effective for parents to declare to their child that they will not allow her to starve to death, no matter what steps they must take to make this happen. It is vital for parents to tolerate their daughter's inevitable fury over this mandate and yet insist on hospitalization. In this case, it appears the daughter will comply since she has a history of complying with their treatment wishes. According to the case history, she attended therapy to keep her parents "off her back."

Parents can use different types of leverage successfully but often a firm mandate for hospitalization is enough and is preferred. When further influence is needed, parents can refuse the child privileges, eg, refusing to pay college tuition or car payments until their child is no longer on death's doorstep.

The psychiatrist also can use her alliance with Suzanne to help her understand what must and will happen. In 2 years, the outpatient treatment appears not to have moved the patient in the right direction. It is a good use of the therapeutic alliance to help hospitalize the patient. The psychiatrist and medical team should discuss at length with Suzanne their treatment recommendations and reasons for the hospitalization. Ideally, the treatment team helps the patient to understand the necessity of inpatient care so that she willingly agrees to follow their recommendations.

Usually, though, the cognitive distortions around body weight and shape which are diagnostic criteria for anorexia nervosa impede a measured reasoning process [6]. In addition, starvation clouds thinking. Data shows processing and attention deficits in patients with anorexia nervosa [7,8]. In fact, there is anatomical change in these patients' brains such as increased ventricular size [9]. Since the parents' role in treating their daughter's illness is critical, it may be helpful to describe these cognitive changes to Suzanne's parents so they can be firm in their pleas with their daughter.

Under circumstances similar to this case, we rarely hospitalize a patient involuntarily. We avoid involuntary hospitalization because we would always prefer voluntary treatment. We often send patients to inpatient eating disorder units such as the ones at University of Iowa and Columbia University. By law we cannot send patients across state lines involuntarily. I must emphasize the importance of specialty eating disorder programs over general medical or general psychiatric units.

While it will be hard for the patient to go through this, refeeding must be the first priority. The patient's emotional upset is far less damaging than starvation. The psychiatrist's fear about the need for nasogastric feedings is unfounded; most patients do not require such measures on a specialty unit and such measures are usually avoided. While insight and understanding are important for someone with anorexia to move towards greater richness and meaning as she leaves starvation behind, refeeding must come first.

If Suzanne still declines a voluntary hospitalization and the patient's medical status is ominous, involuntary hospitalization is appropriate. This is especially true with a relatively young patient who will likely do well with inpatient treatment. While laws vary from state to state and over time, they tend to support paternalism in such circumstances. Consultation with a legal expert or ethicist may help the physician determine the correct path in a specific instance [4]. The subject of competency in anorexia is complicated and controversial, again underlining the importance of achieving voluntary hospitalization.

Amidst such controversy, it helps to recall the Hippocratic Oath's instruction to "first do no harm." Sending this girl home inevitably to starve is doing harm. Hospitalizing her will not be pleasant for anyone but will "do no harm" and may well do much good.

The principle of beneficence requires physicians to care for those who are unable to take care of themselves, as uncomfortable as it may be in our society where liberty and autonomy are treasured values. Hospitalization aims to help the patient regain her autonomy—an autonomy that the anorexia nervosa, not the physician, has taken away.


  1. Goldner EM, Birmingham CL, Smye V. Addressing treatment refusal in anorexia nervosa: clinical ethical and legal considerations. In: Garner DM, Garfinkel PE. Handbook of Treatment for Eating Disorders. New York, NY: The Guilford Press; 1997:450-461.
  2. Rich LM, Caine MR, Findling JW, Shaker JL. Hypoglycemic coma in anorexia nervosa. Case report and review of the literature. Arch Intern Med. 1990;150:894-5.
  3. Smith J. Hypoglycaemic coma associated with anorexia nervosa. Aust N Z J Psychiatry.1988;224:448-53.
  4. Goldner EM, Birmingham CL. Anorexia nervosa: methods of treatment. In: L Alexander-Mott, DB Lumsden. Understanding eating disorders: anorexia nervosa, bulimia nervosa, and obesity. Hove, England: Taylor & Francis; 1994:135-137.
  5. Lock J, LeGrange D, Agras WS, Dare C. Treatment Manual for Anorexia Nervosa: A Family-Based Approach. New York, NY: The Guilford Publications Inc; 2001.
  6. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington DC: American Psychiatric Association; 1994.
  7. Strupp BJ, Weingartner H, Kaye W, Gwirtsman H. Cognitive processing in anorexia nervosa. A disturbance in automatic information processing. Neuropsychobiology. 1986;15:89-94.
  8. Dodin V, Nandrino JL. Cognitive processing of anorexic patients in recognition tasks: an event-related potentials study. Int J Eat Disord. 2003;33:299-307.
  9. Hendren RL, De Backer I, Pandina GJ. Review of neuroimaging studies of child and adolescent psychiatric disorders from the past 10 years. J Am Acad Child Adolesc Psychiatry. 2000;39:815-28.

Jennifer Bremer, MD, is a child and adult psychiatrist trained at the Massachusetts General Hospital and McLean Hospital, Belmont, MA. She is currently the medical director of the eating disorders program at the University of Chicago, where she has struggled with a number of cases similar to this one.

Commentary 2

by Roy Lubit, MD

Dr. Feinberg was not taking a sufficiently active and assertive role in the treatment of Suzanne. Suzanne should be hospitalized. Suzanne is not at all like a diabetic who slips out of control 1 day. Suzanne is more like a diabetic who denies having diabetes than like a typical diabetic who gets a bit sloppy. Suzanne was not really analyzing and measuring the consequences of her refusal of treatment against an internal set of values and goals. There is no indication that she understood the precarious medical situation she was in, that she could have died, that she was undoubtedly doing severe harm to her body and brain, or even that she had an illness. She may well be delusional about her weight and believe that her weight is in the normal range.

Dr. Feinberg was reportedly concerned that hospitalizing her would lead to nasogastric feedings and damage to the therapeutic work. There are problems with this assessment. First, hospitalization would not necessarily lead to nasogastric feeding. Nasogastric feeding against Suzanne's will would require an evaluation of her competence to refuse. Similarly, patients who are admitted to the hospital for medical or psychiatric problems have the right to refuse treatment. To override their refusal a forensic evaluation is needed.

In addition, Suzanne was not making progress in therapy. Suzanne did not appreciate the nature of her illness despite 2 years of therapy. She went to therapy but had not really engaged and does not appear to be on a path in which she would be able to really appreciate and work on her illness. There was not much work to be set back. Moreover, patients with anorexia nervosa often do not make progress in therapy until refeeding has begun and the clouding of their thinking from malnutrition subsides.

Even though Suzanne is no longer a minor, given her precarious condition and the reasonableness of hospitalizing her, the wish of her closest relatives (her parents) that she be hospitalized is material.

As a side issue, there is no indication that Dr. Feinberg obtained a consultation to help with this case. She needs assistance since it is going poorly. There is also no indication that she has experience and training in this area. If she is not highly trained in this area her need for consultation is that much greater.

Dr Lubit is an assistant professor in the Department of Psychiatry at Mount Sinai School of Medicine. He is board-certified in child, adult and forensic psychiatry. He is coauthor of the chapter on Ethics in Psychiatry in the upcoming edition of the Comprehensive Textbook of Psychiatry.

Commentary 3

by Robert D. Orr, MD,CM


Suzanne suffers from anorexia nervosa, a chronic condition which carries some risk of life-threatening complications. However she has little insight into the condition or the dangers. An acute complication has now arisen, and her parents want her primary physician or her psychiatrist to declare her incompetent so that she may be involuntarily hospitalized and treated. Her primary physician is uncertain and requests a consultation from her psychiatrist. Dr. Feinberg, her long-standing psychiatrist, is concerned about Suzanne's safety, but she is reluctant to honor her parents' request, fearing that her participation in involuntary hospitalization might threaten her 2-year relationship with Suzanne.

Question: Is it ethically permissible, or even obligatory, to involuntarily hospitalize this patient to protect her from a potentially life-threatening condition?

Patient autonomy has gained prominence, even predominance, in contemporary medical ethics. This focus on the patient's right to self-determination has led to a consensus that it is rarely justified to impose treatment on an unwilling patient if certain conditions are met. It is almost always ethically required to allow a patient to make her own decisions if (a) she has been given adequate information to make an informed decision, and (b) professional recommendations have been made, as long as (c) she has decision-making capacity, and (d) she is not being coerced by others. It is permissible for professionals or family to try to persuade the patient, but it is not permissible to manipulate (eg, by overstating the benefits or understating the risks) or to coerce (ie, to threaten).

Are there exceptions to these criteria? It is generally accepted that a patient may sometimes be treated involuntarily if she presents a danger to herself. It is not uncommon to admit an elderly patient to a long-term care facility over her objection if it is determined that she can no longer safely care for herself. It is, however, often a difficult matter of clinical judgment to predict when a patient's current or future decisions present sufficient danger that the benefit of involuntary treatment outweighs the harm of abridged freedom. In addition, we often allow a patient to make a poor choice which presents some risk, as long as the patient understands and accepts that risk.

Primary physicians and psychiatrists not infrequently have to decide if a patient has sufficient decision-making capacity to allow autonomous decisions that carry some risk. This case narrative says "Suzanne Martin was not incompetent to make medical decisions. She could understand the information she was given; she could analyze and measure the consequences of her refusal of treatment against an internal set of values and goals; and she could give back her decision in a coherent and consistent way." Using these criteria, some might believe that Suzanne has the capacity to refuse treatment. However, it is not entirely clear that she can "analyze and measure the consequences" because of her ongoing denial (see below). It is important to note that "capacity" is a characteristic of the patient.

It might be argued that this patient's denial has led her to make an irrational decision. Rationality (or irrationality) is not a characteristic of a person, but of a decision. An irrational decision is one that is not consistent with the patient's own goals and values. Thus a frail patient who chooses to decline nursing home admission and stay at home, placing herself at risk of a fall and fracture, is making a rational decision if she acknowledges and accepts the risk. A person of the Jehovah's Witness faith is making a rational decision if he decides to forego potentially life-saving blood transfusion based on his eternal values. However, a young man in the ED with meningitis who refuses antibiotics but says he doesn't want to die is making an irrational decision, because the choice he is making is not consistent with his goals and values. When an irrational decision has dire consequences, it is ethically justified to override that decision and treat the patient involuntarily.

Suzanne's refusal of admission cannot be considered a suicidal decision, at least not in the classical sense, since the suicidal patient wants to die. Suzanne does not want to die. She is refusing hospitalization because she believes she is not at risk. This could be interpreted as an irrational decision if her goal is to live, but her choice presents danger of death. Whether it is justified to override her autonomy and treat her involuntarily is a judgment call revolving primarily around the seriousness of the risk.

Dr. Feinberg must make a difficult decision. She must balance the physiologic benefits of involuntary admission with the harms such an action might bring to the therapeutic relationship. There comes a time when the balance tips toward the obligation to protect the patient from her own irrational decisions, but it is often difficult to determine when that time has been reached.


(1) Since this is the first metabolic imbalance of Suzanne's illness and it has now been corrected, it would be acceptable for Dr. Feinberg to honor Suzanne's refusal of admission if (a) she believes continued weekly out-patient counseling will provide sufficient oversight and treatment, or (b) she has an alternative treatment plan that is acceptable to the patient. If however, she deems this collapse to be the first step down a potentially fatal course, it would be justified to involuntarily admit her for treatment.

(2) If Dr. Feinberg wants to try to maintain her relationship with Suzanne, but also feels the danger point has been reached, another option would be for her to request a second opinion from another psychiatrist, or even to defer entirely to another psychiatrist for this critical decision.

Robert Orr, MD, is director of clinical ethics at the Center for Bioethics and Human Dignity, in Bannockburn. The Web site is

The people and events in this case are fictional. Resemblance to real events or to names of people, living or dead, is entirely coincidental. The viewpoints expressed on this site are those of the authors and do not necessarily reflect the views and policies of the AMA.