Ethical Challenges in the Treatment of Patients With Severe Anorexia Nervosa

Ms. A, a 35-year-old patient, was first diagnosed as having the binge-eating/purging subtype of AN in her early teens. Treated on seven occasions, she was eventually able to live and function, albeit at a low weight. After maintaining this low weight for 8 years, she began having abdominal pain due to gallstone pancreatitis for which she underwent multiple abdominal procedures and surgeries and, as a result, she began excessive use of pain medications. She also suffered from depression and anxiety. Her supportive family set strong boundaries regarding the need for treatment. In addition, her family controlled her finances.

At the time of inpatient admission, she weighed approximately 65 pounds (54% of her ideal body weight). She refused to eat and would not accept nasogastric tube (NGT) feeding. Her treating psychiatrist petitioned for inpatient certification, and also requested court-ordered medications and involuntary NGT feeding. Ms. A was subsequently certified and the court approved the petition for court-ordered medication and NGT feeding. Ms. A required physical restraint in order to insert an NGT to initiate feeding on several occasions. The certification was renewed at 3 months, uncontested by the patient. Ms. A was restored to her ideal body weight. During the weight-restoration process, she was transitioned to 100% oral nutrition, and discharged to outpatient treatment. Four months later, Ms. A returned for a short hospital stay as a voluntary patient after having lost 10 pounds. She was readmitted approximately 1 year thereafter, once again being severely ill. Following a prolonged hospital stay, which included certification, she was discharged to outpatient care, maintained her ideal body weight, and was engaged in gainful employment.

Although this patient had a long history of eating disorder and several protracted hospitalizations, she was nonetheless eventually able to achieve and maintain a reasonable level of recovery. Certification was helpful in that it kept her in treatment when her body weight was so low that her cognitive distortions did not allow her to voluntarily restore her weight. She was fortunate in having access to one of the few programs willing to accept and capable of managing both her dire medical and psychiatric conditions under involuntary status. In addition, her treatment was helped by having family who were able to maintain strong boundaries for her and support the treatment team. She was well aware that had she eloped from treatment to escape the certification, her family would not have continued to support her financially.

Mr. B was a 33-year-old man with restrictive AN, severe obsessive-compulsive disorder (OCD), recurrent major depression, and chronic suicidality. Obsessional and perfectionistic traits were noted from early childhood. As a high school freshman, he began to exercise excessively, eliminated fats, and created food rules. He was subsequently hospitalized several times for AN and participated in family therapy. Despite graduating as high school valedictorian, he was unable to complete college due to eating disorder and OCD symptoms. Over the next 2 decades, he spent countless months in medical hospitals and in inpatient and residential eating disorder settings, repeatedly gaining weight, only to relapse. In his early thirties, following a long downward spiral, when his family finally threatened to call 911 if he did not enter treatment, he was admitted to an inpatient eating disorder program. After he refused to eat, an NGT was placed. A court mandate to ensure ongoing treatment was granted on grounds of grave disability, and he spent the next 10 months in involuntary inpatient and residential eating disorder care. He continued NGT feeding, and finally sustained his weight with oral food, agreeing to eat only to avoid being administered olanzapine against his will. After months of weight restoration, he underwent a renewed cycle of partial hospital program (PHP) treatment, during which he proceeded to lose nearly a pound a day, followed by readmission to residential treatment to once again achieve target weight, and then another 2 months of PHP, which yielded no improvement. Treated with intranasal ketamine and subsequently with intravenous ketamine, he obtained no benefits in alleviating his OCD, depression, or suicidality. Despite his many treatments, including the year-long court-mandated treatment, he never meaningfully changed his eating-related attitudes, thoughts, or behaviors. He showed absolutely no motivation for recovery.

Discharged to live with his parents, Mr. B refused psychiatric medications, stopped eating altogether, and drank only water, stating, “I don’t want to die, but my eating disorder is in charge.” A team consisting of a physician, therapist, and dietitian cared for him at home. His longstanding, very apprehensive outpatient therapist worried that she would be forced to have him detained against his will. A formal competency evaluation performed by a local psychiatrist found Mr. B to possess decisional capacity. He did not want to die but did not seem to be meaningfully attached to life. As Mr. B clearly refused to return to formal treatment for his eating disorder, his clinicians suggested a palliative care consultation to support comfort and dignity. He found the staff to be extremely kind.

After almost 6 weeks without food, he began accepting anxiolytics and antiemetics. Still plagued by obsessional thoughts, and still driven to lose weight, he noted, “Just because you aren’t eating doesn’t mean it’s all good now.” He spent increasing amounts of time connecting positively with his parents and siblings and described deeply feeling that “peace comes from God.” After about 8 and one-half weeks without any food, spontaneously vomiting daily, and feeling much weaker, he was placed on home hospice care and began to accept low-dose morphine for pain and distress. Two weeks later, Mr. B passed away surrounded by his family. Even as they grieved deeply, his parents expressed enormous gratitude for the care he received and for the way they had been able to reconnect with him. They were reassured that the cause of death listed on his death certificate would be AN and malnutrition, not suicide.

This patient benefited from the support of a family that was consistently caring and able to financially provide extensive amounts of top-quality medical and psychiatric care throughout his courses of treatment. They also coerced him into accepting treatment in his thirties, at which point he endured many months of involuntary treatment, including NGT feeding. Although the patient’s weight was restored in the short term, these treatments had no impact on his long-term course. Eventually, his family was able to accept that active recovery-focused treatment seemed futile, and honored his decision to seek palliative and, ultimately, hospice care.

Most physicians have little experience in assessing decisional capacity in patients with AN. Although AN is not synonymous with lack of decisional capacity, individuals who have eating disorders frequently lack this type of capacity due to the effects of starvation on the brain (19–21). In patients diagnosed as having AN, the patient’s competence is usually only doubted in the singular narrow area of self-nutrition and/or body image (21). Individuals with severe AN can appear rational in all areas although still lacking the ability to fully appreciate the gravity of their own condition and how the risks and benefits of treatment decisions apply to them (20). Nevertheless, standardized tests frequently fail to identify lack of capacity in patients with AN (22, 23). Although several models for assessing competence have been employed, and others are available, studies involving capacity in patients with AN have compared “clinical judgment” with findings using the more specific MacArthur Competence Assessment Tool for Treatment (MacCAT-T), which assesses patients’ capacities in four areas: understanding information relevant to their condition and the recommended treatment, reasoning about the potential risks and benefits of their choices, appreciating the nature of their situation and the consequences of their choices, and expressing a choice (23).

In some of these cases, clinicians and legal authorities often have little difficulty in determining whether coercive or involuntary interventions are necessary and justified. However, in other cases, determinations of the mental capacity of patients to make their own health decisions may not be so clear-cut. Even patients with SE-AN may retain the capacity to judiciously reflect on their conditions (24). Vigorous debates exist among experts regarding how to assess and think about mental capacity in patients with AN, and findings in research studies are mixed (25, 26). Some patients with AN retain capacity regarding most life functions but show selective cognitive impairment in that their eating disorder distortions conflate an objective assessment. Other patients with AN have a greater awareness that their eating disorders’ compulsions are irrational but irresistible. Yet at the same time, they may feel that their suffering is intolerable, maintain that they no longer want to be treated, and stipulate that they can no longer fight their ego-alien disease, even if it means dying. In such cases, the harm caused by involuntary treatment may override the slim chance that such patients will recover in a meaningful sense.

At certain points in their trajectories, patients, families, and their clinicians may reason that further recovery-focused treatments are unlikely to be successful or of sufficient benefit to justify condoning involuntary methods. Furthermore, there is a justifiable reluctance on the part of eating disorder professionals to undertake any treatment that bears little hope of advancing a patient’s quality of life, directly opposes the wishes of that patient and family, and simply extends a life of suffering, even if the patient has diminished capacity (27, 28). Whether further recovery-oriented treatment would be perceived as futile should therefore be judiciously considered on a case-by-case basis and applied cautiously when considering a particular treatment intervention, at a particular time, for a particular patient. McKinney (29) has proposed a framework that balances patients’ wishes with legitimate concerns about allowing an individual with severe AN to refuse further treatment. McKinney opined that treatment-refusal decisions regarding further episodes of care should be made 1) at a time when the patient is more likely to have decisional capacity (i.e., between episodes); 2) the patient must have a realistic expectation regarding outcome (i.e., the patient must know that refusing nutrition will lead to death); 3) the patient’s decision to die must be on the basis of a realistic assessment of current quality of life, and the low probability that current or future treatment will succeed; and 4) patients must be consistent in communicating their desires (29). Each case must be considered in a nuanced and thoughtful manner.

When clinicians, patients, and families are at odds or unclear as to how to proceed, seeking a formal ethics committee consultation from local health institutions may be of help, along with seeking legal input from a program’s legal counsel. However, hospital ethics committee members may have had little experience with severe and enduring eating disorders, and, it has been observed, may come with their own biases (30).

When patients refuse ongoing treatment, and clinicians believe, after much deliberation, that insisting on additional recovery-focused treatment is likely to be futile, how should clinicians proceed? Remaining treatment options include harm reduction and palliative care approaches, which, it should be emphasized, are not mutually exclusive. Their components can be blended. Harm reduction models help patients to reduce risky behaviors, even if they cannot be eliminated. For patients with AN, this might entail eating less than necessary to recover, and maintaining a weight that is psychologically tolerable although below what might be considered healthy or ideal, but not so low as to be life-threatening. Harm reduction might also focus on reducing the frequency of purging, laxative use, and excessive exercise (31, 32). Typically, harm reduction approaches include early discussions and agreement about “nonnegotiable aspects of treatment,” such as expectations for medical monitoring and care, defining when hospitalization might be indicated, and clear, reasonable next steps if these expectations are not met.

Palliative care approaches, ordinarily provided by teams of practitioners, prioritize quality of life and the prevention and relief of suffering through symptom reduction and psychological and spiritual support that focuses on therapeutic alliance, compassionate witnessing, interpersonal attunement, empathic holding, and nonjudgmental positive regard. Thus, palliative care is “life affirming” (33). A primary tool in the palliative care approach is the “goals of care” conversation, through which clinicians aim to establish goal concordance with patients by essentially developing a shared vision of what might be achieved and how to go about treatment. Key to goal concordance is noncompulsory care. Goals may shift over time, as patients improve or decline, and care plans adjust accordingly.

Significantly, palliative care does not focus on preparing for or hastening death. It also fully supports the concurrent or future renewed implementation of further treatments that are aimed at recovery or harm reduction if they are available and acceptable (23, 33). Palliative care approaches have been recognized as appropriate for patients with psychiatric disorders at certain points of clinical care, usually when patients have stopped responding to and/or stopped participating in recovery-oriented care but still hope for alleviation of suffering due to their disorders and circumstances (34–36). Even when not called palliative care, a good deal of common-sense, every day, so-called supportive psychiatric practice embodies principles of palliative medicine (34). Palliation includes prescribing medications to help patients reduce cravings, anxiety, insomnia, and pain. Patients with AN might agree to maintain minimal nutrition, but refuse to engage in weight-restoration activities (33).

For instances in which patients’ conditions decline to the point that they are life-threatening, and the patients refuse active treatment, palliative care may transition to hospice care, which is specifically designed to provide patients and their families with support at end of life, as illustrated in Case 2.

Ethical principles assist us in determining the best course of action with regard to patients with AN. The vast majority of patients with AN should be offered high-quality, restorative-informed care. But for a minority of patients (e.g., those with SE-AN) other treatment options need to be considered. Clinicians are obliged to realistically assess each patient’s potential for recovery or ability to engage in harm reduction and palliative care approaches, and be cognizant of the wishes of the patient, family, and treatment team. In addition, the burden on caregivers and stewardship in the expenditure of health care resources should also be considered when deciding whether involuntary treatment, harm reduction, palliative care, or end-of-life care be recommended for a particular patient.

The field needs better information to guide decisions regarding which patients are most likely to be helped or harmed by coerced and involuntary treatments. To these ends, we advocate for long-term follow-up studies that embrace lived experiences of patients and their families who have undergone various treatments, including those of families who have lived through the deaths of loved ones as a result of AN. We also advocate that palliative care interventions should be reserved for patients who have relapsed multiple times, who consequently believe that conventional, active recovery-oriented interventions are unlikely to benefit them, and who, in turn, decline such recovery-oriented treatments. Large studies, including meta-analyses, might also permit calculation of treatment effect sizes. Such information could better equip clinicians, patients, families, and courts to make more appropriate ethically informed, patient-centered decisions.