Competing Interests in Pediatric Eating Disorders Patients

White Space

Parental Authority, Best Interests, and Staff Safety

Joseph Shapiro

White Space

White Space

She can’t be as bad as they’re describing, I thought, as I took sign-out on the 11-year-old eating disorder patient I would be caring for during the next week.

White Space

White Space

Case Overview

AR was initially admitted to the inpatient child psychiatry unit (CPU) for symptoms of obsessive-compulsive disorder, including obsessive counting, repeated hand movements, and insomnia. During two weeks on the CPU, it gradually became clear that AR’s obsessive compulsive disorder (OCD) was complicated by a severe eating disorder. She was restricting her calories, taking only 500-600 per day; she lost a kilogram in her first week of hospitalization. When CPU staff attempted to discuss her nutrition, RR became agitated and perseverated aimlessly about her weight. It was clear her OCD was unlikely to improve in the setting of malnutrition, so she was transferred to the adolescent medicine service for nutritional rehabilitation.

She was immediately non-compliant with the eating disorder protocol, which calls for the majority of her calories to be given via nasogastric tube feeds. RR was frequently violent and aggressive toward staff. She made multiple escape attempts. On one occasion, she kicked an intern in the chest while trying to get out of her room.

Despite hearing some of these things, I was skeptical this patient could be so remarkably different from the myriad of combative eating disorder patients I had taken care of in the past. As I started my first night shift, I was told she had been calm all day. I was advised to avoid the room unless the situation devolved. That did not take long. As soon as AR’s overnight feeds started, she became agitated and the intern escalated to me for assistance. I came into her room to find a cachectic, but vigorous, pre-pubescent girl screaming, cursing, and thrashing as two nurses, a tech, and an intern held her to the bed. All were wearing masks with face shields to protect them from her attempts to spit in their faces. Each time they lost focus on their hold for even a second, her hand would shoot to her nasogastric tube to try and rip it out.

I stood there, stupefied, watching this fiasco until one of the nurses got my attention. I fired off a series of incredulous questions: was this really the plan? How have we handled this situation on previous nights? The nurse tells me the night team has had to put AR in restraints almost every night, for weeks.

Typically, when restraints are necessary for children we will consider medical sedation, such as lorazepam or even haloperidol, as an adjunct treatment. The goal is to calm the patient and reduce the physicality necessary to keep them safe. I asked the nurse if we have been using any sedatives in AR’s management. The nurse tells me the team has not been giving sedatives because the family is refusing; they are worried about the side effects and do not like seeing her in the groggy, altered state these medications sometimes induce.

I agree to place the restraints, but soon learn this is far from a solution to the problem. Getting them on is a nightmare; AR repeatedly manages to get free and lash out. One of the nurses is punched in the chest, and we are all scratched repeatedly. Once the restraints are in place, they do not work effectively; she is so thin they cannot be tightened and locked into place properly. RR still requires multiple staff members to hold her at all times.

This scene goes on all night, hour after hour, our forcefully restraining this small girl. The nasogastric tube feed finishes as the sun rises, and we are able to let her go. Knowing that she is not actively being fed, AR is able to relax. I sign out and head home to sleep, but not before emailing the attending physician on service to make sure she has an accurate accounting of the night’s events.

The following night I am given an escalation plan that involves appropriate use of sedative medications before physical restraints. Unfortunately, AR demonstrates little to no response to sedatives. Despite adult doses of lorazepam and haloperidol, she remains violent, agitated, and requires physical restraint for prolonged periods. After another night of frequent restraining, I ask the day team for a new plan for overnight management. When I return to work, I am told the new plan is daytime feeds with behavioral health staff on hand. Without the NG feeds running, she is mostly calm and cooperative overnight. For the next few weeks, AR undergoes a daily procedure of being sedated and placed in 5-point restraints so she can be given a nasogastric feed as fast as possible to meet her daily caloric requirements. She gains weight, and her behavior very slowly improves. Still refusing all solid foods, she starts taking Pediasure by mouth and the nasogastric tube feeds are discontinued. Arrangements are made for transfer to an outside facility for further treatment of her eating disorder.

White Space

Discussion

Issues Related to Parental Authority
The care of this patient raised difficult ethical issues for our care team. Throughout the majority of AR’s many-week hospital stay, she required frequent and forceful physical restraint. Many adolescents with eating disorders strongly resist changes because doing so may be viewed as giving up one’s identity, [1,2] or being out of control [3] which can be a precursor to power and control issues. [4] While normalizing eating patterns and facilitating weight gain is a core therapeutic objective for eating disorder patients, [5] many providers at all levels of AR’s care team expressed concerns about the physical and psychological harm using restraints may have been causing her. However, her behavior was such that it was impossible to provide her with nutritional rehabilitation via nasogastric tube without the use of restraints; we decided that her severe malnutrition was putting her in imminent physical danger (arrhythmias, electrolyte abnormalities, etc.) and that without nutritional rehabilitation, her psychiatric health was not likely to improve. Hence, in balancing the principles of beneficence and non-maleficence, we determined that using physical restraints in order to provide her care, though potentially harmful, was justified in order to prevent the very real, and serious risks to her life. An ethics consult was obtained. The ethics consult recommendation supported treating the patient against her will, but also recommended the following: First, in a quiet moment the doctor should recognize this situation is painful and infuriating to her (AR). Second, the doctor should express a lament for having to go this far, and apologizes for the use of force. [7]

Attempts were made to mitigate the trauma of the physical restraint by designing an appropriate medical sedation plan. Initially, when this plan was presented to the family they refused and insisted on avoiding sedatives. After the difficulties of this approach and the harm it was causing our patient became clear, sedatives were made a prominent part of her treatment plan. The delay in this decision represented an attempt to balance parental authority and the principle of beneficence. [8] This delay was an example of counterpoise iatrogenesis in which harm results from an attempt to balance two or more competing ethical obligations. [9]

Numerous ethics frameworks exist to aid clinicians in deciding when to override parental decision-making. [10] Diekema suggests the Harm Principle, wherein healthcare providers are justified in overriding a parent’s decision when that decision “significantly increase[s] the likelihood of serious harm as compared to other options.” [11] This framework is useful in evaluating the issue of medical versus physical restraint in this case as it is put forth in the context of evaluating parental refusals of treatment. Diekema proposes eight conditions that should be met to justify overriding parental authority. Our case easily meets most of these criteria, but there is some ambiguity in evaluating the benefit to burden ratio of refusing medical sedation. Diekema suggests the chosen intervention must be, “significantly more favorable,” than the parent’s desired option. [10] Although the favorability of a combined medical sedation/physical restraint approach became clear as AR’s care progressed, it was not immediately obvious how difficult and traumatic it would be to physically restrain her. One may have reasonably assumed at the outset of her care that physical restrain alone could be successful.

The difficulty of balancing parental authority with patient best interest and harm prevention was a recurring theme throughout our patient’s hospitalization. It was particularly relevant in determining patient disposition. Relatively early in the hospitalization, it was agreed that AR would be best served in a med-psych unit with the capacity to provide eating disorder treatment. Multiple facilities that met this description were identified and presented to the family, and they were asked to approve the transfer. The family deemed the options given to them to be unacceptable for reasons including the aesthetics of the facility, the age of the other patients (older than our pre-teen patient), and the facility’s perceived lack of expertise in treating eating disorders. The family identified a geographically distant facility that they felt was perfect for their daughter’s care, but this treatment program was far enough away that it required transport by airplane. The family’s insurance company initially refused to pay for either transporting the patient, or her stay at the desired facility.

There was significant disagreement and discussion among members of the care team and our hospital case managers about whether the family’s requests were reasonable and to what extent we should be attempting to honor them. This debate, and the associated negotiation with the insurance company, significantly prolonged the length of the hospitalization, and represented a harm to the patient. While awaiting disposition, AR received very limited therapeutic intervention outside of NG feedings; the acute care unit on which she was hospitalized had little capacity to engage in specific psychotherapy techniques necessary to treat an eating disorder. [12] Eventually, the family’s insurance agreed to cover the program at the distant facility, the family paid for her transport on air ambulance out of pocket, and she was transferred.

Difficulties with Ensuring Caregiver Safety
Perhaps the most complicated and persistent ethical issue in this case was balancing the need to ensure caregiver safety with our responsibilities to the patient. [13] At our hospital, it is a policy for security staff to not physically engage with a patient unless they observe the patient to be physically aggressive toward staff or dangerous to themselves. Early in her hospitalization our patient figured this out and proceeded to manipulate her care by remaining placid and calm when security staff were present but resuming her violent and aggressive resistance whenever they left.

Establishing a therapeutic relationship with adolescent eating disorder patients poses a number of challenges that arise from the effort of patients to exert control as a manifestation of their disease. [4,14,15] AR ascertained that attending physicians were the ultimate decision-makers in her care, so she would censor herself in their presence. As a result of this manipulative behavior, nurses, patient techs, and junior residents were often the only witnesses to, and victims of, her outbursts. Because of this dynamic, the residents and nurses frequently disagreed with attending physicians about her plan of care. These frontline providers repeatedly expressed a feeling they were being asked to go above and beyond their responsibilities to provide AR’s treatment.

 While the attending physicians involved were sympathetic to the resident’s and nurse’s concerns, they were hesitant to take action for two reasons. First, they were not witnessing many of the most dramatic and violent sequences of her hospitalization firsthand. Second, it was unclear what options, if any, were available to alleviate the need for residents and nurses to be performing physical restraint at the bedside. Eventually, some resolution was achieved when we decided all tube feeding would be done during the day (when attendings are present), and whenever violent behavior became intractable, a rapid response team was dispatched from the psychiatry unit to intervene. After this intervention was instituted, it was necessary to use it on a near daily basis for the final weeks of her hospitalization.

White Space

Conclusion

The difficulty of creating a treatment plan that was acceptable to the parents, and ensured the safety of the patient and all caregivers involved was almost intractable in this case. It is somewhat unique in pediatric medicine (as opposed to adult medicine, or psychiatry) to have such difficulty ensuring the safety of the care team.

As physicians, we are drilled extensively on our duty to our patients. The principles of ethical patient care are reinforced often throughout our training, but less often are we instructed on our responsibilities to trainees and hospital staff. In an increasingly complex healthcare system, situations like this one are frequent and likely to reoccur.A 2016 New England Journal of Medicine study found 100% of ED nurses reported verbal assault, and 82.1% reported physical assault within the last year. [16] As physicians, we must be cognizant that attempts to fulfill traditional fiduciary or ethical obligations to patients and families may be at odds with ensuring the basic safety of employees and trainees under their supervision.

White Space