We conducted this study to investigate the attitudes of physicians in Japan, the UK, and the US toward refusal of treatment for AN. To our knowledge, this is the first report of an international investigation on the propensity of physicians to choose compulsory treatment in cases of refusal of treatment for eating disorders. This study revealed that in Japan, the US, and the UK, compulsory treatment tends to be the prevalent choice in cases of life-threatening malnutrition, regardless of the patient's age or duration of illness. Compulsory treatment was chosen more often when family members requested treatment than when family members left the decision to the individual. The results also indicated a significant difference in the tendency to choose compulsory treatment among physicians in the three countries.
Refusal of treatment in life-threatening cases of malnutrition poses the ethical dilemma of whether the physician should prioritize the protection of the patient's life or the patient's right to self-determination . From the perspective of the four principles of medical ethics , it can be analyzed as an ethical issue of comparative consideration between the principle of good conduct and the principle of respect for autonomy. In the present study, physicians from Japan, the UK, and the US often chose the policy of compulsory treatment in life-threatening cases of poor nutrition, regardless of the patient's intention. The background of this attitude seems to be the idea of prioritizing the protection of life as the medical interest of the patient; that is, prioritizing the ethical obligation of beneficence over the ethical obligation of respect for autonomy. This idea has been reported in several articles on ethical considerations in cases of treatment refusal .
There is no consensus on whether patients with AN are able to make decisions regarding treatment [7, 8]. However, one reason that the duty of beneficence may take precedence over the patient's self-determination is the presumption that patients with AN who are undernourished are not in a condition to make sound decisions [9, 10]. In a situation where the patient cannot make appropriate decisions, it is common for family members to speculate on the patient's wishes on behalf of the patient. In fact, the responses in the present study indicated a greater tendency for selection of compulsory treatment when the patient’s family wanted treatment than when they did not. This may be because the medical practitioner believes the patient is not competent enough to make sound decisions, and therefore follows the opinion of a family member who is the surrogate estimator. Conversely, over half of the physicians in Japan, the UK, and the US responded that they would choose compulsory treatment even if the patient’s family did not wish to initiate it. It is probable that in such situations, the patient's family did not consider the patient's best interest to be a priority, and the medical care provider considered the patient's best interest to be the protection of life. Regarding the refusal of treatment by patients experiencing life crises, Giordano  advocated for “legitimate use of prudence, recognition of the value of life, and common sense” and stated that “if there is a fairly good chance that the patient will thank you for rescuing her, then you should rescue her.” It has been reported that although patients with AN may refuse treatment, they are glad to have received coercive treatment after recovery . It could be said that this is a strong form of paternalism .
The possibility that cognitive and affective biases might influence clinicians’ decision-making in cases of patient with AN refusal of treatment cannot be ignored . For example, self-serving concerns about the reputation of their peers and colleagues for the decisions they make may lead many physicians to be very conservative in their clinical decisions. Concerns about being sued for negligence may also be a factor in their choice to make conservative decisions (especially in the US).
In the present study, the proportion of physicians who preferred involuntary treatment was significantly higher in Japan than in the US and the UK. This may be due in part to the fact that awareness of patient self-determination occurred late in Japan  and attitudes are more paternalistic, emphasizing the ethical obligation of beneficence . Moreover, Japanese law makes it easier to provide inpatient treatment when a patient refuses treatment despite having life-threatening malnutrition. In Japan, a person who is not in a condition to be hospitalized voluntarily, even if there is no risk of self-injury or other harm, can be hospitalized after examination by a doctor and with family consent. The tendency for a patient’s family to influence the treatment plan in Japanese medicine has been reported previously . Japanese culture is more family-centered than the more individualistic British and American cultures, and this may be another reason why the choice of compulsory treatment is prevalent in Japan, particularly if the patient’s family members request treatment.
In the present study, the number of physicians who chose compulsory treatment for young patients was significantly lower when the decision was left to the patient than when the patient’s family wanted to initiate treatment in the UK and the US. The reason for this may be that if the decision to initiate treatment is left to the patient, compulsory treatment may break down the therapeutic relationship and make it difficult to continue treatment in the future. Although young patients with acute disease are less likely to be in mortal danger than are older patients, ensuring that treatment can be continued may be in their best interest. In fact, it has been reported that compulsory hospitalization may only have a limited therapeutic effect .
The difference between Cases A and B was the age of the patient and the duration of the disease. Case B was more severe because of the longer duration of the disease. Although it has been reported that some patients appreciate having received coercive treatment after recovery , some severely ill patients have been reported to oppose coercive treatment even after they have recovered . For patients with severe AN with a long disease duration, the long-lasting fear of obesity may have been pathologically internalized as an identity; thus, refusing treatment may have reached the point of being a belief for the patients. In the present study, the proportion of physicians in all three countries who did not choose compulsory treatment for older patients with long disease duration was higher than that of physicians who did not choose compulsory treatment for younger patients. This trend was also noted in the choices of physicians if family members left the decision to receive treatment to the patient; however, the result was not statistically significant. This trend may indicate that in the case of severe and enduring patients with long disease duration, their morbid attitudes may have been interpreted as beliefs related to their identities, and as a result, physicians may have respected the patient's self-determination. Furthermore, even if a life-threatening AN crisis is averted for patients with severe and enduring disease, the crisis is likely to recur . This may be a reason why some respondents in the present study considered involuntary treatment to be futile from a long-term perspective.
In the UK and the US, where trends are relatively similar, we examined attitudes toward refusal of treatment among physicians with more or less years of experience as clinicians, and among physicians who saw more or fewer patients, and found no significant differences. Attitudes toward refusal of treatment reflect ethical values, and these results may suggest that these values may be in place from the beginning in each physician as an individual, rather than being developed by their experience as physicians.
Strengths and Limitations
Ethical issues in clinical practice are difficult to solve with normative theory alone. When responding normatively to ethical issues in clinical practice, it is important to consider what should be done after understanding the current situation through empirical data. This is the first international survey of physicians' attitudes toward refusal to treat patients with AN and the data presented here could be used as reference when considering treatment refusal in patients with AN from the perspective of empirical bioethics . This study has some limitations. First, the attitude toward a fictitious vignette case may differ from the attitude when confronted with an actual case of refusal of treatment in the course of clinical practice due to mental conflict. Second, since it was conducted using simulated cases, the effect of the inpatient facilities and ethics policies from the respondents’ institutions on their individual responses is unknown. Third, owing to the small size of the study sample, the generalizability of the survey is difficult to estimate. Fourth, since the survey was conducted using questionnaires delivered by mail in Japan, but web-based questionnaires were used instead in the UK and in the US, it is possible that differences in the survey method may have affected the results. Fifth, although a dual translation was used, the translation of the vignette cases may have altered some nuances and therefore potentially biased the responses. Despite these limitations, this study is significant in that it is the first international survey conducted to compare the attitudes of medical practitioners toward refusal of treatment by patients with eating disorders.
In the future, it will be necessary to investigate whether physicians in practice recognize the mental capacity of patients with severe AN, what they emphasize in assessing mental capacity, what specific treatment they provide after choosing compulsory treatment, and what type of treatment is provided when compulsory treatment is not chosen.