The qualitative interviews were semi-structured and the interview was conducted using a topic guide, which served as a springboard for wider, more flexible and unstructured narratives and discussions. Participants were asked to talk about their own experiences of treatment, and the interviewer encouraged elaboration of experiences relating to the three main foci of the study given above. Questions in the topic guide relating specifically to compulsory treatment are listed in Box 1. As will be seen the questions ask about whether it is acceptable to ‘make people have treatment’ or ‘to be treated even if they don't agree’. The concept of coercion and compulsory treatment was thus left to participants to interpret as they saw fit and was not defined in terms of legal processes.

Patient records were not accessed, so all information was obtained from the participants themselves. Participants' self-reported Body Mass Indices (BMI) ranged from 12.4 (dangerously underweight) to 28.4 (overweight, technically ‘pre-obese’), with a mean body mass index of 17.7 (below normal range) and a median BMI of 17.65. By their own accounts, the participants were at various stages of illness, treatment and recovery at the time of interview.

Of the 29 participants, eight were inpatients in mental health units although at the time of interview none was detained under the Mental Health Act 1983.¹ Eighteen participants were either day patients or outpatients utilising mental health services at the time of interview. One participant was waiting to have treatment. One participant had been discharged by the eating disorder service after declining an offer of inpatient treatment for low weight. One participant had chosen not to accept treatment as she had a previous aversive experience of inpatient treatment.

Of the eight inpatient participants, five described themselves as having been admitted without free choice, either owing to parental pressure or under the implied or overt threat from the mental health professionals of a Mental Health Act 1983 ‘section’ (compulsory detention order for the purposes of assessment or treatment) if they did not comply. Two of these had subsequently been placed on a Mental Health Act 1983 Section 3 (a compulsory detention order for treatment of a mental disorder) during the course of their admission and both had been recently discharged from the Section 3 order at the time of interview. One of these two patients had also experienced being detained using the Mental Health Act 1983 during a previous admission. Only three of the eight inpatients, therefore, described themselves as having made a free choice to be admitted to hospital for the current admission, and one of these three participants described a previous inpatient admission to a different unit to which she had not given consent.

Only three of the 18 day patient and outpatient participants said they had made a choice to enter and remain in treatment on their own. Six participants described being coerced into having treatment against their will. A further nine described either shared decisions concerning treatment made together with doctors and parents (with varying degrees of pressure from professionals and parents ranging from encouragement to ultimatums), or decisions about treatment made by doctors and relatives on their behalves with their tacit agreement. It is important to note that the Mental Health Act 1983 did not enable compulsory outpatient or day patient treatment, so no formal compulsory treatment under mental health legislation would have been possible.

A concept that is receiving increasing research attention is that of ‘perceived coercion’ (Bindman et al., 2005; Guarda et al., 2007). This is the idea that psychiatric patients often experience coercion whether or not formal legal means are used. Such coercion includes ‘leverage’ when a patient may be ‘encouraged’ to behave in a certain way through a threat of the use of more formal coercion or the loss of some benefit (Appelbaum & Redlich, 2006). The participants in this study described such perceived coercion. Indeed their experience of formal legal compulsion was much less frequent than that of other forms of compulsion.

The participants' frequent accounts of experiences of lack of choice and the use of leverage are ethically problematic when contrasted with the relative infrequency of legal compulsion they experienced. On the one hand, the use of leverage can be an effective and flexible way of enabling acceptance of treatment while avoiding the stigmatizing and potentially traumatic experience of formal legal compulsion, and may not be experienced negatively; on the other hand, the use of leverage and other non-legal forms of coercion leave already vulnerable patients at risk of unethical treatment and loss of autonomy and rights without the protection of legal procedure, advocacy and right of appeal that formal legal compulsion brings. It has been argued that it is unethical and discriminatory towards those with mental disorders to use any form of coercion or act in a patient's best interests without consent unless he or she clearly lacks capacity (Szmukler, 2001); and furthermore, that coercion is counterproductive in anorexia nervosa (Rathner, 1998). Studies suggest that patient perception of ‘coercion’ or experience of formal legal compulsion is not subsequently associated with a poorer engagement with treatment or poorer therapeutic relationship with mental health professionals (Ayton et al., 2009; Bindman et al., 2005; Greenberg, Mazar, Brom, & Barer, 2005). The limited research also does not clearly show that formal compulsory treatment is associated with poorer short or longer term clinical outcome (Ayton et al., 2009; Carney et al., 2008; Ramsay et al., 1999; Watson et al., 2000). Even if there is no clinical harm, however, the issue of compulsion in treatment for anorexia nervosa is still ethically problematic. From the ethical point of view, a traditional stance is that patient autonomy should be respected even if unwise or foolish decisions are made, and the justification for acting in a patient's best interests instead of according to his or her wishes is the pivotal issue of capacity (Buchanan & Brock, 1989).

What is striking from this study is that the issue of capacity was not the central one for participants in the context of compulsory treatment. It is true that many participants did view treatment refusal by people with anorexia nervosa as problematic because the anorexia compromises decision-making. But this was not the main issue in their consideration of the rights and wrongs of compulsion. The main issues centred on their relationships with those involved in the treatment: health professionals and sometimes parents. The participants did not appear to resent leverage or ‘informal’ compulsion if it is carried out in the context of a trusting relationship and perceived as care and help, and may not even experience it as coercive.

These results have implications for the ethical analysis of compulsory treatment in the context of anorexia. One approach, we suggest, is that, on the accounts of those with anorexia, the central issue is not the question of whether and when a person lacks capacity to consent to, or refuse, treatment. Instead the central issue concerns the context and relationships involved in good and compassionate psychiatric care. A greater objective restriction of choice may be experienced as less negative, and indeed ultimately as good care, if conducted within a trusting and supportive relationship and environment. Such an approach might be seen as a ‘virtue ethics approach’ in which the focus is on how carers can act with compassion. It is also consistent with recent trends in bioethics that highlight the importance of trust in decision-making (O'Neill, 2002).

An alternative response to the accounts reported here is to retain the emphasis on autonomy and capacity in thinking about the ethics of compulsion but to suggest that more subtle and broader difficulties in making treatment decisions may need to be considered (Charland, 1998). Key questions are: when and why a patient may fail to make autonomous decisions because of the influence of the mental disorder, and whether help and even coercion to engage in treatment might increase autonomy (Charland, 2002; Tan & Hope, 2006; Tan et al., 2006). Conversely, a patient may technically lack capacity, for example being impaired in his or her ability to retain information, but broader factors such as an ability to express underlying values may enable him or her to continue to make valid choices with the appropriate support (Department of Constitutional Affairs, 2007; Tan et al., 2006).³

These two approaches are not mutually exclusive, as poor relationships and difficult or hostile settings tend to disempower patients with respect to whether they feel able to make decisions for themselves and can also cause them to resist or refuse help; while strong, trusting relationships and supportive treatment settings tend to facilitate patients in making their own decisions and help them to feel more actively involved in their care. Indeed, the ethical justifications voiced in these results suggest that it is a complex mix of a virtue ethics and nuanced capacity approach that is used by patients in thinking of their experiences of, and attitudes to, compulsory treatment and coercion. The analysis of this study and results of a previous study suggest that in anorexia nervosa issues of control and choice can have a strong influence on whether patients are able to make decisions. This is because patients suffering from anorexia nervosa often struggle to remain in control of themselves and their lives and, as a result, resist help, while at the same time feeling out of control with respect to their eating behaviour and wishing to have help (Tan et al., 2003c).