‘When someone is taken into hospital dehydrated and malnourished, we have to make a choice. You may say “I don't have to do anything. Thank God I'm not a doctor, I'm a philosopher and my job is to think, not to act”'.

This snatch of a conversation is related by the author in an ironic vein. Yet it illustrates the gulf between the ethicist and the clinician which is the main theme of this book.

Simona Giordano is a lecturer in bioethics at the School of Law in Manchester. She holds a Doctorate of Philosophy from the University La Sapienza in Rome. From the acknowledgement in her book, it appears that she worked with patients in the private psychiatric clinic Villa Rosa in Viterbo, Italy. Her book purports to be the first full philosophical study of ethical issues in the treatment of anorexia and bulimia nervosa.

The author presents the principle of ‘weak paternalism’ which requires some explanation. In the sphere of clinical decisions, paternalism is when the doctor intervenes against his patient's manifest wishes in order to protect her welfare. The term weak paternalism is used in a technical sense meaning that the paternalistic intervention should occur only when the patient's autonomy is impaired. Strictly speaking it is not so much the paternalism that is weak, but the patient's autonomy that is so impaired that her welfare is at risk in the absence of a therapeutic intervention.

Autonomy is the person's right of self-rule and is generally supported and defended in liberal societies. The restriction of autonomy is justified only if it is likely that the individuals will do serious harm to themselves, or if they deny themselves important benefits (e.g. health). Weak paternalism is thus the justification for a non-consensual intervention when it can be shown that the person's autonomy is impaired.

Although autonomy seems to be a very similar concept to that of ‘mental capacity’ proposed by the UK Law Commission in 1995, the author shrinks from accepting that impaired autonomy is due to a mental illness. After all, mental illness does not necessarily impair a person's autonomy. A patient's lack of competence (capacity) in reaching treatment decisions should not be presumed but rather it should be assessed, and this assessment will need to be repeated at different times.

The author distinguishes two kinds of autonomy. The first, named substantive autonomy, is satisfied when the content of the person's action is deemed rational, that is, the majority of people would act similarly in similar circumstances. The second kind is the formal or procedural conception of autonomy. It is satisfied when the person's process of reasoning and deliberation is judged appropriate to her decision-making. The person requires an assessment of her understanding and her ability to ‘balance the costs and benefits of proposed alternatives’ (rather than the result of the choice). This capacity for decision-making is relative to the specific decision and to the time it has to be made. This second kind of autonomy is characteristic of the legal approach to decision-making capacity in the UK.

This is not a dull book. At times it may be taxing to follow the tight-knit arguments but the reader's attention is soon revived by a series of radical and controversial asides. For example:

According to the author, the diagnosis of an underlying mental illness is no justification for coercive treatment. This also applies to the patients with anorexia nervosa. Yet when it comes to the crunch she inclines towards the need to save the patient's life, adopting weak paternalism because she can accept that the patient's behaviour is not truly autonomous. She describes eating-disordered behaviour as far too irrational: ‘it is impossible for one to sacrifice her health and even her life for the sake of “thinness”'. The ethics of care and treatment of the person with an eating disorder therefore relies on a better understanding of the disorder.

With this aim in mind the author has conducted a compelling analysis of the psychological mechanisms whereby the anorexic patient's autonomy is likely to become impaired. They consist of a disturbance of body-image, a faulty awareness of signals of hunger and satiety, and cognitive distortions (‘I'm different, 300 calories a day is plenty for me'). These faulty mechanisms compromise the process of deliberation and thus rule out eating-disordered behaviour from being autonomous. This may provide an ethical justification for non-consensual intervention.

This book achieves a high level of scholarship in its reviews of the literature on philosophy, ethics and the law relevant to eating disorders. It suffers a serious lapse, however, when it comes to the interpretation of empirical studies on the compulsory treatment of patients with anorexia nervosa. The small number of these studies makes it all the more important to report the findings accurately.

It is not true that there have been no studies of treatment outcome comparing compulsorily and voluntarily treated patients (Reference Griffiths, Beumont and RussellGriffiths et al, 1997; Reference Ramsay, Ward and TreasureRamsay et al, 1999). It is a serious misinterpretation to state that the short-term weight gains in compulsory patients will be followed by higher long-term mortality, or that suicide is more likely (Reference Ramsay, Ward and TreasureRamsay et al, 1999). It is erroneous to state categorically that compulsory treatment compromises the relationship with the therapist (Reference Serfaty and McCluskeySerfaty & McCluskey, 1998).

In conclusion, the practising clinician may gain only limited practical help from this book when it comes to the non-consensual treatment of anorexic patients, with one exception. This exception concerns the author's clear guidelines on how to assess the individual patient's autonomy (or mental competence/capacity) when accepting or refusing medical treatments. But the most important reasons for studying this book carefully are for its tightly argued philosophical and ethical discourses.