Respect for autonomy is an important moral principle within medical ethics. However, the question of whether the normative importance of respect for autonomy is derived from other moral principles (such as welfare) or has independent moral value is debatable. In this paper it is argued that the normative importance of autonomy is derived from both welfare and non-welfare considerations. Welfare considerations provide two types of reason to respect autonomy, one related to the role of autonomy in creating welfare and one related to its role in constituting welfare. In addition, autonomy seems to have normative importance that is unrelated to welfare considerations. This type of normative role is difficult to defend within medical ethics, because most non-welfare justifications of autonomy work for only a proportion of the autonomous decisions that patients make and give no clear guidance on how to respond to autonomous yet welfare-reducing treatment requests. A recent account of autonomy (Stephen Darwall’s “demand” account) provides a nuanced defence of autonomy that does not rely on welfare considerations. Darwall’s approach seems to work well within medical ethics and provides a principled explanation of how to respond to autonomous patient requests for treatment options that may not be in their best medical interests. It is argued that to fully respect autonomy within a medical consultation, practitioners must consider non-welfare autonomy as well as instrumental and intrinsic welfare-related autonomy.
Statistics from Altmetric.com
Competing interests: None.
Provenance and peer review: Not commissioned; externally peer reviewed.
↵ii At one extreme is the view expressed by Raanon Gillon6 that autonomy is primus inter pares—“autonomy is … what makes morality—any sort of morality—possible … autonomy is morally very precious and ought not merely to be respected but its development encouraged and nurtured.” In the same vein, John Harris7 argues that “where concern for welfare and respect for autonomy are incompatible one with another, concern for welfare must give way to respect for autonomy.” A similar view is expressed (though not without serious reservation) by Callaghan:8 “autonomy is then de facto given a place of honour because the thrust of individualism is to give people maximum liberty in devising their own lives and values.” By way of contrast, while commentating on Gillon’s paper, Dawson and Garrard9 argue that “this claim [that autonomy is first among equals] makes little sense if the principles are to retain their prima facie nature.”
↵iv Typical examples taken from the UK General Medical Council’s Good medical practice include “Patients must be able to trust doctors with their lives and wellbeing … make the care of your patient your first concern … act quickly to protect patients from risk …”12
↵vi In fact, all these examples occurred in my individual general practice while I was writing this paper.
↵vii There may be more—for example, autonomous decisions based on aesthetic considerations.
↵ix Another possible explanation is that this kind of activity represents acts, whereas allowing the daughter to eat what she wants is an omission.24 However, given the philosophical problems with the acts and omissions doctrine, coupled with the difficulties in drawing a clear line between acts and omissions, this seems a less than satisfactory explanation of the difference.
↵x For the same reason, a doctor would not be expected within a consultation to endorse a patient’s racist opinion (even if that was an autonomous opinion expressed by a patient) or adopt a religious view that was contrary to the doctor’s core spiritual values or adopt an aesthetic standpoint that was uncomfortable to the doctor. And so on. Indeed (for example), in the situation in which a patient expresses racist opinions, the doctor might have to contest these views for fear of the patient interpreting silence as endorsement of racism.
↵xi This paper glibly assumes that it is easy to differentiate between autonomous and non-autonomous choices; of course, this is rarely easy in clinical practice. However difficult this may be in practice, the argument advanced in this paper depends only on the existence of some choices that are both autonomous and non-welfare-promoting—and this seems non-controversial for the reasons given earlier in the paper.
↵xii Though not necessarily an over-riding reason—if, for instance, the welfare gain to this particular patient produces a large welfare loss to other individual(s).