Respect for autonomy: Consent doesn’t cut it

As we all know, respect for a patient’s ability and freedom to make decisions about healthcare matters that concern them (i.e. ‘respect for autonomy’) is necessary for the ethical conduct of clinical practice. In contemporary liberal societies, the slogan ‘the doctor knows best’ is no longer afforded much currency. Indeed, even when the spectre of paternalism rears its head in specific, unusual situations, for example, during the COVID-19 pandemic, it tends to be accompanied by a sense of unease among ethicists, legal scholars, patients, and healthcare practitioners. In most jurisdictions worldwide, informed consent is seen as the application of the principle of ‘respect for autonomy’, a perception that has been supported by now classic discussions of patient autonomy in medical ethics and law. The roots of the alignment of informed consent and respect for autonomy can be found in late-1950s US case law, which employed the language of autonomy to introduce the concept of informed consent to clinical medicine. These developments in law permeated medical law, ethics, education, and Western clinical practice in the 1970s. Today, the initial legally motivated links between the concept of autonomy and informed consent have become conceptually entrenched in the domains of healthcare and biomedical research, and the conflation of these two concepts is now a widespread assumption in law and clinical and research ethics frameworks.3–6 In short, the courts, regulators, clinical ethics teams, and biomedical researchers have assumed that when an individual gives valid consent, it is autonomous. This assumption is deeply problematic with wideranging ethical, legal, and well-being implications. From a theoretical point of view, informed consent is the standard mechanism through which a patient exercises their liberty at law, giving permission for, and setting the limits of, bodily interference.3–8 By contrast, according to Feinberg, the concept of autonomy can refer to the capacity to govern oneself, the authentic exercise or achievement of selfgovernment, a value that one instantiates such that one should be afforded equal standing, or the value that paternalism, coercion, and other malign influences fail to respect. The concept of informed consent fails to adequately capture any of these interpretations of autonomy. The practice of informed consent does not take into account whether an individual has rationally responded to their values, desires, or motives or whether these values are truly theirs. In addition, standard requirements for informed consent are such that some individuals who should arguably be afforded the opportunity to make claims to autonomy are denied. Further, the giving of valid consent does not imply that the decision has not arisen from normatively significant external influence. Finally, because the capacity conditions for valid consent are framed in purely cognitive terms, consent does not account for important relational or embodied (i.e. non-cognitive) factors that have been shown to affect or constitute one’s capacity for autonomy.12–15 The main upshot of this conceptual confusion is that a patient can give valid consent and thereby be perceived to have had their autonomy respected, yet still fail to make an autonomous treatment decision (i.e. because, for instance, they’ve failed to satisfy certain autonomy conditions beyond those required by informed consent). At this point, defenders of consent as the mechanism for respecting autonomy may argue that these problems are purely theoretical and to be expected if we rely on ideal theory, and that, when it comes down to non-ideal, concrete situations, informed consent is our best bet for living up to the spirit – albeit not the letter – of the principle of respect for autonomy. I am hugely sympathetic to the idea that although medical ethicists can develop clear, principled analyses of what ought to happen in clinical contexts, the demands of clinical reality are such that it can be of little use to patients and practitioners to point to the ideal. If the issue were purely a theoretical one, then I’d be content to accept consent as a proxy for respect for autonomy in the interest