‘Waiver or Understanding? A Dilemma for Autonomists about Informed Consent’

Gopal then went on to sketch an argument that the use of the waiver is itself justified in these terms as an exercise of autonomy, in that the subject themself is shaping through their own choice (itself perhaps informed in certain respects), and in the light of their own values, the conditions under which they live. Their later acts may perhaps be less autonomous, but this can be seen as consistent with their original act’s having been autonomous (and need not imply that one can autonomously surrender one’s autonomy entirely, perhaps by selling oneself into slavery).
But why might someone want to waive disclosure? Isn’t it just straightforwardly unreasonable? Gopal described three cases in which that seems not to be the case: a person may have a particular (not unreasonable) view on which information they see as relevant, and a desire to avoid information overload; a person may have a terminal condition and be worried that information will put them off trying a treatment that may in fact benefit them; or they may be altruistic and have the same worry about accepting a treatment that may benefit others.
by Roger Crisp
At a recent New St Cross Ethics seminar, Gopal Sreenivasan, Crown University Distinguished Professor in Ethics at Duke University and currently visitor at Corpus Christi College and the Oxford Uehiro Centre, gave a fascinating lecture on whether valid informed consent requires that the consenter have understood the relevant information about what they are being asked to consent to. Gopal argued that it doesn’t.
The upshot is that, if autonomy provides a sound basis for allowing a person to waive disclosure, then it cannot also be used to support a requirement on informed consent that it rest on understanding that disclosure.
He began by elucidating the two medical contexts of informed consent — treatment and research – and the two duties involved – those of disclosure and securing consent. He then noted that in many jurisdictions, subjects of randomized controlled trials (RCTs) can waive disclosure. This shows that understanding disclosure is not, within these jurisdictions anyway, required for valid consent (on the assumption that the RCTs in question do require consent).
Gopal ended by formulating two further arguments that can be interpreted as independent of autonomy. The first rests on the simple value of allowing people to choose (which might be based, for example, on the Millian idea that individuals are the best judges of their own interests). The second appeals to considerations of institutional or regulatory design. The usual standard for information in RCTs is that which would be required by the ‘reasonable person’. On the face of it, it seems more morally ‘pure’ to tailor the information provided to each potential subject, but this is of course impractical. The upshot is that most consent forms contain much more information than most people want. Allowing a waiver enables the process to come closer to providing what individuals actually want.
Gopal next outlined a second, autonomy-based argument: that by forcing the person to receive the relevant information, or not to participate, one is violating their autonomy. Autonomy can’t plausibly be said to consist in forcing someone to make a decision under conditions they themselves reject.
Gopal’s arguments, all of which, as he noted, allow that there might be other reasons for not allowing such waivers, are powerful, and — importantly — based on assumptions internal to the autonomist’s overall position. But what if the autonomist refuses Gopal’s elucidation of their notion of autonomy? Then we will be left asking exactly what autonomy consists in. One possibility here, emerging from recent work on ‘conceptual engineering’, is that, rather than debating which is the correct conception of autonomy, we allow that there may be different conceptions worth considering, and then ask which conception it would be best for us to use, either generally or in particular circumstances. Once we make clear what the ultimate point of RCTs is – in particular, to promote human health – that might provide a pragmatic argument in favour of allowing a ‘diachronic’ conception of autonomy in line with Gopal’s arguments.