Statistics from Altmetric.com
- clinical ethics
- quality/value of life/personhood
- allowing minors to die
- allocation of health care resources
Work done within the realm of what is sometimes called ‘descriptive ethics’ brings two questions readily to mind: How can empirical findings, in general, inform normative debates? and How can these empirical findings, in particular, inform the normative debate at hand? Brick et al 1 confront these questions in their novel investigation of public views about lives worth living and the permissibility of withdrawing life-sustaining treatment from critically ill infants. Mindful of the is-ought gap, the authors suggest modestly that their data ‘may be useful when constructing guidelines for clinical practice’. In this commentary, I consider some potential limitations on the generalisability of these data in the context of policy and clinical practice, limitations which may in turn serve as directions for future empirical research.
The authors’ study presented participants six case scenarios, five of which involved the possibility of withdrawing life-sustaining treatment from critically ill infants. The results revealed that “[a]lmost all respondents agreed that at some level of quality of life, life may be of no benefit or worse than death for an infant” and that “participants significantly favoured withdrawal of treatment for the most severe cases.”
One salient detail in these scenarios is that they involve withdrawing treatment— discontinuing treatment keeping the infant alive—as opposed to withholding treatment—never initiating the treatment to begin with. As we know from Savulescu and colleagues’ earlier work,2 however, medical professionals “sometimes find it harder to stop a treatment (withdraw) than to refrain from starting the treatment (withhold),” a preference termed withdrawal aversion. If, as seems plausible, withdrawal aversion is a cognitive feature exhibited by the general public, then we should expect public attitudes about the ethics of withholding and withdrawing life-sustaining treatment in the cases described to differ. Herein lies the first limitation to the generalizability of Savulescu, Brick, and Kahane’s data: Given withdrawal aversion, the public may consider withholding treatment to be permissible/obligatory at a higher level of quality of life than the level at which withdrawing treatment is permissible/obligatory. This point risks being abstracted away in the authors’ Figure 1, adapted from Wilkinson,3 which depicts a single “lower threshold” at which “it is obligatory to withdraw/withhold treatment.”
Another relevant feature of the study is the fact that participants received little by way of background for the patients described in the case scenarios. Were the infants hospitalised as a result of a progressive congenital condition? Were they perfectly healthy prior to an infectious disease? Did they experience some kind of traumatic injury? Whether or not these kinds of details turn out to be morally relevant, it seems that they may well influence folk intuitions about the cases.
Consider, for example, two modified versions of the “Possible Relational Capacity” case. The first involves an infant born with severely impaired cognition, requiring artificial nutrition and hydration, who is hospitalised for respiratory distress and placed on ventilator. The second involves a previously healthy and cognitively typical infant who ends up in the same clinical condition after suffering a traumatic brain injury in a motor vehicle accident. Despite the fact that both now have an equivalent absolute quality of life, their quality of life relative to baseline is drastically different, which may influence evaluations of the cases.
Of course, how the distinction between relative and absolute quality of life influences judgments about the ethics of withdrawing life sustaining treatment is an empirical question. But one can imagine it cutting both ways. Perhaps respondents will be less accepting of withdrawal from the previously healthy infant, on the grounds that the prior richness of his identify and relationships conveys greater value on his continued life. Or perhaps, conversely, respondents will be less accepting of withdrawal from the infant with the congenital disability, on the grounds that his quality of life has remained more-or-less unchanged, whereas the previously healthy infant has experienced a significant disruption to his life. Regardless of where the data falls, the point here is that public views about quality of life and treatment withdrawal in the abstract may be limited in their applicability to particular cases.
Furthermore, given that the case scenarios involve infants—approximately 6–18 months of age, by stipulation—it is worth asking what influence patient age has on views about quality of life and the ethics of treatment withdrawal. Do attitudes about infants reveal any broader attitudes? If so, what kinds of attitudes? If not, what makes infants unique with respect to these questions? Again, it is a potentially interesting empirical matter to sort out how age considerations affect the relevant judgments. On the one hand, the public might tend to think that death is worse for the very young, given that it deprives them of a greater amount of potential life. On the other, one could note that infants are less developed in terms of social identities, interests, and life plans, which suggests that death may actually deprive them of less. This suggests a third, if obvious, potential limitation of the data: Attitudes about quality of life and the ethics of withdrawal may change according to patient age.
The limitations of Savulesu, Brick, and Kahane’s data noted above are meant not as objections to their methodology, but rather as points of caution for those wishing to apply these data in the context of clinical practice, health policy, and law. Further empirical investigation into whether and how attitudes on this issue may vary light of withdrawal aversion, relative quality of life, age, and other factors may help to address these limitations and further elucidate how public views can inform the normative debate.
Thank you to Jennifer Blumenthal-Barby and Bryanna Moore for reviewing a draft of this commentary.
Contributors RN is the sole author.
Funding Baylor College of Medicine, Center for Medical Ethics & Health Polic.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.