Statistics from Altmetric.com
As a neonatal neurologist, I serve families facing tragic decisions in which they must balance trade-offs between death and life with profound disability. I often find myself in complex discussions about future outcome, in which families sort through in real-time what information they value most in making such a choice. Will he laugh? Will he be in pain? Will he know how much he’s loved? In this month’s feature article, Brick et al share the results of an online survey aimed at assessing public views on when a life is not worth living, in an effort to inform ongoing legal and clinical debates about withdrawal of life-sustaining treatment for children. These findings raise important questions about how we define and measure outcomes that matter to parents as they make these tragic decisions.
It is challenging to interpret these findings in the absence of context of how decision-making for infants occurs in real-time. The nature of this study and its objective required that cases be substantially simplified. As the authors acknowledge, the prognoses and function of the children involved in these cases were often deeply contested. Prognostic uncertainty is common for children with significant neurological compromise. Conditions that cause profound neurological impairment are often rare, and include conditions with a wide range of outcomes. Despite a growing and robust literature in prognostication, there is no single biomarker that predicts future outcome in most conditions that impact a child’s brain. Outcomes are mediated by an individual child’s genes and environment, and the paediatric brain fortunately has a tremendous capacity for plasticity.
The ability for clinicians to adequately offer prognostic predictions is further limited by how data about outcomes are presented in the literature. There are a variety of reasons an outcome might be selected for a given study—to understand the impact of an intervention, to help families and clinicians anticipate the future, or to help guide clinical decision-making. Outcomes required to facilitate complex decision-making with families may differ from those required to measure intervention efficacy. Many important studies about outcome in the face of neurological injury rely on composite outcomes, which lump together outcomes that may carry different weight to decision makers. For example, landmark trials in paediatric neurology and neonatology have often relied on a primary composite outcome of death or moderate or severe disability. Moderate and severe disability are themselves composite outcomes, inclusive of some combination of motor, cognitive, visual or hearing impairment.1 Effectively, in trials that utilise the composite outcome of death or severe disability, death is equated with isolated blindness. Further, isolated blindness is equated with blindness in conjunction with quadriplegic cerebral palsy and profound intellectual disability. Few would argue that decision makers value these outcomes equally.
All of the cases presented in this article would have been classified as ‘severe disability’ in most clinical trials. And yet, it is clear that respondents valued certain attributes differently from others. Respondents valued the capacity for basic relationships. Most people thought that treatment ought continue in the possible relational capacity scenario. These findings suggest that, at a minimum, studies of outcome should include some description of a child’s capacity of a back-and-forth relationship with those around them. Future work should explore what types of relationships stakeholders consider and value in this context.
Respondents also prioritised the presence of pain and the ability to experience pleasure. These attributes are rarely measured or described in the medical literature around prognosis, and are challenging to assess in children with severe disability. How does one reliably assess pain severity for children who have limited ability to speak or move? How does one gauge the pleasure such a child may experience from being held by a loved one? Measures used to assess infant and child pain often rely on behaviours like crying, body movements and facial expressions. These features may be altered or absent in infants with neurological impairment.2 Existing literature would further suggest that, a priori, providers perceive that infants with neurological compromise experience less pain than other infants.3 Assessing well-being in the setting of severe disability is equally complex. Clinicians, families and patients with neurodevelopmental disability may consider well-being and quality of life differently from each other, and existing literature raises the possibility that physicians are more likely than parents to view death as preferable to severe disability.4 5 An important next step of this study will be to explore differences and similarities between the public, parents, patients and clinicians.
Respondents had modest agreement about how to approach cases at each end of the outcome spectrum. The cases in the middle, however, were more divisive, even with simplified scenarios that lacked the uncertainty present in the real-world. It seems likely that this disagreement would be amplified had cases included uncertainty around current function and expected outcome. As noted by the authors, this finding appears to support the threshold view, which recognises that there are certain scenarios in which either treatment continuation or withdrawal is morally permissible.6 This practical framework allows clinicians and parents to weigh these awful decisions in the context of individual values and preferences, and recognises the inherent moral and prognostic uncertainty that exists. It also underscores the responsibility of clinicians who partner with families in these tragic choices, when there is no morally agreed on answer. How clinicians predict and communicate about future outcome, especially information about pain, pleasure and relational capacity, is likely to impact the choices parents make. It is imperative that we define how to best measure and communicate decision-relevant outcomes to inform these discussions.
This important article raises novel questions in a persistent debate. Ethicists, clinicians and researchers should use this work to consider how we might better measure outcomes that matter most to families.
Contributors ML drafted and revised this manuscript.
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.