Death and best interests

Introduction

In this paper I have considered the problem of how we can assess the interests of critically ill children who will survive only while aggressive medical support is continued. If aggressive medical support is withdrawn, the child will die shortly afterwards. This is an important topic because when decisions are made for children by the courts, the standard is that decisions should be made in the best interests of a child. My claim is that this is not a coherent way to consider how children in this situation should be treated.

Medical decisions are usually made for a child by their parents, ¹ but on occasions when the treating clinical team and parents are unable to agree on the medical treatment, the courts are asked to decide how the child should be treated. The Children Act requires that when decisions must be made for a child ‘the child's welfare shall be the court's paramount consideration’. ² The phrase ‘the child's welfare’ seems similar to the idea of the child's interests, though not necessarily the child's best interests.

However, in an example from a recent case in the UK, a court was asked to decide on the further treatment of a terminally ill child (MB) aged 18 months with spinal muscular atrophy (SMA), when the family and clinicians could not agree. ^3,4 In SMA, although babies are usually normal at birth, the nerves which control the muscles degenerate so that children progressively become paralysed. The disease is incurable and relentless. Some forms of SMA are more severe than others. MB had the most severe form and in this form respiratory failure causes death within the first year unless the child is ventilated artificially. MB had been ventilated but the treating clinical team believed that further treatment was ‘cruel’ and so aggressive treatment should be withdrawn. If this happened, MB would die immediately. MB's parents disagreed with this course of action and wanted aggressive medical treatment to continue. It was not possible to reach a compromise. In the absence of agreement between MB's parents and the treating clinical team, an application was made for judgment in court.

The judge described (from an earlier case ⁵ ) the process by which the decision should be made:

As a dispute has arisen between the treating doctors and the parents, and one, and now both, parties have asked the court to make a decision, it is the role and duty of the court to do so and to exercise its own independent and objective judgment.

An objective test of interests

Mr Justice Holman describes that there is ‘… an objective approach or test. That test is the best interests of the patient.’ In a liberal society it is broadly left up to individual members of society to make their own assessment of what is in their own interests. This is the approach underlying Mill's Harm principle. Or these are the liberties which are protected by the International and European Conventions on Human Rights which emphasize the freedom of citizens to make their own choices. The fact that respect for autonomy is taken to be the most important of the four principles of medical ethics (or ‘the first amongst equals’ ⁷ ) again underlines that ideas of best interests come second to approaches that recognize the importance of a person's autonomy, and that we rely on autonomous individuals to choose what they regard (subjectively) as being in their interests. Perhaps, another way of describing this is that the most important or significant interest for people in a liberal state is their interest in the freedom to make their own considered choices. It is wrong to be more specific than that, within a liberal society. I use these examples to demonstrate that although we may believe that we have a clear idea of what may be in another's interests, it is rarely, if ever, tested or put into practice. Best interests come second to respect for autonomy. There is neither any agreement about what a person's best interests are, nor an established process through which an assessment of someone's interests could be made. As an example of this, although the Mental Capacity Act ⁸ (MCA) makes use of the idea of best interests (the guidance notes to the MCA note ‘… all steps and decisions taken for someone who lacks capacity must be taken in the person's best interests' ⁹ ). The Act itself does not provide a clear definition of what best interests are. For example, the guidance notes state ‘… the notion of “best interests” is not defined in the Act’ and ‘The factors in this section do not provide a definition of best interests and are not exhaustive’. ⁹

However, incompetent children do not have capacity and so cannot make an autonomous choice, and for them an equivalent of an autonomous choice is a choice made by their parents. I have argued elsewhere that the justification of parental authority is different from (and less strong than) the basis of respect for an autonomous choice of a competent adult. ¹⁰ Nevertheless, choices must be made for children and parental decisions can be over-ridden (on occasion) by decisions which appeal to a child's interests. In some cases it may be easier to come to an agreement where there is dispute over how a child should be treated. The easier cases will be where those in dispute agree to what would be in a child's interests, and just disagree on how those interests will best be achieved. An example here may be where divorcing parents argue over the custody of their children: the parents may agree that their children need love, security and to be with their siblings among other factors, but just disagree on who would be best placed to meet those needs for their children. The more difficult cases are where there is no agreement on what would be in a child's interests. Here, the disagreement is not over the practicalities of how to meet a child's interests, but what those interests are. This is where the absence of an objective test of best interests or a recognized procedure to derive what could be regarded as a child's best interests is felt most.

Furthermore, not only do we not have an objective test of best interests we do not even have a clear framework within which best interests can be assessed. Various broad outlines to the way that we should assess another's interests have been proposed (among which are an objective list of what is good for someone; an assessment of desire fulfilment – an approach often used in economic evaluations; or a modification of desire fulfilment to take account of what may matter over a whole life, or to reflect that some desires may be more valued or valuable than others). There is extensive disagreement regarding even which is the best framework within which interests should be considered. ¹¹ Furthermore, this matters practically: if different approaches are taken when considering how terminally ill patients should be treated, then this alters how the child (or adult) should be treated. ¹² This difficulty is reflected in the way that the MCA is not prescriptive about best interests.

I have not argued that we will not be able to develop an objective test of best interests, merely that we have not yet done so, nor does there seem to be a clear way in which we could work towards developing an objective concept of best interests.

Best interests around the time of death

I will now consider how we can talk about a child's interests in the situation where a child will die shortly after withdrawing aggressive medical interventions. Perhaps, the question could be phrased as: are there situations in which to die is in someone's best interests? There are two broad ways in which it can be successfully argued that to die is in a person's best interests.

First, where this is based on the autonomous decision of the patient (or based on their wishes before they became incompetent). It is recognized that a competent (or previously competent) person may refuse intervention even where such a refusal results in the person's death. As an example, following a brainstem stroke a lady continued to require ventilation, which she asked should be withdrawn, though her physicians argued that ventilation was in her interests. The court considered the case and agreed that ventilation should cease. When ventilation was withdrawn she would die. ¹³ Or as another example, there is a well-recognized right of adult Jehovah's Witnesses to refuse blood transfusion even though it may result in their death. There are perhaps two ways to understand these cases. First, to claim that a person's own assessment of their interests overrides (or is more correct than) any other objective or impersonal assessment of a person's interests. In these cases we are acting in a person's best interests, but the only person able to assess their own interests is the person themselves. Or secondly, in the absence of an impartial assessment of best interests, we must recognize that a person's autonomy to choose how their life should go is more important than another's conception of best interests.

A second sort of case where it may be argued that death would be in a person's best interests is where the only life possible for a person is one of suffering: full of pain or torment. It is possible to conceive of situations of such pain or distress that death would be a benefit – the classical philosophical example is of a person trapped in a burning vehicle, who will inevitably die, but whose suffering may be relieved by hastening his death, perhaps by being shot. Or in a real-life example, Dax Cowart, an American who was severely burnt recounting his first words to a rescuer as ‘I asked him for a gun. He said why. I said, can't you see I am a dead man? I am going to die anyway. I have got to put myself out of this misery.’ ¹⁴ However, this is not a realistic scenario within the situation we are considering (the medical service of a sophisticated country) where if distress and pain are present they will usually be relieved by drugs.

It is unlikely that either of these situations will be present in the cases which I intend to consider: critically ill children who will die soon after aggressive medical treatment is withdrawn. Children are not competent to make autonomous choices concerning their treatment, which excludes them from the first group. Good medical practice demands that suffering is relieved. Pain is relieved with analgesics, and anxiety and distress can be treated by sedating or anaesthetizing a child. It is unlikely (and unacceptable clinical practice) that a child in a critical care unit of a hospital would be suffering and they would certainly not be suffering to the extent that death would be a benefit. If death would not be a benefit by removing suffering, then they will not fall into the second group.

Perhaps, we can start off by considering whether the dead have interests? I accept that those who will die when treatment is withdrawn are not yet dead, but the situation I have described is where withdrawal of aggressive medical treatment will be followed by death quite promptly. This is a very different situation from other situations where aggressive medical treatment is withdrawn (or limited) with the prospect that a child (or adult) may then spend a reasonable amount of time pursuing activities they enjoy instead of being burdened by unpleasant medical treatment. In the latter case it may be argued that it is in the child's interest to limit aggressive medical treatment because they will be able to spend time with people they love, doing activities that they enjoy.

In the situation in critical care, we cannot appeal to the interest of the child in having a good (albeit shortened) time between limitation of treatment and death and so the question of whether the dead have interests becomes more important. This question has been discussed extensively. ^15–17

A standard objection to the idea that the dead have interests is that there seems to be no-one who can be harmed or benefited. Notwithstanding that objection, there is some sense in which those who are dead can be harmed (as an example where evidence that they behaved badly becomes available only after they have died: the exposure of Robert Maxwell's fraudulent behaviour) or benefited (where evidence of good conduct becomes available only after death, as an example the legend of Gelert). Accepting the idea that there is some way in which a dead person may be harmed or benefited, it seems difficult to conceive of a situation in which the interests of a dead person could be compared on the same scale as the interests of a living person. The interests of a dead person will always seem inconsequential alongside those of the living. It may be ideas such as these which underlie the recent changes in transplant law to make salvage of organs for transplant from the brain dead more likely. ¹⁸ Or, although we may feel disgust in the situations where survivors of a tragedy have cannibalized the flesh of their fellows who have died, we can understand their actions even if we disagree with them. ^19,20 These examples reflect the way that the interests of the living deserve so much greater attention than the interests of those who are now dead.

Consider the situation where the two preconditions are met: first, given that pain and distress are relieved so that death does not remove suffering; and secondly, where it is clear that a child will die shortly after treatment withdrawal. If these are present and if it is agreed that the interests of the dead are much reduced (or almost absent) compared with the living, then it is impossible to claim that a decision to withdraw treatment where the inevitable consequence is prompt, death is to act in a child's best interests. Especially if, because of the irrevocability of the choice to withdraw treatment, we add in ‘There is without doubt a very strong presumption in favour of a course of action which will prolong life…’. ²¹

Tony Bland was a young man who was left in a persistent vegetative state following the Hillsborough disaster. The court was asked to consider whether it would be acceptable that artificial feeding could be stopped and symptomatic treatment given should it be appropriate. The court agreed with this course of action, and in judgment Lord Mustill noted that ‘The distressing truth which must not be shirked is that the proposed conduct is not in the best interests of Anthony Bland, for he has no best interests of any kind’. ²² My claim is that some critically ill children present similar difficulties with the assessment of their interests.

And to further complicate matters, I have described that in good medical practice pain or anxiety would be relieved with analgesics or hypnotics. In sufficient doses these drugs may have the effect of effectively anaesthetizing a child so that the ‘good’ or positive parts of life are reduced alongside the bad parts. A child may be in distress or pain and it would then be in that child's interests to allow them to die, but drugs can relieve the unpleasant symptoms. When sedative and analgesic drugs are given, they may remove the awareness that the child has of her surroundings, or her parents or siblings. This approach may reduce a life to an existence (with perhaps no interests or certainly dramatically reduced interests).

Although I have separated the idea of a life and of an existence, I have not clearly described what distinguishes the two concepts. The concept of an existence (separate from a life) is perhaps best seen in a person who is declared brain dead after formal testing following devastating brain injury: there is no sign that his brain works at all. Air is forced in and out of his lungs through a ventilator so his chest moves; he is still warm; his heart continues to beat. He is biologically alive but biographically dead. A live body continues to exist, but he (the person) is legally dead. And as well as this, there is no sense in which anything that happens could mean anything to him now. To move too far from the paradigm of the brain-dead individual requires that we consider what is valuable in a life, or what makes a life worth living or flourish and that is not my subject here.

If we can never claim it is in the best interest of a child to die, can we re-phrase the question and say it is not in a child's best interests to continue to ventilate her, or to treat the child aggressively? My response is that this is just sophistry. A claim to act in someone's best interests is demanding. I may claim to act in a child's interests having considered only my one proposed course of action. But a claim to act in someone's best interests demands that we have considered all the other reasonable approaches and have chosen the best comparing one course with another. In some situations a child will benefit from continuing treatment and so it is appropriate to continue. An assessment of a child's interests or a claim to act in their best interests makes sense in this situation. But, the claim that I want to make is that for at least some children (children who are in the same sort of state as Anthony Bland), a decision to limit treatment cannot be in their best interests. To phrase this in a different way, best interests will work in one way only: it will justify a decision to keep treating a child. It will not work the other way, in that there will not be situations in which it will be in the child's best interests to stop treating the child (in the situations where a child lacking capacity is not suffering).

An example

In order to demonstrate the difficulties with interests around the time of death, consider a situation where there are two children ventilated in an intensive care unit. One, Anna, has suffered a devastating neurological injury. She has no evidence of conscious neurological function at all, merely some automatic brainstem responses, which no one is persuaded are evidence of any conscious mental life that means anything to Anna herself. Because she has insufficient neurological drive to breathe, she needs to be ventilated using a machine. Her parents are emphatic that she should continue to be ventilated: they cannot bear the thought that she might die. Her parents accept that she is in a coma and that she has no conscious life and that she does not interact with them or recognize them. They understand she has no life of any meaning to Anna.

Boris has also suffered a devastating neurological injury. His brain injury is less severe though. He is conscious, but is in chronic and persistent pain. He also requires invasive ventilation. His pain is unrelievable by any drug or intervention though many have been tried. His discomfort is added to by the interventions needed for nursing care (to suction down into his lungs to relieve the secretions which build up) and the need for turning to avoid pressure sores which exacerbate the spasms in his limbs. His parents are emphatic that he should continue to be ventilated: they cannot bear the thought that he might die. His parents accept that he has very limited conscious life and that he does not interact with them or recognize them. They understand that he has a life of very limited meaning to Boris (and that life is unpleasant), but beg that you sedate and analgese him as much as you can.

If we are acting in a child's best interests we cannot claim that it is in Anna's best interests to stop treatment: she seems to have no interests at all. This may be too quick: Anna's parents have interests in Anna and how she is treated. And perhaps (had Anna been competent) Anna may have (or have had) some interest in the way that her parents would remember her after her death. And in a further complication had Anna been competent prior to the episode of neurological injury then she may have had an interest in the way that she survived, perhaps she had an interest in not surviving as a shell of her former self, or perhaps she had staunch pro-life views which would have demanded that she should be treated with full aggressive medical therapy. I am not sure what the correct response to the claims that Anna (had she been competent) could or would have had an interest in how she should be treated after a neurological catastrophe. I can side-step them by describing that Anna is a baby who has never been competent, and she is now in a state where she will never be competent. If we agree that she does not have interests, then to say that it is in her best interests to stop treating her is a strange way to describe the situation.

But we could say that it is in Boris' interests to stop (because he has sufficient residual mental ability to suffer). We could claim that because Boris is suffering, if he is allowed to die, then he will be better off. The responses to these two cases seem to be wrong: it is counter-intuitive that it is acceptable to stop treatment in the less severely damaged child rather than the more severely damaged child.

An added complexity here is that these two situations could co-exist within one child: if a child was in Boris' condition it would usually be possible (by using sedatives, anaesthetic agents or analgesics) to overcome the problems that Boris had. The consequence of this is often that the side-effects of the treatment may be to reduce Boris to the sort of state that Anna was in: but in a drug-induced coma (and this does happen in practice). ²³ So, although I described a case in which Boris' suffering was unrelievable by drugs, this is unlikely to be the case (at least in the short term: in the longer term children may become tolerant of sedative and analgesic drugs and their original problems return and are unresponsive to increasing the doses of the drugs).

If we accept the conclusions of the example (that when these decisions are based on best interests then it is acceptable to withdraw therapy from the less damaged child but not the more damaged child) and that this runs counter to our intuitions, what response is appropriate?

As a first response, perhaps it should just be accepted that the application of sound principles in some situations will produce outcomes which are counter-intuitive. For example, there is a strong intuition that an act which produces a bad outcome is worse than an omission which produces the same event. Rachels argues that this is not so. ²⁴ That we should treat children in their best interests is such a strong principle that this counter-example should not reduce our commitment to it. Even though this individual example seems wrong, it is right that we should treat children (and other incompetent individuals) in their best interests. The idea that we should treat people in their interests should not be overridden or overridden only in clearly defined circumstances (perhaps as an example of a situation where it is acceptable to override the principle that we should treat an individual in their interests is where a person's own autonomous choice runs counter to what others would see is in their interests as in the examples given earlier of an adult Jehovah's Witness or Ms B).

However, the response which I believe to be correct is to accept that in these circumstances it doesn't make sense to talk of the child's interests and certainly not of the child's best interests. I am not arguing against the claim that when we deal with children in most situations we should consider their interests, but against the idea that an assessment of interests around the time of death in these circumstances makes sense. I want to make clear my deep unease when claims are made that we can act in the child's best interests in these situations. Or if we maintain the fiction that we should decide in a child's best interests (in these situations) then, if my arguments are sound, I cannot foresee any situation in which it is possible to decide that we should limit aggressive treatment (and if so, we will impose treatment on children in situations where – if they were adults – they would be able to refuse it, regardless of the further problem of the implications for resources needed for this sort of approach). If my argument is correct, then the approach which is used by the courts to determine the treatment of critically ill children in these circumstances is wrong and an alternative approach needs to be developed.