1. Quality of life

Charlie was severely ill, on life support in intensive care. His mitochondrial problem caused severe muscle weakness and he was completely paralysed. He could not open his eyes, and was deaf. There was also evidence that the disorder was affecting his brain: an electrical recording of his brain showed very abnormal disordered electrical patterns with frequent subclinical seizures, a pattern referred to as a “severe epileptic encephalopathy”. It was very difficult to know whether Charlie was aware of anything around him. However, other children with similar electrical patterns of brain activity, (at least where that pattern is persistent), have minimal or no awareness. In that situation, life support machines were keeping Charlie alive, but were arguably not benefiting him in any meaningful way.

Of course, ‘quality of life’ assessments are fraught and difficult. There can be a range of different views about quality of life. Julian has highlighted the philosophical complexity of assessing wellbeing as well as the controversial nature of a judgement that someone else’s life is “not worth living”. There are very good reasons to be extremely cautious about making such judgments and to pay particular attention to the views of parents and to those of people with disabilities. However, taking great care about such assessments does not mean that they should never be made, nor that any view of parents about their child’s quality of life should be accepted uncritically. Imagine, for example, that parents of a child with Down syndrome decided that they did not wish their child to have surgery for a treatable abnormality and would prefer that the child die. They might point to the lack of philosophical agreement on theories of wellbeing, or even to the views of some philosophers that a life with moderate intellectual disability is not a life worth living. Yet, it is clear to almost all readers, I suspect, that it would be wrong to go along with these parents’ views about quality of life. As has been pointed out several times in this book, pluralism is not the same thing as relativism. While there may not be a single ‘right answer’ to the question of how we should think about wellbeing, or what would be a life worth living – there can be wrong answers.

In 2013, in my book “Death or disability? The Carmentis machine and decision making for critically ill children”, I focussed specifically on the question of quality of life for children with severe brain problems.¹ I looked at when, if ever, a child’s brain problem would be so severe that intensive care should not be continued or provided, even if parents wished it to be. In that book, four years before the Gard case, I argued that the combination of very severe cognitive, movement and sensory problems along with ongoing painful illness or medical interventions would be the clearest case where treatment would not be in a child’s interests.² Why think this? Drawing on the metaphor that we described in chapter 3 of a set of scales, this situation seems to be one where a child’s condition removes all, or almost all of the benefits of life for the child.

This seems clear, whichever philosophical theory of wellbeing we adopt. On a hedonic theory (life goes well if we experience more pleasure than pain), it did not appear to his doctors that Charlie was able to experience any pleasure at all. He could perhaps be aware of his parents’ presence through touch or smell, but he could not hear their voice, and could not open his eyes to see them. If we adopt a desire based theory, it seems that Charlie was unable to develop any desires or preferences – because of his severe brain dysfunction. There are different lists of objective goods, however, many of the things that are present across different lists are closely related to learning, awareness and cognitive function. For example, such lists often include communication, knowledge, aesthetic appreciation, the development of deep reciprocal relationships, the achievement of goals or ambitions. Yet, none of these appeared to be available to Charlie.

Here the combination of different forms of severe disability is highly relevant. Someone with blindness and deafness, but who has normal learning and movement, can learn to appreciate the world through their other senses. There can still be enormous benefits and value of their life (for them)– as famously evident in the life of Helen Keller. Alternatively, a child with severe learning disability may still be able to develop deep and valuable relationships with those around them, they may be able to communicate non-verbally, and they may experience a range of sensory pleasures, as well as pleasure in learning and gaining new skills. There is also important evidence from those with severe physical impairment – even, profound forms such as quadriplegia or locked-in syndrome. Such individuals often report high levels of personal wellbeing, despite their physical limitations.³ These are all examples of humans’ capacity to adapt to illness or impairment and to find value in life despite adversity. However, the combination of severe or profound impairments of different types makes it very difficult and perhaps impossible to compensate or adapt. That is why I have argued that in this situation doctors would potentially be justified in declining to keep a child alive.

Charlie’s condition, by the time that nucleoside therapy was being contemplated, had left him in a state of profound weakness, with some evidence of very abnormal brain function and sensory impairment. His condition, then, was much more severe than some of the cases discussed earlier in this book. Jonathan Simms, the adolescent with Creutzfeldt Jakob disease discussed in Chapter 5, was able to recognise and respond to his family, listen to music, and enjoy watching games of soccer on the television.⁴ Infant MB, with profound muscle weakness from spinal muscular atrophy (chapter 3) was thought to be able to appreciate and gain pleasure from stories that were read to him, or watching DVDs.⁵

Importantly, Charlie’s parents in the court hearing accepted that Charlie’s current quality of life was not good – they did not wish to prolong his life if it could not improve.

2. Suffering

I have argued that in Charlie’s case, there was little benefit to him in keeping him alive. But what of the other side of scales? If there is no reason to think that a child is in pain or discomfort, it is hard to see what would be wrong in prolonging their life with medical treatment (setting aside any issue of limited resources).

The medical professionals who cared for Charlie were concerned that he was suffering, and would be harmed by continuing life support. (As far as was known, the nucleoside treatment had few side effects). That seemed in large part because they were aware of the unpleasantness of keeping children or adults alive in intensive care. Despite the best efforts of doctors and nurses to minimise invasive procedures and tests, and use drugs and other measures for symptoms, being on life support in intensive care is often painful and uncomfortable. However, as noted by Julian, and other commentators, there are plenty of children who receive treatment, sometimes very prolonged treatment in intensive care. Extremely premature infants, sometimes with relatively low chances of survival, may be ventilated and have multiple invasive procedures over a period of 3 months or more. Ethicists and doctors do not normally seem to be worried about the suffering inflicted on children by being in intensive care.

But the situation of most children in intensive care is different from the situation that faced Charlie. Even if unpleasant, the pain and discomfort of life support is worth enduring if a child is going to get better, and ultimately be able to be discharged home. For example, in a recent cohort, 36% of extremely premature infants born at 22-24 weeks gestation survived to at least age 18 months, (and more than 50% of survivors had no neurosensory disability).⁶ But where there isn’t any real chance that the child will improve, the patient has to endure the side effects of intensive care without any benefits.

In Charlie’s case, there was some disagreement about whether he was suffering. His parents spent hours at his bedside and felt that they knew him well. They did not believe that he was in pain. It is crucially important for doctors to listen to parents, and not to dismiss their views. They will usually know a child, and what he or she likes or doesn’t like, experiences or doesn’t experience, better than anyone. Yet, one of the sad features of Charlie’s illness was that it rendered him almost completely paralysed. He could not indicate his needs, he could not even cry if he were uncomfortable. The nurses and doctors looking after Charlie would have been aware that the other children in the intensive care unit on life support show signs of pain and discomfort at least some of the time. They would have suspected that Charlie would have similar periods – though because of his paralysis, they may not know for sure when.

Charlie’s situation, then, was different from that of the brain-dead children discussed in chapter 2 (infant A and Jahi McMath). In those cases, there is very good reason to think that they had or have zero capacity for suffering. While there may be no benefit to them of continuing somatic support, there is equally no harm.

For Charlie, there were two apparent possibilities.

One possibility was that Charlie’s brain was so severely affected that he did not even experience pain. Perhaps he was genuinely in something akin to a persistent vegetative state or brain death. (He did not apparently fulfil criteria for brain death). If that were the case, it might not be harmful to Charlie to continue intensive care and to receive nucleoside therapy. At the same time, if his brain were that severely affected by MDDS that he experienced no pain at all – it would be even more improbable that nucleoside treatment would help.

However, another very real possibility was that Charlie was at least intermittently aware of his surroundings, perhaps in something like a “minimally conscious state”. If that were the case, there was a real worry that some of the time he would be distressed, agitated or unhappy but paralysed. Neuroimaging studies suggest that severely brain injured adults in a minimally conscious state are able to perceive pain.⁷ With no easy way to detect pain in Charlie, it would be hard to console him or treat his pain.

3. Uncertainty

The court found that nucleoside treatment for Charlie would be futile. However, as Julian persuasively argues, this appears to be a mistake. There is a key difference between there being no evidence that something will work, and there being clear evidence that it will not work. Brain scans did not show, indeed could not show “irreversible” damage to his brain. The question of whether brain damage is reversible is a prediction about future brain function and whether it will improve. It can’t simply be read from a scan. It depends on experience from past patients who have received similar treatment. But of course, there weren’t any of those.

It was not possible to be certain that nucleoside therapy would not lead to any improvement in Charlie. It would not have been “futile”. On the strictest definitions of futility, there are very few situations in medicine where it is possible to be certain that treatment absolutely cannot work.

However, everyone involved in assessing Charlie (including Professor Hirano) agreed that it was very unlikely that nucleoside therapy would lead to any significant improvement. Why did they believe this?

All previous children with Charlie’s severe condition had died. Charlie had progressively and inexorably deteriorated over time due to MDDS. By January 2017, there was apparent evidence of severe epileptic encephalopathy. Nucleoside treatment for Charlie’s condition had never been tried in animal models, never been tried in other children, and, importantly, had (in other forms of MDDS) never been reported to improve brain problems. Even in children with the milder muscle-only form of MDDS, nucleoside treatment wasn’t a cure, it wasn’t a magic bullet. It led to some improvement in muscle strength, but most children remained dependent on breathing support on ventilators. It seemed extremely unlikely that in Charlie’s much more severe situation, with established severe brain problems, that the treatment would make any difference at all.

But so what? Even if the treatment was unlikely to help Charlie, the alternative, as bluntly noted by Professor Hirano, was death. Julian points out that faced with certain death many adults might be willing to take the gamble of several months of intensive care and an experimental treatment, even if the odds of benefit were tiny. He contends that this would be a reasonable choice by an adult and potentially in their interests. Accordingly, we should think it a reasonable option for Charlie Gard, and potentially in his interests.

In Chapter 3, we discussed some of the challenge of assessing best interests in the face of uncertainty. People have different attitudes to risks, and place different weights on the best or worst-case scenario. Whether or not it is worth undergoing an intervention that has only a small chance of benefit depends a lot on how unpleasant or risky it would be (which would take us back to the suffering question, discussed above). It may also depend on someone’s perspective. From a parent’s point of view, they are focussed on the child in front of them, and face the devastating prospect of losing their child. The statistics are, in a sense, meaningless to them. Their child will die, or will recover. It is completely understandable that they would seize any chance of avoiding that terrible loss.

Yet from the professionals’ point of view, and from the point of view of society, it is important to step back from an individual child to think about what the implications would be of allowing parents to subject their children to very unpleasant treatments that have only a very small chance of helping them. If doctors or judges allow those decisions, that will mean that many children will suffer the side effects and harms of treatment for only a very few children to see any benefit. As noted in Chapter 3, if the chance of Charlie benefiting from nucleoside therapy were 1 in 10,000, that would mean that similar treatment choices would harm 9999 children for every child that they help.

This argument has some overlap with other so-called ‘collective action’ problems (for example, climate change, voting or littering). While the impact of any individual is small, the total harm will be substantial if everybody takes that action. For example, you may be wondering whether to go to the trouble of visiting the polling booth on election day. Perhaps you feel that it is important which party is elected, but that it does not matter if you decide not to vote – the chance of your individual vote making the difference between one party winning or losing in an election is remote. However, if everyone reasoned the same way, that would make a very large difference.

A particular vivid hypothetical example, described by philosopher Derek Parfit, is a thought experiment he called the “Harmless torturers”. Parfit imagined that a torture machine that administered a large and painful electrical charge was replaced by many machines operated by 1000 different torturers who each administered an imperceptible charge (1/1000^th of the previous amount).⁸ None of the individual torturers was harming anyone perceptibly, but together they predictably cause terrible pain. We shouldn’t allow them to administer their tiny shocks, because we know what will happen if they do. Low probability treatments in intensive care might seem very similar to “harmless torture”.

Of course, the decision faced by parents of a dying child is not the same. They do not wish to harm their child. Nor do their choices combine or interact with others in the way that votes combine, or the mini-shocks add up. But it is similar in the sense that a decision that might be understandable and rational from the point of view of any individual, becomes unwise and obviously harmful when seen as a general rule or principle.

This is not a mere theoretical point about statistics and abstract philosophical theories. In the intensive care unit where I work, every week there are some infants whose outcome is uncertain, but who I and my colleagues judge are unlikely to survive. Reflecting back on similar infants over the past year, many of the infants in that category did die, often after many weeks or even months of invasive and unpleasant treatment. However, some of the infants also survived.⁹ In the face of uncertainty, those parents asked us to continue intensive care, and because of the degree of uncertainty we provided that treatment. But sometimes the chance of benefit seems even smaller and the risks overwhelming. Several years ago, I cared for an infant who had been in intensive care on a ventilator for 6 months. She had deteriorated despite maximal treatment, and sadly we had exhausted all ideas about further treatments to help. Specialists from a number of different departments reviewed her, and none of them, nor any of my colleagues believed that she would survive. Her parents, though, would not agree to withdraw treatment. The infant lingered for weeks more in the intensive care unit. Despite our best efforts to provide pain relief, she had become tolerant to many of our medications – she often appeared uncomfortable or in pain. Eventually, as predicted, she deteriorated further, and died, after more than 8 months in the intensive care unit.

If it were justified for doctors to refuse to provide treatments that have a real prospect of harming children and only a low chance of benefit – that obviously raises difficult questions about how small a chance of benefit would count. There is no straightforward answer to that question. Should it be 10% or 1% or lower? It will depend on how harmful the treatment is (again circling back to the question of suffering above). In my book ‘Death or Disability’ I adopted the legal standard of clear and convincing evidence.¹⁰ In the courtroom, as in the NICU, it is rarely possible to be 100% certain – but decisions still have to be made. If there is clear and convincing evidence that treatment would harm an infant or child, it should not be provided, even if parents request it. My view in the Gard case, was there appeared to be such evidence, and it was both clear, and to those who reviewed it in detail, convincing.

4. Reasonable disagreement

The recurring theme throughout this book has been that of reasonable disagreement, and the importance of acknowledging and respecting differing viewpoints about difficult and contested questions of value. Where there is reasonable disagreement, (and where there are not questions of limited resources) we have argued that parents’ wishes should be respected.

In the case of Charlie Gard, there were such differing points of view. There were some professionals in Italy and in the US who supported the option of nucleoside treatment and were prepared to provide the treatment if Charlie were transferred to their centre. There were also some ethicists who defended that option and felt that the court decision was flawed. Julian has articulated a strong case in favour of allowing Charlie to undergo such treatment.

However, some of the conditions of reasonableness that we have set out in this book did not seem to be met at the time that Charlie’s case was being actively debated in the courts. It wasn’t clear whether Professor Hirano had all of the necessary information about Charlie’s current condition to reach an informed decision about what would be best for him. In the High Court hearing, he admitted that it was difficult for him to assess Charlie from a distance

“Perhaps, if I were there, I would support [the Great Ormond St plan to withdraw treatment]. Not seeing the child, not seeing progression, it’s difficult for me to make an assessment.” Para 99¹¹

While Professor Hirano was able to provide the nucleoside treatment, he noted that he would defer a decision about provision of life support to “ICU people”. There were no submissions, however, from intensive care specialists in the US. All of the intensive care experts from the UK (and all the other medical specialists who had reviewed him in person) had concluded that continuing treatment and nucleoside therapy was not in his interests.

It was also unclear whether Professor Hirano or, at a later stage, the other specialists (for example from the Vatican hospital in Rome) who provided support for the requested treatment were motivated by values that are not relevant within the context of UK legal decision-making. For example, in Chapter 7 we argued that “reasonable dissensus” would not be present if professionals defended treatment on the basis of a belief that parents should always be the ultimate decision-makers about treatment or on the basis of a strong sanctity of life view.

This is not to cast aspersions on the views of these experts. It is possible that they knew all of the relevant facts and were genuinely motivated by concern for Charlie’s interests. However, given that they were offering views that were in opposition to all of the medical team caring for Charlie, as well as all of the external experts who had examined him – it would have been vital to present evidence to substantiate the reasonableness of their views. It would have been important to be clear about the values underlying their opinions. That evidence was never, to my knowledge, presented.