Some jobs nobody wants.

Telling a parent that their child is dying has to be up there with the worst jobs in the world, the flip side perhaps to telling them that you can save their child’s life. Telling parents that their very sick child cannot be helped, and worse still, you believe that keeping them alive is pointless and possibly damaging, telling them you’d like to switch off the machines keeping their child alive, must be amongst the most soul-destroying jobs in the world.

Recent newspapers have been full of the difficult case of the British boy Charlie Gard, the latest in a series of court cases in the UK when parents and doctors have disagreed about medical treatment for a child. Charlie Gard is a 9-month-old boy with the rare neurodegenerative disorder severe encephalomyopathic mitochondrial DNA depletion syndrome.

He is dependent on life support and has been in intensive care at Great Ormond Street Hospital for Children in London, UK, since October, 2016. In such disputes, typically, doctors regard life support treatment as “futile” or “potentially inappropriate”.

Parents, by contrast, want treatment to continue. In the current case, High Court Judge Mr Justice Francis has recently rejected the request of Charlie Gard’s parents for him to travel to the USA for an experimental medical treatment, nucleoside therapy. On April 11, 2017, Justice Francis ruled that it would be lawful and in Charlie Gard’s best interests to withdraw artificial ventilation and provide palliative care. Charlie Gard’s parents have appealed this ruling.

When doctors and the courts consider cases like this one, they often focus exclusively on the best interests of the patient. In some cases, however, it is uncertain whether or not treatment would be in the interests of the patient. Indeed, there could be stronger and clearer arguments to limit treatment on the basis of finite and scarce medical resources. Although it feels brutal, keeping this little boy alive inevitably means less money to spend on keeping other children alive.

The different ethical reasons that justify a decision not to provide treatment might come together, or they might come apart. If treatment would be both affordable and in the child’s interests, it should unquestionably be provided. If it is neither affordable, nor in the child’s interests, treatment should not be started or should be stopped.

Where there is uncertainty about the benefits and costs of treatment, parents’ views are crucial. But sometimes the picture is more mixed. Perhaps treatment is in the interests of the patient, but unaffordable within a public health system. In the case of Charlie Gard, the parents have raised money over the internet through crowdsourcing to enable him to be taken to the USA for medical treatment. That would mean that the resource issue is not relevant. Perhaps for him treatment would be affordable, but contrary to Charlie Gard’s best interests?

One way of thinking about what would be in someone’s best interests is to imagine a set of scales. On the right side of the scales are the reasons in favour of a course of action, on the left are the reasons against. If it were a question of weighing a small chance of a positive outcome against an empty scale, the balance would be tipped in favour of treatment, even if the chance (or magnitude) of benefit were tiny.

But there are often substantial negatives in the balance.

Although health professionals do their best to provide pain relief, sedation, care, and comfort to severely ill children and babies, that ability is finite and imperfect. Children on long-term ventilation often seem uncomfortable at least part of the time, they endure needles and invasive procedures, and might be distressed and unable to communicate the source of their distress.

It is possible to argue that the small chance (perhaps one in 10 000) of benefit would outweigh the negatives of treatment in intensive care. However, a shift in perspective casts that argument into doubt. Charlie Gard’s condition is extremely rare, but imagine that there were a sudden epidemic of mitochondrial DNA depletion syndrome affecting thousands of newborn babies. Would it be ethical to artificially ventilate for months thousands of infants to achieve some measure of improvement in one infant? Setting aside any consideration of resources, it seems wrong to subject thousands of infants to invasive and unpleasant life-support treatment to benefit one child. That implies that this chance of recovery is too slim to make treatment plausibly in the current child’s interests.

The reason why these decisions come to the court at all is because parents do not have an absolute right to make medical decisions for their children. Parents are given broad discretion about how to raise their children, for example, how to feed them, how to educate them, and whether or not to immunise them. Parents will not always make the best choices, but for the most part the state will not interfere or intervene. However, where parents’ decisions run a substantial risk of causing serious harm to their child, their decisions must be challenged, if necessary in a court.

When it comes to experimental treatment, there can be different reasonable views among health professionals about how to weigh up the chance of benefit against the burdens of the treatment. In the face of such disagreement, the decision properly belongs to the parents. Assuming the treatment is affordable, and the parents want it, it should be provided. However, when no health professionals think that the experimental treatment is worth pursuing, parents’ request for treatment should not be granted.

In the case of Charlie Gard, just one expert in the USA was prepared to provide experimental treatment. However, the expert admitted that the treatment had never been tried in a child with established encephalopathy and that benefit was “unlikely”; in his ruling, the judge clearly thought that this possibility did not represent a reasonable treatment option.

Decisions about life-sustaining treatment for critically ill children are fraught and difficult for all involved. Parents are, rightly, at the heart of the decisions that are made in intensive care. Their views about treatment are important, and their wishes are usually followed. However, there are limits. Sadly, reluctantly, doctors and judges do sometimes conclude—and are justified in concluding—that slim chances of life are not always better than dying. Providing comfort, avoiding painful and unhelpful medical treatments, supporting the child and family for their remaining time: sometimes that is the best that medicine can do, and the only ethical course.

Pending appeal Charlie Gard continues to be kept alive.