Decisions about medical treatment for critically ill children open up a world of competing sorrows. The case of the 11-year-old Canadian boy with leukemia, whose father refused a second round of chemotherapy for him, is no exception. The plight of the boy and his father has sparked a nation-wide debate.

The right of parents to decide on medical treatment for their children flows from a presumption that they will act in their children's "best interests". Parents have an obligation to provide their children with the necessities of life, which include the medical treatment needed to protect their lives and health. For a variety of reasons, however, parents can refuse treatment for their children that medical professionals and others regard as essential.

The right to refuse treatment for one's child is narrower than if the parents were deciding for themselves -- competent adults have an almost absolute right to refuse, which is much broader than what they can refuse for their child.

It's been reported the young boy has also refused treatment. Whether that refusal is legally valid depends on his capacity to provide an informed refusal -- the flip side of an informed consent.

A valid informed consent or refusal requires the person to be competent and informed, and their decision voluntary. Competence is often in doubt with children's own treatment decisions. In judging competence, courts take into account the child's age, experience and intelligence; independence from their parents; and the reasons they give for refusing the treatment. Courts have increasingly found mature minors competent to decide about treatment, but that is highly unlikely with an 11-year-old. Ethically, however, the child's wishes should still be taken into account.

When, as in this case, there is an irresolvable conflict about treatment between parents and doctors -- unless the situation is an emergency -- a court order is required.

In deciding whether to take decision-making power away from the parents, courts look at the nature of the illness from which the child suffers and whether or not it is life-threatening, how likely treatment is to be effective, and the suffering the treatment involves.

Under child-welfare legislation, to displace his parents as decision-makers, the court must first find that he is neglected or abused and in need of protection. Failure to provide a child with necessary medical care, when this endangers the child's life or health, is child neglect within the statute of the province of Ontario.

Such a finding is devastating to loving, caring parents and adds to their distress and trauma -- they feel as though they are David facing Goliath. We have ethical obligations to try to reduce their suffering, and one possibility would be a special provision in child-welfare legislation to allow intervention in these cases without needing to find the parents guilty of neglect. As well, even when a court or child-welfare agency is ethically justified in interfering with the parents' decisions, it has obligations to do the least harm possible.

In the past, courts sometimes made the treatment decision themselves, usually by endorsing the doctors' recommendations, or they gave decision-making power to the doctors. More recently, they tend to name a person -- frequently a child-welfare authority or, in one case, the Saskatchewan Minister of Social Services, or sometimes another relative of the child - to replace the parents as the decision-makers.

Moreover, in the past, courts took away all decision-making power from parents.

Today, they tend to take away only the right to decide on medical treatment -- as shown in the Vancouver sextuplets case, in which blood transfusions were ordered despite the parents' refusal -- and only for as long as is necessary to give the child treatment.

That implements the ethical principle of using the least harmful, least restrictive and least intrusive means reasonably available to deal with situations in which we, as a society, have ethical and legal obligations to intervene to protect children, but, in doing so, harm the parents.

That principle also means that imposed treatment should be the least harmful, least restrictive and least intrusive it is feasible to use.

The other side of the coin is our ethical obligation not to be over-interventionist when this would inflict serious suffering that is not compensated for by the promise of adequate benefits. This can sometimes happen in treating children dying of cancer.

Some parents want all possible treatments, including experimental ones, for their dying child. I have had immensely distressed nurses contact me about such cases, describing the treatment as "torturing the child".

It is extremely difficult to accept that there is nothing that we can do to prevent a child dying. This can cause us to lose a proper perspective on the suffering/benefit ratio of the treatment given to the child -- or even of that we choose not to give. Research has shown that children dying of cancer receive inadequate treatment for pain or other symptoms of serious physical distress -- such as breathlessness or vomiting -- because of fear that the treatment could "harm" them.

These conflict-over-treatment cases are almost always very hard to call and good facts -- which, as the public, we don't have in this boy's case, at present -- are essential to good ethics and law. But in fulfilling our obligations as a society to protect this child, we should err on the side of giving him the best chance at life.