• Newborns and minors
• withdrawal/withholding treatment
• best interests
• neonatology
• intensive care
• clinical ethics
• allowing minors to die
• donation/procurement of organs/tissues
• prolongation of life and euthanasia
• quality/value of life/personhood

When is death preferable to continued existence? This question all too often faces people, such as me, who work in intensive care. We have the technical ability to sustain organ function for long periods of time—sometimes, it seems, almost indefinitely. However, in the face of severe illness, particularly severe irremediable brain injury, doctors and family members sometimes wonder whether that is the right thing to do.

Several articles in this issue engage with this difficult question. Neurosurgeon Stephen Honeybul and colleagues (Editor's choice, see page 707) provide a thought-provoking insight into the practical dilemmas facing those working in neurosurgical intensive care. They ask whether it is possible to identify a subgroup of adult patients after head injury whose prognosis is sufficiently poor that they should not be treated. The authors apply an outcome prediction model to 5 years of data from patients with head injury in Perth. Patients in the highest risk category had a 58% chance of survival with surgery. However, 88% of survivors were either severely disabled or in a persistent vegetative state. Providing treatment for this group had substantial resource implications, with an average hospital length of stay of 120 days, and only six out of 43 able to be discharged home. Honeybul et al argue that neurosurgical rescue therapy for this group of patients may be inappropriate on utilitarian grounds (because of the enormous cost of treatment), but also because the most likely outcome would be unacceptable to many patients. However, they also point to the considerable personal and psychological pressure to ‘rescue’ patients.

Patients with severe head injury who die often develop signs of raised intracranial pressure, and may be diagnosed as being ‘brain dead’. A number of those in Honeybul's poor prognosis group may have fallen into this category. But are they actually dead? No, according to Seema Shah (see page 719) writing with Franklin Miller and Robert Truog. This paper continues Miller and Truog's previous criticism of the conceptual coherence of current definitions of death. They have long argued that it is ethically acceptable to remove the organs of patients with a diagnosis of ‘whole brain death’ or ‘total brain failure’, where those patients or their proxies have agreed to donation. However, this is not because of a metaphysical fact of the patient's death—rather, it is because the patient would be neither harmed nor wronged by organ retrieval. Previously they have taken their argument to suggest rejecting the so-called dead donor rule.1 In this paper, however, Shah, Miller and Truog take a different tack. They present a pragmatic argument: it is only a matter of time before the conceptual problems in current definitions of death reach the public arena; the best course of action may be to acknowledge that current definitions of death are justified ‘legal fictions’. This would enable donations to continue, while giving the general public time to come to terms with the value judgements inherent in death determination.

Yes, these patients are dead, argues John Lizza (see page 743). John Lizza's paper is a reply to Franklin Miller and Robert Truog's previous paper in JME ‘Decapitation and the definition of death.’2 Lizza uses decapitation thought experiments to generate a reductio ad absurdum of Miller and Truog's cardiorespiratory definition of death. Lizza draws on extraordinary experiments involving head transplantation of monkeys (the transplanted monkey heads exhibited some evidence of consciousness). He imagines that a human (Waldo) underwent such a procedure. Should we identify Waldo with the transplanted head, or with his artificially sustained headless former body? Lizza argues in favour of neurological criteria for determining death, as this provides a plausible answer to the thought experiment, but is also consistent with the beliefs and values most of us share about human beings, and about the treatment of patients with ‘total brain failure’.

In intensive care, death is an option because life-sustaining treatment can be withdrawn. Could death be preferable to continued life for patients who are not dependent on life support in intensive care? Can suicide be rational? One reason for opposing liberalisation of laws relating to assisted suicide is the belief that suicidal behaviour invariably reflects mental illness. On page 723 Stephen Ginn and colleagues present data from a postal survey of UK general practitioners and specialists on their views about suicide. Seventy per cent of doctors indicated that they believed that suicide could be rational (ie, occur in the absence of mental illness). However, a significant proportion of those who believed that suicide could be rational were opposed to laws allowing physician-assisted suicide, implying that doctors' opposition to assisted suicide must be on the basis of other concerns.

What could lead a rational patient to prefer death to continued life? On page 727 Marianne Dees and colleagues from the Netherlands provide a rich picture of patient perspectives on this question. The authors conducted in-depth qualitative interviews with 31 patients who had requested euthanasia or assisted suicide. Eleven of the 31 patients did not have their requests granted, mostly because of the absence of a physical diagnosis. The authors explored patients' experience and understanding of ‘unbearable suffering’, a prerequisite for legal euthanasia or assisted suicide in the Netherlands. Patients described a wide range of different types and sources of suffering. Interestingly, continuous severe suffering was described particularly by patients with psychiatric illness (whose requests for euthanasia were all denied). Also, of relevance for debates about assisted suicide, although some patients described loneliness or being a burden on others as contributing to their sense of unbearable suffering, these factors were relatively minor in comparison with emotional and existential causes of distress. Physical suffering was less important in patient accounts than a pervasive sense of hopelessness.