Prompted by the increasing practice of organ transplantation, and thus the need to procure donor organs that are as fresh as possible, many countries have modelled their legal definition of death on a US law passed in 1981 after extensive debate and thoughtful input from a specially appointed president's commission of experts.

The law seems admirably straightforward: “An individual who has sustained either (1) irreversible cessation of circulatory and respiratory functions, or (2) irreversible cessation of all functions of the entire brain, including the brain stem, is dead.”

In practice, unfortunately, physicians know that when they declare that someone on life support is dead, they are usually obeying the spirit, but not the letter, of this law. And many are feeling increasingly uncomfortable about it.

In particular, they struggle with three of the law's phrases: 'irreversible', 'all functions' and 'entire brain', knowing that they cannot guarantee full compliance. They do know that when they declare a death — according to strict clinical criteria, the principles of which are outlined in the original report of the president's commission — that the person is to all intents and purposes dead. But what if, as is sometimes the case, blood chemistry suggests that the pituitary gland at the base of the brain is still functioning? That activity has nothing to do with a person being alive in any meaningful sense. But it undermines a claim that all functions of the entire brain have ceased. As do post-mortem observations that relatively large areas of tissue can be metabolically active in different brain areas at the time death is declared.

The criterion of irreversibility raises the question of how long one should wait to be sure that no function will re-emerge. Is the six hours recommended in the commission's report sufficient? Physicians who have been required by circumstance to wait much longer have occasionally observed a brainstem-mediated reflex — a cough, for example — up to 36 hours after they would have declared death.

The problem is that death is not a phase transition whereby a person stops being alive and becomes dead in an instant. It is a long process during which systems, networks and cells gradually disintegrate. At some point, the person is no longer there, and can never be made to return. But the kind of clear, unambiguous boundary assumed in the 1981 law simply does not exist.

Ideally, the law should be changed to describe more accurately and honestly the way that death is determined in clinical practice. Most doctors have hesitated to say so too loudly, lest they be caricatured in public as greedy harvesters eager to strip living patients of their organs. But their public silence was broken on 24 September at an international meeting that included physicians, transplant surgeons and bioethicists at the Italian Festival of Health in Viareggio. The meeting concluded that lawmakers in the United States and elsewhere should reconsider rigid definitions of death, and called for a wider public debate.

The time has come for a serious discussion on redrafting laws that push doctors towards a form of deceit. But care must be taken to ensure that it doesn't backfire. Learning that the law has not been strictly adhered to could easily discourage organ donation at a time when demand for organs already vastly exceeds supply. Physicians and others involved in the issue would be wise to investigate just how incendiary the theme might be, perhaps in contained focus groups, and design their strategy accordingly.

Few things are as sensitive as death. But concerns about the legal details of declaring death in someone who will never again be the person he or she was should be weighed against the value of giving a full and healthy life to someone who will die without a transplant.