When a 39-year-old Belgian woman suffered a stroke and fell into a coma, doctors concluded that she was unlikely to regain consciousness and, after a time, diagnosed her condition as persistent vegetative state (PVS). One of the criteria on which they based their decision was her inability to blink or track a moving object with her eyes. It was only when they discovered that the stroke had damaged a cranial nerve, preventing her from opening her eyes, that they realised their error. If they opened her eyes for her, she followed their instructions. Having regained full consciousness soon after her stroke, she revealed she had overheard all the bedside discussions as to whether it was worth keeping her alive. At no point had she wanted to die.

Others might not be so lucky. Research suggests that many patients left to die after being diagnosed as in PVS might have eventually recovered. They may have been in a twilight condition called minimally conscious state (MCS), which has until now proved difficult to identify. In recent years, MCS has muddied the waters further on what it means to be alive, and confused the debate over when is the right time to pull the plug. Now, thanks to a new test, doctors may finally be able to save those who still have a chance.

Until two years ago, patients who remained unconscious for a protracted length of time were diagnosed as PVS. They were regarded as "hopelessly damaged and permanently unconscious", and a court could consider petitions to withdraw their food and water, allowing them to starve to death and their organs to be removed for transplantation. Then, in 2002, MCS was introduced to describe a new condition where patients might smile when a certain friend entered the room. Or they might spontaneously utter a swear word, or obey a simple command.

Those fleeting glimpses of sentience meant that the medical community now viewed them differently: these patients could conceivably recover. Although they might stick at this point on the spectrum, or slide back into PVS, they might also regain a more consistent level of consciousness that would allow them to communicate. Reports of people "emerging" from vegetative state are, by contrast, extremely rare - in medical circles, PVS and futility go hand-in-hand.

So at a stroke, hundreds of thousands of patients were rescued from the threat of euthanasia and removed from the pool of potential organ donors. Terry Schiavo, a 39-year-old Florida woman at the centre of a right-to-die battle is a case in point.

Unfortunately for both doctors and patients, MCS is no easier to diagnose than PVS. But a paper published recently in the Archives of Neurology could change all that.

Stephen Laureys of the University of Liège, Belgium, and colleagues scanned the brains of patients diagnosed as MCS and PVS, as well as healthy controls, using positron emission tomography (Pet), and detected consistent differences in their responses to loud clicks. The perception of sound in healthy people involves the coordinated communication of many brain areas, and whereas they saw similar, widespread patterns of activation in the MCS and control groups, those patterns were absent in the PVS patients. To put it bluntly, the disintegration of their brains appeared to be much more advanced.

Laureys' findings merely reinforced the diagnoses the patients had already received, based on standard clinical criteria including the eyeblink test. But, he says: "We clearly showed differences. I think this is the first step [to an objective test]."

If doctors had such a test at their fingertips, it would reduce the risk of misdiagnosis. It would also, he believes, provide evidence to support their decision that a patient might, in the case of MCS, benefit from therapy.

Other researchers around the world are pursuing the same goal, but the Belgian group is the first to publish its conclusions in a peer-reviewed journal. The findings could have profound implications for many thousands of patients.

Although the treatment that restores consciousness doesn't exist, there are intriguing clues that it could, one day.

Deep brain stimulation is a popular and effective surgical treatment for Parkinson's disease and epilepsy among other disorders, which involves feeding current to electrodes implanted deep in the brain. In the early 1990s, Japanese researchers showed that it could also provoke a strong electrical response in the brains of PVS patients. The patients didn't show any outward improvement, probably due to the degree of disintegration of their brains. But Laureys believes that his findings suggest that such therapies could help MCS patients, even pushing them over the line that divides them from communication.

That prospect is some way off and for now the question remains: what use is a test that saves a person's life, if that life is unlikely ever to involve meaningful interaction? "Even if we demonstrated there was function in the brain, but in spite of everything we could do the patient was still not able to indicate to us any evidence of thinking, where does that place us?" says Keith Andrews of the Royal Hospital for Neurodisability in London, author of a 1996 study which showed that 40% of the PVS patients in his hospital had been misdiagnosed, mainly due to blindness or visual impairment. Yet he defends the need for the new definition, and indeed was a member of the working party that ratified it. "Without terminology you can't carry out research," he says. "Only when it's got a name can you begin to collect data."

Others argue that the MCS label is so broad as to be meaningless.

"The upper limit of it, the border with severe disability, is intrinsically arbitrary," says Alan Shewmon, a paediatric neurologist at the University of California, Los Angeles. "The term creates the impression of a diagnostic entity that is different not only from PVS but also from severe disability in terms of ethical implications, yet there is no such distinction in reality, only a continuous spectrum."

There are also fears that PVS patients re-categorised as MCS might be worse off than before, because MCS describes a twilight zone between consciousness and unconsciousness and doctors don't know how to treat it. In the case of PVS, they can suggest withdrawing a feeding tube, but can't go ahead without a court declaration. According to Anthony Cole, a spokesman for the British pro-life Medical Ethics Alliance, no such declaration is needed when it comes to MCS. There are only the guidelines laid down by the British Medical Association, which he says are controversial anyway.

Andrews says it's probably only a matter of time before the issue of removing a feeding tube from an MCS patient comes before a court. When it does, it will be watched closely by the same interested parties who watched the first such case involving a PVS patient, Karen Quinlan.

On one side will be those who argue that patients with no better quality of life than their PVS counterparts are being denied the right to die; on the other, disability rights groups who feel that doctors want to kill patients who are still alive and have the potential to recover.

The decision is rarely straightforward. Eight years ago, Dominique Toussaint suffered a basilar artery thrombosis which left him paralysed, except for the ability to move his eyes up and down, and still fully conscious. Before long he learned to operate a specially adapted computer using just those vertical eye movements, and began to bombard media outlets, powerful organisations, leading political figures - anyone he could think of - with letters in which he pleaded to be allowed a dignified death.

His demands fell on deaf ears (though the French President sent him a pretty card and his best wishes) and, in desperation, he refused all food and treatment - "to make people realise that I still existed". Paralysed, he couldn't perform the lethal acts for himself, and the ultimate humiliation came when those caring for him simply ignored his request and forced him to live.

The experience turned him into, as he puts it, "a true militant for human rights". Eventually he came under the care of a new medical team, who unanimously agreed to stop giving him the two daily insulin injections that controlled his diabetes - a step both he and they knew would be lethal - if it was his genuine wish.

Taken aback by his change in fortune Toussaint asked his children, 11 and 13, what they expected of him. They replied that they wanted their father, invalid or no invalid, and he promised them he would live - a promise he intends to keep, at least until they have finished their education.

"Study after study indicates that medical professionals have a much lower opinion of quality of life of people with disabilities than the general population, including people with disabilities themselves," says Stephen Drake, a research analyst at the American disability rights organisation Not Dead Yet. He argues that without a clear statement from the patient, an advance directive or something similar, the presumption should always be that they want to live.

Locked-in patients are intellectually intact and able to express themselves. Journalist Jean-Dominique Bauby, for example, wrote his memoir, The Diving Bell and the Butterfly, when he could only communicate by blinking one eye. They are far removed from those in the MCS, but according to Laureys provide the evidence that life should be the default option.

When surveyed, only four of the 300 members of the French Association of Locked-In Syndrome (ALIS) said they would opt for euthanasia. The rest, like the Belgian woman who was misdiagnosed, desperately wanted to live.

So Toussaint, it seems, is in the minority. But then all he asks is that his personhood be recognised, along with his right to choose. The most striking thing he says is this: "Life seems more bearable now that I have agreed to it freely."

How Karen won the right to die

In 1975, 21-year-old Karen Quinlan was admitted to a New Jersey hospital in what was described as a "coma of mysterious origin", probably as a result of taking a combination of barbiturates, benzodiazepines and alcohol. The coma persisted, and after doctors refused Karen's parents' request to remove her from the ventilator, her parents started legal proceedings and won. The court decided that Karen's right to privacy now included her right to be removed from the machine.

The Quinlan case marked the first of a series of such decisions concerning patients whose "permanent unconsciousness" meant that their continued treatment was considered futile. Some years earlier, in 1968, Harvard bioethicist HK Beecher had argued that death should be linked to the permanent loss of consciousness so as to permit the removal of organs while they were still usable. Without such a link, he wrote, "The curable, the salvageable, can thus be sacrificed to the hopelessly damaged and unconscious who consume the time and space and money better devoted to those who could be helped."

With the introduction of the minimally conscious state (MCS), some fear that doctors are now asking society for permission to intervene to save patients whose neurological state barely differs from those in whom the right to die was established - that is, PVS patients.

Further reading

Brain Death and Disorders of Consciousness: Proceedings of the Fourth International Symposium on Coma and Death, 2004
New York: Kluwer Academic/Plenum Publishers (in press)

The minimally conscious state: definition and diagnostic criteria
by Joseph Giacino et al, Neurology, 2002: 58, 349-353

Ethical problems created by the hopelessly unconscious patient
by HK Beecher, New England Journal of Medicine, 1968: 278, 1,425-1,430