Special Reports

IN 1996, Lin Russell and her daughter Megan were bludgeoned to death near their home in the English countryside. Michael Stone, a man with a history of mental problems, was subsequently convicted of murdering them. The case caused a public outcry after newspapers reported that Stone had asked unsuccessfully for psychiatric help before the murders.

The British government now hopes to protect society by locking up hundreds of people judged to have a dangerously severe personality disorder (DSPD)—even though some of them may never have committed a violent crime. "We need to create new powers to allow for the detention of this small group of exceptionally dangerous people," said Home Secretary Jack Straw last week.

Mentally ill people can already be detained under the Mental Health Act, but only if doctors believe they can be treated. The government now intends to close this loophole and lock up psychopaths, even if they cannot be treated.

Dubious diagnosis

Frank Dobson, the Secretary of State for Health, says the plan simply extends the existing Mental Health Act. "We don't think there's a precedent here," he says. But Nottinghamshire solicitor Paul Bacon, who was a member of an earlier Home Office and Department of Health working group looking at psychopathic disorders, disagrees. "It's a fundamental change from what we've known," he says.

But can you diagnose if someone is suffering DSPD? Psychiatrists say that up to 10 per cent of the urban population in the Western world has a personality disorder of some kind, but only a fraction of these are classed as having a "severe" disorder. And only a tiny proportion of these are considered dangerous.

The British mental health charity MIND says that "personality disorder" is a highly controversial label, which has been misused to categorise people that society finds difficult or different. "The definition of personality disorders is very much open to debate," adds Graham Fanti of the British Association of Social Workers.

Three different recognised medical diagnoses come close to the DSPD label: antisocial personality disorder, dissocial personality disorder and psychopathy. And there is some research linking these conditions to dangerous behaviour. However, Ronald Blackburn, professor of clinical and forensic psychology at the University of Liverpool, says that categories such as these are not clinically useful. And DSPD is a new medical-legal term, so no one knows how these results may apply.

Reliable diagnosis is even more difficult when the severity of the disorder also has to be judged. Jeremy Coid, a professor of forensic psychiatry at St Bartholomew's and the Royal London School of Medicine, says that the definition of severity is arbitrary. "The inherent dilemma is where you draw the cut-off line," he says. "Where does abnormality begin and end?"

When Dobson and Straw announced the government's plans last week, Coid summarised the present state of research on forecasting whether a prisoner would reoffend. In these studies, researchers gave prisoners psychiatric tests before they were released and predicted whether they would offend again. The ex-prisoners were then tracked to see if they went on to commit further crimes.

The results of the studies varied greatly. Some studies were able to predict who would offend again with 100 per cent accuracy. But others could only pinpoint 36 per cent of the future reoffenders. "The tests may work better for some people than others," says Coid. On average, Coid found that these studies predicted who was going to be violent 72 per cent of the time. "That means that you'd have to lock up ten people to get it right seven times," says Coid. "It also means you'd lock up three people who didn't need to be."

Coid's research also revealed that more than 40 per cent of prisoners who were not thought to be dangerous went on to commit violent crimes. "So you'd be releasing four dangerous people out of ten," says Coid. But this research was largely carried out in North America and the techniques may not be effective in Britain. "There has not been any significant research on prediction in this country," says Coid.

The government is planning research to improve the forecasts. But Coid, pointing to the complex interaction between individuals and their environment, says that the forecasts will never be infallible. "These results are pretty good," he says. "Everything has a margin of error."

In practice, decisions about whether to lock someone up will be made by assessment teams with members from a number of different professions, including probation officers and psychiatrists. Anyone who is judged to be a risk to society may be held indefinitely, although there will be a system of reviewing cases regularly.

Most experts are not optimistic about predicting how dangerous people will be. "Risk assessment is a risky business," says Gerard Bailes, a consultant forensic psychologist at the Norvic Clinic, a specialist psychiatric centre in Norwich. "The problem is that clinical risk assessment has been pretty well discredited," says Blackburn. "People who try to predict future violence on clinical judgment invariably produce large numbers of errors."

Celia Taylor, a forensic clinical psychiatrist and fellow at the Institute of Criminology, Cambridge, strongly agrees. She estimates that as many as three out of four detainees would not commit another violent crime if they remained at large. "You'd have to shut up a lot of people who would not commit violence in the future to catch those who would," she says. "And you'd have to shut them up for years."

The government, however, remains confident that it can reliably identify the estimated 300 to 600 people with DSPD now in the community (see Diagram). The government plans to set up about 50 separate centres, each holding up to a dozen people, to detain them. Dobson says that they will provide the highest standards of care in these centres. The government estimates that doubling the time that this target group is locked up would prevent nearly 200 violent crimes every year.

But the proposals raise serious civil rights questions. Not only do they single out one group of people on the basis of who they are rather than what they've done, but they impose an extremely severe penalty: indefinite loss of liberty. Some of the 600 people being threatened by the government may never have committed a violent crime—they have simply, "come to the notice of the authorities".

"Here we've got a proposal that you detain people when they've not even been accused of a wrongful act," says Bacon. "It's discrimination of the worst possible kind." Bacon is also concerned that people in prison or psychiatric hospital will find it difficult to prove that they are no longer dangerous. "How can you prove you'd not be dangerous in the community," he asks, "when you're not in the community?"

Bacon's concerns are shared by John Wadham, the director of the civil rights campaign Liberty. "People should only be detained on the basis of what they have done, rather than what some experts think they might do in the future," he says. "That goes to the heart of what the criminal justice system is about."