Law: How 'junk science' has been put on trial: Tim Cornwell in Washington reports on the controversy surrounding so-called 'expert' evidence which, critics say, is damaging to the American legal system

'EXPERTS reveal common drug causing deformed babies,' was how the National Enquirer, most infamous of the American supermarket tabloids, told it in 1979. 'Thousands of babies are born with hideous birth defects,' the story revealed, caused by a common pregnancy drug - not thalidomide, but Bendectin.

It was a formidable scoop. Four years later Bendectin (marketed in the UK as Debendox) was pulled off the market under a flood of lawsuits. It had been prescribed to 33 million American women since 1957 - and was soon blamed by the parents of hundreds of children with warped or missing limbs.

But of about 30 published studies, involving over 130,000 patients, not one found a statistical link between Bendectin and birth defects. All but a handful of personal injury claims have been thrown out by the courts. The overwhelming weight of established medical opinion is that Bendectin was safe.

'Junk science' is blamed by critics of the US legal system for removing a drug, of proven effectiveness in combating morning sickness, from US pharmacy shelves. It shows how so-called experts can fuddle juries and judges, and destroy a successful American product, they say. This summer, more than a decade after claims against Bendectin first surfaced, the Supreme Court will pass judgment on two of the remaining cases. The parents of Jason Daubert, 19, born with three fingers missing, and Eric Schuller, now 12, born without a left hand, launched a law suit against Merrell Dow Pharmaceuticals Inc in 1984.

The families are seeking to overturn judgments in lower courts that barred their cases even reaching a jury. But their children's injuries - and even whether Bendectin was to blame - are not the issue any more. The court is instead being asked to rule on the standards of science in US courtrooms.

The current Supreme Court tends to inch rather than leap in a conservative interpretation of US law. This time the snail race is being watched with particular fervour, by the US business and medical establishment on one side, and the powerful plaintiff's Bar on the other.

The Daubert case has plunged the court into grandiose debates over the meaning of scientific truth and the right of judges to screen evidence. In 1991, lawyers for the Daubert and Schuller families had eight experts waiting to testify in a California court. Drawing on animal, in-vitro and chemical experiments, and by 're-analysing' existing human studies, all would be called to show that Bendectin was associated with birth defects. One of the experts was a senior science adviser at the US Environmental Protection Agency. Another, Shanna Swan, was a statistician at the University of California. All had solid scientific backgrounds. But the eight admitted that none of the published studies linked Bendectin to birth defects; and that they had themselves not published any studies on Bendectin.

The California appeals court threw out the parents' suit on a defence motion to dismiss it, and suggested a 'publish or perish' approach to expert testimony. It singled out work by Swan, who in the witness box takes studies that conclude Bendectin is safe and using statistical 're-analysis' turns them on their head to suggest it is not. The 're-analyses', the appeals court noted, were unpublished, escaped the peer review process, and were generated solely for litigation. Experts should only testify on the basis of scientific techniques that are 'generally accepted', the court said.

The appeals court quoted in its judgment a book, Galileo's Revenge: Junk Science in the Courtroom. Its author is Peter Huber, former law clerk to Supreme Court Justice Sandra Day O'Connor, a Reagan appointee. He is now. senior fellow at the Manhattan Institute, a conservative legal think-tank set up in 1978.

His book is spiced with 'junk science' stories. A psychic who claimed she lost her powers after a CAT scan was awarded dollars 1m by a jury in 1986 (the case was reversed on appeal). Suits earlier this century claimed cancer was caused by trauma like being hit in the chest by an umbrella handle. Mr Huber sees the modern version of 'junk science' at work in claims involving side-effects from electro-magnetic fields created by power lines, to supposed allergic reactions to newsprint.

For defendants in America, the only way to cut off law suits, he says, is to stop them before they go to the jury. Otherwise the risk of a massive jury award - and escalating court costs that cannot be reclaimed - bring enormous pressure to settle.

Merrell Dow itself offered dollars 120m to settle a group of Bendectin lawsuits, though the offer was rejected and the case eventually dismissed. In 1983 the company withdrew Bendectin from the market, having run up millions of dollars in fees defending the drug. It explains why the American Medical Association, the American Insurance Association and the National Association of Manufacturers have showered the Supreme Court with briefs supporting the lower court.

Six Nobel Prize winners and the stately New England Journal of Medicine have joined them to uphold the 'peer review' process as the arbiter of good science. 'When there is a very solid consensus - peer-reviewed published literature - for someone to go to court and say that's wrong, without offering evidence, is irresponsible and damaging,' says Dr Arnold Relman, who edited the journal for 14 years.

A ruling that excludes 'junk science' could go well beyond personal injury suits. It could affect many expert witnesses who challenge scientific consensus: from psychiatrists who testify on the 'future danger' of a prisoner facing a possible death penalty, to faith healing 'experts' backing Christian Scientists who refuse to let their children see a doctor, or even economists who testify in shareholder lawsuits. 'We have to articulate standards,' Mr Hubert says.

The Association of Trial Lawyers of America (Atla), with some 60,000 members primarily representing injured plaintiffs, has weighed in on the other side, citing 'grave concern' about increasing a judge's power to edit evidence. 'Our view is that it doesn't matter what the consensus is,' says Barry Nace, the Daubert family's attorney and coincidentally president-elect of Atla. 'You don't shut someone off for expressing an opinion.'

Carl Sagan, astronomer, Pulitzer Prize winner and author of popular science books, is among a series of US scientists who have also filed briefs cautioning against a strong definition of science. 'Majority science today may be minority science tomorrow,' Sagan's counsel told the court.