An estimated 16 million people use them, and surveys suggest that four out of five are satisfied with the results. Sales have almost doubled since 1999 to £191m, fostered by celebrities led by Prince Charles. There are at least 1,000 registered practitioners and an uncountable number who are unregistered. Can so many people be wrong?

The answer is yes, according to two experts who have launched the latest assault on herbal medicine, the practice of treating ailments with either a single "natural" plant, refined and standardised, or a concoction individually tailored to a patient's needs.

Dr Peter Canter and Professor Edzard Ernst from the Peninsula Medical School at the University of Exeter called yesterday for the sale of herbal medicines to be banned unless evidence of their efficacy can be shown. Where there are known risks and no evidence of benefit, there can be no justification for allowing them to remain on to the market, they said.

Herbal medicine, once restricted to specialist outlets patronised by elderly women and bearded youths in sandals, is now available on every garage forecourt and branch of Boots. Curiously named nostrums such as echinacea, for colds and flu, and Devil's Claw, for muscle and joint pain, are walking off the shelves. Boots has seen sales more than double in four years, from £17m in 2002 to £47m in 2006.

Yet there are risks. Some herbs are toxic and have damaged organs such as the kidney and liver; some are misidentified and wrongly included in remedies; some are adulterated with prescription drugs such as steroids or polluted with agro-chemicals; some interact in dangerous ways with orthodox medicines. The risks are regularly highlighted by the UK Medicines and Healthcare Products Regulatory Agency (MHRA), which has issued a dozen alerts about individual products in the past year.

In the first systematic review of individualised herbal medicine – the sort where a customer goes into a herbalist's shop, describes their symptoms, and the herbalist then makes up a preparation of several different herbs – the two researchers found no evidence that the complex Chinese, Ayurvedic or western European potions worked. Of 1,300 studies that they were able to locate, only three were randomised clinical trials comparing the herbal preparations with placebos. Of those three, two found that the individualised preparations were no better than placebos, and the third – of patients with irritable bowel syndrome – found that they were less effective than the standard herbal treatment, a mixture of 20 Chinese herbs. The results are published today in the Postgraduate Medical Journal.

Professor Ernst, a professor of complementary medicine who has been a thorn in the side of the alternative medical lobby, argued that the findings demonstrated that medical herbalists could no longer claim that their tailored remedies worked. What astonished him, he said, was the failure of the herbal medicine establishment to respond to their request for evidence. "We wrote to 15 professional organisations asking if they had any evidence and they came back with nothing. That was the most amazing thing for me," he said.

Dr Canter said: "A drug in mainstream medicine doesn't get used on patients until its efficacy has been demonstrated. Why should double standards apply in herbal medicine? If there is no evidence of benefit and there are serious risks, we should just stop it."

Advocates of alternative medicine are used to charges that they practise magic not medicine. They answer that much orthodox medicine is not evidence-based – for example, most antibiotics have never been tested in randomised trials. Most drugs were discovered by accident, not design, and much orthodox treatment is based on anecdotal evidence, not systematic review. The randomised controlled trial is a relative newcomer to medicine. Orthodox medicine also carries its own risks: the powerful painkiller Vioxx, for example, was withdrawn from sale after it was linked with heart attacks.

Fears about the side effects of conventional medicines and disappointment at their failure to heal the chronic ailments that fill surgeries in the 21st century have caused millions to turn to herbal and other alternative remedies. Complementary medicine has done orthodox medicine a service by reminding doctors of the power of care, the "soothing hand on the brow". The consultations are detailed and lengthy, patients feel taken seriously, and the focus is on results, not scientific explanations. Magic is acceptable if it accomplishes what it promises.

At least half of all GPs refer some patients to complementary practitioners paid for by the NHS, according to the Smallwood report commissioned by the Prince of Wales Foundation for Integrated Medicine in 2005. The GPs do so for conditions for which conventional medicine has little to offer, such as stress and back pain, on the grounds that they have to do something and they are being pragmatic.

Professor Ernst is impatient with this view. "You can call it pragmatism, but I call it bad science," he says. He is impatient, too, with the powerful lobby for alternative medicine. "The marketing is clever and relentless. It starts with Prince Charles and ends with 40 million websites."

The growth in interest in alternative medicine is also driven by increasing wealth – it exactly parallels the rise in sales of luxury cars – and by a desire for "natural" remedies, as opposed to chemically based ones. But this rests on a misunderstanding. Professor Ernst cited a survey of pregnant women that showed that they wanted to avoid chemicals and so preferred plants. "They didn't realise that plants contain chemicals," he said. "The gullibility of the public is monstrous."

Since his appointment as professor of complementary medicine – the world's first – at the University of Exeter almost two decades ago, he has attempted to bring scientific rigour to the discipline, so that patients can distinguish what works from what doesn't. And there is good evidence from randomised trials that certain herbal remedies are effective – St John's wort for depression, hawthorn for congestive heart failure, gingko biloba for dementia, and half a dozen others, he said. He cautions not to throw the baby out with the bathwater and condemn all herbal medicine as ineffective.

But this raises another problem: is it safe for them to be sold over the counter, as at present, without the involvement of a trained practitioner?

Professor Ernst said: "It is no good having a seriously depressed person taking St John's wort and being stimulated enough to go and commit suicide. That is not acceptable. If someone has congestive heart failure for which they are being prescribed conventional drugs, and supplements these with hawthorn, there is a danger of overdose, and they could run into all sorts of medical problems."

Medical herbalists who issue diagnoses and make up remedies have varying levels of training, or none at all, and might miss warning signals of serious illness or be unaware or incapable of dealing with serious reactions, he added.

A European directive is coming in that will require herbal medicines to demonstrate that they are safe, though not that they are effective. But those that have been in traditional use for at least 30 years are exempt.

Professor Ernst says that it would be better to do trials to determine what works before regulating, he said. An application by his own department two years ago for government funding to test herbal medicines was turned down. "The regulation of nonsense is bound to be nonsense," he said. "Without evidence, regulation makes no sense."

His strictures sail over the head of the herbal medicine establishment. Jill Davies, chair of the Association of Master Herbalists, said yesterday that herbal medicine was a vocation that did not need scientific backing as proof of its efficacy.

"I have been in practice for 30 years and I find it works very well," she said.

"You can't have a successful practice and pay your mortgage unless patients get better. That can't be just the placebo effect. People are not stupid. I am not against trials to test herbal medicine, but who is going to fund them?"

The good and the bad

Herbal remedies that work (as shown by clinical trials):

* St John's wort – effective in the treatment of mild to moderate depression

* Hawthorn – helps with congestive heart failure

* Gingko biloba – effective in the treatment of dementia

* Devil's claw – for muscular-skeletal pain

* Echinacea – can help treat the common cold

* Saw palmetto – can help treat benign prostate hyperplasia (enlargement of the prostate not due to cancer)

* Valerian – for treating insomnia

* Salix (from willow bark) – for the treatment of pain

* Kava kava – a tranquilliser (banned in UK because it has been linked with liver damage)

* Andrographis paniculata – can help treat the common cold (banned in UK)

Some remedies that have been warned against:

Aristolochia

In August, the MHRA warned against Xie Gan Wan pills, which commonly contain aristolochia, which in turn can cause kidney failure and cancer

Caterpillar fungus

In July, the Hong Kong Health Department found that the traditional Chinese product called Royal Medic No 1 Chinese Caterpillar Fungus contained a high bacterial count

Senecio

In May, experts proposed a ban on traditional Chinese medicine products containing senecio plants, which cause liver damage

Heavy metals

In February, figures revealed that at least six cases of heavy metal poisoning by Ayurvedic medicines had been recorded in the past three to four years

'Natural' aphrodisiacs

In May, Australian regulators found that Excite for men and Ultimates for women, sold as libido-boosting herbal medicines, contained Viagra.

(Por Jeremy Laurance, The Independent, 04/10/2007)