Friday October 5 2007
No evidence that tailored herbal medicine does any good
Herbal medicine that is tailored to the individual is “futile”, and may “do more harm than good”, The Guardian and the BBC reported on October 04 2007.
These, and other newspaper stories reported that a study has found there to be no evidence that customised herbal medicine, where the herbalist prepares a selection of herbs tailored to the individual’s description of their symptoms, actually works.
The stories are based on a review of all the suitable existing studies that compared customised herbal treatment with other therapies.
This study concentrated on the type of herbal medicine where several herbs are mixed together to match an individual’s requirements. There is evidence that certain herbs, such as St Johns' Wort, have beneficial effects when used for specific purposes. However, there are risks that these herbs can interact badly with prescription medication, other natural remedies, or that people with certain medical conditions can react badly to them.
Where did the story come from?
Drs Guo, Peter Canter and Edzard Ernst from the Universities of Exeter and Plymouth carried out this systematic review. No information is provided on how this university-based review was funded. The study was published in the peer-reviewed medical journal Postgraduate Medical Journal.
What kind of scientific study was this?
The study was a systematic review of the research into individualised (customised) herbal medicine for use as a treatment in any condition.
The researchers carried out a search for all the randomised controlled trials that assessed customised herbal medicine. They also contacted experts in the field and 15 professional bodies to seek any additional studies that had not been published.
They then assessed the results of any suitable studies that compared customised herbal medicine (defined as treatment that had been specifically tailored for individual patients) with either placebo or standardised treatment (defined as a combination and choice of herbal medicines that was not customised).
The researchers expected that any studies they found would be quite different from each other (for example including different populations, or using different preparations of herbs etc.). Therefore, a decision was made not to combine the results of the studies they found into one overall measure of how effective or not the treatment would be. Instead, they intended to describe the results of high quality studies individually.
What were the results of the study?
The researchers identified 1,345 articles through their literature search and contact with professionals. Among these were three randomised, placebo-controlled trials; one of which was complete but unpublished while the other two were ongoing studies. The reviewers considered these three trials to be of moderate to good quality in terms of the methods they used.
The first of the three studies assessed both standardised herbal medicine and customised herbal medicine. It found that herbal medicine in general was better than placebo at reducing the reported symptoms of irritable bowel syndrome. However, when the data for the two groups were analysed separately against placebo, the standardised treatment was more effective in improving the symptom score than the customised treatment.
The second study found no difference between customised herbal treatment and placebo for treating symptoms of osteoarthritis.
The third study found no difference between customised treatment and placebo for chemotherapy-induced blood toxicity in people with early-stage breast or colon cancer.
What interpretations did the researchers draw from these results?
The reviewers conclude that the available studies offer no evidence that customised herbal treatments are effective for any condition. They point out that the lack of evidence for an effect, the potential for side effects and the potential for herbs to interact with each other or with other drugs mean that the use of customised herbal medicine cannot be recommended.
Considering the widespread use and long history of herbal medicine, they express concern that they could only find, despite their efforts, three randomised controlled trials assessing the treatment. The evidence base for herbal treatments rests mainly with studies of single, standardised herbal extracts and they note that, for this reason, “claims by herbalists who use the individualised approach that their practice is evidence based are disingenuous”.
The reviewers raise concern that all three included studies interpreted their findings over-optimistically and that all three studies had specific weaknesses (including groups that were different at baseline, using unclear analyses or not completing with the numbers of participants required (a lack of power) to detect any difference).
What does the NHS Knowledge Service make of this study?
This was a well-conducted systematic review of the evidence for and against the use of customised herbal treatments for human ailments. This sort of review is generally considered to be the best way to establish how effective a treatment is, and so this should give us the best verdict yet on individualised herbal medicine. Our comments about the quality and conclusions of this review parallel the authors’ own:
- The search for published studies may have missed those that are “hidden in the oriental literature”. However, the researchers contacted experts and professional bodies from all over the world to try to identify these studies.
- Herbal practitioners may find it difficult to fund research into their treatments. However, such studies are entirely feasible and considering the extent and history of the practice of herbal medicine, they are warranted.
- As the news stories and the review authors themselves say, a priority is to minimise the risk to the public from treatments that can cause harm. There is concern about the lack of a standardised approach to diagnosis and prescribing when the herbal combinations are tailored to individuals. Limited knowledge about potential herb-herb interactions, or interactions with other drugs the patient is taking could be dangerous.
There is a difference between having too little evidence to decide whether something is good or not and having really good evidence that it isn’t good at all. The fact that only a few good studies are available may put ‘individualised herbal medicine’ into the former category.
However, there is a noticeable contrast between the rigorous analysis and testing that is required for drugs to be licensed for use in humans and the existing evidence for herbal medicine. Considering our priority to protect people from harm, it would be sensible to limit the use of treatments to those that have proven benefit and to regulate that treatment accordingly.
In an accompanying editorial to this, one of the authors makes an important distinction between phytotherapy (using herbs of proven benefit like St John’s Wort), plant-based therapies sold without consulting a professional and the traditional herbalism which was studied here. He calls for more thought into ways to reduce the damage of irresponsible advice in this area and says “health writers should be reminded that the promotion of nonsense is not entertainment but puts people at risk”.
Sir Muir Gray adds...
The blunderbuss was never very effective. A sharpshooter aims one bullet at a defined target and that is a good principle for medicine too.
Links to the headlines
Tailored herbal medicine 'futile'. BBC News, October 4 2007
Herbalists' cocktails may do more harm than good, say researchers. The Guardian, October 4 2007
Natural medicine on trial: The trouble with herbs. The Independent, October 4 2007
High street herbalists can offer no evidence that their remedies work. The Times, October 4 2007
Herb cures that 'do you more harm than good'. Daily Mail, October 4 2007
Links to the science
R Guo, P H Canter and E Ernst. A systematic review of randomised clinical trials of individualised herbal medicine in any indication. J Postgrad Med 2007;83: 633-637
There are several reviews in the Cochrane library that consider the evidence around herbal medicines (though none specifically consider “individualised treatment”) including:
Liu JP, Yang M, Liu YX, Wei ML, Grimsgaard S. Herbal medicines for treatment of irritable bowel syndrome. Cochrane Database Syst Rev 2006, Issue 1
Liu JP, Zhang M, Wang WY, Grimsgaard S. Chinese herbal medicines for type 2 diabetes mellitus. Cochrane Database Syst Rev 2002, Issue 3
Rathbone J, Zhang L, Zhang M, Xia J, Xiehe L, Yanchun Y. Chinese herbal medicine for schizophrenia. Cochrane Database Syst Rev 2005, Issue 4