In the United States, herbal medicine has grown faster than any other type of complementary or alternative medicine (CAM).1,2 An estimated 60 million Americans use herbal and other dietary supplements; they spend more than $5 billion on herbal supplements3 and $10 billion on nonherbal supplements annually.4
Most physicians are not opposed to their patients' use of supplements and other types of CAM.5 However, as many as 70% of patients do not tell their physicians they are taking herbs or other supplements.4 Thus, it is often necessary to ask about supplement use. Pose the question in a neutral manner: for example, "Patients with your condition often use herbs or other supplements. Have you taken or considered taking such products for your chief complaint or for other reasons?"
Effective communication with patients who use CAM-as well as optimal care-is based on these key elements:
- Information that is scientific and, if possible, evidence-based.
- Recognition of the legal ramifications of the use of CAM therapies (Box I).
- Lack of condescension.
My aim here is to help you apply the methods of evidence-based medicine to evaluate CAM therapies. I examine the most recent data on the pharmacodynamics, efficacy, and safety of 5 commonly used herbs: echinacea, St John's wort, ginkgo biloba, saw palmetto, and black cohosh.
WHERE TO FIND RELIABLE INFORMATION
It is becoming easier to stay abreast of the latest scientific evidence about herbs. However, a discerning and critical approach to all information about herbal products remains necessary.
Books and journals. Several publications on herbs for the physician audience have recently appeared6; however, few are evidence-based and/or clinically oriented.
The Internet. The Internet is a rich source of information; however, keep in mind that it is also a vehicle for spreading myths, hoaxes, and rumors. Several of the more reliable Web sites are listed in Table 1. In addition, although it is far from exhaustive (because of the limitations of voluntary reporting), the FDA's MedWatch Web site is an important source of information about safety issues, such as herb-drug interactions. The address is www.fda.gov/medwatch.
A multimodal approach. Consult more than one source to ensure that your information about the herbs your patients use is as broad and reliable as possible. If you rely only on the published results of clinical trials and guidelines from major medical organizations, you will find too little information available. If you rely only on the FDA MedWatch program, you may be misled by the small numbers of adverse events reported. And if you use only resources tied to the supplement industry-as many Web sites are-you will find numerous unsubstantiated claims.
Patient education. Guide patients to reliable sources of information about herbs. A recent review by Ernst6 of books on herbs for a lay audience found them "more of a risk to the health of the reader than a helpful source of knowledge." Although it is not exhaustive, the FDA's article, "Tips for the Savvy Supplement User," can be printed out from their Web site and made available to all patients who inquire about the use of herbal products. The address is (www.cfsan.fda.gov/~dms/dssavvy.html)
EFFICACY: ASSESSING THE EVIDENCE
Both herbs and standard pharmaceuticals are best evaluated by double-blind, randomized, controlled trials.7 Thus, in theory, there should be no difference in the research meth- ods applied to botanicals and pharmaceuticals to assess their efficacy. Although trials of herbal products have become more common and more peer-reviewed journals are publishing evidence-based reviews of herbs, several obstacles have hindered close scrutiny of many herbal products.
First, the manufacturers of herbal products have little incentive to fund rigorous studies, since the law permits them to sell their products without doing so (Box II).8 Consequently, very few herbs have been evaluated rigorously in randomized controlled trials.
Second, systematic reviews and meta-analyses of these trials usually reveal multiple inadequacies. These include:
Reported effects that are limited and/or require confirmation by larger trials.3
No consideration of long-term effects. Because most clinical trials of herbal products have been short-term and because the adverse effects of many such products develop only after a certain period9 or occur infrequently, such trials reveal little about the products' safety profiles.
Comparison of the herb with placebo rather than with the standard of care, or if the herb is compared with the standard of care, failure to consistently use the correct dosages. Rigorous investigations—those in which herb, placebo, and conventional therapy are studied (to assess the sensitivity of the participants to placebo)—are rare.
Unreliable results-even in well-done clinical trials-because of the lack of standardized manufacturing procedures in the herbal supplement industry (see Box II). The multitude of formulations of a given herb that are available, or batch-to-batch inconsistency of any given formulation, can cause a product to vary from one trial to the next, or even within the same trial.
Publication bias (the more frequent reporting of positive results than of negative results), which leads to an overestimation of the treatment effect of many products.10,11
With these limitations in mind, I summarize the recent evidence on 5 widely used herbs (Table 2).
Commercially available herbal medications are produced from 3 species of the herb: Echinacea augustifolia, Echinacea pallida, and Echinacea purpurea.6
Pharmacodynamics. Depending on the part of the plant used in its manufacture, an echinacea product may contain 6 or more distinct chemical constituents with pharmacologic activity.12 The types of activity include stimulation of the immune system; local anesthesia; and anti-inflammatory, hormonal, antiviral, and free radical-scavenging properties. The nonspecific immunostimulatory activity includes wound healing.12 The specific immunostimulatory effects of echinacea are anti-infective and probably result from activation of phagocytosis, release of tumor necrosis factor, and production of macrophage interleukin-1 and interferon beta-2.12 Echinacea is not thought to be directly bactericidal or bacteriostatic.13
Evidence of efficacy. A systematic review of 26 clinical trials examined the effect of echinacea on immunomodulation.14 Of these 26 trials, 18 were randomized and 11 were double-blind. Thirty of the 34 outcomes in the treatment groups were identified as "superior to controls" by the authors of the original studies. However, the review found that only 22 of the 34 outcomes were "reasonably demonstrated."
In another review of the results of 5 randomized studies of the immunomodulatory activity of different preparations of echinacea, 2 of the 5 studies demonstrated statistically significant evidence of immunomodulatory activity in the treatment groups (significant changes in phagocytic activity).15 The 3 other studies were unable to be compared because of differences in evaluation methods.
Most double-blind, randomized, controlled trials of echinacea have evaluated its effectiveness in the prevention and treatment of upper respira-tory tract infections (URTIs). In one such trial, no prophylactic effect of echinacea could be demonstrated, although more patients in the group that received echinacea believed they had benefited from the treatment.16
In other randomized trials of echinacea, researchers found that echinacea did not significantly decrease the incidence, duration, or severity of respiratory tract infections compared with placebo.17 Other double-blind, randomized, controlled trials have found significant diminution in severity and duration of symptoms-especially at higher doses-in patients with colds or flu-like symptoms.18
More recently, a Cochrane review of 16 randomized clinical trials of echinacea for treatment of URTIs was done.19 The authors concluded that the evidence was insufficient to recommend a specific echinacea product for treatment of URTIs. They also noted that, although publication bias may have distorted the results, echinacea might be efficacious. The trial data, however, are weak and inconclusive.
Caveats. The safety and tolerability data suggest that echinacea is a relatively benign substance.20 Adverse effects are rare.21
Echinacea products should not be combined with immunosuppressants because of the possibility of pharmacodynamic drug interactions. Echinacea is therefore contraindicated in patients with multiple sclerosis, autoimmune disorders, AIDS, or tuberculosis.22 Because of insufficient safety data, echinacea products are best avoided during pregnancy and lactation, although one study found no evidence of adverse pregnancy outcomes after echinacea consumption.23 Most researchers also recommend that echinacea not be used for longer than 8 weeks24 because no long-term data are available.
Recommended dosages and formulations. Because a variety of dosages and formulations have been used in clinical trials, the most effective dosage and formulation are not known. The most commonly used preparations in the United States are liquid extract of E purpurea root and echinacea tea. Typical dosing for the extract (for treatment of a URTI) is 3 mL every 3 to 4 hours for the first day or two, then 3 times daily for the subsequent week. Suggested dosing for the tea formulation is 6 to 8 oz 4 times daily for the first 2 days, titrated down to once or twice daily on days 3 through 7.25