Why should we as family physicians take on the difficult and perplexing task of learning about herbal medicines? There are several reasons. First, with approximately 60 million adult Americans regularly using herbal supplements,¹ the principles of patient-centered and comprehensive care require that we ask our patients about this practice. The recent American Medical Association recommendations on alternative medicine include the suggestion that all physicians become better informed about the use of unconventional therapies.² In my own practice, I have found that my willingness to discuss the use of herbal medicines has greatly added to the feeling of collaboration with my patients and to my understanding of their health beliefs and practices.

The second reason to inform ourselves about herbal medicine is that, as noted in the article by Zink and Chaffin in this issue of American Family Physician,³ herbal supplements and conventional medications may have significant interactions and, unless we inquire about the use of herbs and begin to document such interactions and complications, we will not be able to deal with these complexities responsibly. For example, two case reports of bleeding complications in patients taking ginkgo biloba have been published⁴; since ginkgo is known to be an inhibitor of platelet activating factor, these reports are of some concern. Should patients taking warfarin or aspirin be taking ginkgo? Should patients taking vitamin E, which also may have antiplatelet activity, be taking ginkgo?

Another example involves St. John's wort, which is thought to have both monoamine oxidase inhibition activity and an effect on serotonin pathways.⁵ Does this mean that St. John's wort should not be used in combination with conventional antidepressants? Is there a risk of serotonin syndrome if this herb is used in combination with selective serotonin reuptake inhibitors (SSRIs)? A recent letter to American Family Physician described a case of lethargy in a woman who used paroxetine (Paxil) in combination with St. John's wort.⁶ Does this represent a cause-and-effect relationship, or is it coincidence?

These pressing questions will not be answered until we routinely begin to inquire about patients' use of herbs and collect data on outcomes and complications. Many elderly patients are taking ginkgo in addition to aspirin; many patients use St. John's wort in combination with SSRIs, in hopes of lowering their dose of the conventional medication and reducing side effects. What should we recommend in these situations?

Individual published case reports will not adequately answer these questions. The post-marketing data collection system that prompted the withdrawal of dexfenfluramine (Redux) does not exist for herbal products, and the first step in creating it will be routinely documenting data on use of herbal medicines. The next step, of course, will be a move by the U.S. Food and Drug Administration (FDA) and the manufacturers of herbal medicines to establish a coordinated system for data collection on both the adverse effects and the beneficial outcomes of these products.

The final reason for taking on the task of teaching ourselves about botanical medicines is the many clinical situations in which they might prove useful. Herbs might provide an effective alternative when conventional medicines have not been well-tolerated because of side effects. For example, a patient with benign prostatic hyperplasia who does not tolerate the alpha blockers or finasteride (Proscar) might receive relief from his symptoms by taking saw palmetto: a recent European multicenter study⁷ of 1,098 men with benign prostatic hyperplasia showed that saw palmetto and finasteride had similar benefits on both subjective and objective outcome measures.⁷ Although some concern exists that saw palmetto may affect prostate-specific antigen levels, this effect has not been proved in any of the major trials, nor have any other significant adverse effects been reported.

Another common situation in which botanical medicines could be useful is when the indications for use of a conventional medication are borderline, but both the physician and the patient feel that some treatment is indicated. An example of this situation might be a 40-year-old man with a family history of heart disease but no other significant risk factors and a low-density lipoprotein (LDL) cholesterol level of 162 mg per dL (4.20 mmol per L) despite maximal diet therapy. This patient might be a good candidate for a trial of garlic supplements, which have been shown to reduce both total cholesterol levels and LDL cholesterol levels in a number of studies.⁸ Here again, the risk is insignificant, because no major adverse effects of garlic have been reported in any of the clinical trials, and the use of a botanical medicine expands the options available to the physician and the patient.

More research on the utility and safety of herbal medicines is desperately needed. The Office of Alternative Medicine of the National Institutes of Health (NIH) is currently co-funding major trials of St. John's wort for the treatment of depression and ginkgo biloba for the treatment of dementia, but this effort only represents a beginning. In 1997, the Commission on Dietary Supplements Labels, appointed by President Clinton to advise the administration on how to regulate herbal products more appropriately, recommended the establishment of a “natural products drug review procedure,” and a joint industry/FDA effort to track adverse effects in post-marketing surveillance.⁹ As physicians and teachers in search of a rational, evidence-based practice of primary care, it behooves us to advocate for a greater role for the FDA, the NIH and the herbal supplement industry in collecting data on outcomes of herbal medicine use, both adverse and beneficial.