Herbal Medicines and Perioperative Care

THE article by Lee et al.¹ is an important addition to the anesthesia literature. Several survey studies have documented the widespread use of herbal products by patients who undergo surgery, but there are few controlled clinical trials of the efficacy or adverse effects and virtually no outcome studies of the effects of herbal medications on surgical patients. The biologic properties of several herbs can interfere with bleeding and recovery from anesthesia, but clinical data on herbal adverse effects are largely anecdotal. Lacking clinical trials, the clinician must rely on case reports and an understanding of the underlying biology of the herb.

Outcome studies of herbal products are difficult to perform for several reasons. First, it is difficult to accurately assess herbal medicine use in individual patients, even when patients are asked about usage directly, either because of their reluctance to disclose herbal use to anesthesiologists or because they do not consider many herbals significant enough to mention.² A physician understands that herbs may be therapeutic, harmless, or dangerous, but many patients assume that natural always means harmless. Therefore, it may be difficult to know which patients are taking herbal products. Studies suggest that more than 30% of the surgical population has used herbs, with higher use by patients with acquired immunodeficiency syndrome or cancer and by people residing in the Western part of the United States.^3–5 Second, although a physician may know the herbal medications a patient is taking, it is virtually impossible to assess the concentration of active ingredient in any herbal product. Herbal medications, which are classified under the Dietary Supplement Health and Education Act of 1994, are exempt from preclinical animal studies, controlled clinical trials, and postmarketing surveillance. Therefore, for many herbal medications, the identity of the active ingredient, the proper dose to achieve an effect, and metabolism and disposition are often unknown. One study found an order of magnitude difference in the ginsenosides, depending on the brand of ginseng.⁶ We have also observed this variability with ginseng. Depending on how ginseng is grown and extracted, it may contain different ginsenosides.⁷ Two of the clinically important effects of ginseng include hypoglycemia⁸ and interference with drugs such as coumadin,⁹ but the extent to which these effects impact care is unknown. Finally, several of the adverse effects attributed to herbs may only become apparent perioperatively. For example, case reports and laboratory studies show that some herbs such as garlic, ginseng, and ginkgo interfere with coagulation.¹⁰ However, these effects may be detected only during acute blood loss. Drug interactions with several herbs such as kava and valerian, which are mild sedatives, could be expected to potentially complicate general anesthesia.² On the other hand, many herbal products, including St. John's wort, change the metabolism of immunosuppressants and cancer chemotherapy by stimulating cytochrome P₄₅₀3A4.¹¹ These effects, clinically important outside of the operating room environment, are unlikely to directly influence anesthetic care.

Despite anecdotal evidence and model populations, little is known about the outcome of patients taking or discontinuing herbal medications before anesthesia and surgery. The study by Lee et al.¹ represents one of the early attempts to objectively quantify outcome differences for patients taking herbal medicine. As such, the authors applied validated outcome methodologies to herbal medicine. The overall conclusion, that the authors could not define specific outcome changes, is encouraging.

Several caveats to this study should be highlighted. First, the study evaluated traditional Chinese medicine. There are more than 12,000 identified herbals, of which more than 500 are commonly used in China.¹² Chinese medicines are complex concoctions of various herbs. Such concoctions are given in therapeutic packages for a variety of disease states as well as perioperatively. Second, without standardization of the concoctions, it is difficult to know what patients are actually receiving. Finally, the study includes a large number of surgical procedures and a relatively small number of patients. Infrequent but highly important events may have been missed, as occurs in drug allergy. Some herbs, such as aristolochia (aristolochic acid), present serious but somewhat rare problems.¹³ Despite these caveats, it seems that the magnitude of the perioperative adverse effects associated with herbal usage is relatively modest. Most of the events fall within the parameters of clinical practice, and absolute attribution to drug interaction is not possible. Based on their biologic properties, if any effect were to emerge as that of an herbal remedy, it would be coagulation, but there is only one description of a probable event, and its severity is ranked as mild.

Although this article is reassuring to clinicians and patients, we should be aware that herbs are drugs whose interactions could manifest in the perioperative period. Thus, ginseng can decrease blood sugar and interfere with the action of warfarin. Hepatotoxicity with ka-va^14,15 has banned its use in Europe, where liver transplant has been necessary in more than 20 patients. Therefore, despite the reassurances of this study, it seems prudent to recommend discontinuation of most herbals before surgery. However, given the reality of modern practice in which many patients are not seen until shortly before or on the day of surgery, the results of this study are encouraging to patients and physicians.

Jonathan Moss, M.D., Ph.D.

Chun-Su Yuan, M.D., Ph.D.

Department of Anesthesia and Critical Care, The University of Chicago Hospitals, Chicago, Illinois. [e-mail]