A decade ago, few ethicists were interested in complementary/alternative medicine (Cam). This changed when consumers started turning to this form of healthcare. Now a large proportion of the population uses some form of Cam: in the UK the figure is about 25%, in Germany it's 75%, and in the US it's around 60%. Faced with this "landslide vote", it is no longer possible to ignore Cam and its implications.

For medical ethics, Cam poses a number of challenges; informed consent is perhaps the most obvious. The principle of informed consent means that patients need full information about a treatment before receiving it. During recent years, the rules of informed consent have become stricter as patients' rights have become more acknowledged. Simply consulting a healthcare professional used to pass as a tacit agreement to receive the treatment that that professional deemed necessary. Today therapists must have the patient's consent before diagnostic or therapeutic procedures can be carried out. This could be a verbal or written agreement.

Informed consent, however, is more than just agreement, it's also about information. According to the Department of Health, the data that healthcare professionals need to provide includes "information about the benefits and risks of the proposed treatment and alternative treatments". What exactly does that mean for Cam?

Let's assume you consult a Cam practitioner with a chronic problem such as recurring headache. After an examination and some questions about your medical history, your therapist will suggest a treatment. Before she starts you should receive, according to the rules of informed consent, full information on the following: the nature of your headache, the likelihood of it disappearing on its own, the chances of the proposed therapy helping, its risks, other therapies that might be considered for headaches, their likely success rates and risks.

Little is known about informed consent obtained by UK practitioners. A laudable exception is a new survey of 150 British chiropractors. One third did not advise patients of other treatment options for their condition. Only 23% reported always discussing serious risks of treatments, and only 14% obtained written consent for a treatment that carries the risk of a major side-effect. This contravenes the UK General Chiropractic Council's code of practice. A glaring gap thus seems to exist between what practitioners should do and how they behave in practice.

But perhaps it is a little unfair to blame Cam practitioners for the discrepancy? Not only would the fullness of this information take a considerable amount of time; in many instances it might even go well beyond their competence. What's more, if they tell you all the risks, they may never see you again.

Informed consent is essential for all healthcare providers. This also means that GPs have to inform you about all Cam options that are supported by sound evidence. If, for instance, you suffer from poor circulation in your legs, most GPs will prescribe drugs, walking, exercise and smoking cessation. However, the rules of informed consent demand that your GP also tells you that ginkgo biloba, a herbal remedy, has been shown to improve this condition. I know of only a few GPs who would do this. The issue of informed consent about Cam could create havoc in all branches of medicine. Nevertheless, I am in favour of giving patients the fullest possible information. If nothing else, it improves healthcare, empowers the patient, and keeps the legal profession at bay.

· Edzard Ernst is professor of complementary medicine at the Peninsula medical school at the universities of Exeter and Plymouth.