The Pharmaceutical Journal
Vol 274 No 7352 p679
4 June 2005

This article
Reprint   Photocopy

PDF 85K, Acrobat Reader

Medicines guilty until proven innocent

In the 11th article in a series on complementary medicine, Edzard Ernst says that even though complementary medicines have been used for hundreds of years, good evidence of safety and efficacy are required

Complementary medicine series

Edzard Ernst, MD, FRCP, is professor of complementary medicine at Peninsula Medical School, Universities of Exeter and Plymouth, and editor-in-chief of the journal Focus on Alternative and Complementary Therapies

In law, one of the most fundamental principles is “innocent until proven guilty”. In medicine, one overriding rule indicates the exact opposite. How can we explain this contradiction?

Imagine a new drug is produced to treat depression. The manufacturer is likely to claim it is “innocent”, ie, free of serious side effects. But should we blindly trust such assertions? Of course not — we have to insist on evidence. And only if the data show beyond reasonable doubt that there are no major safety problems can we accept the manufacturer’s claim of innocence and allow the drug on the market. Any other course of action would jeopardise the health of the nation. In other words, applying the principle of “guilty until proven innocent” is essential for health care to be safe.

Is complementary medicine innocent?

The other claim a manufacturer of a new medicine would make is “drug X is effective”. Again, until such data are on the table, we cannot assume innocence; that is, we cannot take the claim of efficacy for granted. So again we are applying the principle of “guilty until proven innocent”.

On closer inspection, the contradiction of the legal principle (innocent until proven guilty) and the medical principle (guilty until proven innocent) is only an apparent one. Both promote the same overriding concept, namely to protect those who are vulnerable. In law the most vulnerable party is the defendant; in medicine it is the patient. Rather than being opposites, both principles pursue similar aims.

There are few reasons why the medical principle “guilty until proven innocent” should not apply to all areas of medicine — and that includes complementary medicine (CM). Yet some relevant differences do exist between CM and conventional medicine. In CM, we are not normally dealing with a new drug or treatment, but with therapies that have been around for hundreds of years. One could, therefore, argue that a long history of usage amounts to some kind of demonstration of innocence. If a treatment has survived for centuries, we can assume that it is both safe and effective — or can we?

Historical evidence

Rational thinkers will point out that a long history of use does not constitute proof.¹ Side effects of a herbal remedy could, for instance, be delayed for such a long time that one simply does not associate the cause with the effect any longer. Thus they could have been overlooked for hundreds of years. Furthermore, interactions with modern treatments that were not available 100 years ago, could occur today. Historical evidence might, therefore, lull us into a dangerously false sense of security. And, of course, there is no medical intervention, however ineffective, that does not produce its own “fan club” of convinced followers who believe it works. Therefore, testimonials are often of limited value, and effectiveness can only be demonstrated in proper clinical trials.

Even if a long history of use does not constitute proof, it could still be used to indicate that efficacy and safety are a little more likely.² This is precisely why regulators tend to be more liberal when legislating for complementary medicine. They try to account for the fact that we are dealing with treatments which have been field tested for hundreds of years on millions of people.

But, the liberal stance must have its limits. As soon as alarm bells start ringing, the whole situation requires a thorough reassessment. If, for instance, serious adverse effects start being reported in association with a herbal remedy, we must look into the matter with all the necessary rigour. A few years ago, when St John’s wort was first linked with powerful herb-drug interactions,³ this is precisely what happened. Initially there was little more than a mere suspicion. But dozens of research projects subsequently investigated the possibility of interactions in depth and elucidated two mechanisms through which the interactions are mediated. Today we know that St John’s wort should not be combined with a range of drugs, including warfarin and the oral contraceptive pill⁴ (see Panel). As a consequence, herbal medicine has become a little safer. Omitting the responsibility of thorough re-assessments would not have been liberal — it would have been irresponsible.

Conclusion

The rule “guilty until proven innocent” applies to all forms of health care, including CM. The debate cannot be about the principle but about the often vexing question, what should be considered proof? For the sake of public safety, we should not sacrifice reason for political correctness, inertia or powerful opinions.

References

1. Ernst E, De Smet PAGM, Shaw D, Murray V. Traditional remedies and the “test of time”. European Journal of Clinical Pharmacology 1998;54:99–100.
2. Ernst E. The “improbability” of complementary and alternative medicine. Archives of Internal Medicine 2004;164:914.
3. Ernst E. Second thoughts about safety of St John’s wort. Lancet 1999;345:2014–6.
4. Mills E, Montori VM, Wu P, Gallicano K, Clarke M, Guyatt G. Interaction of St John’s wort with conventional drugs: systematic review of clinical trials. BMJ 2005;329:27–30.