This site is intended for health professionals only

There are only three categories of treatments

Edzard Ernst puts forward his suggestions for a three-part classification system for medical interventions

I propose a simple (I'm sure that some would claim it to be simplistic) view of therapeutics. Medical interventions, CAM or otherwise, can be divided into three categories: those which (with reasonable certainty) generate a net benefit, those which don't achieve this goal, and those where reasonable certainty does not exist.

The first type is defined through well-documented efficacy, e.g. walking exercise for intermittent claudication, or horse chestnut seed extract for varicose veins, or garlic for hypercholesterolaemia (see previous blogs). As no treatment is entirely without its risks, we always have to consider adverse effects. But, in the three cases mentioned above, these are minor. It follows that the benefits outweigh the risks – in other words, these therapies generate a net benefit and therefore they belong to the first category.

Defining the second type of medical intervention can be more complex. These are therapies where the risks outweigh the benefits. Obviously, even treatments with proven efficacy can fall into this category, if their risks are disproportionate. Conversely, treatments with only relatively few risks but no proven efficacy can also belong to this group. And, of course, treatments with considerable potential for harm and no proof of efficacy would also fall into this bracket. Examples for the latter scenario are chelation therapy for intermittent claudication, which can cause serious adverse effects and has been shown repeatedly to be ineffective, or upper spinal manipulation.

The third type of treatment is probably the largest of the three categories. These are interventions where we either are not sure of their efficacy, or of their safety, or both. In such circumstances, risk-benefit assessments will be too unreliable for prudent advice. We therefore have to conclude that the treatment in question is not associated with a demonstrably positive risk-benefit profile. To recommend such a therapy for routine use would therefore be irresponsible.

What does all of this mean? As GPs or scientists, our aims are 1) to employ, where we can, the first category and to enlarge it through rigorous research, 2) to avoid the second category altogether, 3) to decrease the size of the third category through well-targeted clinical investigations.

In theory, all of this sounds simple. Sadly, in practice, it often is not.

Professor Edzard Ernst is professor of complementary medicine at the Peninsula Medical School, University of Exeter

Click here to read more Edzard Ernst. Edzard Ernst