Many people seem to think that the value of a therapy is determined by its efficacy: a treatment that is highly efficacious must be better than one that is less efficacious. Others seem to believe that it is the safety of a therapy which matters most: a treatment that causes no or few side effects must be good, one that has many is bad.
Such notions might appear logical, but they are mistaken. Things are usually more complicated. Some treatments can cause extremely serious side effects but are still extremely valuable. An example would be chemotherapy; it often causes all sorts of awful problems but, if it saves cancer patients’ lives, it cannot be bad.
Other treatments might be virtually free of side effects, but they are nevertheless rubbish. Take crystal healing, for instance; it is hard to imagine that it causes any side effects but, as it also does not cure anything, it cannot possibly be a good therapy.
To determine the real value of a therapeutic intervention, we need to consider more than its efficacy alone and more than its safety alone. Obviously, we must look at the balance of the two factors.
When a new drug comes on the market, it has been tested thoroughly for efficacy; we therefore can be fairly sure that it works. But initially we know relatively little about its safety; in particular, we know little about possible rare side effects. Such knowledge requires data not just from the few hundred patients who took the drug when it was tested in efficacy trials, but we need data from a few hundred thousand patients.
To generate this information, drugs are monitored for side effects while they are used in routine practice. Should this ‘post-marketing surveillance’ throw up any serious problems, the drug might be withdrawn from the market.
But this only applies to conventional medicine. In alternative medicine things are different, sometimes dramatically different. As the value of any therapy is determined by its risk/benefit balance, we would ideally want to know the efficacy and the safety of alternative therapies too. Yet we often don’t know enough about either.
Alternative therapies have not been tested for efficacy before they come on the market; they usually were in use long before we had the idea of licensing and regulating drugs. Consequently, we have little or only incomplete knowledge about their efficacy.
On the safety side of the equation, things are even worse. There is no post-marketing surveillance of alternative therapies, and all we know about their risks comes from the occasional case report published in the medical literature. This means that under-reporting of harms is huge, and our data are just the tip of the iceberg.
It follows that any attempt at evaluating a risk/benefit balance of alternative therapies is highly problematic. We usually know too little about both determinants to even begin a reasonable estimation. All we can do in this situation is rely on rough estimates.
If any given therapy generates no benefit because it is not efficacious, we can be sure that its risk/benefit quotient can never be positive. Dividing any finite number for risk, however small, by zero gives an infinitely large figure. We can furthermore assume that, for any therapy that is only marginally efficacious and thus generates only a small benefit, even a very small risk would result in an unfavourable risk/benefit balance.
Finally, we can say that an alternative therapy that is known to cause serious harm, the benefit would need to be substantial for its risk/benefit balance to come out favourable.
And what about those alternative therapies for which we have not enough information to attempt even such rudimentary analyses? Alternative practitioners and their followers tend to think that we must give them the benefit of the doubt. This is a dangerously misguided view.
In the interest of our patients, we ought to consider any intervention to be inefficacious until we have good evidence to the contrary. Similarly, any therapy must be considered unsafe until the time we have sound data showing it is not unduly harmful. Giving alternative therapies the benefit of the doubt is therefore not an option.
Such talk is alarmist, claim fans of alternative medicine. After debating with them ad nauseam, I now have this challenge for them: show me your list of alternative therapies that demonstrably are associated with a favourable risk/benefit balance. Considering that there are more than 400 different alternative therapies and that most of them are used for a wide range of conditions, such a list could potentially be very long indeed.
But I will be modest: if you can list more than a dozen alternative therapies for specific conditions, I promise to never write about the risk/benefit balance of alternative medicine again.
Edzard Ernst, emeritus professor at the University of Exeter, is the author of Homeopathy: The Undiluted Facts and the awardee of the John Maddox Prize 2015 for standing up for science. He blogs at edzardernst.com.