Research on therapies that have no known mechanism of action

Steve Simon


Several people on the IRBForum have made comments along the lines that research on certain alternative medicines (such as homeopathy) has no scientific merit because there is no known mechanism by which these therapies can work. Since there is no scientific merit, there is no balance between the risks and benefits. This leads directly to a conclusion that it is unethical to conduct a study on these alternative therapies.

An argument can (and should) be made that this is a poor use of scarce resources. There are many more promising therapies than homeopathy that have gone untested, so a study of homeopathy diverts money from these more promising alternatives. But to say that investigations of therapies with no known mechanism of action are always unethical is perhaps a bit too extreme.

First, such an argument is paternalistic. It presumes that the researcher always knows what is in the best interest of the patient and the researcher alone decides what risks a patient should be willing to endure. Many of these alternative therapies are currently being embraced by thousands of people and who are we to tell them that they should not endure in a research setting something that they routinely seek out for their daily care?

Paternalism or partnership? Patients have grown up-and there’s no going back. Coulter A. Bmj 1999: 319(7212); 719-20. [Medline] [Full text] [PDF]

Second, the concept of mechanism is somewhat subjective. Often you can find a mechanistic explanation for why a certain exposure might increase your risk of disease and a competing mechanistic explanation for why a certain exposure might decrease your risk of disease. It also depends on what your current beliefs are. Someone who believes in the concept of Qi will be able to postulate mechanisms for all sorts of alternative medicines. Someone who believes in God will have no problem with a mechanism for prayer studies. The mechanistic explanations that you are willing to accept for intellectual development will vary depending on whether you believe that IQ is determined entirely by your genes, entirely by your environment, or some mixture. For that matter, the mechanisms that you will accept depend on whether you believe that intelligence is a single dimensional construct that we can measure validly.

Third, some therapies have been adopted prior to the identification of a mechanism of action. Only later was a mechanism discovered. I don’t have any specific examples, but I’m sure that some readers will provide me with good examples.

Fourth, presence of a mechanism should be only one of several considerations being evaluated to gauge scientific merit. You might also want to consider whether there is any animal data to support the use of a certain therapy. Animal models are imperfect, but they do prove useful in preventing many totally hopeless therapies from being tested on humans. Is there credible observational data to support an alternative therapy? Observational data is also imperfect, especially in an area like alternative medicine where there is a lot of enthusiasm among proponents. But credible observational data makes the case for developing good randomized studies all the more compelling.

Finally, if a therapy is currently widespread but without any scientific basis, then a research study will still have merit if it produces convincing evidence to discourage the use of this valueless therapy.

The bottom line is that having a mechanism for a therapy strengthens the scientific merit, but absence of a mechanism does not imply that the research is wholly without merit.

You can find an earlier version of this page on my original website.