Reasons why ineffective treatments can appear to be effective

In so-called alternative medicine (SCAM), there are few notions more difficult to argue against than this one:

I KNOW IT WORKS BECAUSE I HAVE EXPERIENCED IT MYSELF!

Yet, it is this notion that we most often encounter when discussing with proponents of SCAM. To argue against it often gets an emotional response, not least because the SCAM proponent feels that we are implying that (s)he is either lying or stupid or both. Therefore, it is important, I feel, to keep on addressing this issue. So, let me once again explain why ineffective treatments often appear to be effective.

To make this less boring, let’s consider a concrete case, one where nobody (well, almost nobody) will get emotionally involved. Let’s consider a patient who has been suffering from recurring pain of an undiagnosed origin and experiences benefit after crystal healing, a form of SCAM for which there truly is not a jot of evidence.

Why then does (s)he experience less pain after consulting her crystal healer?

There is a myriad of reasons, and in this post I will borrow some from the literature of psychology:

1. The patient might think his/her pain is less frequent without actual change; recall bias.
2. The healer might have provided an explanation for the pain which the patient experiences as a symptomatic improvement.
3. The healer could have convinced him/her that his/her pre-treatment pain was worse than it truly was; recall bias.
4. The patient might report improvement in accord with what (s)he believes to be the healer’s hope and expectation; social desirability.
5. The patient is prone to preferentially remembering improvements as opposed to worsening; recall bias.
6. Patients might interpret ambiguous changes in symptoms as indications of improvement.
7. Many symptoms disappear or improve on their own; natural history of the disease.
8. Many symptoms fluctuate; natural history of the disease.
9. Extremes become less extreme on re-testing; the phenomenon of ‘regression towards the mean’.
10. Patients employ several therapies simultaneously and later misattribute an improvement to one treatment.
11. Patients expect to improve; the placebo-effect.
12. Patients are conditioned to improve; the placebo-effect.
13. Improvement owing to enthusiasm of receiving a new intervention; the novelty effect.
14. Improvement owing to enthusiasm of receiving an exotic therapy.
15. Improvement due to the compassionate care and attention received.
16. Improvement due to the time spent with the healer.
17. Improvement due to the effort and resources invested by the patient.

All these phenomena (and several more, I am sure) work in concert and can generate a clinical outcome that makes ineffective therapies and even slightly harmful treatments appear to be effective. In any single case, it is impossible to decide what precisely brought about the improvement. The only way to make sure that the specific effects of the treatment (in my example the crystal therapy) was involved is to conduct one (better several) controlled clinical trials.