Let them have placebos!

I hope the New Year is treating you well. I had a fairly clean break from work, with many family commitments and work to do around the house. Anyway, now I'm back into it, and with a number of blog post ideas that have been germinated over the last two weeks. First up is the placebo effect.

In Dan Ariely's wonderful book Predictably Irrational, he devotes an entire chapter and more to the placebo effect. It works like this: because we are told enthusiastically that a particular medicine or procedure will help cure what ails us, our body reacts to make it so (at least, to some extent). Ariely lists a number of medical (and other) applications where the placebo effect has been shown to work.

On Christmas Day the Wall Street Journal published this article, reporting on a Finnish study (gated) that demonstrates that the positive effects of a type of arthroscopic knee surgery (specifically, partial meniscectomy) are a placebo effect. The study reported:

In the intention-to-treat analysis, there were no significant between-group differences in the change from baseline to 12 months in any primary outcome... the outcomes after arthroscopic partial meniscectomy were no better than those after a sham surgical procedure.

The Finnish study looks very similar to this one from 2002, on arthroscopic surgery for osteoarthritis of the knee, which reported:

At no point did either of the intervention groups report less pain or better function than the placebo group... the outcomes after arthroscopic lavage or arthroscopic débridement were no better than those after a placebo procedure.

These studies (and others that Ariely discusses in his book) raise the question of how often the positive results of surgical procedures are driven largely by placebo effects. Unlike pharmaceutical products, surgical procedures are often not rigorously tested in randomised controlled trials in order to determine efficacy. They continue to be performed because surgeons think they work, because patients exhibit positive responses to them.

Is the placebo effect an example of irrationality though? Does being tricked into thinking that something has positive effects mean that we are irrational? Here, I disagree with Ariely.

A rational person weighs up the costs and benefits of a decision, and proceeds where the benefits outweigh the costs. So, let's start by considering the costs and benefits of surgery (vs. not undergoing surgery) on your sore knee. If you don't have the surgery, then you experience some level of pain and discomfort. If you have the surgery and are affected by the placebo effect, then you experience less pain and discomfort. This is a benefit of having the surgery, and it's still a benefit even if it's only a placebo effect (and therefore not 'real'). I don't think it matters if the benefit arises solely because we are not fully informed that the surgery has no lasting or 'real' benefits.

On the cost side, there is a monetary cost (which in a purely public health system such as New Zealand is borne by the taxpayer, but in private health systems may be borne by the patient or by their insurer) and opportunity costs (cost of foregone income and leisure while in hospital or recovery). There is also the risk of serious or minor medical complications arising from the surgery itself.

So, for the patient themselves, they are weighing up a reduction in pain and discomfort against the monetary and opportunity costs, and the risk of complications. If the benefits outweigh the costs, then the patient should choose to have the surgery.

However, if people are fully rational, we assume that they are fully informed. That is, they know all of the pertinent information that is relevant to their decision. So, a fully rational person would know about the placebo effect, and know whether the surgery is going to confer 'real' benefits or not. In this case, the rational person would not undertake the surgery, because the costs clearly outweigh the (non-existent) benefits.

But wait! Isn't the fully-informed rational person foregoing benefits of the surgery that they could have experienced had they chosen not to be fully informed?

For at least some of those suffering from painful knees, the net benefit of not being fully informed (and therefore choosing to undergo the surgery and receiving the placebo benefit) exceeds zero (which is the net benefit of not undergoing the surgery and continuing to suffer pain and discomfort). So, being fully informed about this possibility they should choose not to be fully informed. It's just as well that people aren't fully rational because otherwise, how would we solve this paradox?

Given that patients aren't fully informed anyway, why not just let people have their placebo surgeries? Some 97% of doctors already prescribe placebo medicines anyway.