Australasian Science: Australia's authority on science since 1938

The Ultimate Placebo

By Ian Harris

The placebo effect is usually invoked with pharmaceutical treatments, but why not surgery?

The introduction of new drugs is rigorously controlled, with effectiveness established through scientific comparison with a suitable control – ideally a placebo. For surgery, introducing new procedures relies on a biologically plausible mechanism and evidence from lab studies and case series, with little regard to possible placebo effects and no requirement for testing against treatment alternatives, let alone against a placebo.

Is surgery largely exempt from any requirement to test against a placebo because the effect is so obvious? This is not a strong argument, as many previous “strong effects” have been shown to be absent on scientific testing.

Because placebo trials of surgery are unethical? Again, no. The use of placebo surgery in trials can easily satisfy the ethical requirements for human research. For conditions that rely on subjective reporting, like pain, such trials provide the most ethical way of establishing effectiveness because they incorporate the least biased method of testing, thereby increasing the probability of finding the true answer and minimising research waste (and possible healthcare waste from introducing ineffective treatments).

Another reason to test surgical procedures against a placebo control is that everything that we know about the placebo effect points to surgery being the ideal placebo. It is expensive, elaborate, relies on high-tech devices, is painful, requires considerable “buy-in” from the patient and is delivered by authoritative figures convinced that it will work.

Furthermore, when surgery is (albeit infrequently) tested against placebo (sham surgery), it often fails to show a clear benefit, even for common procedures accepted as effective.

History provides many examples of common surgical procedures that were later shown to be ineffective. These range from one of the original surgical procedures, bloodletting, to nephropexy (tying down a “floating kidney”), which was once the most common urological procedure and is still being performed today.

For some current surgical procedures there is evidence that they are not superior to non-surgical treatment, but for many more of them there have been no trials comparing them to non-surgical treatments or placebo surgery.

In my own field of orthopaedic surgery, there is currently considerable debate regarding the appropriateness of our most common procedure, knee arthroscopy, with two landmark trials showing it no more effective than placebo in treating osteoarthritis or degenerative meniscus tears.

What drives surgical practice, particularly in the face of such evidence? Largely, an over-reliance on one’s personal perception of effectiveness – a perception easily distorted by the natural history of the condition and logical fallacies that confuse association with causation.

So what’s the big deal? If people feel better after the operation, who cares if it wasn’t directly due to the procedure? I care, and we should all care. To accept the placebo effect as legitimate treatment would mean that we could not identify the specific effectiveness of a procedure, making further scientific enquiry and extrapolation difficult. It also involves deceit, carries a financial cost and involves exposure to potential harm.

But the biggest problem I have with continuing to use untested surgical treatments and relying on the placebo effect is that, once this becomes acceptable practice, we remove the barrier between mainstream medicine and alternative medicine, or any non-science-based healing.

The solution is for surgery to be subjected to the same stringent criteria as any science. It must provide objective, reliable and independent evidence of effectiveness before being allowed to introduce (or continue with) surgical interventions.

This is possible, and it is happening more frequently, with trials of surgery versus non-operative treatment (and sometimes placebo surgery) being reported regularly. There is some evidence of practice change in response to that evidence but, in my opinion, there is more research and education to be done before surgery has the evidence base that justifies its use of resources.

Relying on observational evidence when less biased experimental evidence is available or possible to obtain is not the way to pursue a scientific endeavour.


Ian Harris is Professor of Orthopaedic Surgery at The University of NSW, and author of Surgery, The Ultimate Placebo (published by NewSouth Books).