Issue 9 / 30 August 2010

How does one of the world’s leading medical journals come to publish an article that finds a medical intervention is no better than sham treatment but goes on to recommend it as possible first-line therapy?

That was one of the questions being asked in the medical blogosphere after the New England Journal of Medicine recently examined the usefulness of acupuncture in treating chronic lower back pain.

Here’s one part of the article that generated a lot of ire:


Acupuncture … has not been established to be superior to sham acupuncture for the relief of symptoms of low back pain. As a result, it is not often regarded as the first choice of therapy. However, since extensive clinical trials have suggested that acupuncture may be more effective than usual care, it is not unreasonable to consider acupuncture before or together with conventional treatments…

As several commentators have pointed out, you could make exactly the same argument in favour of sham acupuncture, since it was equally successful in outperforming conventional care.

US infectious diseases specialist Dr Mark Crislip, an ascerbic critic of complementary medicine, was one who launched an attack on the NEJM, describing the article as “total tripe” and “tooth fairy science”.

The 10-12 acupuncture sessions the authors recommended for a patient whose chronic back pain had not responded to conventional treatment constituted an “unethical, expensive, useless placebo,” he wrote.

Obviously, many patients and doctors would disagree with Crislip’s dismissal of acupuncture generally as “prescientific magic”.

In fact, it is estimated that at least 15% of Australian GPs claim Medicare rebates for providing acupuncture services.

But the controversy does raise questions about how journals, and medicine more generally, can best engage with treatments that patients swear by, even if they do not meet conventional standards of proof.

A few years ago, I attended an international medical acupuncture conference where one of the visiting experts told me lack of evidence from randomised controlled trials should not be an obstacle to treatment.

He estimated about 20% of patients with neuropathic pain achieved adequate relief with acupuncture. “Lack of evidence does not mean lack of effect,” he said.

It may be counterintuitive, but, if a patient can afford an intervention that improves symptoms (whether or not that is via the placebo effect) and does not cause side-effects, is a doctor doing the right thing in opposing it?

Some argue a hostile approach might actually do more harm.

It’s unlikely to deter the patient from using their beloved therapy, but it might stop them from telling their doctor such things in future, increasing the risk of therapy interactions and other problems.

Jane McCredie is a Sydney-based science and medicine writer. She has worked for Melbourne’s The Age and contributed to publications including the BMJ, The Australian and the Sydney Morning Herald. She is also a former news and features editor with Australian Doctor magazine. Her book, The sex factory, on the science of sex and gender will be published by UNSW Press later this year.

Posted 30 August 2010

5 thoughts on “Jane McCredie: Acupuncture a placebo? Who cares?

  1. Tony says:

    The claimed benefits of acupuncture or sham acupuncture notwithstanding, the shortage of GPs and the often long waiting time to gain an appointment suggests that such procedures should be referred to the local acupuncturist who will be able to give the same non-medical care as the over-busy GP with medical skills that are in such short supply. Painting might be recommended as excellent therapy for an anxious and distressed patient but the GP would probably draw the line at giving art lessons; surely preferring rather to suggest that the patient seek out the local art club where it is much more likely that helpful therapy will be gained from a tutor who has not needed to spend many years in gaining scientific expertise and whose art skills will be far more likely to be superior anyway.

  2. Simon A Collins says:

    If a placebo does more harm than good, and is not unethically administered, then I am not opposed to it. It would be wrong for people to profit excessively from an intervention no better than placebo (including arthroscopies if, in a given setting they are of no benefit compared with placebo), but if it makes people better without making them worse: great!

    When are GPs going to start offering sham acupuncture and what should it be called? Presumably ‘sham’ takes away some of the placebo effect!

  3. Bruni Brewin says:

    It can be argued that a lot of ‘evidence based research’ is also due to the placebo effect.

    How many times recently have we heard of pharmaceutical trials that showed the ‘sugar pill’ gave the same desired results as the actual drug. It even shows that a patient can get the nosebo effect, i.e. being told this trial medication may cause anxiety and/or nausea, the patient taking the placebo pill believing it to be the real thing, experiences anxiety and/or nausea.

    It amazes me how many doctors put people on Champix or Patches, prior to seeing if hypnosis works. Most people who attend my practice give up smoking in a 1.1/2 hour session. Shouldn’t that be a first course of action and then if it doesn’t work, revert to medication that for some may cause unpleasant side effects?

    I agree, placebo or not – who cares? We are here to help people and should do what is best for the patient.

  4. pique4 says:

    Sham arthroscopies of the knee give better functional results than real arthroscopies of the knee in ostearthritis of the knee. Should arthroscopies of the knee not be funded under Medicare? Of course not! Reductio ad absurdum is the refuge of the sensationalists. It is actually very difficult to do “sham” acupuncture. It is like trying to be “a little bit pregnant”. Thus sham acupuncture has a statistical effect in clinical trials. In “real world” medical acupuncture, where the patient’s out of pocket expense is greater than the tiny Medicare rebate, “real” medical acupuncture has to be way more effective with an NNT as close to 1 as your skill and patient mix allows otherwise you will go out of business. (1 failed acupuncture treatment course generates 11 negative patient contacts and 1 successful acupuncture treatment course generates 4 positive patient contacts)

  5. Stewie Griffin says:

    Jane states “It may be counterintuitive, but, if a patient can afford an intervention that improves symptoms (whether or not that is via the placebo effect) and does not cause side-effects, is a doctor doing the right thing in opposing it?”

    To claim that acupuncture has no side-effects is incorrect. Common (but generally non-serious) side effects of acupuncture include pain, bleeding, bruising, headaches, tiredness (1,2).

    Rare (but serious) side effects of acupuncture include septicaemia, abscess, septic arthritis, Hepatitis B, Hepatitis C,pneumothorax, needle fragmentation, endocarditis, Mycobacterial infection etc. (there are plenty more rare side effects) (1,2,3).

    As such, it is simply unethical to recommend acupuncture at this stage, even if it does offer an excellent placebo, as at the very least, there is a safer alternative placebo (non-penetrative sham acupuncture).

    Thus, the only ethical choice when discussing treatment options with patients is indeed to recommend against trying acupuncture, and to inform the patient that if they wish for the benefits of a placebo, they may opt to try non-penetrative sham acupuncture instead.

    1) Ernst and White (2001). “Prospective studies of the safety of acupuncture: a systematic review”. The American Journal of Medicine. 110, 6. 481-485

    2) Chung, et al (2003) “Adverse effects of acupuncture:
    Which are clinically significant?”. Canadian Family Physician. 49. 985-989

    3) Woo, et al (2010). “Acupuncture transmitted infections”. British Medical Journal. 340. 1268.

Leave a Reply

Your email address will not be published. Required fields are marked *