May 11 2017

Open Letter to FDA on Acupuncture

The FDA recently amended their draft proposal for: FDA Education Blueprint for Health Care Providers
Involved in the Management or Support of Patients with Pain. The draft proposal includes this statement:

“Nonpharmacologic therapies – includes psychological, physical rehabilitative, surgical approaches; and complementary therapies.”

And below further defines, ” Complementary therapies – e.g., acupuncture, chiropracty”


“HCPs should be knowledgeable about the range of available therapies, when they may be helpful, and when they should be used as part of a multidisciplinary approach to pain management.”

While the current problems with opioid overuse are significant and adequate pain management remains a challenge, introducing unscientific methods will not help health consumers and will only waste resources. Introduction of these recommendations for unscientific methods does not reflect scientific consensus, but intense lobbying by proponents who stand to financially gain from their inclusion.

Acupuncture specifically has been studied extensively, with thousands of clinical trials, and yet proponents have not been able to demonstrate that acupuncture is effective for any indication, including pain. In short, acupuncture does not work.

A 2009 systematic review of acupuncture studies for pain (Madsen et al) is typical of the research. They found:

Thirteen trials (3025 patients) involving a variety of pain conditions were eligible. The allocation of patients was adequately concealed in eight trials. The clinicians managing the acupuncture and placebo acupuncture treatments were not blinded in any of the trials. One clearly outlying trial (70 patients) was excluded. A small difference was found between acupuncture and placebo acupuncture: standardised mean difference -0.17 (95% confidence interval -0.26 to -0.08), corresponding to 4 mm (2 mm to 6 mm) on a 100 mm visual analogue scale. No statistically significant heterogeneity was present (P=0.10, I(2)=36%). A moderate difference was found between placebo acupuncture and no acupuncture: standardised mean difference -0.42 (-0.60 to -0.23). However, considerable heterogeneity (P<0.001, I(2)=66%) was also found, as large trials reported both small and large effects of placebo. No association was detected between the type of placebo acupuncture and the effect of acupuncture (P=0.60).
A small analgesic effect of acupuncture was found, which seems to lack clinical relevance and cannot be clearly distinguished from bias. Whether needling at acupuncture points, or at any site, reduces pain independently of the psychological impact of the treatment ritual is unclear.

Nothing has changed in the years since. The most rigorous trials for acupuncture, ones in which subjects are properly blinded and adequate controls are use, show no difference between verum (true) acupuncture and sham or placebo acupuncture. For example, a 2013 review of acupuncture for low back pain concluded:

“Compared with no treatment, acupuncture achieved better outcomes in terms of pain relief, disability recovery and better quality of life, but these effects were not observed when compared to sham acupuncture.”

This pattern is typical – when acupuncture is compared to no treatment, which is an unblinded comparison, there is a difference. When compared in a properly blinded fashion to placebo acupuncture, there is no difference. The only scientific interpretation of this pattern of results is that acupuncture provides no specific benefit, the results are placebo effects only.

Thousands of studies in acupuncture clearly show that acupuncture points do not exist. They have no basis is physiology or anatomy. Further, and unsurprisingly, needle location does not affect treatment outcome. In fact, needle insertion does not affect outcome, when compared to simulated insertion (such as poking the skin with tooth picks). Since the operational definition of acupuncture is inserting needles into acupuncture points, it is fair to say that acupuncture does not work.

By officially recommending acupuncture as a treatment for pain the FDA is essentially “approving” the efficacy claims for acupuncture.

Would the FDA, however, approve a drug if the most rigorous clinical trials all showed no difference in efficacy between the drug and a placebo, and only showed a difference in subjective outcome in unblinded comparisons?

Acupuncture proponents are essentially arguing that both their drug (acupuncture) and the placebo of their drug work by some unknown mechanism. This turns scientific logic on its head, however. The standard interpretation, the one on which the entire FDA drug approval process is based, is that in order to prove efficacy you must demonstrate a clinically significant difference between treatment and placebo in a properly blinded comparison.

Acupuncture has failed to demonstrate such efficacy after thousands of studies, and it therefore seems unlikely it ever will.

I urge the FDA not to abandon its scientific principles under lobbying pressure from special interest groups, or out of desperation from dealing with an intractable problem.

24 responses so far