to the editor: I was disappointed that in the case of acupuncture for chronic low back pain, the authors cited a narrow selection of the available literature: a few recent articles that suffered from being poorly interpreted or of poor quality.¹ This gives the incorrect impression that acupuncture is an effective intervention for chronic low back pain, and led the authors to ascribe an A level of evidence and recommend as a first-line therapy an intervention that has no evidence of effectiveness compared with placebo, and very real risk of harm.^1–5

The Vickers article cited by the authors is one meta-analysis of many on this subject. Another meta-analysis using similar inputs came to an opposite conclusion.^1,5 Chronic low back pain is certainly a condition ripe for demonstrating false positives because of poor research design and/or bias, given that an approximately equal initial response to any therapy has been well demonstrated.^1,2,6

I also dispute the authors' assertion that “In general, acupuncture is extremely safe, with the most common risk being transient mild discomfort.” The reality is quite different, as demonstrated in a 2011 review that failed to show any benefit from acupuncture and detailed the possible harms, including some deaths and other major adverse effects such as septic arthritis, spinal epidural abscess, retroperitoneal abscess, necrotizing fasciitis, and infectious aortic aneurysm.²

As a whole, the available evidence shows unequivocally that acupuncture is not an effective therapy for any condition.^1–5 When an intervention has no evidence of effectiveness beyond placebo, no risk of harm—however small—justifies its use or recommendation by family physicians, especially for a condition that is well known to respond equally as well to any treatment offered.^1,2,6

in reply: We thank Dr. Wright for his critique of our conclusions about the effectiveness and safety of acupuncture for low back pain.

Regarding effectiveness, it is often possible to find individual trials or systematic reviews that do not demonstrate benefit, even for well-established therapies. We based our recommendations on a preponderance of the most recent evidence, which strongly supports the effectiveness of acupuncture for chronic back pain. A recent review commissioned by the Agency for Healthcare Research and Quality on noninvasive treatments for low back pain confirmed that there is moderate evidence of effectiveness for acupuncture.¹ This level of effectiveness is equivalent to that for nonsteroidal anti-inflammatory drugs and superior to that for long-term opioid therapy. A recent review from the National Institutes of Health also included acupuncture as an effective complementary and integrative treatment for pain management.²

Regarding safety, the 2011 review that Dr. Wright cited included multiple examples of case reports of adverse effects from acupuncture. Evidence on safety needs to be based on large-scale prospective surveys, not on single case reports. In one rigorous survey, 12 prospective studies of more than 1 million acupuncture treatments showed the risk of a serious adverse event to be 0.05 per 10,000 treatments, and 0.55 per 10,000 patients.³ Systematic reviews and surveys have clarified that acupuncture is safe when performed by appropriately trained practitioners.^4,5 Every effective treatment, including acupuncture, has the potential for adverse effects. The proper approach is to evaluate the frequency of serious adverse events and balance this against the evidence of benefit. In the case of acupuncture for chronic low back pain, this risk-benefit analysis clearly favors including this approach on the list of potential therapies.