Needle (or Laser) the Knee for Relief from Chronic Osteoarthritis Pain?

By David Kiefer, MD

Synopsis: In people with chronic knee pain from osteoarthritis, needle and laser acupuncture fared no better than sham laser acupuncture for a variety of parameters at 12 weeks and 1 year.

Source: Hinman RS, et al. Acupuncture for chronic knee pain: A randomized clinical trial. JAMA 2014;312:1313-1322.

This randomized clinical trial was designed to evaluate the immediate and delayed efficacy of acupuncture for knee osteoarthritis. The authors justify their methodology based on cited prior work, including a meta-analysis supporting acupuncture for osteoarthritis pain and the fact that laser acupuncture (a low-intensity laser) may be more effective on myofascial points than needle acupuncture. The structure of this trial was meant to improve on past methodology, most notably by comparing needle and laser acupuncture and by having the participants’ agreement to have acupuncture occur after randomization, simultaneously making the control group unaware of the clinical trial. The researchers hoped that the agreement-after-randomization approach would improve on issues related to recruitment (only people with positive attitudes toward treatment participate) and outcomes (expectations can influence results) when the control group is aware that the other group was receiving acupuncture.

The study’s randomization involved 282 adults in Australia with inclusion criteria (see Table 1) sufficient for a diagnosis of osteoarthritis as per guidelines from the United Kingdom (National Institute for Health and Care Excellence). Of the total study participants, 71 were randomized to no treatment (the control group), 70 to needle acupuncture (13 declined to actually receive the treatment), 71 to laser acupuncture (12 declined treatment), and 70 to sham laser acupuncture (nine declined treatment). An intention-to-treat analysis was used to analyze the data, so even those individuals who declined treatment were included in that group’s analysis as if they had received the relative treatment.

The needle acupuncture used was a combination of Western and Chinese style, delivered by eight family physicians with acupuncture certification (Australian Medical Acupuncture College). The acupuncture sessions lasted 20 minutes and were delivered once or twice weekly for a total of 8-12 times. The acupuncturists treated the patients "according to usual practice," using standardized points both near and distal to the affected knee. The acupuncturist could also choose "other points" at their discretion, a balance between trying to standardize treatment for research and clinical purposes, and a more individualized approach that may be more typical for traditional Chinese medicine (TCM) practitioners. The initial treatment could use six points (four on the affected limb and two others); subsequent sessions could vary from that as per the acupuncturist’s discretion. For the laser acupuncture, an Acupak laser machine or a red non-laser light were shined on similar points to those used for needle acupuncture. It was previously established that it is impossible for the patient to differentiate between true and sham laser treatments.

The primary outcomes were two self-reported pain and function measures for osteoarthritis. First, using a 0-10 scale, average knee pain over the previous week was recorded. Second, the Western and McMaster Universities Osteoarthritis Index (WOMAC; 0-68, higher numbers indicating worse function) was used to measure physical function. Numerous secondary outcomes were recorded, including average knee pain on walking, average daily activity restriction, WOMAC pain subscale, health-related quality of life, physical and mental component summary scores, and global change in pain and function (5-item ordinal scale). All of these parameters were measured at baseline, 12 weeks, and 1 year.

The baseline demographics, symptom duration, medication use, and past osteoarthritis treatment type were similar across the four groups, except for the fact that the sham laser acupuncture group had slightly more people with symptoms longer than 10 years’ duration. There are many permutations of the data analysis. For example, Table 2 shows the three intervention groups at 12 weeks’ outcomes as compared to the control group. Needle acupuncture seemed to fare better than laser acupuncture or sham laser, but, interestingly, when these data were further analyzed and comparisons made between the three intervention groups, the needle acupuncture results were not statistically different than either the laser acupuncture or the sham laser group; presumably the lack of a difference from the sham laser group was because the benefits compared to the control group were, according to the authors, of a "clinically unimportant magnitude" or of a "clinically irrelevant magnitude." Furthermore, the benefits ("better") in Table 2 all disappeared at the 1-year mark. Adverse events were described as "few, mild, and transient," the details of which were provided in an electronic table separate from the article itself.

Commentary

The potential clinical relevance of this study’s conclusions is high. First, acupuncture is commonly used and recommended as a part of integrative treatment plans for several medical conditions, most notably postoperative nausea¹ and low back pain.² Second, pain is a common complaint for which people turn to integrative medicine, including acupuncture.³ The support or refutation of acupuncture’s use for chronic knee pain from osteoarthritis would help to appropriately allocate people’s time and health care dollars in the pursuit of symptomatic and functional improvement.

The authors of this study draw parallels between their results and those of other clinical trials; of note, most other clinical trials have compared needle acupuncture with sham needle (not laser) acupuncture. One difference, though, is a lack of statistically discernible difference when true acupuncture is compared to sham acupuncture, which they attribute to their relatively small numbers of study participants. That said, they still int to the fact that their findings, or findings that might have emanated from larger trials, were minimally clinically relevant. Their methodological improvements speak in favor of the believability of these results, the significant numbers of participants declining treatment, a criticism. True, the statistics were clear, the intention-to-treat analysis sound, but those dropouts are hard to ignore (even the authors brought it up). Another concern is the length of the visit (20 minutes) and the number of visits (8-12) in this study protocol. An enhanced clinician visit during acupuncture therapy in a seminal paper yielded profound positive effects on irritable bowel syndrome,⁴ but those visits were longer than 20 minutes. That same study showed that curt office visits are little better than no office visit. Perhaps these results could have been augmented with a little more QT (quality time) with the acupuncturist. And, with respect to number of visits, acupuncture will likely not bring about healing immediately, especially in cases of chronic conditions. The visit number in this protocol may be on the low side for chronic knee pain.

It would have been very interesting to see a sham needle acupuncture arm in the context of these results. Some studies on the use of osteopathic manipulation for pain find results, even for sham maneuvers, likely the result of a patient response to physical contact. Laser acupuncture lacks that aspect of the patient-provider interaction, mediated instead through a machine and lights (I can’t resist referencing the imaging of the Wizard of Oz and his smoke and light display, not meaning, of course, to minimize the fact that laser acupuncture has been shown to be efficacious in some studies), perhaps explaining why, in Table 2, only one variable was affected by laser therapy.

When considering the efficacy and clinical magnitude of the results, the authors strongly vote against acupuncture as a therapeutic option for chronic knee pain from osteoarthritis. A more balanced view of these results would also take into consideration the risks of this therapy (low, other than time and cost) and the other options (and their risks and benefits) for this condition. In addition, an increase in clinical time with the acupuncturist, as discussed above, may have fortified the results presented here. So, in conclusion, we have a safe treatment option, which may or may not be efficacious for chronic knee pain from osteoarthritis, leading me to advocate for research that clarifies and addresses this study’s shortcomings. In the meantime, I would consider needle acupuncture as an option for people with knee pain who are acupuncture willing or motivated.