Inspired by Wang et al 2026.[1]
EA – electroacupuncture
OA – osteoarthritis
RCT – randomised controlled trial
MA – manual acupuncture
DSG – diclofenac sodium gel
MCPjs – metacarpophalangeal joints
VAS – visual analogue scale
MCID – minimum clinically important difference
OTC – over the counter (used for medications available to buy without a prescription)
QDS – quarter die sumendum (“to be taken 4 times a day”)– key to acronyms
This is a moderately large (n=120) comparative RCT of EA in hand OA compared with topical diclofenac sodium gel. It was published earlier this month in the journal Mayo Clinic Proceedings (IF 6.5), which has been published continuously since 1926 and is sponsored by Mayo Clinic – ‘the largest integrated, not-for-profit medical group in the world’ (according to its own website).
This could be the first trial using EA in hand OA, although it is not the first RCT. I highlighted one from the UK back in 2020 (see CMCj OA and PHN 2020).[2] That one was sham controlled and probably under powered (n=70).
In the acupuncture group, MA and EA was applied 3 times a week for 4 weeks. In the DSG group, a 4cm strip of gel was applied 4 times a day for 4 weeks targeting affected joints.
The points used in the acupuncture group were Ashi points (tender points close to affected joints), Baxie points (4 points in the web spaces between the MCPjs), SI3, and TE5.
The text of the methods suggests that EA was applied at all the listed points, but then specifically describes Baxie 1 being connected to Baxie 4. There is a diagram in the Supplement apparently, but I cannot access that at present, so I have asked the boss (Zhishun Lui) who is a familiar name on this blog.
The needles used were 0.30x40mm, which is quite thick for points in the hand and wrist.
EA was applied at 10Hz for 30 minutes and an intensity of 0.5 to 2mA.
The primary outcome was the change from baseline in the average overall finger joint pain intensity over the previous 48 hours as assessed on a 0 to 100mm VAS at the 4-week point (end of the treatment phase). Outcomes were also measured at 8 and 16 weeks.
The MCID on the VAS pain was taken to be 15mm.
The majority of patients had OA in both hands, and their mean age was 62. The baseline VAS was ~60 and this dropped to ~29 in the EA group and ~43 in the DSG group at 4 weeks.
The paper presents the absolute values of the VAS at week 4 under the heading primary outcome in their Table 2, which I prefer to change values, although the primary outcome in the methods specifically refers to change from baseline. The differences between the groups are very highly significant no matter whether you compare absolute or change values, so it doesn’t really matter. However, the difference between change scores exceeds the MCID and the difference between the absolute values at 4 weeks does not.
Readers will know that I am firmly against the application of MCID to any sort of group mean difference. MCID can only be relevant when applied to the change from baseline in a single individual or group receiving the same intervention.
This is an open study with a subjective outcome, so it cannot account for differences in non-specific effects; however, from a pragmatic perspective, it is useful to be able to compare the outcomes between EA and one of the most commonly used conventional treatments (DSG or equivalent is first line treatment for hand OA in the UK and US). The authors suggest, on the basis of their results, that EA should be considered a reasonable alternative treatment, but 6 hours of treatment by a clinician in a hospital is clearly a lot more expensive than the self-application of an off-patent gel. Based on current UK prices of OTC DSG, I estimate that the 224g required for 4 weeks of qds application would cost ~£40. Six hours of outpatient treatment in a hospital is likely to be 15 to 20 times more costly, and that does not include other societal costs in terms of travel and time off work.
So, whilst this is a useful comparative effectiveness RCT, we still need a cost utility analysis from the UK NHS perspective. To be candid, considering the needles used, I would also recommend some form of feasibility assessment in the West. I would guess that the dropout rate in the EA group would be over 20% in the UK, whereas it was under 7% in this trial based in Beijing.
References
1 Wang W, Kang J, Shi J, et al. Efficacy of Electroacupuncture Compared With Topical Diclofenac Sodium Gel for Patients With Hand Osteoarthritis: A Randomized Controlled Trial. Mayo Clin Proc. 2026;S0025-6196(25)34683-5. doi: 10.1016/j.mayocp.2025.12.018
2 Barnard A, Jansen V, Swindells MG, et al. A randomized controlled trial of real versus sham acupuncture for basal thumb joint arthritis. J Hand Surg Eur Vol. 2020;45:488–94. doi: 10.1177/1753193420911326
The BMAS Blog 
You must be logged in to post a comment.