Myofascial Trigger Point–Augmented Acupuncture vs Conventional Acupuncture for Early- to Mid-Stage Knee Osteoarthritis: A Randomized Clinical Trial [Letter]

Dear editor

We read with interest the randomized clinical trial by Zhang et al published in the Journal of Pain Research.¹ The study compared conventional acupuncture with the same regimen augmented by individualized myofascial trigger point (MTrP) needling. Although the prespecified primary outcome did not differ significantly between groups, some exploratory findings suggested potential benefit. We would like to raise three points that may help clarify the interpretation of these results.

The Comparability of Baseline MTrP Characteristics Between Groups Remains Unclear

Both groups received 10 sessions of conventional acupuncture. At each session, the augmented group received one additional needle insertion at each of three individualized MTrP sites selected before treatment. Previous research found that MTrP prevalence varied across examined muscles in patients with mild-to-moderate painful knee osteoarthritis.² In a randomized trial by Ma et al, participants were required to have at least one active or latent MTrP.³ The present trial did not require an MTrP for enrollment and did not report baseline MTrP characteristics for both groups. We suggest clarifying whether every participant in the augmented group had three sites meeting the prespecified criteria, how participants with fewer than three eligible sites were managed, and whether MTrP number, muscle distribution, and tenderness grade were recorded using the same criteria in both groups. This information would help define the population to which the findings apply.

Some Gait Outcomes Require Clarification

In Table 1, “single-leg support time” and “swing phase” sum to exactly 100% in both groups. In Table 3, their changes are exact opposites at every time point. Under conventional gait definitions, the stance and swing phases of the same limb together constitute a full gait cycle. Single-limb support is only part of stance, which also includes double-limb support.⁴ We therefore suggest clarifying the operational definition of “single-leg support time” and checking the variable names and device output fields used in the tables. Participants also walked at a self-selected pace, and differences in walking speed may affect other gait parameters.⁵ These results should therefore be interpreted with caution.

The Sample-Size Calculation Does Not Appear to Align with the Analysis of the Primary Outcome

The published protocol used a one-sided critical value to estimate a requirement of 48 participants per group.⁶ The results paper compared the primary outcome using the Pearson χ2-test and explicitly stated that all tests were two-sided. Methodological literature indicates that, with other assumptions held constant, two-sided testing generally requires a larger sample than one-sided testing.⁷ We suggest that the authors clarify the basis for the sample size and recalculate it according to the test used in the results paper.

We thank the authors for conducting this study. Clarifying the participants’ MTrP characteristics, the definitions of the gait measures, and the sample-size calculation would help readers interpret the clinical meaning of the findings and inform future trials.

Disclosure

The author has no financial or personal relationships to report in this communication.