Transcutaneous Electric Nerve Stimulation for Analgesia During Outpatient Endometrial Biopsy

We are writing to bring to your attention a potential oversight in the statistical analysis in the study by Wu et al1 in the January 2025 issue. Their randomized controlled trial assessed the efficacy of transcutaneous electric nerve stimulation (TENS) for pain relief during outpatient endometrial biopsy. The results of their study confirmed that TENS did not significantly reduce pain during endometrial biopsy, but it was associated with high patient satisfaction and minimal side effects. We would like to offer suggestions to enhance the statistical analysis and interpretation. In the Methods section, the authors describe that visual analog scale pain scores were collected at speculum placement, tenaculum placement, 5 minutes after biopsy, and 15 minutes after biopsy. The repeated measurements of the same outcome event at each time point were not independent but correlated. One observation value would be affected by another observation value. Repeated measures analysis of variance (ANOVA) takes into account the time effect in addition to the main factor (active TENS or placebo TENS). It would allow the authors to compare whether there was an interaction effect between the treatment of the main factor (active TENS or placebo TENS) and the time factor. However, repeated measures ANOVA was not mentioned in the article. Given that the study involved the same patients being measured on multiple occasions, the absence of this analysis could undermine the validity of the conclusions drawn. The use of repeated measures ANOVA is crucial in such cases, because it accounts for the correlation between the repeated measures, providing a more accurate assessment of within-patient effects.2 The failure to conduct this analysis may lead to an underestimation or overestimation of the treatment effects, which could have significant implications for the interpretation of the results. We suggest that the authors consider conducting repeated measures ANOVA to supplement their findings or, if possible, provide an explanation for not including this analysis in their original study. We appreciate the authors’ contribution to the field, and we believe that their work could be further strengthened by addressing this statistical concern.