Dear Editor,

We are writing to provide a thoughtful critique of a recent article published in your esteemed journal, titled “Pilot study on efficacy and patient satisfaction with 1,450-nm laser boosting treatment after a thread lift procedure” [1]. We appreciate the efforts of the authors in comparing 2 groups of data and their inclusion of graphs and raw numerical data; however, we would like to highlight our concerns regarding the statistical analyses conducted in the study.

In Table 2, the authors compared the pre- and post-treatment groups according to the score system, utilized the Wilcoxon signed-rank test for the analysis, and provided p-values. However, we observed that no other statistical analyses were conducted, even though 8 other sets of data needed comparison. Statistical analysis plays a critical role in research; therefore, we believe the credibility of the article would be greatly enhanced if the authors could include the results of further statistical analysis to substantiate their conclusions.

Our team also attempted to analyze the data provided in Fig. 2, 3 using the Wilcoxon signed-rank test and encountered a problem. Using a p-value <0.05, the minimal sample size required was at least 7 samples with a critical value of 2. For a sample size below 6 samples, the critical value will be zero, making it impossible to reject the null hypothesis of having no difference between two groups. With a sample size below 5, achieving a p-value <0.05 is not feasible. For the Wilcoxon signed-rank test, we can ignore cases where the difference is zero, while for all other cases, we can assign their relative rank. Therefore, we suggest that the authors consider enrolling more cases to achieve the minimal number of cases required to achieve statistically significant results, thereby enhancing the credibility of this paper.

Furthermore, the conclusion that patients had less pain and discomfort after a thread lift (TO) and laser treatment than after a TO only appears counterintuitive. Our team could not find any explanation in the discussion section regarding the mechanism or rationale behind this observation. We propose that this may be due to the placebo effect, where the action of the medical procedure itself may have contributed to pain and discomfort relief on the treated side.

Additionally, Fig. 2A provided only the average scores between the TO and thread lift and laser treatment (TL) sides. While it showed that, on average, the TL side had a faster reduction of pain and discomfort than the TO side, it is challenging to conclude that the results are statistically significant. In Fig. 2B, C, the use of the difference in scores between the TO and TL sides was strange and made it hard to reach a conclusion. We suggest that the authors directly use the visual analog scale scores instead of the difference between the scores (Fig. 2B is the score at postoperative day 0 minus the score at postoperative day 3; Fig. 2C is the score at postoperative day 0 minus the score at postoperative day 7). Furthermore, there appear to be printing mistakes in Fig. 2B, C, as all the scores were postoperative day 5 minus postoperative day 3.

In summary, we commend the authors for presenting a very well-written paper; our team enjoyed reading it.

Sincerely,

Dr Lee Kar Wai Alvin, Dr Chan Kwin Wah,

Dr Lee Cheuk Hung

Ever Keen Medical Centre, Hong Kong

The authors have nothing to disclose.