We thank Drs Yan, Wang, and Zhou for their interest in our study (Yang X, Li C, Bai D, Su N, Chen T, Xu Y, Xianglong H. Treatment effectiveness of Fränkel function regulator on the Class III malocclusion: A systematic review and meta-analysis. Am J Orthod Dentofacial Orthop 2014;146:143-54). They thought that there was a mistake about including a trial with a small sample conducted by Waheed-Ul-Hameed, but we must point out that it was not.
First, the purpose of a meta-analysis is to increase the power of detecting a real effect as statistically significant by combining several small samples. If a trial was excluded because of its small sample size, how could a meta-analysis be made? In addition, the inclusion process was conducted according to previous protocol, which was based on the instructions in the Cochrane Handbook , and the sample size was not in the inclusion criteria.
Second, the authors considered that the trial should not be assessed as “high quality” because of the small sample size. Actually, we had already presented the items and criteria for quality assessment with the Newcastle-Ottawa scale for cohort studies in Table I, and sample size was not included in the quality assessment system.
Third, the authors said that too few participants would certainly cause the probability of type II error (risk of false-negative) increasing. However, too few participants could cause not only type II error but also type I error (risk of false-positive). Definitely, the chance of type I and type II errors cannot be avoided but can be controlled by regulating the value of α and β statistically, and the increase of causing type II error would decrease the probability of type I error. Most importantly, what we have done was a systematic review of interventions, but not diagnostic accuracy; therefore, the risks of false-negative and false-positive should not be overestimated.
Fourth, the authors mentioned that there was a wide confidence interval of the trial conducted by Waheed-Ul-Hameed. Again, as we have already presented clearly in Table IV about the evaluation system, the confidence interval was assessed in the GRADE analysis, and the quality of evidence was evaluated by the accumulation of several different items, including the confidence interval in the GRADE system.
Finally, the inclusion process and quality assessment were conducted by 2 authors independently, and any disagreement between them was discussed with the third author for consensus. The process met the standard of inclusion criteria according to the Cochrane Handbook , and the kappa score was used to measure the agreement.