Dear Sir,

We have read and appreciate your commentsReference Chandelia ¹ regarding the article entitled “A meta-analysis of re-treatment for intravenous immunoglobulin-resistant Kawasaki disease”. Regarding the issue that the potential risk of publication bias might possibly occur, especially towards positive outcomes, a funnel plot would help out with the analysis of the bias. The relatively small sample number of the selected publications would likely result in publication bias as discussed in the manuscript.Reference Yang, Liu and Huang ² Indeed, although over the past 10 years the incidence of Kawasaki disease as well as coronary artery lesions have significantly increased, there were a limited number of randomised-controlled trial studies investigating the re-treatment for children with intravenous immunoglobulin-resistant Kawasaki disease. Whether glucocorticoids should be used routinely for patients with intravenous immunoglobulin-resistant Kawasaki disease is still in need of studies with a large sample size.

We used fixed effects model in this meta-analysis, where the possible moderators such as the study design characteristics of the original articles might affect the results. Statistical tests showed that there was heterogeneity among the four selected articles; however, the heterogeneity made out by the statistical test could not explain the nature of heterogeneity, whether it was clinical heterogeneity or statistical heterogeneity. Random effects model would make the analysed results applicable to a larger set of studies. More rigorous original studies with a large sample size would definitely help us with the analysis.