We thank Dr Yamamoto for his comments. This recent article is an updated meta-analysis for our previous study.[2, 3] Due to the small number of included studies and patients, the conclusion of the former meta-analysis might not be powerful and sufficient. Given more relevant studies in this hot topic published in recent 3 years, we decided to perform an exhausted meta-analysis to obtain high-level evidence. We are now convinced that preoperative infliximab treatment does not increase the early post-operative complications in patients with ulcerative colitis. Not only because the current meta-analysis included 13 studies with 2933 patients but also since were the outcomes of the different subgroups in line with the overall conclusion.
The main limitation of this meta-analysis was the retrospective nature of all the individual studies. These observational studies varied in reporting the characteristics and outcomes. So, we could not get the complete information we wanted in each study. However, we listed them as much as we could in the supplemental tables. As mentioned in the discussion section, the heterogeneity in the meta-analysis could be partly due to the slightly different definition of infectious and non-infectious complications among the included studies. However, heterogeneity was only obvious in the analysis of infectious complications, and there was no publication bias detecting in this meta-analysis. We failed to divide infectious and non-infectious complications into single type for synthesis such as anastomotic leak, pelvic abscess and ileus etc. Early studies did not consider infliximab's durability of biological activity. In our sensitivity analysis, infliximab use within 12 weeks prior to surgery did not increase the risk of total and non-infectious post-operative complications, but decreased the infectious complication risk. It strongly supported our conclusion.
We agree that more confounding variables should be taken into account in the design of future prospective studies, and the reported outcomes should be unified standardised.