Dear Editor, We have read the article entitled ‘Needling after trabeculectomy - does augmentation by anti-metabolites provide better outcomes and is Mitomycin C better than 5-Fluoruracil? A systematic review with network meta-analyses’ (Halili et al. 2020) with great interest. The authors had concluded no significant difference in the efficacy of using either mitomycin C (MMC), 5-fluorouracil (5-Fu) or no antimetabolites for bleb needling after trabeculectomy, which most doctors may disagree according to the clinical practice. However, there are a few points that need to be clarified before we can state that MMC or 5-Fu application does not outweigh those without antimetabolite usage. One included article retrospectively compared 26 eyes that underwent needling revisions for trabeculectomy failure (Fagerli, et al. 2003). On the one hand, as the individual data presented in this article, four patients had multiple times of needling after the original trabeculectomy. They received the injection of MMC and 5-Fu subsequently. Instead of excluding these cases or analysing them separately, they had abruptly allocated these cases to the MMC group with the presumption that MMC had a more potent effect than 5-Fu. For the current meta-analysis, the inclusion of these inaccurate data reduced the credibility of comparison. On the other hand, the no antimetabolite group in this article had only three cases. In the current meta-analysis, the authors directly compared 13 cases in the MMC group and 10 cases in the 5-Fu group with the 3 cases without antimetabolite application, as Fagerli et al. did in his article. Comparison between small sample sizes data may lead to considerable bias, to a certain degree. As the random effect model was used, it increased the weight of studies with small sample sizes, which magnified its influence on the meta-analysis results. Therefore, ambiguous application of medication and small sample may lower the value of comparison with the inclusion of this article. Another included research focused on topical application of MMC with microsponge before needling revision (Ghoneim & Abd El Hameed 2011). Compared with no antimetabolite group, topical use of MMC brought about significant efficacy in bleb needling. Based on the study design, although the concentration of MMC can be predetermined, the exposure time will be short as irrigation with the balanced salt solution before needling will wash out the antimetabolites. The doses of MMC were therefore inaccurate, as well. In comparison, for the other articles included, antimetabolite augmentation was accomplished by subconjunctival injection of fixed doses within the study. The duration of the antimetabolic effect was prolonged. Though mentioned in the meta-analysis, such difference in drug administration may introduce inaccuracy in assessing MMC’s antifibrotic impact. Selection bias was seen in the included studies. In retrospective studies, patients were selected for antimetabolite injection based on the severity of bleb failure and the judgement from glaucoma specialists (Fagerli et al. 2003; Kim et al. 2019). Under this consideration, those tend to suffer needling failure may be alleviated by antimetabolites and reach similar IOP control and success rate as those only receive needling and have lower chances of failure. The selection bias may hide the potential difference. For the current meta-analysis, all the studies included for the direct comparison of MMC/5-Fu with no antimetabolites were not robust enough to come up with the conclusion. For this reason, we cannot reasonably compare the application of antimetabolites based on the included studies. To sum up, we thought it with low credibility for the comparison of with/without antimetabolites in the current meta-analysis. In agreement with the authors, we call for more extensive randomized studies on this topic.