As I am not a clinician, I will only comment on the methodology of this meta-analysis.

Three issues need some clarifications:

1. Some of the included studies use data from the same national cohorts. Please clarify whether participants in those studies were repeated counted. If this is indeed the case, some adjustment to the weighting of those studies needs to be considered. Otherwise, those studies would be given greater weight than they deserve.

2. Please give more details about how the absolute risk reduction (ARR) was obtained. Because the included studies are not randomised controlled trials, I suppose the adjusted risk ratios are used in the meta-analysis.

When we calculate the absolute risk reduction, we need to have risks of both groups, which can be directly obtained from RCTs. However, for observational studies, some assumptions need to be made.

1. Since this meta-analysis aims to see whether real-world evidence is consistent with results from RCTs, I think a brief summary of results from RCTs or meta-analyses of RCTs would be helpful.