Reply to Q. Wu et al.

Abstract

We thank the reviewers for bringing this error in methods to our attention. We appreciate that double counting of participants in the control arm of trials with multiple intervention arms is not recommended. We have re‐analyzed our data and found very minor changes to the estimates presented in the abstract and results section. Importantly, these changes have not altered the conclusions of our paper. A supplementary corrected file has been provided that details the changes in our analyses and the revised Figures 2A–D and 3A–D.