Overall Statement

The manuscript were well written and brought a novel topic. Some minor revision are required here and there, though the much of the improvement needed in methods. The author also need to add conclusion at the end of the manuscript.

Overall strength & Impact

- Strength: the study bring a novel topics and used publicly available data.
- Impacts: The study bring a new angle on the importance of BCG vaccines and the easy, affordable and safe intervention to reduce Covid-19 in the future. However, much work and studies are needed to verify and improve our understanding of this two seemingly unrelated things (BCG and Covid-19).

Specific Weakness

Major Point

- The first variable, the BCG vaccination policy treated as categorical variable, were too general (lack granularity). The latest data regarding the percentage coverage of BCG each country were probably should be sought, as it increase the detailed of effectiveness BCG policy in each country. The reason to choose covid-19 related death (second variable) were backed with adequate reason. However, the reason sourcing data from google were unexplained, this is important as the real time data of covid-19 were provided by many institution (JHU CSSE, Worldometer, WHO etc). It also may trigger any suspicion on conflict of interest with the private company that host the data.
- The study methods did not mention any specific correlation used, whether it is pearson, spearman or kendall's method. The author also did not mention variable were treated as numerical or categorical explicitly.
- Some parts of the results should be written in methods, some should reseved in discussion
- No conclusion provided in the text. Although Medrxiv did not provide specific format, most journal require conclusion written in its own section. 
- The author should explore and explain some of the study limitation and possible confounding variable.

Minor Point

- No keywords were provided, although medRxiv didnt not require specifically any of those.
- The methods section did not explain the selection or specific requirement of countries that will be included in the analysis. Instead the author explained these in the result section.
- The study data were publicly available, however due to unclear statistical method description, it is still needs additional explanation from the author to reach a point that the study fully reproducible. The author did not provide any supplementary files or explanation regarding the mentioned Matlab Scripts. We  suggest the author used a free-publicly availabel code repository (e.g Github, Gitlab, Bitbucket) to make the code available to the general public.
- Figure 1, the point (or circle) representing the upper-middle & high income country with universal BCG coverage were not labelled, thus it give additional hurdle to read the the figure. The correlation methods that derived the p value were unexplained.
- Absence of Point labelling were presence in table 2 and 3 as well.
- The scatter plot  could be improved to showed the trend of covid-19 death along the gradient of countries income and BCG policy. 
- The label for X axis and the units on X grid were too close.