Impact of oral anticoagulation on clinical outcomes of COVID-19: a nationwide cohort study of hospitalized patients in Germany - new article chosen as more suitable for review.

This is an interesting paper that uses a data set from a German insurance company to investigate the impact of long term oral anti-coagulation prior to hospital admission with a diagnosis of COVID-19 on outcomes. It is a resourceful use of the data set to investigate trends established in other countries during a period when much research is rapidly being produced on a topic.

The article does much well - it defines the data set, the limitations and benefits of such a cohort. Including retrospective data, limitations of the coding and lack of follow up imaging/testing. The authors attempt to mitigate errors in analysis with rigorous statistical analysis and testing. I am unable to comment further on the statistics performed by multivariate analysis as this is outside my knowledge base.

Minor areas for the authors to consider - it might be useful to have primary outcome measures re-defined in the results section. In the 3rd paragraph of the discussion the authors re-introduce the background of the paper and context in which they are working. This could be better placed in the introduction. They suggest that oral anti coagulation is commonly a long term medication however in those with a DVT or PE it is common to treat for 3-6 months and therefore this may impact further on the outcomes - they might consider analysis according to indication for anti coagulation.