A critical review here. False positives are a big issue for this type of test.
Oxford’s Blavatnik School of Government is tracking government responses to Covid-19 on a by country basis. The data is downloadable in several formats and there is a working paper describing how it is collected and what it means.
Bendavid et al report the results of a “representative sample” recruited from Facebook of anti-body testing.
Some interesting back of the envelope calculations by Kevin Drum.
Help provide social distancing information for US communities.
Jakub Steiner draws our attention to an interview with a US expert about plans there.
Some information about who is doing the testing and where although they apparently aren’t yet done. This is 100% sample of an antibody test in San Miguel County, Colorado.
The European bioinformatics institute is putting together a platform for sharing of medical data about Covid-19, some of which is already available.
Andrea Galeotti, Sebastian Hohmann and Paolo Surico have a project aimed at estimating Covid-19 mortality using excess mortality methods. They employing official statistics on total deaths in 2020 and their growth rates in the previous years over the same period. They aim to constantly update the analysis as new data becomes available. They also aim to add robustness checks and alternative methods to accurately determine Covid-19 deaths. Hence, this article will be updated as often as feasible.
They would like to replicate the analysis for as many countries as possible. If you have data or you know where we can access data in a country, please share them. They are happy to replicate the analysis, share the results with you, and update the analysis so that other people can access this information.
If you have the data and you want to perform the analysis yourself, in what follows you will find all material you need. However, we would appreciate if you share the results, with a summary of how you conducted the analysis. In particular, with explicit reference to changes in the codes or methodology. This is to assure that results are comparable across countries.
The article can be found here
The codes and documentation are here