1 comment

[ 4.5 ms ] story [ 14.3 ms ] thread
Media reports over the last two months mention many risk factors for dying from Covid-19. But there is very little reliable data on their relative importance.

This new study does exactly that. It compares all UK Covid-19 hospital deaths with 17m electronic patient records held by the NHS (National Health Service).

The headline conclusions aren't surprising: risk factors are older, gender, ethnic miniorities, co-morbidities like diabetes/asthma. Smoking is a mild risk reduction.

To see how the relative factors compare, check out the graph on p12 of the PDF preprint:

* https://www.medrxiv.org/content/10.1101/2020.05.06.20092999v...

The code is on GitHub:

* https://github.com/ebmdatalab/opensafely-risk-factors-resear...