1/7. So, there is a BIG paper out (not yet peer-reviewed) out looking at the factors associated with #COVID19-related hospital death. Starting with health records of 17 mil UK NHS patients, they studied >5.5K COVID deaths. Graph summarises factors associated with risk of death. 
2/7. Read the full pre-print by @bengoldacre and others here: medrxiv.org/content/10.110…
3/7. Interpreting the graph: A 'hazard ratio' of 1 (along dotted red line) has been set by a reference group under each heading e.g. never having smoked. Dots to the right of the red line are indicating higher risk than the reference group and dots to the left, lower risk.
4/7. Interpreting the graph: The error bars on either side of the dots (like this: -o-) indicate the uncertainty in the data – basically the true value could fall anywhere along the bars. More data might increase the confidence in the results over time.
5/7. If you are among those that have one or more of the factors associated with risk, please remember that you may have other factors in your favour. And... nobody can predict your personal story.
6/7. As a side note, it's good to see the inclusion of a section reporting on patient and public involvement (#PPI), even if there was none!
7/7. Side note number 2. Wow, look at the power in the huge numbers of anonymised NHS electronic health records. We really need to move to electronic records in Ireland as standard. #BigData #EHR #healthresearchmatters
