To me, the most pressing question would be: what's my risk of dying if I catch it. The UK data has probs with that question because (a) the infection rate for COVID-19 is pretty unclear but probably at least 10X the detected & reported cases and (b) it seems the fatality data isn't complete either.
FWIW, the Oz data doesn't have the same problems, and the infection has pretty much run its initial course by now with a handful of new cases per day and only a few still in hospital. In Oz, you can be pretty certain of these things:
- Infection rate pretty much the same for diabetics as non-diabetics. (About 5.2% of COVID-19 cases are diabetics, the same as diabetes prevalance). No surprise there.
- 6.9% of diabetic COVID-19 cases have died, versus 1.4% for the general COVID-19 population. So as a whole, diabetics 4.9X more likely to die of COVID-19 than the general pop.
In the UK, working from the first paper, the figure is 6.4X but that only includes deaths in hospital and there are other reasons as well to think the data is probably incomplete. So the real figure may well be closer to the Oz 4.9X.
Unfortunately I haven't seen any Oz analysis looking at separating out diabetes from type, age, HbA1c, BMI, CV issues etc etc. Hopefully somebody gets on to that soon - it's not rocket science if you have access to the complete records - but maybe the numbers are too small for good confidence intervals; only ~100 deaths ...
Just another thing about the NHS England papers - it surprised me that they didn't look at PAD/CAD as comorbidities, as far as I could see.