I was impressed by this simple observation from Nobel Laureate Romer, so thought I would bodge it into a model of my own.
Suppose Omicron adds onto delta, rising 15% a day (until accumulated infections slows its spread). Like this: 1/
Suppose Omicron adds onto delta, rising 15% a day (until accumulated infections slows its spread). Like this: 1/
https://twitter.com/paulmromer/status/1471886846175191041
Now suppose it is, in terms of hospitalisations, about as dangerous as Delta. That is - about 10 days after a case number, the daily hospitalisation number is 2% of that. You would expect two curves like this 2/
Now, just take out the lag and look at the ratio of day X's hospitalistions over that *same* day's cases, you get this (which is what we are appearing to see right now - 900 hospitalisations, 90,000 cases, a drop to 1%) 3/
And you see a similar pattern, just with a higher peak, if you assume Omicron is half as likely to hospitalise a case 4/
This is why it is so hard for pols to know whether to be phlegmatic about the (so far) failure of hospitalisations to rocket. Even w/o the impossibility of predicting the case numbers. It's at least a month till the hospital numbers might peak, on these simple assumptions 5/5
• • •
Missing some Tweet in this thread? You can try to
force a refresh
