This is a good example of sensational news about the brain hijacking my brain: nytimes.com/2020/06/26/opi
I read this piece and (out of concern) immediately forwarded it to a bunch of friends in the US. But then I realised: I needed to do the numbers! A thread about lessons learnt.
Conversation
The article says 'a third of hospitalised cases in Wuhan'. My brain glommed onto the 'THIRD' and glossed over the rest of the sentence. A third seems like a lot!
The article then goes on to describe all the horrible things to have happened to these people. God. Time to share!
1
Actually, there are two bits of data that I have to dig up, because the article reports:
1) a 3rd of all hospitalised cases in Wuhan
2) 85% of French ICU cases.
These numbers seem super high on their own! It was what got me to share indiscriminately! But: what's the context?
1
What is the % of hospitalisations in Wuhan?
And what is the % of ICU cases in France?
Wuhan's numbers are difficult to find, and the French numbers are also based on a model. But the goal here is not to get exact numbers, it is to get ballpark figures to use as an anchor.
1
Wuhan's numbers are underreported and bad. So I took a paper from The Lancet, which reports numbers for all of China; the % infected that are hospitalised looks to be between 1.1% and 18.4% depending on age. thelancet.com/journals/lanin
If we take 18.4%, a third of that is about 6%.
1
This stat article, on the other hand, reports 15% hospitalisation rate amongst those infected: statnews.com/2020/03/10/sim
If we assume 1/3 display neurological symptoms, that means 5%.
Ok so this is bad, but at least it's in context now. On to France.
Replying to
This paper in Science reports 3.6% of infected are hospitalised, and of those 19% go to the ICU. So, if we accept that 85% of the ICU patients have neurological symptoms, that is around 0.006% with neurological symptoms.
Of course, the wrinkle is that age matters as well.
2
1
The numbers here aren't accurate, but they give us an anchor. We have bounded this to 0.006%-6% of infected cases. This is bad, but it FEELS less bad than the original NYT article makes it out to be!
I was so alarmed by the NYT piece that I shared it to 7 different chat groups.
1
1
After pausing to do the math, I had to message all 7 groups with the numbers in context, and to apologise.
What did I learn? News pieces may report facts, and yet still make them seem a lot more scary than they actually are!
I don't know why I still bother reading news. Sigh.
3