I'm revisiting my model of #COVID19 spread that I did over a month ago and see what I got right, what I got wrong and what changes I would make

For this to make sense please read this old thread if you haven't already https://twitter.com/gummibear737/status/1280547634382606339

Let's start with my initial prediction

I think it holds up pretty good

East/SE Asia is still at low levels

North East US and Western Europe have not seen a significant second wave

Rest of US/Eastern Europe has seen increases but not exponential https://twitter.com/gummibear737/status/1280547677718093829?s=20

But some Asian countries have started rising:

Japan
Philippines

Yes, but not significantly compared to EU/US

When your strategy is to crush the virus, you are really just delaying the inevitable

H/T @JordanSchachtel for this perfect illustration of this point

What about lockdowns?

People say lockdowns don't work but technically they do reduce cases

I think saying lockdowns are pointless is more accurate

Both Israel and Romania shut down early

Really made no difference - bet they wish they hadn't

Also, some countries are seeing new cases rising

Yes, and I will explain this very shortly as I adjust the model

Strangely enough, we are not seeing deaths resulting from these new upticks

So what needs to be adjusted?

1) Burnout theory should be applied to interconnected local communities such as cities+suburbs - not to states or countries
2) Must account for seasonality, but not the way you might think
3) Crushing the virus not just an Asian thing
4) Masks? Meh

It's important to understand that COVID19 mainly affects large urban areas and their interconnected suburbs

The perfect example is the NY, NJ, CT Tri-State area

These counties are all interconnected by millions of people commuting into NYC every day

The Tri-State area has 59% of the 3 state's population, but accounts for 88% of deaths

Tri-State deaths per million: 2,453

Deaths per million of the remaining 41% of the states: 516

Covid is going to have to make its way through every major urban area for the US to burnout

This is why we see cases again rising in Spain

Their deaths were mainly in the Madrid region

There are lots of major urban centers in Spain with very low infections percentages

Thankfully deaths are not rising - I can't really understand why

The brilliance of what Sweden did is they established a baseline of sacrifices they were willing to make: shut down major events +basic social distancing

Then they let the virus burnout based on what they considered acceptable lifestyle

What a beautiful Gompertz curve they have

Next is seasonality

By seasonality, I mean that Rt is increased Oct-April so you see an increase

This isn't really the case now because the virus is new and can still spread even in more difficult summer conditions because there is such a large reservoir of susceptible people

Eventually, it will probably settle into a seasonal pattern like other coronaviruses

What this means for us is that burnout achieved during the summer may not hold when winter comes and Rt of virus increases

We won't see a big second wave but we are likely to see something

Third, is the idea that only East/SE Asia can crush the virus

It seems there are western democracies that can crush the virus

Examples: Germany, Austria, Finland

It is possible to replicate the South Korean model

Germany 111 deaths/M

Not sure it's advisable to try...

A number of EU countries have managed this feat (blue)

Mostly countries with very disciplined cultures or small island nations (really helps to be an island!)

Green: burnout

Red still burning

Only a few US states have managed to do this but mostly due to low urban demographic

Finally, we get to masks

Common sense says they should help at least a little

There are studies which show this, but many show they don't work

No significant RCT has proven that masks work

EU countries have crushed COVID without masks so I will say they probably don't work

And that's it

If you enjoyed please RT the first tweet

Thanks for reading this far!

I forgot to add this excellent thread by @trvrb which explains why 20% infection + human mitigation is enough for burnout to be achieved

It's all about the Rt and getting it to or below 1 https://twitter.com/trvrb/status/1291860659118804992?s=20