Wednesday 6 May 2020

Model on outbreak stages classification: May updates

It has been 20 days since I last posted my analysis on the pandemic using a test count vs positive rate model. Here are some quick updates, particularly from those countries that I posted last time.

For graphs below, the "new parts" are indicated by a thicker line, starting from Apr 15 up to 5/6 May.

For further reference this is my first analysis

Classification of outbreak stages by variation of test count and positive rate

*

Stage 2->3



Pretty scary, right?

Strictly speaking, the positive rate that we are using is not the positive rate respective to the group we tested. Instead it is the ratio between confirmed case and testes case on that day.

In wealthy countries, tests are supposed to be done now within hours since we do not rely on DNA tests anymore, so such statistics would somehow reflect the positive rate of the day. In less developed countries however, this is less reliable.

Theoretically, a "positive rate" of above 1, even after smoothing, is possible though not sustainable. It means some delayed tests returned positive. Apparently this is not the case for Ecuador here. In the raw stats, both the test count and the confirmed count spikes quite often -- a big spike that cannot be wiped out by taking week average.

But such fault does not change the trend after all. We can see that Panama enters stage 3 with their extended tests. Meanwhile Ecuador could be still in stage 2 considering that the last spike only occurred 1 week ago (11183 -> 22719 on 27/4), we are still experiencing the drop as the spike moves out of affection. The daily confirmed case shows no signs of improvements, so I doubt things are under control for now.

Mexico is actually another country with spikes in their raw data so that the positive rate rockets above 1. The more worrying news is, they failed to expand the testing scale to cope with the emerging cases!

Stage 3->4



Once countries enter stage 3 it is almost sure that the countries did take proper measures, so as long as they do not lift the lockdown too early things will go improve.

Italy's 50000 daily test count is enormous given its size, but also understandable given how bad the outbreak was. It is certainly working and the positive rate drops vertically. She's in the end of stage 3 and will enter stage 4 once the testing capacity is in excess.

Indonesia is seeing similar improvement. Early spikes are due to religious events, but there was never an uncontrolled outbreak. With their testing capacity gradually improving they will enter stage 4 soon, too.

Stage 4



These are models for reference -- what happens after gaining full control over the virus?

Estonia is a rare example in Europe that avoided a full-scale outbreak. She's able to slowly put out the virus with constant testing.

New Zealand, who announced an early lockdown, is close to a full elimination. She's now ready to re-open again, but the test is not going to stop anytime soon to prevent any accident.

Updates on large countries



Canada was under the risk of favoring weather for the virus, but she proved that proper antivirus measures definitely helps regardless of the environment. I would say it is stage 2 to 3.

US is definitely improving. Not necessarily nationwide but at least for most states, including New York. We expect similar development for the rest of the states, but things may change if the governor decides to lift the ban early. US is surely in stage 3.

Russia: I mentioned last time that it is weird to see the outbreak with extensive tests -- the test count is even comparable with US! But now we know why is that the case. It is the poorly treated foreign labours, similar to what happened to Singapore (but well, Singapore provided no data). Russia is in stage 1 to 2.

Japan and UK

Two countries that are not tackling the virus like the rest of the world, so I was really curious what would happen next after I wrote my last analysis. Here are the result.



UK: well, apparently their limited hospital capacity did not limit their will to test more. They drop along the "isotherm" (green line) meaning that they have a constant rate of daily confirmed case, but the outbreak potential is slowly declining because there are less hidden patients now. Unfortunately the damage has been done, and UK is one of the most severely infected country in Europe. UK is in stage 3 and it takes some time before they reach stage 4.

Japan: probably the only well-developed country that does not really like testing. Although the emergency state order is not as restrictive as in other countries, the society slowly adopted the right measure and the outbreak somehow dies down. Everything seems to work in wonder, but I would not say Japan is out of danger really...

*

So that's all for me today. I don't think another update is necessary, but we will see. With outbreaks currently occurring in South America, the globe's economic nightmare is far from over.

Ref.

[1] Total confirmed cases vs total tests conducted, retrieved 6 May 2020.

No comments:

Post a Comment