COVID-19: Data, analysis, research links & related discussion

Forums › Health and fitness › COVID-19: Data, analysis, research links & related discussion

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | ... | 21

1702 posts

Uber Geek
+1 received by user: 771

Lifetime subscriber

#2460544 13-Apr-2020 14:15

Quick update + added HK

One day does not make a trend - but brief observations anyway!

AU now dropped to SK level of new cases vs NZ flattened out. AU real or just an Easter lull? NZ still trailing AU by a few days, despite their more liberal lockdown?

HK also dropped significantly yesterday (to Taiwan levels). They had a surge in cases at a similar time to NZ, AU, albeit a lower level. HK surge was also import driven. For all three, the surges at that time likely reflect the large surge in cases in the US & Europe etc.

mattwnz

20515 posts

Uber Geek
+1 received by user: 4795

#2460545 13-Apr-2020 14:20

frankv: It's concerning that S Korea and Taiwan can't eliminate the disease, even after 6 weeks of very low infection rates.

It would be interesting to see where the cases are occurring. Resthomes? Other institutions where the spread is slow?

PolicyGuy

1820 posts

Uber Geek
+1 received by user: 1769

ID Verified

Lifetime subscriber

#2460550 13-Apr-2020 14:40

Here's the latest graph from Chris McDowall | The Spinoff

Edit: spelling

tdgeek

30048 posts

Uber Geek
+1 received by user: 9455

Trusted

Lifetime subscriber

#2460560 13-Apr-2020 15:03

The update still seems to relate todays cases with yesterdays tests. Id love to know the median lag between a test and a 1pm update. IMO todays cases are from yesterdays test "results"

frankv

5705 posts

Uber Geek
+1 received by user: 3666

Lifetime subscriber

#2460620 13-Apr-2020 16:05

tdgeek:
The update still seems to relate todays cases with yesterdays tests. Id love to know the median lag between a test and a 1pm update. IMO todays cases are from yesterdays test "results"

The tables posted to the MoH website are as at 9am, so typically only show a handful of results for the current day. That's why I exclude the current day from my graphs. So, by the time we get the 1pm announcement, the numbers are already at least 8 hours old. Often the website isn't updated until mid to late afternoon.

I'd hope that there would be some automatic way that new results get added to those tables. However, based on my previous interactions with MoH IT, I think it's more likely done manually. In which case, someone at MoH arrives at work at 8am, and concatenates CSVs received in emails that have arrived overnight from the DHBs. And I'd hazard a guess that those CSVs are mostly completed the afternoon before.

So I agree... what we're getting at 1pm is pretty much yesterday's numbers.

frankv

5705 posts

Uber Geek
+1 received by user: 3666

Lifetime subscriber

#2460651 13-Apr-2020 16:37

Geekzone support

Support Geekzone with one-off or recurring donations Donate via PressPatron.

mattwnz

20515 posts

Uber Geek
+1 received by user: 4795

#2460685 13-Apr-2020 17:30

I wonder what the breakdown in cases, including new cases is, in resthomes, and other institutions?. eg where the bubble is a lot larger and can include resthome staffs own bubbles.

Apparently they also don't test people when they move between resthomes, whether or not they are showing symptoms. Resthomes are like land based cruiseships IMO, where they are filled with higher risk older people, in close proximity to one another with staff. In the US resthome virus numbers are a major issue.

Also I noticed from the facebook Q&A that the MOH did say that some tests are giving false negatives, so some of the probables are those that give false negative with the test, but the symptoms would suggest they are probable. II imagine some of those are with people recently returned from overseas, or in affected resthomes ,so they expect exposure has occurred. But my concern is how many positive cases they have missed, by the giving negative results, as the test isn't 100% accurate as they have said. It probably isn't that much of a problem yet, while we are still in level 4 lockdown, but if people are going to start mixing with people again, it could be a problem.

tdgeek

30048 posts

Uber Geek
+1 received by user: 9455

Trusted

Lifetime subscriber

#2460695 13-Apr-2020 17:48

mattwnz:

I wonder what the breakdown in cases, including new cases is, in resthomes, and other institutions?. eg where the bubble is a lot larger and can include resthome staffs own bubbles.

Apparently they also don't test people when they move between resthomes, whether or not they are showing symptoms. Resthomes are like land based cruiseships IMO, where they are filled with higher risk older people, in close proximity to one another with staff. In the US resthome virus numbers are a major issue.

Also I noticed from the facebook Q&A that the MOH did say that some tests are giving false negatives, so some of the probables are those that give false negative with the test, but the symptoms would suggest they are probable. II imagine some of those are with people recently returned from overseas, or in affected resthomes ,so they expect exposure has occurred. But my concern is how many positive cases they have missed, by the giving negative results, as the test isn't 100% accurate as they have said. It probably isn't that much of a problem yet, while we are still in level 4 lockdown, but if people are going to start mixing with people again, it could be a problem.

Thats the key, the only key. Thats where the distancing etc has to be maintained. We can't have any "yay its over, grab some beers"

tdgeek

30048 posts

Uber Geek
+1 received by user: 9455

Trusted

Lifetime subscriber

#2460729 13-Apr-2020 18:50

frankv:
tdgeek:

The update still seems to relate todays cases with yesterdays tests. Id love to know the median lag between a test and a 1pm update. IMO todays cases are from yesterdays test "results"

The tables posted to the MoH website are as at 9am, so typically only show a handful of results for the current day. That's why I exclude the current day from my graphs. So, by the time we get the 1pm announcement, the numbers are already at least 8 hours old. Often the website isn't updated until mid to late afternoon.

I'd hope that there would be some automatic way that new results get added to those tables. However, based on my previous interactions with MoH IT, I think it's more likely done manually. In which case, someone at MoH arrives at work at 8am, and concatenates CSVs received in emails that have arrived overnight from the DHBs. And I'd hazard a guess that those CSVs are mostly completed the afternoon before.

So I agree... what we're getting at 1pm is pretty much yesterday's numbers.

Just watched Dr Ashley's Q+A on Facebook. https://www.facebook.com/minhealthnz/videos/2663131057340406/

A question was, how do todays results relate to the test numbers? Typically, todays cases (Monday) are mainly Saturday tests, that get tested on Saturday and Sunday, and results on Monday. This obviously depends on where and when and what time, but broadly todays Monday cases will come from Saturday, and maybe some early Sundays tests. Id tend to look at Saturday = Monday and ignore the consistent overlap as a rough guide.When testing goes back to consistent daily numbers, you can simplify it by using Sunday=Monday, just for convenience, as long as daily tests are near enough to consistent

ANglEAUT

altered-ego
2436 posts

Uber Geek
+1 received by user: 842

Trusted

Lifetime subscriber

#2460829 13-Apr-2020 22:52

Eat your heart out: https://avatorl.org/covid-19/

Please keep this GZ community vibrant by contributing in a constructive & respectful manner.

frankv

5705 posts

Uber Geek
+1 received by user: 3666

Lifetime subscriber

#2461141 14-Apr-2020 14:53

HP

Shop now for HP laptops and other devices (affiliate link).

PolicyGuy

1820 posts

Uber Geek
+1 received by user: 1769

ID Verified

Lifetime subscriber

#2461152 14-Apr-2020 15:20

Here's the graph for Tuesday 14th April from Chris McDowall | The Spinoff

tdgeek

30048 posts

Uber Geek
+1 received by user: 9455

Trusted

Lifetime subscriber

#2461171 14-Apr-2020 15:50

PolicyGuy:

Here's the graph for Tuesday 14th April from Chris McDowall | The Spinoff

If the Active continues the trend, Active would be zero by the 22nd, the last day of Level 4. Stats are great! But why not? We are in lockdown, infected cases are busily recovering, new cases are existing situations smouldering away as transmission affected the close contacts of the existing bubbles. Elimination doesnt sound a distant unrealistic target, based on my back of an envelope numbers. If that did happen, or near enough, it would be a major global headline.

frankv

5705 posts

Uber Geek
+1 received by user: 3666

Lifetime subscriber

#2461190 14-Apr-2020 16:02

tdgeek:

PolicyGuy:

Here's the graph for Tuesday 14th April from Chris McDowall | The Spinoff

If the Active continues the trend, Active would be zero by the 22nd, the last day of Level 4. Stats are great! But why not?

Yes, but if you look at my graph, you can already see that it's beginning to flatten out.

Remember that new cases identified today or tomorrow or next week may not recover for another 10 days (or more, if hospitalised).

So elimination is I'd say at least 10 days, and more like 20 or 40 away.

tdgeek

30048 posts

Uber Geek
+1 received by user: 9455

Trusted

Lifetime subscriber

#2461213 14-Apr-2020 16:25

frankv:

Yes, but if you look at my graph, you can already see that it's beginning to flatten out.

Remember that new cases identified today or tomorrow or next week may not recover for another 10 days (or more, if hospitalised).

So elimination is I'd say at least 10 days, and more like 20 or 40 away.

Yep, thats why I added Stats are great. What should happen is that Actives will still reduce, but slower, but when we catch up to the higher case numbers, and they recover, Actives will take a dive, the flatten out again at a low level, as only the more recent low days will are left.

The unknown question is how many unknown Actives are there in bubbles, that just have a "cold". That's the next challenge.

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | ... | 21