Tag Archives: virus

SARS2 and Animals

Posted by xaminmo
This CNET Article discusses research on SARS-CoV-2 in several common animals, as well as a bit of history on the virus.  The article also discusses the animal origins of this virus, and a quick summary to date. A limited number of animals have been tested for and proven to be infected by the virus.  No proof of humans catching it back from animals has happened.  Very little study has occurred on the communicablility, and it was initially thought to be no risk.
 
The jist of the research paper is that adults and juvenile cats can get it the same as humans; young cats and ferrets can get it just in the upper respiratory tract (sinuses, tonsils), but not the lower respiratory tract. Dogs can technically get it, but are not very susceptible. It does not stay in them long. Ducks, pigs, pidgeons, etc are not susceptible at all. 
 
Another group did computer modeling of 253 animals’ ACE2 receptors to see what other animals we should investigate as possible transmission vectors.  
  • Human, Flying Fox, Horseshoe Bat, Lynx, Civet, Cat, Swine, Pangolin, Cow, Buffalo, Mustela (ferrets, weasles, etc), Goats, Sheep, ACE2 were clustered with humans.
  • Mice, birds, reptiles, etc were not, and mice were proven not susceptible.
  • Civet and Bat have been implicated in SARS1 sources, and Pangolin and Bat for SARS2 sources.

     SARS2 ACE2 Phylology Chart

This chart shows the ACE2 receptors that conserve the same binding sites as humans. They suspect that 50% and above “could” harbor the virus, but that birds generally are not a reservoir for betacoronaviruses. We see in other research that dogs, at 90%, are poor carriers, and clear the virus within 4 days. Swine were not actually susceptible, and ferrets were not able to get it into their lungs. The mechanism of those differences is unknown.

SARS2 ACE2 Phylology similarities


Ivermectin kills SARS-CoV-2 in vitro

Posted by xaminmo
It looks like an anti-viral drug blocks a nuclear protein import function, which halts SARS-CoV-19 replication. In cultures, a single treatment caused a 93% reduction in 24H, and 99.98% reduction at 48H.
 
It’s part of the WHO essential drugs list, and already FDA approved. That bypasses several early and time consuming steps for releasing treatments. Next steps are finding out whether this will work in humans, and what dosage is needed.
 
Somewhere in there, I hope there is a balance between keeping people safe, and killing the virus so quickly that we don’t build any immunity.
 
https://www.sciencedirect.com/science/article/pii/S0166354220302011

But the flu…

Posted by xaminmo
Too much write-up to not share here, but someone had asked why we cared about COVID19 deaths right now, considering influenza had around 730k hospitalizations, and 61k deaths.
 
I’ll take a stab at it. The 730k and 62k look like preliminary estimates of the 2017-2018 flu season in the US. Flu season is October through February, or 5 months. To translate that into a hospital burden, we need to know that mean influenza hospitalization is 5.4 days. That totals 3,942,000 hospital days over what was the highest flu season in the last several years. That amounts to 17,600 hospital days per calendar day.
 
By comparison, the average COVID19 hospital days is 10 per patient, which is almost double what influenza needs. In the first three weeks of US COVID19 hospitalizations, we had a cumulative 400 hospitalizations. 10 days per hospitalization, so that’s 4000, over 21 days, so 190 hospital days per calendar day. Almost nothing. Who cares, right?
 
In the fourth week, we had 2200 additional hospitalizations. Ok, that’s something. 2200 times 10 is 22k, divided by 7 days (a generous suppression of the 7th peak day), and we get 3143 hospital days per calendar day. That’s something, but it’s still only a 5th of the flu.
 
In the fifth week, we had 17,200 additional hospitalizations. That puts us at 24,571 hospital days per calendar day. Again, generously ignoring that the last day in the week is about 4x the first day, we still are 50% over influenza’s peak highest peak week in the last 10 years.
 
We’re finishing up week 6 on Monday, April 6. On that day, we’ll have 48k cumulative hospitalizations, or 28,200 additional from the prior week. That’s 282,000 additional hospital days committed, or 16x influenza’s highest peak. Any errors in my math or assumptions simply go away at this point.  Even if I’m off by half or double, we have blown past the flu by a gargantual margin.  Also, we have not reached the peak yet for COVID19.
 
Since we do not have any sort of partial herd immunity, and because its incubation period is about 3x that of the flu, without quarantine or treatment to reduce the serious and mortal cases, this sweeps through the nation in such a way that it does not slow down until we have 60% immunity, a quarter of the nation is sick at one time, and about 9.8 million US citizens die, mostly over the course of 2 months. It’s just more than our system can handle.
 
Back to the death portion of the influenza equation, the deaths happen over a certain span, such as week 48 to 23 in 2016-2107, or 42 through 22 in 2017/2018. 2017-2018 was a particularly rough year, though 2018-2019 had a longer flu season, week 41 through week 37. 2019-2020 is not fully recorded yet, and will be muddied by COVID10 this year. Either way, flu season is about 32 weeks, with a peak of 12% of the deaths in one week. It’s pretty consistent that way.
 
That means the worst week in the worst seasonal influenza cycle in the last decade had around 1100 deaths per day for one week, and everything else is an pretty steep, inverted bell curve. COVID deaths happen around day 20, and detection happen around day 12. That means we have 8 days of latency in the statistics.
 
The simple and accurate way to look at that is to project the death rate ahead by 8 days, and that’s what you will actually have. Anything further, and you might have a big fall off of cases that you cannot see yet. Also, +1 day because the official numbers always report the next day, with the WHO posting at 4am CDT, and JHU posting at 8pm CDT, and some other stats posting at 4pm EDT.  Alternatively, you can look at today’s hospitalization cumulative counts, and that is what your death count will be in 8-9 days.  That’s just how is has aligned, and is not because all of these people die.  As things slow down, that shortcut will no longer apply.
 
Well, April 2 formal number was 1100 deaths just for that day. We have already reached the peak influenza deaths per day, and our hCOV19 infection rates are still growing by 15% per day, and death rates still growing by 20% per day. That’s compounding, like a bad payday loan. So, we look ahead 8 days, and that’s 48k deaths, with 10k on April 11. That is nine times the peak influenza deaths per day.
 
Also, the deaths and hospitalizations don’t just suddenly stop. It’s not 10k one day, and zero the next. We’re on board for at least 100k deaths this year from COVID19. That assumes we have no further flare-ups. No new outbreaks. No new hot zones. None of the hospitals overload and have their death rates go up by 2-4x… You know, like NY/NJ/MA area which is already at capacity, and planning for who to refuse service to because they simply do not have beds, equipment, and staff to handle.
 
Deserving of mention is ventilator days, but that’s about 10 days for both patient types. Influenza is about 60% death from ICU and COVID19 is about 86% death from ICU, but that doesn’t matter so much as the raw numbers above.
 
What does matter is that if we listened to everyone who said “it doesn’t matter”, or “we’re overreacting”, or “there’s nothing we can do”, or “the flu is worse”, then in about 4 weeks, we will have a single day that had as many new cases and as many new deaths as the entire influenza season combined. At that point, the naysayers would say “Oh golly, why didn’t THEY do something about this!“
 
So, what’s happened is that they have done something about it. They have chosen a reaction level which skirts the edge of how many people will refuse to comply, vs how many people can die without their re-election campaigns being affected. If it were up to the medical experts, then restrictions would have happened weeks earlier.  It’s not unpredictable, nor a surprise.  We knew China was not being forthcoming.  We knew this was coming.  The naysayers simply suppressed mitigation.
 
The WHO and the CDC are not able to control disaster funds on that level, nor financial packages to help prevent everyone from losing their houses in a lockdown. They also don’t have the authority. They had to wait for the governing bodies to make those decisions, and grant additional power by delegation. Some countries were faster, and better at reacting. We see those with low infection and death rates, and they are still open for business.
 
Others, who did like us, are suffering pretty heavily right now, many of them simply unable to report how many people died from the disease, because it overwhelmed their systems. We will have to count the deaths, cremations, burials, etc. after the fact, and subtract the normal amount year to year to find the excess deaths. Not every death will be from the disease. Many will be people who could not get medical care, or supplies, or suicides from the stress.
 
A common argument is “but Italy has an older population” and a variety of other issues. Yes, they do. That accounts for why their death rate is more than double what China’s was (both went into hospital overload). In the unconstrained mode, doubling happens every 2.2 days, so that is really not a good argument.
 
Some sources for various claims above:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4603361/ – Mechanical ventilation demand estimates for an influenza pandemic
 
https://www.ncbi.nlm.nih.gov/books/NBK63484/ – Length of hospital stays for the flu
 
https://www.cdc.gov/flu/about/burden/index.html – Number of deaths per year from the flu in the US
 
https://gis.cdc.gov/GRASP/Fluview/PedFluDeath.html – pediatric mortality by week, which I’m using as a proxy for all mortality.
 
If you want to know what the healthcare community is doing to prepare for and handle the hot zones that exist now, and the ones that will sprout up over the next couple of weeks, check out this site:
https://www.elsevier.com/clinical-solutions/covid-19-toolkit
 
This site is my favorite for easy to understand visualization of the spread of the disease. A straight line 45 degrees up on a log scale means it’s not getting better, and each unit up is 10x the previous unit. Horizontal means no more spread.
http://91-divoc.com/pages/covid-visualization/
 
If you want to play with the raw data yourself, you can get daily detail reports, as well as tables showing each day in a single row for confirmed cases, and confirmed deaths. Check here for JHU’s data:
https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data
 
Similar data for testing rates, hospitalizations, ICU usage, ventillator usage, etc is here:
https://github.com/COVID19Tracking/covid-tracking-data/tree/master/data
 
Unfortunately, there is no formal estimate on how many unreported cases there are out there, but it is probably somewhere between 3x and 10x the official counts. That dilutes the relative mortality rates to somewhere around 0.2% and 1.0%, compared to influenza which is somewhere around 0.05% according to the sources above.
 
EDIT: Fixed hospital-days math, and corrected spelling/grammar.

CV19 April 2

Posted by xaminmo

CV19 Pandemic Update for 2020-04-02

I did not make screenshots of my favorite graphs, but their URLs are at the bottom of this post. Deaths and confirmations log chart is very visibly bending in the right direction. We still have a ways to go.

Executive Summary:

  • Worldwide, 930,725 confirmed cases and 49,661 deaths, which is within 1% of the 1-day estimate.
  • US is 243k infected, and 5926 dead, which is also within 1% of 1-day estimates.
  • China probably has ~400k unconfirmed cases, and ~16,500 unconfirmed deaths.
  • We need our confirmed and death exponents to fall to as close to 100% as possible.
  • US confirmed exponent is stalled in the 114% range. The death date exponent has not yet fallen. I think NY and CA hot zones are affecting it.
  • TX confirmed exponent is averaging in the 115% range. The death rate exponent just fell into the 1-teens for the first time.
  • US deaths per confirmed case continues to increase, and will do so until as far as 47 days after the last new infection.
  • Many places in the US extended their stay-at-home plans to May 4.

1-day Estimate for April 2
non-China 932,543/48,756; Italy 115,572/13,925; USA 241,947/5,843; Texas 4,979/81; Maine 303/5
SD Adjusted USA 256,985/5,982; Texas 5,331/83

Actual April 2
Non-China 930,725/49,661; Italy 115,242;13,915; USA 243,453/5,926; TX 5,069/77; ME 376/7

1-Day Estimate for April 3
Non-China 1,018,824/56,704; Italy 120,107;14,719; USA 277,775/7,382; TX 5,900/90; ME 411/8
SD Adjusted USA 287,625/7,557; Texas 6,139/100

7-day Estimate for April 8
non-China 1,917,505/109,549; Italy 169,991/23,851; USA 745,176/25,239; Texas 15,874/306; Maine 656/26
SD Adjusted USA 707,819/21,874; Texas 14,357/317

Commentary:

Day to Day deaths are still climbing nationally. Some states have begun slowing, but CA and NY have a lot going on. It may take some time for them to dig out of this.

After lockdown, confirmed cases take 9-12 days to be affected, and deaths another 8-11 days after that. Anywhere that is overloaded will have a higher death rate, which may offset this prediction

As expected with the slow-down in spread, our mortality rate keeps climbing. 1.44%, 1.55%, 1.67%, 1.75%, 1.84%, 2.06%, 2.23%, 243%

Inflection point for outbreaks with an R0 of 2.2 is 60% of the population becoming immune. Divide that by whatever you think our detection rate is. I think we get about 38%, so maybe 75m people in the US showing as confirmed. In other words, we are clamping down on the spread. It is not slowing due to going through the bulk of the population. If we all started mingling again tomorrow, we would see a big surge in cases 2 weeks later.

Average is 6 days to onset and 14 days from onset to death for those who do not survive. Social distancing data lags by 3-5 days. Statistics lag by 1 day. Detection lags by 9-12 days. Impact delay of social distancing on confirmations MAY BE 9 days. The correlation is a little fuzzy, but using 9 days got the estimates within 10% over 7 days.

Some good research is going on that promises to improve mortality rates, but that will probably be more for the next batch of people. Even if it is solved now, logistics of getting it deployed are not instantaneous.

Spreadsheet is updated, and downloadable here:
https://drive.google.com/file/d/1vocCN445AZyVBBLsv0kJR8ZDP9DM0UST/view

My favorite charts:
http://91-divoc.com/pages/covid-visualization/
https://www.unacast.com/covid19/social-distancing-scoreboard

Additional Data Sources:
https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data
https://github.com/COVID19Tracking/covid-tracking-data/tree/master/data
https://ourworldindata.org/covid-testing

China Underreporting Claims:
I do not think this is willful deception, so much as, they are only reporting what they can confirm. It just happens that they can confirm only a small subset of the infections prior to March.

China cremation rates, I lost the URL for the older data, but current data is here:
http://mzj.wuhan.gov.cn/tjxx/387166.jhtml
2019    Number of cremated remains    JZH014221000    56,007
It varies by about 2500 per quarter, so roughly 5k per month in Wuhan province.

The number of urns and cremations are very high, and “the funeral homes are unable to keep up.”
https://www.bloomberg.com/news/articles/2020-03-27/stacks-of-urns-in-wuhan-prompt-new-questions-of-virus-s-toll
https://english.alarabiya.net/en/features/2020/03/31/Delivery-of-5-000-urns-undermines-China-s-coronavirus-official-death-toll-Report
http://www.asianews.it/news-en/Wuhan,-endless-queues-for-ashes-of-coronavirus-dead-cast-doubts-on-numbers-49673.html

There are multiple sources indicating people who were not checked into a hospital were not counted.
Journalistic sources include
https://www.rfa.org/english/news/china/cremate-02142020105822.html
https://www.scmp.com/news/china/society/article/3050311/its-pneumonia-everybody-china-knows-about-many-deaths-will-never
https://www.who.int/bulletin/volumes/84/3/news30306/en/

I did not save teh ones from February where the WHO felt they were underestimating because they had no way to count those otuside of the hospitals. They said 4-8 times higher rates were likely.

This exclusion has been common in other countries, including Italy, Iran, etc:
https://www.wsj.com/articles/italys-coronavirus-death-toll-is-far-higher-than-reported-11585767179
https://www.space.com/iran-coronavirus-graves-satellite-images.html

 


CV19 April First

Posted by xaminmo

CV19 Pandemic Update for 2020-04-01

Executive Summary:
* Worldwide, 850,244 confirmed cases and 43,493 deaths, which is 22% lower than predicted 7 days ago.
* China probably has ~400k unconfirmed cases, and ~16,500 unconfirmed deaths.
* US is 213k infected, and 4757 dead, matching the social-distancing estimates, and 55% under un-adjusted 7-day predictions.
* US confirmed exponent continues to fall, but death exponent has not yet fallen. Maybe by Monday if NY & WA not too overloaded.
* US deaths per confirmed case continues to increase, and will do so until as far as 47 days after the last new infection.
* Many places in the US extended their stay-at-home plans to May 4.

Numbers for 03-31 1-day projection was within 1% as usual, but on the low side like the last 5 days. Except Italy.
1-Day Projection: NON-China 768,398/38,363; IT 105,957/12,464; US 185,835/3,595; TX 3,547/55; ME 299/5
Actual Numbers: NON-China 775,208/38,798; IT 105,792/12,428; US 188,172/3,873; TX 3,809/54; ME 303/5

Numbers for 04-01 7-day projection on the distancing formula was within 10%, which is impressive.
7-Day Distancing: US 232,880/4,603; TX 5,875/72
4-day Distancing: US 237,600/4,513; TX 5,003/68
3-day Distancing: US 231,486/4,464; TX 4,761/68
2-day Distancing: US 227,620/4,475; TX 4,583/55
Actual Numbers: US 213,372/4,757; TX 4,355/66

Numbers for 04-01 7-day projection on the averaging formula are 55% low for the US, 32% low for Italy, and about 22% low for non-China. When things are as expected, it comes up within 25% for a 7-day projection. This means the US and Italy are improving, and the rest of the world as a whole is still doing poorly at containing it.
7-Day Projection: WW 1,169,645/61,852; PRC 82,134/3,338; NON 1,087,512/58,514; IT 157,313/19,560; US 477,863/6,398; TX 9,131/162
4-day Projection: WW 1,097,956/54,118; PRC 82,357/3,321; NON 1,015,599/50,796; IT 132,717/16,052; US 333,315/5,929; TX 7,263/90; ME 468/5
3-day Projection: WW 1,017,269/50,218; PRC 82,428/3,321; NON 934,841/46,897; IT 124,403/14,738; US 278,237/5,179; TX 5,427/83; ME 426/5
2-day Projection: WW 964,689/48,019; PRC 82,366/3,318; NON 882,322/44,702; IT 115,671/13,966; US 240,828/4,723; TX 4,543/69; ME 350/5
Actual Numbers: WW 932,605/46,809; PRC 82,361/3,316; NON 850,244/43,493; IT 110,574/13,155; US 213,372/4,757; TX 4,355/66; ME 303/5*
* PRC numbers are likely 1/6th of the actual due to creamatorium activity.
* Maine did not update on 4/01.

Hospitalizations were 7-day projected at 68,262, but came in at 31,142. The hospitalization chart estimated 4770 deaths and we had 4700.
7-Day US Testing: POS 134,035 NEG 728,160 PEND 100,028 HOSP 68,262 DEAD 4,770 Overrun: 04-06 335k
4-Day US Testing: POS 150,376 NEG 785,328 PEND 83,576 HOSP 64,012 DEAD 5,886 Overrun: 04-06 342k
3-Day US Testing: POS 164,152 NEG 817,199 PEND 77,376 HOSP 44,840 DEAD 4,592 Overrun: 04-09 408k
2-Day US Testing: POS 197,908 NEG 966,946 PEND 80,606 HOSP 36,789 DEAD 4,604 Overrun: 04-11 349k
ACTUAL April 1 : POS 210,770 NEG 939,190 PEND 59,687 HOSP 31,142 DEAD 4,700 Overrun: 04-14 336k

1-day Estimate for April 2
non-China 932,543/48,756; Italy 115,572/13,925; USA 241,947/5,843; Texas 4,979/81; Maine 303/5
Adjusted USA 258,180/5,970; Texas 5,400/84
Tweaked USA 256,985/5,982; Texas 5,331/83

7-day Estimate for April 8
non-China 1,917,505/109,549; Italy 169,991/23,851; USA 745,176/25,239; Texas 15,874/306; Maine 656/26
Adjusted USA 679,037/18,955; Texas 15,401/282
Tweaked USA 707,819/21,874; Texas 14,357/317

Day to Day deaths are still climbing at 126%/day. They should start falling short by Monday. After lockdown, confirmed cases take 12 days to be affected, and deaths another 10 days after that. Anywhere that is overloaded will have a higher death rate, which may offset this prediction

As expected with the slow-down in spread, our mortality rate keeps climbing. 1.44%, 1.55%, 1.67%, 1.75%, 1.84%, 2.06%, 2.23%.

Inflection point for outbreaks with an R0 of 2.2 is 60% of the population becoming immune. Divide that by whatever you think our detection rate is. I think we get about 38%, so maybe 75m people in the US showing as confirmed. In other words, we are clamping down on the spread. It is not slowing due to going through the bulk of the population. If we all started mingling again tomorrow, we would see a big surge in cases 2 weeks later.

Average is 6 days to onset and 14 days from onset to death for those who do not survive. Social distancing data lags by 3-5 days. Statistics lag by 1 day. Detection lags by 9-12 days. Impact delay of social distancing on confirmations MAY BE 9 days. The correlation is a little fuzzy, but using 9 days got the estimates within 10% over 7 days.

Some good research is going on that promises to improve mortality rates, but that will probably be more for the next batch of people. Even if it is solved now, logistics of getting it deployed are not instantaneous.

Spreadsheet is updated, and downloadable here:
https://drive.google.com/file/d/1vocCN445AZyVBBLsv0kJR8ZDP9DM0UST/view


CV19 Update Sun 03-29

Posted by xaminmo

CV19 Pandemic Update for 2020-03-29 (Sunday Night).

Most of the numebrs are slowing down vs projections. On 03-24, our growth rate dropped from 135% to 123%, and the last 2 days have been in the 1-teens. TX lags by a couple days in that regard. We are stubborn. There are reports in the worst hit places of reaching a tipping point, but mostly in the highest density areas.

1-Day Projection: WW 736981/34604; PRC 82101/3302; NON 654880/31302; IT 98859/10999; US 145164/2596; TX 3112/35; ME 265/1
1-Day Distancing: WW 736981/34604; PRC 82101/3302; NON 654880/31302; IT 98859/10999; US 144607/2494; TX 2934/37; ME 265/1
2020-03-29 WW 720117/33925; PRC 82122/3304; NON 637995/30621; IT 97689/10779; US 140886/2467; TX 2792/37; ME 253/3
Reminder: PRC numbers are likely 1/6th of the actual due to creamatorium activity.

Social distancing rates have not been updated in a few days, but we are tracking similarly between the two formulae.

Found another error, so the estimates for testing and hospitalization were too high. One of my cells had a $ in it, so it did not change as I added new data.

1-Day US Testing: POS 126,269 NEG 659,435 PEND 70,178 HOSP 23,397 DEAD 2,585
20200329 POS 139,061 NEG 692,290 PEND 65,549 HOSP 19,730 DEAD 2,428

Fewer hospitalized now means fewer dead in 2 weeks.

2020-03-30 (Monday) Projections
1-Day Projection: WW 785,602/37,588; PRC 82,245/3,309; NON 703,357/34,279; IT 103,200/11,592; US 163,395/3,004; TX 3,175/46; ME 303/3
1-Day Distancing: US 164,555/2,970; TX 3,269/44

+1 Day US Testing POS 157,461 NEG 783,890 PEND 74,222 HOSP 24,311 DEAD 2,932

We are definitely slowing down. Look at the April 1 projection trends.

2020-04-01 7-day Projection April 1
7-Day Projection: WW 1,169,645/61,852; PRC 82,134/3,338; NON 1,087,512/58,514; IT 157,313/19,560; US 477,863/6,398; TX 9,131/162
4-day Projection: WW 1,097,956/54,118; PRC 82,357/3,321; NON 1,015,599/50,796; IT 132,717/16,052; US 333,315/5,929; TX 7,263/90; ME 468/5
3-day Projection: WW 1,017,269/50,218; PRC 82,428/3,321; NON 934,841/46,897; IT 124,403/14,738; US 278,237/5,179; TX 5,427/83; ME 426/5

The distancing formula has been tracking more consistently over the longer term.
7-Day Distancing: US 232,880/4,603; TX 5,875/72
4-day Distancing: US 237,600/4,513; TX 5,003/68
3-day Distancing: US 231,486/4,464; TX 4,761/68

Hospitalization rates are going down, while positive and total testing rates are going up.
7-Day US Testing: POS 134,035 NEG 728,160 PEND 100,028 HOSP 68,262 DEAD 4,770 Overrun: 04-06 335k
4-Day US Testing: POS 150,376 NEG 785,328 PEND 83,576 HOSP 64,012 DEAD 5,886 Overrun: 04-06 342k
3-Day US Testing: POS 164,152 NEG 817,199 PEND 77,376 HOSP 44,840 DEAD 4,592 Overrun: 04-09 408k

I am just giving the general overrun rate above. That is typically within 1 day of when the TX projections look bad, and tracking kinda bad vs super bad is a lot of effort.

Testing rates slowed a little, but not much.

Our mortality rate keeps climbing. 144, 155, 167, 175. TX just started the climb, 122 yesterday, and 133 today. Probably the spring break people. Average is 6 days to onset and 14 days from onset to death for those who do not survive.

Some good research is going on that promises to improve mortality rates.

Inflection point for outbreaks with an R0 of 2.2 is 60% of the population becoming immune. Divide that by whatever you think our detection rate is. I think we get about 38%, so maybe 75m people in the US showing as confirmed.

Social distancing data lags by 3-5 days. Statistics lag by 1 day. Detection lags by 9-12 days. Impact delay of SD on Confirmations MAY BE 9 days, or it may be longer. The correlation is hard to suss out, but it may be that it will become more evident as we get more days of data.

Spreadsheet is updated, and downloadable here:
https://drive.google.com/file/d/1vocCN445AZyVBBLsv0kJR8ZDP9DM0UST/view


CV19 US Hospitalization Rates

Posted by xaminmo

CV19 Pandemic Update!

US has stepped up testing rapidly, and so while social distancing is at -40%, we still show +1% of the new model prediction.

03-26 Thursday
1-day Projection: WW 524523/24175; PRC 81731/3289; NON 442792/20886; IT 79988/8254; US 80513/1257; TX 1582/19 non+IT 1.2%, US 5.1%
1-Day Distancing: WW 524523/24175; PRC 81731/3289; NON 442792/20886; IT 79988/8254; US 82170/1244; TX 1632/20 TX 5% high, US 2% low
Actual Numbers: WW 529591/23970; PRC 81782/3291; NON 447809/20679; IT 80589/8215; US 83836/1209; TX 1563/21

The rapid rise in US testing (127% per day average) may numerically counter the decrease in disease spread from social distancing (-40%). I hope that is factored into the policy plans, because it may look a lot more spooky over the next week than it really is.

03-27 Friday
1-Day Projection: WW 601509/27192; PRC 81903/3297; NON 519605/23895; IT 87309/8995; US 106851/1552; TX 1988/29
1-Day Distancing: WW 601509/27192; PRC 81903/3297; NON 519605/23895; IT 87309/8995; US 103554/1554; TX 1971/27
1=Day US Prototype: Positive: 89,618 Negative: 486,863 Pending: 66,880 Hospitalized: 13,923 Dead: 1,471

I added the Number of Tests and number hospitalized in the US to the spreadsheet. This is somewhat complete for the US and TX, but other countries are hit or miss. Those checkpoint earlier in the day than other stats, but that is fine. Those are much better numbers to track hospital overload than trying to extrapolate from confirmed cases, but much less data. I am still tweaking it all.

Midday Thursday, the US had 10131 hospitalized, and 1163 dead from COVID-19. The rates of increase of these are still climbing at the typical 135% per day. There is not yet a clear correlation between social distancing and death or hospitalization rates, but I am still testing ideas.

Since I only have that for the whole US, I am still tracking confirmed, death, and projecting with social distancing in the main sheet.

2020-04-01 7-day Comparison
7-Day Projection: WW 1169645/61852; PRC 82134/3338; NON 1087512/58514; IT 157313/19560; US 477863/6398; TX 9131/162
7-Day Distancing: WW 1169645/61852; PRC 82134/3338; NON 1087512/58514; IT 157313/19560; US 232880/4603; TX 5875/72
7-Day US Prototype: Positive: 134,035 Negative: 728,160 Pending: 100,028 Hospitalized: 68,262 Dead: 4,770 Overrun: 04-06 295-335k

The old formula pushes Texas overrun slightly further out for late-stage events:

04-06 29k TEXAS ICU OVERRUN STAGE 1
04-07 36k TEXAS VENT OVERRUN STAGE 1
04-12 146k TEXAS ICU STG3
04-13 178k TEXAS VENT STG3 & Hosp STG1
04-16 400k may be the inflection point for Texas if 14% like China, but our death rate is lower than theirs or the world.
04-17 532k TEXAS HOSP STG3
04-20 1.2m Inflection point for R0 of 2.2 is 60% with detection rate of 39.08

The new distancing prototype formula looks like this after tweaking:

04-15 2929 TEXAS ICU OVERRUN STAGE 1
04-17 1787 TEXAS VENT OVERRUN STAGE 1
05-02 2929 TEXAS ICU OVERRUN STAGE 3
05-03 1787 TEXAS VENT OVERRUN STAGE 2
05-04 26633 TEXAS HOSPITAL OVERRUN STAGE 1
05-09 4550 TEXAS VENT OVERRUN STAGE 3
05-09 400k may be the inflection point for Texas if 14% like China, but our death rate is lower than theirs or the world.
05-16 26633 TEXAS HOSPITAL OVERRUN STAGE 2
06-05 47441 TEXAS HOSPITAL OVERRUN STAGE 3
05-21 1.2m Inflection point for R0 of 2.2 is 60% with detection rate of 39.08

The growth in hospitalizations puts overrun at 04-06. I do not know what to believe.

Here are the milestones from 04-23 data for comparison to track our progress flattening the curve:

04-06 33k TEXAS ICU OVERRUN STAGE 1
04-07 45k TEXAS VENT OVERRUN STAGE 1
04-10 150k TEXAS VENT/ICU STG2 & Hosp STG 1
04-13 366k TEXAS VENT/ICU STG3 & Hosp STG 2
04-15 660k TEXAS Hosp STG 3

Milestones are based on guesstimates: TX 28.7m pop; 2.9 beds per 1000 in TX (83230); 32% unoccupied (26633) / 11% are ICU (2929); 6.7% ventilators (1787); 10% limited function ventilation (2663) / 5-15% conf need hosp(532k-177k @ overrun); 2-10% ICU (29k-146k); 1-5% need vent (36k-178k)

Testing has ramped up substantially, and the US has a capacity around 350k tests per week, or 50k per day. The increased testing, and the shift in social distancing makes these numbers more fuzzy. I tweaked my multipliers for regression testing.

Our hospital load may be 31.86% of the world average for the same number of confirmed. Korea had 8652 confirmed from 316664 tests on 03-20 with 94 deaths. Their death rate is 1.42%. Ours is 1.44% with 579k tests, 83836 confirmed, 1209 deaths. The world rate is 4.52% I suppose that the death to confirmed case is tied to the test percentage.

Inflection point for R0 of 2.2 is 60% of the population, times the percentage of infected people who get confirmed. It was estimated elsewhere that China at 4.02% mortality rate per confirmed was 14% tested. That means our hospital load will be 35.82% for the same number of confirmed, or that our confirmed cases are 39.08% of our total cases.

TX has 28.7 million, and 39.08% of that is 1.12 million as our inflection point.

Unknown accuracy because testing inputs are changing, and I am just making things up.

Social distancing data lags by 3 days. Statistics lag by 1 day. Detection lags by 9-12 days. Impact delay of SD on Confirmations is 9 days.

https://drive.google.com/file/d/1vocCN445AZyVBBLsv0kJR8ZDP9DM0UST/view


Social Distancing Dashboard

Posted by xaminmo
This site uses anonymous mobile phone data to calculate the change in miles traveled, and the change in number of mobile phone encounters. It is an earlier predictor of how well social distancing is working, without having to wait 12 days for the confirmed cases stats to change, nor worry about changes in testing frequency or methodology.
 
It’s really a green-light instantaneous dashboard kind of thing. They don’t expose trending data, or even numerical data, so you cannot chart, trend, or predict off of this.
 

CV19 no big deal

Posted by xaminmo
A good question I saw and responded to is “Hardly anyone is sick, and hardly anyone has died. Why is everyone all panicky about CV19? What’s missing?”
 
TLDR: It spreads exponentially, and while you have no symptoms. Listen to the experts if you don’t understand. Don’t be a party to manslaughter.
 
NARRATIVE: The missing part is that it’s an exponential spread. It’s like the old puzzle, there’s one lily pad on a pond, and every day, each lily pad becomes two. It takes 30 days to cover the pond. When is the pond half covered? 29 days.
 
Today, the absolute numbers look very mild right now, and that’s what we like. We want them to stay mild, but any action we take will not have an effect for 12 days. Also, number reporting lags by 1 day. 13 days before the end looks like just any other day in paradise. We’re not doubling every day. We’re doubling every 2.2 days.
 
For the US, we are 10 days from overloading hospital capacity, and 14 days from overloading even reserve, crappy capacity. When we overload the hospital capacity, the death rate quadruples. That’s why it became such a panic.
 
Luckily, the declaration of pandemic was on the 11th, which just started to show up in the numbers yesterday. We’ll know by late Thursday if the declaration of national emergency made an impact. The goal is to keep the number of people in the hospital below the threshold where a lot of extra people die. Right now, that looks so far off, but if we didn’t drastically slow the spread, that would start to look grim for the worst survivable cases around April 5, and by April 15 would be just letting the really sick asphyxiate. I say “didn’t” rather than “don’t” because the time to take action has already passed.
 
It’s easy to ignore when 85% of the people infected simply are not counted, and when only 10% of the counted people are at risk of dying from a hospital overload. However, that amounts to 2-5 million people in the US potentially dying from this. With that many, chances are one of them would be a close friend or family member. If it happens to you, then you would definitely care, and you would not care that it was because a bunch of people didn’t understand exponential math.
 
The panic is because a bunch of us care even when it’s not someone close to us, or we can see that it could happen to us. There are a lot of asymptomatic carriers, spreading the disease. So “I feel fine, I don’t need to quarantine” has already lead to deaths, and will lead to many more deaths.
 
It wouldn’t be God’s Will, or bad luck. It would be a willful choice of people to ignore the experts because the non-experts didn’t understand, and therefore decided the experts were not actually experts. Reckless action leading to the death of others is called manslaughter, and negligent action leading to death is called negligent homicide. Purposefully infecting someone would be called murder.
 
Lag times per Feb 7 study in JAMA from Wang et al:
Median time from first symptom to dyspnea was 5.0 days
to hospital admission was 7.0 days
to acute respiratory distress syndrome was 8.0 days.
For survivors, the median hospital stay was 10 days.
https://jamanetwork.com/journals/jama/fullarticle/2761044
 
Lag times per Jan 22 report by China National health:
Median days from first symptom to death were 14 (range 6-41)
70 year old or above (11.5 [range 6-19] days)
below 70 year old (20 [range 10-41] days.
https://onlinelibrary.wiley.com/doi/full/10.1002/jmv.25689?af=R
 
The average is 12 days to show up in the stats, and 20 days start to finish, so that’s what most stats focus on. Policy changes consistently take 12 days to show up in the stats. I’m not sure where that first started, but you can look at the raw numbers and see it. It is self evident.

SARS-2 March 24 Update

Posted by xaminmo

First, the update:

03-24 Tuesday
1-day Projection: WW 425953/18600; PRC 81595/3283; NON 346031/15382; IT 69104/6744; US 57310/731; TX 916/10
Actual Numbers: WW 417966/18615; PRC 81591/3281; NON 336375/15334; IT 69176/6820; US 53740/706; TX 955/12

03-25 Wednesday
1-day Projection: WW 461807/21005; PRC 81686/3288; NON 381238/17782; IT 74856/7654; US 66137/903; TX 1203/16

Infection spread appears to be reducing in US and TX, even though TX was above projections for yesterday.
US shows 134% instead of 135%, and the last 3 days were 130%, 131%, and 123%.
TX shows 135% instead of 147%, and the last 3 days were 108%, 121%, and 126%.

This implies that the declaration of pandemic may have had positive effect.

The numbers from tonight or tomorrow will start reflecting any changes caused by the declaration of national emergency.

Milestones would have pushed out a day, but my methodology was poor. I now show the number for the condition in the left of the note, not the estimated number for that day. Also, I dropped stage 2 here, and just show worst case reserve (smallest number) and best case reserve (largest number). I do not have proper numbers to separate stage 1 (over standard) vs stage 2 (over minimum reserve).

04-05 29k TEXAS ICU OVERRUN STAGE 1
04-06 36k TEXAS VENT OVERRUN STAGE 1
04-10 146k TEXAS ICU STG3
04-11 178k TEXAS VENT STG3 & Hosp STG1
04-14 400k may be the inflection point for Texas
04-15 532k TEXAS HOSP STG3

Milestones are based on guesstimates, because the exact number of ICU beds and placement of ventillators is proprietary data that is hard to find and confirm. My estimates are based on: TX 28.7m pop; 2.9 beds per 1000 in TX (83230); 32% unoccupied (26633) / 11% are ICU (2929); 6.7% ventilators (1787); 10% limited function ventilation (2663) / 5-15% conf need hosp(532k-177k @ overrun); 2-10% ICU (29k-146k); 1-5% need vent (36k-178k)

Testing infrastructure may fail before the inflection point, leading to a false decrease in numbers reported. If that happens, we may not numerically reach the inflection point, which is 60% of the population for R0 of 2.2. Remember, we only confirm 12-15% of the actual cases, as many are mild or even asymptomatic, but are still infectious to others.

I expect these to move further out each of the next few days, and then it will probably look like it levels off a bit for several days before reducing further.

We are not out of the woods. Hospital usage is 1-4 weeks. Average infection course is 20 days with no hospitalization, but once people are bad enough to need O2 or ventilation, it takes longer to recover well enough to not need it anymore.  If we were to fall to 112% today (unrealistic extreme) and stay there, the milestone dates become 04-23, 04-25, 05-07, 05-09, 05-16, 05-18. That spreads out enough that we may gain an extra couple of days on each due to early cases resolving (people do not stay in the hospital forever).

The current milestones for the US, since TX is about 9% of the US:
03-31 29k -> 322k US ICU OVERRUN STAGE 1
03-31 36k -> 400k US VENT OVERRUN STAGE 1
04-05 146k -> 1622k US ICU STG3
04-06 178k -> 1977k US VENT STG3
04-08 400k -> 4444k US inflection point
04-09 532k -> 5800k US HOSP STG 2

These are MUCH more fuzzy, since it is not exactly 9%, and ICU, Vent and bed capacities vary. That adds 1-2 days uncertainty. This also will be affected by any changes in the numbers as discussed above.

Here are the milestones from 04-23 data for comparison to track our progress flattening the curve:
04-06 33k TEXAS ICU OVERRUN STAGE 1
04-07 45k TEXAS VENT OVERRUN STAGE 1
04-10 150k TEXAS VENT/ICU STG2 & Hosp STG 1
04-13 366k TEXAS VENT/ICU STG3 & Hosp STG 2
04-15 660k TEXAS Hosp STG 3

NOTE that the very first milestone was 04-02 based on 570 ICU beds, and was abandoned as a predictor.

When testing and behavior are nonchanging:
Model is +/- 25% per week
That is +/- 1.5 days for stage 1
That is +/- 2.5 days for stage 2
That is +/- 3.5 days for stage 3

Inflection point for R0 of 2.2 is 60%.
Current reporting rate is 12-15%.
Assuming we keep testing by same criteria, the Tx inflection point is 400k confirmed.
We may not reach this numerically due to testing/infrastructure failure.
eg, we may reach stage 3 while our confirmed rate is much lower, but still around the same days.

This is all bistromath, and both reporting rates, and trends are changing daily at this point.

Also note that the numbers lag by up to 24 hours based on reporting systems.

Also Also note that it still takes around 12 days for action changes to affect trends, since people infected today will not be detected for a while.

Also Also Also note that there are many people who have NO symptoms at all, but are still infecting others. This, along with poor testing, explains the 12-15% confirmation rate.

https://drive.google.com/file/d/1vocCN445AZyVBBLsv0kJR8ZDP9DM0UST/view