COVID-19

Health, Fitness, Insurance, ...
Locked
Tyler9000
Posts: 1688
Joined: Fri Jun 01, 2012 11:45 pm

Re: COVID-19

Post by Tyler9000 »

Neil Ferguson, one of the authors of the original Imperial College study that predicted 500k deaths in the UK alone, has now revised his predictions and believes less than 20k people may die. And of those 20k, 2/3rds probably would have died by the end of the year anyway because they were old and sick.

The original paper predicted that a full quarantine was required for 18 months until a vaccine was required. But notably, "Ferguson said community testing and contact tracing wasn’t included as a possible strategy in the original modelling because not enough tests were available." Now the UK is on the verge (within days or weeks depending on who you ask) of having millions of take-home antibody tests, which basically blows up the original model.

Add to that the competing Oxford research that indicates the infection rate was also based on incomplete data and the disease may be way less deadly than they first thought, and things are looking up.

Lesson learned: science is great, but models are only as good as the inputs and assumptions.

User avatar
Ego
Posts: 4918
Joined: Wed Nov 23, 2011 12:42 am

Re: COVID-19

Post by Ego »

The Numbers Problem

Asymptomatic cases

Iceland: After doing extensive testing of the population they estimate 50% of the cases are asymptomatic
Japan: After testing every Japanese person repatriated from Wuhan they found 30.8% of the cases asymptomatic
Diamond Princess: The CDC reports that 46.5% of the passengers and crew from the ship were asymptomatic

Important note: These are positive tests. This does not include those who were previously infected and recovered, thereby testing negative but having antibodies.

This is important because up to now we've been using the case fatality rate provided to us from Wuhan. If these asymptomatic rates prove to be true, the true fatality rates will be much lower.

Did we decide to shut down the economy based on faulty data?

George the original one
Posts: 5369
Joined: Wed Jul 28, 2010 3:28 am
Location: Wettest corner of Orygun

Re: COVID-19

Post by George the original one »

Ego wrote:
Thu Mar 26, 2020 1:04 pm
Did we decide to shut down the economy based on faulty data?
No, because we're already seeing that the disease, faulty data or not, is shutting down ICUs.

jacob
Site Admin
Posts: 12885
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: COVID-19

Post by jacob »

Ego wrote:
Thu Mar 26, 2020 1:04 pm
This is important because up to now we've been using the case fatality rate provided to us from Wuhan. If these asymptomatic rates prove to be true, the true fatality rates will be much lower.
I don't know who "we" are, but the US experts (CDC et al) have been using a death rate of 1% from the get go. Same assumption from Denmark. If South Korea did all their contact tracing and testing correctly, then their numbers suggest 1.4%. The Italian city of Vo (pop 3000) which was fully isolated and fully tested had a death rate slightly over 1%. And finally the Diamond Princess also had a death rate slightly over 1% if we assume that everybody there eventually got it. Taiwan is also on the ball and has a death rate of 0.8%.

These are not difficult numbers to find and I'd expect modellers would understand the issues with using the wrong denominator even if the public doesn't. It's all the journalists/laymen that presume that reported CFR is the real CFR w/o wondering why Italy > 10% and China ~ 3.8% or SK ~ 1%, not checking to see how widespread testing was in any given country. I see this mistake over and over everywhere. It's probably the biggest misunderstanding of the math. (#2 being the exponential growth delay. #3 being the threat multiplier effect.)

The more pertinent point, I think, is that whatever social measures taken are actually making a difference (as they would be expected to after about 10 days---median incubation + median time to hospitalization). An epidemic is an adaptive system composed of the virus and its targets. As people change their behavior, the numbers change. For example, the greater the social distancing, the longer the doubling rate. In particular, just because the mayor, governor, or president is acting like an idiot doesn't mean that 100% of the population is copying it. Conversely, even if leaders are making the right choices, it doesn't mean that everybody follows them. Some will go farther and some will not go far enough, but regardless, the numbers change.

The difference between do-nothing and simple social distancing can push the overload date forward by weeks and/or create a difference in the final death toll by a factor 10.

I will also say/predict that as these measures begin to work and prevent the worst, there will be a lot of people who won't understand "why we had to go to such lengths to prevent something that turned out not to be a big deal." That kind of blindness is a political death trap that's sure to be exploited.

Add: Should also note that some places COVID deaths are not being counted or counted as something else. Russia has a strangely high number of pneumonia deaths this year even if the test numbers are low. There are reports out of WA that not everyone is being tallied because it was too late for a test.

theanimal
Posts: 1610
Joined: Fri Jan 25, 2013 10:05 pm
Location: AK
Contact:

Re: COVID-19

Post by theanimal »

New paper showing heart damage from infection.

https://www.ahajournals.org/doi/10.1161 ... 120.046941

Abstract:
Coronavirus disease 2019 (COVID-19) is a global pandemic impacting nearly 170 countries/regions and more than 285,000 patients worldwide. COVID-19 is caused by the Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2), which invades cells through the Angiotensin converting enzyme 2 (ACE2) receptor. Among those with COVID-19, there is a higher prevalence of cardiovascular disease and more than 7% of patients suffer myocardial injury from the infection (22% of the critically ill). Despite ACE2 serving as the portal for infection, the role of ACE inhibitors or angiotensin receptor blockers requires further investigation. COVID-19 poses a challenge for heart transplantation, impacting donor selection, immunosuppression, and post-transplant management. Thankfully there are a number of promising therapies under active investigation to both treat and prevent COVID-19.

thrifty++
Posts: 1115
Joined: Sat May 23, 2015 3:46 pm

Re: COVID-19

Post by thrifty++ »

Re numbers, an NZ based university study predicted that if we did nothing, here, it would kill about 1.67% of the population. This is about 80,000 people. https://cpb-ap-se2.wpmucdn.com/blogs.au ... PM-006.pdf

I guess the problem is that no one really knows for sure. Its all so uncertain. I guess government responses will adjust proportionately as new information is unveiled.

steveo73
Posts: 1432
Joined: Sat Jul 06, 2013 6:52 pm

Re: COVID-19

Post by steveo73 »

Tyler9000 wrote:
Thu Mar 26, 2020 1:01 pm
Lesson learned: science is great, but models are only as good as the inputs and assumptions.
People need to understand science isn't something that is set in stone in lots of cases. If you believe models are reliable predictive tools you do not understand science today.

Good data and a well defined process may lead to accurate models but once something is complex all bets are off.

steveo73
Posts: 1432
Joined: Sat Jul 06, 2013 6:52 pm

Re: COVID-19

Post by steveo73 »

George the original one wrote:
Thu Mar 26, 2020 1:11 pm
No, because we're already seeing that the disease, faulty data or not, is shutting down ICUs.
To me it appears that both comments are true. The virus mightn't be that bad but lockdowns are required to manage it. We need to take precautions to ensure people don't die needlessly. At the same time hopefully the virus isn't that bad and we can keep the mortality rate below 1% in most locations.

CS
Posts: 644
Joined: Sat Dec 29, 2012 10:24 pm

Re: COVID-19

Post by CS »

@steveo73
Hospitals the US can barely (and not always) handle the flu, with a mortality of .2%. 1% is five times that.

steveo73
Posts: 1432
Joined: Sat Jul 06, 2013 6:52 pm

Re: COVID-19

Post by steveo73 »

thrifty++ wrote:
Thu Mar 26, 2020 1:52 pm
I guess the problem is that no one really knows for sure. Its all so uncertain. I guess government responses will adjust proportionately as new information is unveiled.
Yep. I'm still a fan of lockdowns and social distancing. I think the economy can just suffer for a while. The economic cost is though huge. People are worried about their jobs and we are going to see massive amounts of unemployment. I think it was Ego that mentioned increased rates of suicide. There are going to be side effects like that.

black_son_of_gray
Posts: 451
Joined: Fri Jan 02, 2015 7:39 pm

Re: COVID-19

Post by black_son_of_gray »

There is also a "distribution problem," in the sense that while we have these large categories of "mild", "severe" (hospitalization), and "critical" (ICU), we don't know what the underlying shape of that distribution is.

If we were to plot "number of people" on the Y-axis and "disease severity" on the X-axis, we might assume the distribution looks like some smooth right-skewed function with one giant peak in "mild" and a long tail into the nastier outcomes. But I don't see why there couldn't be several peaks in the distribution, perhaps related to some factors known or unknown (e.g. gender, age, ancestry, preconditions, certain less obvious SNPs, microbiome, blood type, pollution/smoking/vaping, etc.). Said another way, the 80% "mild" bucket seems to include people with little to no noticeable symptoms and also people who have a horrible 2 weeks hacking their lungs out with sub-hospitalization pneumonia...and maybe those presentations of infection are actually two separate buckets. And then there's the apparently "mild" cases that might just drop dead on the street from viral myocarditis. If we could better understand the various underlying factors and stratify people out into better defined risk-buckets, that could go a long way towards allowing certain categories of people back out into the economy.

TL;DR: I'd really like to see more resolution than simply "mild," "severe," and "critical."

7Wannabe5
Posts: 6648
Joined: Fri Oct 18, 2013 9:03 am

Re: COVID-19

Post by 7Wannabe5 »

@Tyler9000:

That article is a pretty wishful thinking interpretation of what Ferguson actually said which also included revised estimate of Ro as more like 3 than 2.2. What the WSJ made of it is even worse. The only way the total death burden will remain that low is if extreme lockdown is followed by months of extreme tracking, testing and continued social distancing.

steveo73
Posts: 1432
Joined: Sat Jul 06, 2013 6:52 pm

Re: COVID-19

Post by steveo73 »

CS wrote:
Thu Mar 26, 2020 2:22 pm
@steveo73
Hospitals the US can barely (and not always) handle the flu, with a mortality of .2%. 1% is five times that.
I get it. It can still be a catastrophe. I said earlier in this thread something like your health care system + population demographics + ability of people (including the government) to stop the spread of this will influence how bad this is.

Italy for instance is a disaster zone. I expect that these same effects will happen elsewhere. At the same time some countries will have recorded mortality rates less than 1%. I expect the mortality rate to be more like .5% in a lot of countries where more testing has occurred.

Lastly people dying unnecessarily to me simply isn't good so you have no option but to manage a potential health crisis at the expense of the economy. My parents are at much higher risks of dying if they contract this. My dad has stated he doesn't want to die unnecessarily and that to me is the point.

jacob
Site Admin
Posts: 12885
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: COVID-19

Post by jacob »

"Herd immunity: A rough guide".

Given a reproduction number R0, the herd immunity happens at (R0-1)/R0 (given random interaction chains ultimately connecting everybody). For nCov, this would be around (2.5-1)/2.5 = 60% (Germany's official prediction) which would be the expected number of people ultimately infected w/o a vaccine or functional isolation. Some countries are figuring on only 10% and if so, they're clearly making presumptions about flattening the curve far enough to develop a vaccine and/or engaging in tracing and isolation whether it's individuals or specific cities.

User avatar
jennypenny
Posts: 6636
Joined: Sun Jul 03, 2011 2:20 pm
Location: Stepford USA

Re: COVID-19

Post by jennypenny »

IlliniDave wrote:
Wed Mar 25, 2020 10:18 am
I'd call and see if you can arrange a curbside pickup of sorts, at least avoids one of you having to go inside the office.
That's what we've decided to do. She is going in on Thursday for the whole day. They will come out to get her, be extremely vigilant in handling her (because they know of my son's condition), and bathe her right before we pick her up. Still, we've been under quarantine long enough to know we don't have the virus so taking her to the vet means we could be exposed. I'm considering grocery shopping on Tuesday so we limit our risk to one calendar day and can start the clock again.

Sis is now sick and her symptoms match COVID (she's local and DD lives with her). Between that and the unplanned trip to the vet next week, I feel like the walls I worked so hard to build are crumbling. :(

steveo73
Posts: 1432
Joined: Sat Jul 06, 2013 6:52 pm

Re: COVID-19

Post by steveo73 »

jacob wrote:
Thu Mar 26, 2020 2:32 pm
Some countries are figuring on only 10% and if so, they're clearly making presumptions about flattening the curve far enough to develop a vaccine and/or engaging in tracing and isolation whether it's individuals or specific cities.
Maybe. Italy has an infection rate of < 1% based on the data. It's probably massively understated but is 10% really a low figure for other countries.

CS
Posts: 644
Joined: Sat Dec 29, 2012 10:24 pm

Re: COVID-19

Post by CS »

@jp

Dang, so sorry to hear about your dog, and sister and DD! Man, when it rains... I hope they figure everything out for your dog, and your sister doesn't have it and gets well.

7Wannabe5
Posts: 6648
Joined: Fri Oct 18, 2013 9:03 am

Re: COVID-19

Post by 7Wannabe5 »

((jp))

I feel like those who are writing off the victims of this epidemic as mostly likely to die soon are way off base. The most frequent underlying issue listed is high blood pressure. I know people in their 60s with high blood pressure who bike 30 miles no problem.

Also, many of the “theories” being tossed about make no sense mathematically. For instance, how can you have 50% of a population asymptomatically infected before anybody dies? This would only make sense if it took months to die after contracting the disease. Yes, you could have a number of initial deaths written off as flu or other cause, but you still would not have the exponential increase in daily deaths once clear identification was in place.

For example, here are the daily Covid deaths in Michigan f over the last week-2,2,3,7,9,19,17. Wow, look the epidemic must be a hoax because already dropped from 19 to 17, huge percentage!, right? WRONG!!! Put these numbers/points on a graph and it becomes clear that all this is ONLY indicative of the difference between exponential growth leading to one or another horrendous amount of tragic death before April 9th.

1taskaday
Posts: 461
Joined: Wed Dec 04, 2013 11:45 am
Location: England

Re: COVID-19

Post by 1taskaday »

Jennypenny I can only imagine the stress that you are under trying to protect your son...every move you make outside your home makes you second guess yourself.

I am the same with my elderly parents when it's my turn to care for them.

All carers have been cancelled due to risk.

It is so difficult...all siblings have different ideas on how stringent to be.

It's gone from some shouting in the open window to them (one is completely deaf and the other one halfway there) and passing dinners in this way to others arm looping them on walks.

If anyone brings it to them they are gonners.

Tyler9000
Posts: 1688
Joined: Fri Jun 01, 2012 11:45 pm

Re: COVID-19

Post by Tyler9000 »

Is the Coronavirus as Deadly as They Say?
Current estimates about the Covid-19 fatality rate may be too high by orders of magnitude.

(Dr. Bendavid and Dr. Bhattacharya are professors of medicine at Stanford.)

https://www.wsj.com/articles/is-the-cor ... _lead_pos5

Locked