Open Side Menu Go to the Top

03-28-2020 , 03:00 PM
Pretty much all successful novel disease vectors evolve to get less deadly for their hosts over time. The impression I got was pretty much everything he said was basic fact, and he wasn't really breaking any new ground.

In a "natural" setting a new disease like this would spread through a population and potentially cause massive morbidity and mortality, before it and the hosts evolved over a much longer time period than 3 months into some sort of manageable co-existence. This is what is believed to happen with diseases like malaria, smallpox, measles, polio, flu etc.

Pre industrial revolution a disease like this that spread through and preferentially killed elderly (of whom there were virtually none) and wasn't particular deadly (<<1%) for the vast vast majority of people would have been a nothing burger I imagine.
Coronavirus
$25m Guaranteed WPM on CoinPoker
Join the action now
Daily Rewards • Splash Pots • CoinRaces
Coronavirus
03-28-2020 , 03:05 PM
Also there may be some genetic variety in humans that makes us more at risk. As more of those die than everybody else it makes the virus less deadly over time
03-28-2020 , 03:10 PM
Quote:
Originally Posted by Kelhus100
Pretty much all successful novel disease vectors evolve to get less deadly for their hosts over time. The impression I got was pretty much everything he said was basic fact, and he wasn't really breaking any new ground.

In a "natural" setting a new disease like this would spread through a population and potentially cause massive morbidity and mortality, before it and the hosts evolved over a much longer time period than 3 months into some sort of manageable co-existence. This is what is believed to happen with diseases like malaria, smallpox, measles, polio, flu etc.

Pre industrial revolution a disease like this that spread through and preferentially killed elderly (of whom there were virtually none) and wasn't particular deadly (<<1%) for the vast vast majority of people would have been a nothing burger I imagine.
Ya the key words here are over time. Of course that's true. Mutations are essentially random (no feedback mechanism). The "successful" (doesn't kill quick, highly transmissible) ones will survive. The "unsuccessful" ones (kills too quick to pass on, makes people too sick) die out...over time. But any singular mutation from this point on has as much likelihood of being more severe as less severe. The next major variation could be even more severe and aggressive than the previous one. We shouldn't give it that opportunity with a herd immunity strategy. We should try to get the reproduction rate below 1 and wait for it to die. It's probably too late for that, though.

Last edited by Wittgenheiny; 03-28-2020 at 03:16 PM.
03-28-2020 , 03:21 PM
Dude. I was a biology major. You aren’t teaching me anything. Knowledge of Bio 101 facts are built into what I am saying. You are getting too bent trying to find a gotcha because I am using imprecise shorthand conventions that I assume everyone gets.

Mutations are random, but generally less deadly ones are selected for over time.

Last edited by Kelhus100; 03-28-2020 at 03:27 PM.
03-28-2020 , 03:34 PM
Bad Sign?

Don't know if this is a trend, a "statistical anomaly" or a fluke, but I've been watching the Coronavirus numbers, (i.e. reported cases and deaths), being reported on CNN and MSNBC. (MSNBC cites John Hopkins / NBC News as the source for their numbers.)

Up until today, the death rate here in the U.S. had been running in the range of 1.25 to 1.5 percent. The [current] numbers are now 1,930 deaths out of 115,968 confirmed cases. That equates to a death rate of 1.66 percent so the death rate appears to be increasing. (The worldwide death rate is running around 4.72 percent.)

Because everybody has not been tested, it's impossible to know how many (what percentage) of the population has been infected; but less than 200,000 confirmed cases - out of a population of over 300 million - is not a [statistically] large number. Doing what the experts are recommending, (i.e. social distancing and isolating at home), is the surest way of keeping these numbers (and percentages) from growing.
03-28-2020 , 05:05 PM
Just read something very interesting.

I found a local article discussing how countries are counting their corona-casualties:

All countries on the list below of course count people who tested positive for corona and die in a hospital.

Italy: People who die in care centers & at home are also counted. If someone dies & was not yet tested, a post mortem test is done.
France: France at the moment ONLY reports people who died of corona in a hospital. People dying in care centers or at home are not included in the statistics at this stage.
Germany: Germany reports people who died of corona in a hospital, care center or at home. No post mortem tests are done if the patient died before being diagnosed.
Belgium: Same as Germany.
Netherlands: Same as Germany.
Spain: Some regions are the same as Italy, some regions are the same as Germany.

So taking this into account, the order of countries who are missing the most corona casualties, from missing most deaths to least deaths, should be:
France>Germany=Belgium=Netherlands>Spain>Italy

I feel like given the average age of the people dying (in Italy 51% of all deaths is >80 years old), countries could be missing a lot due to lack of post mortem testing.

It will be very interesting to see after everything normalizes, how much the death rate spiked, and whether the death rate actually decreases for a bit after everything normalizes (since a significant part of the people dying are very old).
03-28-2020 , 05:15 PM
Yes it's clear that how deaths are counted is making a big difference. That's partly a fundamentally real problem, partly incompetence and partly deliberate obfuscation. Those doing the epidemiological modeling will be putting a lot of effort into cleaning the data.

Eventually it will come to down to preferences about how to count some things.
03-28-2020 , 05:28 PM
Quote:
Originally Posted by crazy canuck

Infections doubling every three days does not make sense.
Based on the studies I’ve run across the R0 (# of cases caused by 1 case) estimate is 2.0-2.5 and the serial interval (days from symptom onset in case 1 to symptom onset in directly caused cases) estimate is 3.5-6.0 days. So whiz*whirl^bang = a doubling range of 3-6 days.

Quote:
Many things effect death rate doubling.
I know but the fact is they are doubling at that rate. You can look at the data and see total deaths increasing 100(X) in 3 weeks or less in the US, UK, Germany, France and Spain. My guess is that the doubling rate of transmission somewhat mirrors what we’re seeing with deaths, at least based on these early numbers. So I wouldn’t be surprised to see a death rate slightly below 1% but then again it wouldn’t surprise me if they came back slightly above 2% either.
03-28-2020 , 05:59 PM
Can we make some kind of PDF doc or sticky for people who keep citing flu stats? I'm getting worn out.
03-28-2020 , 06:01 PM
Quote:
Originally Posted by Wittgenheiny
As for the masks, clearly they have some efficacy. It is absurd to suggest otherwise, and the CDC should be held accountable for saying that. I tend to agree they were trying to prevent a panic run on masks. N95 masks filter out 95% of particles 3 microns+ wide. SARS-CoV-2 is 1.25 microns wide. An N95 mask won't get all of them, but it will get a lot of them, and especially the big moisture particles with high viral loads. There is a reason health workers use them--they work.
I’m thinking even wearing a bandanna would reduce the risk of infecting others. No?
03-28-2020 , 06:02 PM
Has there been any discussion on how bad this virus actually is compared to what could have been? I guess, what I'm asking is how unlucky did we get? How often would we expect a virus similar to Covid-19 to happen?
Instead of CFR being 1-4%, could it have been 50%? I know there's been deadlier ones in recent times, but their characteristics made them easier to contain. What if we have another with the same spread and [slow] symptom onset characteristics but 10x deadlier?
03-28-2020 , 06:12 PM
If such virus appeared there would be more extreme response earlier with strict national lockdown and people actually obeying quarantine.
Would end up with even lower overall deaths than current virus probably.
03-28-2020 , 06:12 PM
Quote:
Originally Posted by TooCuriousso1
Has there been any discussion on how bad this virus actually is compared to what could have been? I guess, what I'm asking is how unlucky did we get? How often would we expect a virus similar to Covid-19 to happen?
Instead of CFR being 1-4%, could it have been 50%? I know there's been deadlier ones in recent times, but their characteristics made them easier to contain. What if we have another with the same spread and [slow] symptom onset characteristics but 10x deadlier?
https://www.cbc.ca/news/world/wildli...giFdcme3wKqqWY
03-28-2020 , 06:14 PM
Btw the French clinical data on treatment looks super promising:

https://www.mediterranee-infection.c...ID-IHU-2-1.pdf

"hydroxychloroquine/azithromycine"


Leads one to believe that the 2nd and 3rd waves of this could be less deadly if we can mass produce. Drugs are cheap too.

The video presentation, in French: https://www.youtube.com/watch?v=n4J8...1z6kxSC5CjC6Ow
03-28-2020 , 06:27 PM
Japan is interesting. While they have some cultural advantages when it comes to prevention, there is no widespread testing but now:

Contagion grows: Tokyo, Chiba report roughly 60 more COVID-19 cases apiece

So it's about to get serious there.

Last edited by chytry; 03-28-2020 at 06:33 PM.
03-28-2020 , 06:32 PM
I remember cure Gregory Rigano proposed and Musk shared on twitter turned out to be fraud.

https://www.dailymail.co.uk/news/art...ity-claim.html

Your source, rafiki, does look credible though.
03-28-2020 , 06:35 PM
Quote:
Originally Posted by John21
I’m thinking even wearing a bandanna would reduce the risk of infecting others. No?
I'd say for sure. It certainly can't hurt. Anything covering the exit route of lung moisture which carries the virus is going to catch some % of it.
03-28-2020 , 06:43 PM
Quote:
Originally Posted by ToothSayer
You just can't get away from the numbers; however you run them one of two things happens: a huge death rate OR high infectious with a very high "already infected" rate that even at a low death rate means many hundreds of thousands now certain to die and millions from infections in two weeks.

However you run the numbers, all you can do is lock down.
I’ve read plenty of your posts on this topic and truly appreciate your efforts to provide decent analysis.

I work as part of the hospital supply chain for Interior Health in BC Canada. I’m fairly concerned, and even if I was to end up alright, I don’t want to be a vector for further transmission.

I’m not inclined to think my chance of exposure is low, not only due to working in fairly close proximity to dozens of coworkers, some of whom travel to various hospitals and care facilities, but also due to my responsibility of unloading a supply truck at a main hospital where there are likely infected patients (for privacy reasons the only info I’m formally given is that there are currently 70 cases in Interior Health).

Regardless I’ve decided to wear an N95 mask, safety glasses, and gloves while I’m inside the hospital.

I did this last week and felt significantly out of place partly because no other hospital staff members were electing to wear PPE (I don’t know if that has changed since my last shift), but I think that’s largely irrelevant when it comes to what I should do or what seems reasonable at this point given that I agree with Wittgenheiny.

Quote:
Originally Posted by Wittgenheiny
As for the masks, clearly they have some efficacy. [...] An N95 mask won't get all of them, but it will get a lot of them, and especially the big moisture particles with high viral loads. There is a reason health workers use them--they work.
Anyways if anyone is willing to take a look, are there any opinions of how BC Canada may fare in the coming months?

https://www.ctvnews.ca/health/corona...itish-columbia

Recent developments on the supply chain end is that many types of masks, sanitizers, wipes, and PPE are now restricted when it comes to who they will be supplied to, so if anything imo this suggest that there is at least significant anticipation of a potential spike in the number of cases.

Thanks to everyone who’s contributed to this discussion.
03-28-2020 , 06:45 PM
Quote:
Originally Posted by Kelhus100
Dude. I was a biology major. You aren’t teaching me anything. Knowledge of Bio 101 facts are built into what I am saying. You are getting too bent trying to find a gotcha because I am using imprecise shorthand conventions that I assume everyone gets.

Mutations are random, but generally less deadly ones are selected for over time.
Lol I wasn't trying to get a gotcha, I was just talking. Take it easy.

I agree with you and I'm not sure what Weinstein was implying, if anything. I just do not see the point of him saying viruses trending towards less severe mutations over many many cycles when it looks like most people on earth are going to contract this in the next 6 months-year.
03-28-2020 , 06:48 PM
Quote:
Originally Posted by rafiki
Btw the French clinical data on treatment looks super promising:

https://www.mediterranee-infection.c...ID-IHU-2-1.pdf

"hydroxychloroquine/azithromycine"


Leads one to believe that the 2nd and 3rd waves of this could be less deadly if we can mass produce. Drugs are cheap too.

The video presentation, in French: https://www.youtube.com/watch?v=n4J8...1z6kxSC5CjC6Ow
Yo he cites a Chinese study where they are very certain it's contagious 20 days from positive.

We're still using 14 days up here...
03-28-2020 , 06:54 PM
Quote:
Originally Posted by The Best
I’ve read plenty of your posts on this topic and truly appreciate your efforts to provide decent analysis.

I work as part of the hospital supply chain for Interior Health in BC Canada. I’m fairly concerned, and even if I was to end up alright, I don’t want to be a vector for further transmission.

I’m not inclined to think my chance of exposure is low, not only due to working in fairly close proximity to dozens of coworkers, some of whom travel to various hospitals and care facilities, but also due to my responsibility of unloading a supply truck at a main hospital where there are likely infected patients (for privacy reasons the only info I’m formally given is that there are currently 70 cases in Interior Health).

Regardless I’ve decided to wear an N95 mask, safety glasses, and gloves while I’m inside the hospital.

I did this last week and felt significantly out of place partly because no other hospital staff members were electing to wear PPE (I don’t know if that has changed since my last shift), but I think that’s largely irrelevant when it comes to what I should do or what seems reasonable at this point given that I agree with Wittgenheiny.



Anyways if anyone is willing to take a look, are there any opinions of how BC Canada may fare in the coming months?

https://www.ctvnews.ca/health/corona...itish-columbia

Recent developments on the supply chain end is that many types of masks, sanitizers, wipes, and PPE are now restricted when it comes to who they will be supplied to, so if anything imo this suggest that there is at least significant anticipation of a potential spike in the number of cases.

Thanks to everyone who’s contributed to this discussion.
Ontario has a backlog of 10000 tests right now, because they ran out of reagent. I'm assuming the other provinces are similar. A lot of people are saying the curve is flattening in Canada but I think it's just a function of backlogged tests and test criteria. I don't see any reason at all why Canada would be any different from any other country. Mass transit largely destroys any inherent geographical advantage.

If you to to reddit there are lots of smug Canucks praising da Grorious begoateed Leader. I don't see it, but I hope I'm wrong. We better hope so, because we have 1/3 the ICU beds of the US.
03-28-2020 , 07:06 PM
Quote:
Originally Posted by John21
Based on the studies I’ve run across the R0 (# of cases caused by 1 case) estimate is 2.0-2.5 and the serial interval (days from symptom onset in case 1 to symptom onset in directly caused cases) estimate is 3.5-6.0 days. So whiz*whirl^bang = a doubling range of 3-6 days.



I know but the fact is they are doubling at that rate. You can look at the data and see total deaths increasing 100(X) in 3 weeks or less in the US, UK, Germany, France and Spain. My guess is that the doubling rate of transmission somewhat mirrors what we’re seeing with deaths, at least based on these early numbers. So I wouldn’t be surprised to see a death rate slightly below 1% but then again it wouldn’t surprise me if they came back slightly above 2% either.
You are right, but I was interested in an approximate number to calculate what the IFR might be. It does not matter if it 3 days or 6 days, IFR should be 4%+. See the calculations above.

Quote:
Originally Posted by Former DJ
Bad Sign?

Don't know if this is a trend, a "statistical anomaly" or a fluke, but I've been watching the Coronavirus numbers, (i.e. reported cases and deaths), being reported on CNN and MSNBC. (MSNBC cites John Hopkins / NBC News as the source for their numbers.)

Up until today, the death rate here in the U.S. had been running in the range of 1.25 to 1.5 percent. The [current] numbers are now 1,930 deaths out of 115,968 confirmed cases. That equates to a death rate of 1.66 percent so the death rate appears to be increasing. (The worldwide death rate is running around 4.72 percent.)
Many of the people who are currently sick will succumb to the disease. So we cannot estimate CFR based on these data points. As time goes on, CFR will increase.

Quote:
Originally Posted by rafiki
Btw the French clinical data on treatment looks super promising:

https://www.mediterranee-infection.c...ID-IHU-2-1.pdf

"hydroxychloroquine/azithromycine"


Leads one to believe that the 2nd and 3rd waves of this could be less deadly if we can mass produce. Drugs are cheap too.

Good news for second and third waves, but the drug will be not available for everybody who needs it in the first wave in US/Canada.

India banned all exports. Hungary (also one o the biggest produces) also banned all exports.

https://www.nytimes.com/reuters/2020...-medicine.html

US/Canada imported the raw powder from China/ India and it will take 3-6 months to produce hydroxychloroquine on a mass scale.

Also, China and Korea have been using hydroxychloroquine for a while and patients are still dying.
03-28-2020 , 07:42 PM
Quote:
Originally Posted by chytry
Are you purposely being disingenuous or are you actually this dense? Your data is from the US.



I would love to know what the odds of catching Covid and being admitted to intensive care are for a normal, non-obese 20-40 year old who lives his life, washes his hands as usual, but takes no special precautions. I would guess he is way more likely to: be involved in an auto accident, be a victim of violent crime, be it robbery or just a traditional American mass shooting, or overdose on drugs, etc.
03-28-2020 , 07:42 PM
US numbers are grim. Country will have no choice but to go full lockdown and then 4+ weeks before peaking. I'm not even sure NY is below 1 from the anecdotes we're seeing of people still socializing.

Only way forward for the US unless they're ok with millions dead:

- Total lockdown for 4 weeks
- Get enough masks, thermometers, set up sanitizing and forced distancing at all shops and (still open) workplaces
- Get a huge testing surveillance network in place
- Obsessively contact trace

Since no one seems prepared to do communist-level public indoctrination and shaming to stop virus transmitting behavior (it's almost antithetical to the American spirit). and cops would rather sit on their ass then enforce, this is going to take a long time.

One the benefits of China and Taiwan and Korea is a high IQ social shaming society. The US is almost the opposite in many areas.
03-28-2020 , 07:48 PM
Lots of people are criticizing the design/results of the French study on
"hydroxychloroquine/azithromycine":

https://www.reddit.com/r/COVID19/com...l_effect_of_a/

Quote:
My breakdown

Authors and Conflicts of Interest: This is the same group that published the first HCQ-Azithro paper, so they have skin in the game, namely their reputation. One of the authors is editor-in-chief of the journal, and two more authors are on the editorial board. This is a clear COI, not stated in this or the original paper. Also came across this article regarding the corresponding author Dr. Raoult’s history of previously falsifying data. None of this impacts the actual quality of this paper, just indicating a propensity for confirmation bias.

Design: This is purely a case series with no control group. They try to claim you should compare this series to “the literature which shows that the viral RNA load can remain high for about three weeks in most patients in the absence of specific treatment,” but this is an incredibly unscientific way to try to formulate a control group. To me, this design is pretty egregious in light of their already (self-)published initial open-label nonrandomized trial, as it adds virtually no new knowledge to the literature. Any of the thousands of patients who are taking this combination as a result of their initial study provides the same anecdotal evidence as is presented here. At this point, RCTs or very large cohort studies are needed to advance our understanding.

Population: Admitted patients to ID ward, who were admitted based purely on having a positive NP swab SARS-CoV-2. This is a very strange admission criteria, as many of these patients had very mild symptoms. In fact, 4 of them were asymptomatic! Why were they tested? Why were they then admitted? Unclear. 33 had purely URI symptoms, and only 43 had lower respiratory tract symptoms. 94% of them had a low NEWS score and only 15% of their patients even had a fever, but even with these mild cases, they have the nerve to compare their case series to the initial Lancet paper from March 11 that discussed the clinical course and risk factors for patients with COVID-19, where 28% of patients died, including 94% with fever and 29% with tachypnea >24 (which would immediately put them in red for NEWS) on admission. Apples to oranges, and again, unscientific to try to compare two populations in this new study. They included 6 of the patients from their initial study, which is reasonable in a case series but certainly would have been unacceptable in an RCT.

Outcomes: Mainly disease-oriented outcomes with time to clearance on PCR being their most reported (which also weirdly doubled as their discharge criteria; they kept asymptomatic patients with a positive PCR in the hospital). This disease-oriented outcome is somewhat validated by this Chinese article in the Lancet suggesting viral load as a predictor of severe patient-oriented outcomes. They also changed their primary outcome/discharge criteria partway through the study, reducing number of negative PCR swabs until discharge from two to one. In an RCT, this would be completely unacceptable, but again, okay in a case series. Their more important patient-oriented outcomes look a lot like standard reported numbers from around the world: ~20% need admission (which I’m defining as the 15% needing oxygen; would not admit the rest), 1/4 of those admitted (3/12 in study) need ICU, and 1-2% CFR (1.25% in study).

Echoing another comment, this did show that viral cultures were negative BEFORE negative PCRs, possibly indicating that the PCR is just picking up viral detritus and not live virus and hinting at the possibility that not everyone who is PCR positive is necessarily infectious.

Discussion Section: I would typically not include this in an article analysis, but I think some parts of this are telling as to the authors’ mindset. First, there is zero discussion of the article’s limitations. They seem to be pretending there are none. Second, they make bold and outlandish claims from their case series, including that “this will play a role in controlling the disease epidemic by limiting the duration of virus shedding.” None of their data demonstrates decreasing infectivity. Third, they make claims that chloroquine, HCQ, and Azithro have been prescribed to billions of people [citation needed], ergo demonstrating their safety. This is a popularity fallacy. Finally, they make this bold and incorrect claim: “In conclusion, we confirm the efficacy of hydroxychloroquine associated with azithromycin in the treatment of COVID-19 and its potential effectiveness in the early impairment of contagiousness.” Case series can “confirm” nothing but confirmation bias.

TL;DR: This case series of 80 patients on HCQ-Azithro does not change our understanding of the possible efficacy of the therapy. RCTs (beyond the small Chinese study with 30 patients with no difference in outcomes) and/or large cohort studies would be helpful in determining the utility of this regimen.
Quote:
Disclaimer : I am an engineer, not a PhD doing research. But I can understand what they are doing, as I worked a lot in emergency situations.

You are wrong if you think Raoult et al goal is to publish quality science. He has skin in the game : his reputation is at stake, he basically put his money where his mouth is and this guy is all-in. His goal is to diagnose, treat, and release everyone that comes there.

The guy studied this HCQ+AZ combo for years. He built his hospital to deal with pandemics, and he did that against nearly everyone in France : politics, labs, etc...

His hospital run nearly 33% of tests in France, diagnosed 1500 cases, it has less than 100 beds, and his plan is to compete with whole world for the best CFR.

By the way, how do you want to produce standardized research in a situation where we do not have any randomized set of data about untreated population ? If you want to get a math-grade proof of how the treatment works, you need to test it on people that you know won't recover spontaneously. That mean you need to let people die or suffer terribly in the control groups.

The only other way to get solid insights with a virus where like 90% ppl recover spontaneously is to do a massive trial with thousands of people in the control group. But that would also mean, let people die without treatment.

The rationale about his treatment is that it works (with great efficacity) in vitro, it's a short treatment with HCQ and a quite common regimen with AZ that lots of people get for common upper / lower respiratory infections.
Quote:
The lack of a control makes this garbage. Of 80 confirmed COVID patients with long duration symptoms, the vast majority getting better, some needing oxygen and 3-5% going critical is entirely normal. That's what has happened everywhere.

Why the hell isn't there a control?
Coronavirus
$25m Guaranteed WPM on CoinPoker
Join the action now
Daily Rewards • Splash Pots • CoinRaces
Coronavirus

      
m