|
|
Post by svart on Apr 17, 2020 10:08:04 GMT -6
Hi guys, there is a lot of misconception on what is going on apparently. This is totally understandable, because there are a lot of theories and models spreading, they are interpreted by people with whatever conclusion in mind, there is a debate on how to look at the statistics, and this is obviously a minefield. No field of actually "easier" math is so blatantly misunderstood like statistics and how curves are/can be extrapolated. No matter, what you think is the right interpretation of infection rates, death rates etc., the only safer thing to look at are the numbers that you actually want to know. How many people die and how may it look like for the nearest future. The numbers of actual death cases are, what can be considered much more reliable than the ones of infections or the mathematically drawn rates from those two actually measured numbers. The number of infections can differ a lot depending on where is tested, what is tested, how reliable are the testing methods. (Actually, the last one also applies to the death cases, unfortunately.) The easiest way to make up your own mind is to look at those numbers yourself. The curves that follow are for cumulative cases with time aligned to the outbreak time in the country and unaligned, i.e. at the same calendar dates. I suggest to look at the most reliable number, i.e. the number of deaths, to make up your own mind. All interpretation about what actually causes or would lower these can be flexed into obnoxious levels of contradiction, depending on who is doing it and who you believe knows more about it. The death cases are relatively hard numbers compared to all others. I suggest to change the axes on the graphs to "death" or "death/100k population" and leave them to linear, because this is what everyone understand intuitively in contrary to exponential/logarithmical views. These numbers are, what actually happens. And since it has been asked if the curves actually are exponential or not: If I draw an exponential curve by hand, which I am not very good at doing steadily, it could pretty much look like the US curves. On the beginning it looks a bit like a flat linear, then it becomes steeper. It starts slowly, but then... Real worlsd data is tried to be interpreted with the best mathematical function curve fitting. Since we know, that infections behave like a population growth, which is by nature exponential, it is tried to find the most accurate function coming nearest to the real world data. There are different approaches to do this, that try to minimize the error of the single point of data to the function curve. In the beginning and in the end, the exponential function can be mistaken for beeing linear or roughly simplified for the portions before and after the "knee". By no way the linearizationof the last portion of the graph is a sign to relax. A steady steep linear curve means just the same number of deaths every day. If I am informed right, NYC reached this with around 800 death cases each day. Scary, but still better than exponential. By no means I imply any political meaning. IMHO, all curves look scary still and all improvements that might be seen in the curves by bpwing down are very fragile improvements that can easily change to the worse if curcumstances change, and in the moment I am not happy at all about plans to open up schools next month gradually in my country... Of course, it is no secret, that I am more on the pessimistic side ... or as psychology says: The pessimist might be right, but the optimist definitely has more fun in life. Actually, I would love to be positively surprised in this matter here. Sometimes pictures just say more than thousand words ... Stay safe and healthy everyone. coronavirus.jhu.edu/data/cumulative-casesActually the death rate isn’t that accurate; First the death rate is based on those with confirmed test results and reports from hospitals. Second many deaths at home are being listed as cardiac arrest or even pneumonia. Now it would be quite simple if the testing supplies were available to swab the noses of all those who die during this period and freeze these samples and then afterwards when the demand for use of those nice new Abott machines that can do a 1000 a day ( made in China FYI) could run the samples of those who passed in this time period. We would then be able to have an accurate, well close to accurate death rate in an after action report. This is what happens when you spend your entire life living with lab managers! And to further muddy the waters, some locales are using "suspected COVID-19" as reason for death even when no tests had been administered. NYC added almost 4K deaths to their tally the other day without having tested a single one of them. |
|
ericn
Temp
Balance Engineer
Posts: 16,086 
|
Post by ericn on Apr 17, 2020 10:14:57 GMT -6
Actually the death rate isn’t that accurate; First the death rate is based on those with confirmed test results and reports from hospitals. Second many deaths at home are being listed as cardiac arrest or even pneumonia. Now it would be quite simple if the testing supplies were available to swab the noses of all those who die during this period and freeze these samples and then afterwards when the demand for use of those nice new Abott machines that can do a 1000 a day ( made in China FYI) could run the samples of those who passed in this time period. We would then be able to have an accurate, well close to accurate death rate in an after action report. This is what happens when you spend your entire life living with lab managers! And to further muddy the waters, some locales are using "suspected COVID-19" as reason for death even when no tests had been administered. NYC added almost 4K deaths to their tally the other day without having tested a single one of them. Yeah unfortunately there just aren’t the testing supplies! The average shipment is about 600 pieces. The other new problem is now the CDC is encouraging Universities and Reference labs to develop their own tests like the spit test in NJ. The problem with this is if these are considered as diagnostic tools is the error rate will be inconsistent. |
|
|
|
Post by cyrano on Apr 17, 2020 20:30:55 GMT -6
Just FYI, the test with HCL has been ended, because most patients died... Do you have a link for that? No, sorry, just one of the many things I've been reading. There are several other tests still going on. |
|
|
|
Post by johneppstein on Apr 17, 2020 21:18:16 GMT -6
I think President Trump is "pushing it" because that's what he does - he's positive, on almost every topic, almost all the time. It's complete garbage to suggest that it's being promoted for profit motive - the drug isn't covered by a patent, it is generic. It's widely used in low doses, has a very low risk of side effect *especially* in short term use (as demonstrated by a meta analysis study n=~900k recently done - I think I linked it here). HCL research into SARS goes back to 2007, it's been shown to inhibit viral growth in monkeys. We know for a fact it is an IL6 inhibitor to boot. The testing wasn't started by President Trump. Well, I've been following Trump for a long time - at least 20 years, long before he got into politics, and my take on him is that he never does anything unless he thinks it puts money in his pocket.
Any more than this I think we should take to PM.
On the science stuff, I'd need time to look up your claims. What I have read in the reputable sources is that Quinoline and Hydroxyquinoline are dangerous and the "evidence" that they are at all effective against coronavirii is purely anectdotal.
|
|
|
|
Post by johneppstein on Apr 17, 2020 21:23:34 GMT -6
And to further muddy the waters, some locales are using "suspected COVID-19" as reason for death even when no tests had been administered. NYC added almost 4K deaths to their tally the other day without having tested a single one of them. Yeah unfortunately there just aren’t the testing supplies! The average shipment is about 600 pieces. The other new problem is now the CDC is encouraging Universities and Reference labs to develop their own tests like the spit test in NJ. The problem with this is if these are considered as diagnostic tools is the error rate will be inconsistent. And the Federal Government has been hijacking shipments of supplies that had been contracted for by the states. Then they turn around and say that the states are responsible for testing, not the feds. |
|
|
|
Post by johneppstein on Apr 17, 2020 21:31:15 GMT -6
BTW, you ARE aware that the reason that Trump is pushing it is not because he has any information about efficacy (and probably couldn't understand it if he did), but because he has substantial stock in the company and he's lining his pockets (or trying to) by pushing it. That's a conflict of interest and makes any endorsement of it by his administration extremely suspect and, yes, tainted.
John - your EXTREME hatred of Trump and anything conservative or "right wing" is getting EXTREMELY wearing..... Just so you know, you're annoying the **** out of a lot of people. Maybe you don't care, but just thought I'd be honest and let you know. That works both ways. The EXTREME hatred of anything some people deem "liberal" or "left wing" is EXTREMELY wearing.
FWIW, I don't consider myself to be either "left" or "right" wing - I'm a radical centrist. I look at the facts and call 'em as I see 'em.
Unfortunately these days the center is the most difficult place to be. It does not bode well for the country, or the world.
If Barry Goldwater was still alive people would be calling him a "communist". I miss him. And I think Mitt Romney is one of the bravest men in the country. McCain was, too. Those guys are the true patriots.
The unwarrented hatred I'm getting here is equalled only by the unwarranted hatred I get elsewhere when I oppose gun control.
BTW, in case you haven't noticed, I fully support all the news posted here from legitemate scientific sources. Those sources are neither "right" nor "left". And those sources are, in fact, the most "conservative" in the true sense of the word. And I'm fortunate in that I can actually understand what they say, where the average person's eyes would probably glaze over in seconds.
Why do I support the "shelter in place lockdown?
It's sure as hell not because I like it. It's thoroughly DESTROYED what my be my band's last chance to do anything - with the condition of my health I might not get another year. Most likely I'm screwed.
I support the lockdown because I happen to live in a place that is living proof that it works. We had the earliest known exposure in the entire country, yet our rate of infection has been level for a couple of weeks and is now dropping, according to the most recent figures. That is not an accident. It's because our local government identified the problem early and took the necessary action. Nobody likes it. Pretty much everybody hates it. Lots of people I know are in danger of losing their businesses. The live music business around here is shut down and may never fully come back. But it works, so we go along with it. Relatively few people are dying.
For what it's worth I'm going to turn 70 in July. I have congestive heart failure, evidence of a "silent" heart attack, COPD, weigh about 270 lb at a height of 5'9", and am bordeline diabetic. A couple of years ago I was hospitalized with a lung embolism. I support the lockdown and quarantine measures BECAUSE WITHOUT THEM I'M ALMOST CERTAINLY DEAD. That's not mentioning the minor issues, like arthritis in both hips and two crushed vertebrae from a truck wreck in my touring days. Why do I oppose the sensational pushing of a drug that is KNOWN to cause sudden and unexpected heart failure? Could it possibly be because for people like me it would almost certainly be fatal? Think about it.
And, FYI, this town is probably one of the "reddest" towns in Northern California, if not the entire state. We are home to Travis Air Force Base, which means that we have a large percentage of military and ex-military people here. (It's also the reason that we're the "lucky" place that got the first cruise ship victims dumped on us.)
|
|
|
|
Post by svart on Apr 17, 2020 21:54:37 GMT -6
|
|
|
|
Post by matt@IAA on Apr 17, 2020 22:03:01 GMT -6
It’s hydroxychloroquine. And yes, the evidence of efficacy is anecdotal. That’s why you do trials, to turn anecdotes into evidence.
It’s not dangerous, comparatively. All drugs have side effects but millions of people take HCL for malaria prophylaxis and RA or lupus.
|
|
|
|
Post by johneppstein on Apr 17, 2020 22:43:34 GMT -6
From a political journal. Not scientific.
Again, they put their politics right on their masthead.
Please, I'd like to see some real science, not propaganda.
When I can google a publication and the political bias comes up first thing it's not a scientific, impartial source.
There are many widely recognized scientific publications that are not compromised by politics. Let's see some references from them.
"Science", "Scientific American", "The New England Journal of Medicine" to name just three off the top of my head. Why no references to those?
I haven't read through nearly all of it yet, but it's vetted by recognized people in the scientific community
Scroll down to where it says "Potential therapeutic strategies against COVID-19", the stuff before that will put you to sleep. mFrom the "Conclusions" sectiopn of the article "Remdesivir, Lopinavir, Ritonavir, and Oseltamivir significantly blocked the COVID-19 infection in infected patients." Not one mention of the faddist Quinoline drugs.
This does actually mention Chloroquine and Hydroxychloroquine and gives a pretty fair evaluation - the first one I've seen anywhere that wasn't tainted in one direction or the other.
Sci-Am may not let you read more than one or two articles a month without a paid subscription - any more and you just get the first part. I have just plonked down my $35 for a print and digital subscription and will be glad to copy and paste articles on this subject that might be of interest to readers here on request, if possible.
I really wish that the article above were true. But it's not from a reliable, unbiased source, it's an unabashed political publication.
Incidentally, FWIW, I identified as "libertarian" for quite awhile until things went off the rails.
|
|
|
|
Post by matt@IAA on Apr 17, 2020 22:46:28 GMT -6
www.medrxiv.org/content/10.1101/2020.04.14.20062463v1They got 50 positives out of 3300 for a raw weight of 1.5% infected. They then adjusted this to 2.81% for <<reasons>> (demographic sampling? I don't understand these reasons, happy for someone to explain). When you adjust the raw 1.5% for the false positive/negative rate of their test you get 1.27% (95% CI 0.82% to 1.87%) (using the combined sensitivity number they give, scenario 3). Assuming that is the infection rate, Santa Clara county has 1.928 million people, and 73 deaths. The estimated total number of infections is then 24,485 for an IFR of 0.3%. (95% CI range implies 0.2% to 0.46%). Getting out into speculation land - with that, you'd expect a high R0 - at least 2, implying likely 50% of the population would ultimately be infected at minimum. With a true IFR of 0.3% the final number of fatalities for the US would be 480,000 (95% range implies 320,000-736,000) over the next 12ish months.
|
|
|
|
Post by svart on Apr 17, 2020 22:51:06 GMT -6
From a political journal. Not scientific.
Again, they put their politics right on their masthead.
Please, I'd like to see some real science, not propaganda.
When I can google a publication and the political bias comes up first thing it's not a scientific, impartial source.
There are many widely recognized scientific publications that are not compromised by politics. Let's see some references from them.
"Science", "Scientific American", "The New England Journal of Medicine" to name just three off the top of my head. Why no references to those?
I really wish that the article above were true. But it's not from a reliable, unbiased source, it's an unabashed political publication.
Incidentally, FWIW, I identified as "libertarian" for quite awhile until things went off the rails.
There's a link to the study in the article, if you would stop playing political victim for a moment and read. |
|
|
|
Post by drbill on Apr 17, 2020 23:08:53 GMT -6
johneppstein - it's not just the Covid threads John. It's any thread you have a chance to bash Trump. You take aim and let loose. I swear I could ask you about what the best strings are for my Tele - and you are the guy to know - and you could find a way to bash Trump. Not subtle. Not infrequent. And not appreciated by oh....about 50% of the US who voted for him. The other 50% cheer you on! Creating division which is tearing our country apart - no matter which side you sit on. But somehow, I thought this was a no politics zone?
|
|
|
|
Post by johneppstein on Apr 17, 2020 23:42:13 GMT -6
www.medrxiv.org/content/10.1101/2020.04.14.20062463v1They got 50 positives out of 3300 for a raw weight of 1.5% infected. They then adjusted this to 2.81% for <<reasons>> (demographic sampling? I don't understand these reasons, happy for someone to explain). When you adjust the raw 1.5% for the false positive/negative rate of their test you get 1.27% (95% CI 0.82% to 1.87%) (using the combined sensitivity number they give, scenario 3). Assuming that is the infection rate, Santa Clara county has 1.928 million people, and 73 deaths. The estimated total number of infections is then 24,485 for an IFR of 0.3%. (95% CI range implies 0.2% to 0.46%). Getting out into speculation land - with that, you'd expect a high R0 - at least 2, implying likely 50% of the population would ultimately be infected at minimum. With a true IFR of 0.3% the final number of fatalities for the US would be 480,000 (95% range implies 320,000-736,000) over the next 12ish months. Interesting article, from all appearances unbiased, but not yet vetted - it's a preliminary.
One thing I noticed is that the abstract of the article reports that the level of infection appears to be much higher than the "official" reports.
One thing that I would consider a possible (potentially serious) flaw is that they used Facebook to recruit their subjects, which could bias the overall results. Dunno how that would skew things though.
|
|
|
|
Post by matt@IAA on Apr 17, 2020 23:50:46 GMT -6
Of course it’s biased. Three of the authors have published opeds previously about how it was overblown and they basically recruited a bunch of karens for the study. And the test they’re using apparently has garbage sensitivity. But that doesn’t mean it’s not good science or a useful data point.
|
|
|
|
Post by johneppstein on Apr 18, 2020 0:32:25 GMT -6
johneppstein - it's not just the Covid threads John. It's any thread you have a chance to bash Trump. You take aim and let loose. I swear I could ask you about what the best strings are for my Tele - and you are the guy to know - and you could find a way to bash Trump. Not subtle. Not infrequent. And not appreciated by oh....about 50% of the US who voted for him. The other 50% cheer you on! Creating division which is tearing our country apart - no matter which side you sit on. But somehow, I thought this was a no politics zone? Since when did I say anything about Trump and guitar strings? AFAIK he has no musical ability or interest at all, not that I care. I suppose that could be interpreted as a negative comment, but it's not intended that way.
And yeah, I don't like the guy but there's really nothing political about it. Don't like his dad, either, which goes back to a beef Woody Guthrie had with him way back in 1950. Business, that's a different matter. Matter of fact, I don't believe that Trump himself is really political at all, at least in the traditional sense. He's an opportunist, does whatever he thinks is good for him. You know he used to be a "democrat", right? Actually, he's probably a lot like P.T. Barnum - with a different upbringing he could have been a great showman.
|
|
|
|
Post by johneppstein on Apr 18, 2020 0:36:04 GMT -6
Of course it’s biased. Three of the authors have published opeds previously about how it was overblown and they basically recruited a bunch of karens for the study. And the test they’re using apparently has garbage sensitivity. But that doesn’t mean it’s not good science or a useful data point. Good science strives as hard as humanly possible to be unbiased. Anything else is not good science.
The big problem with many hard science publications is that they can be rather difficult for laymen to read.
|
|
|
|
Post by chessparov on Apr 18, 2020 2:25:02 GMT -6
Well as long as all of you (and your loved ones) stay "safe and sound", through all this...
That's my main hope here at RGO.
I make the motion, we defer any other "regularly scheduled" bickering... To when after bathroom tissue store supplies, have been fully stabilized. Thus easing mounting tensions.
Chris
|
|
|
|
Post by johneppstein on Apr 18, 2020 2:28:04 GMT -6
There's a link to the study in the article, if you would stop playing political victim for a moment and read. Oh, Ok. I'f rather not have to read through a bunch os extraneous verbiage to get to the study. but OK.
That's the Santa Clara Facebook study.
This is the preliminary report on the study; it is not yet vetted. That may or may not be significant, depending of the outcome of the peer approval process.
I question the wisdom of recruiting subjects via Facebook, that alone could skew the results. In fact, the authors admitted that the study is racially biased, with the percentage of African-Americans and Hispanics in the study greatly under represented. Those are the two ethnic groups that show the greatest prevalence of Covid infections nation wide.
Also, Santa Clara went into quarantine ln in mid-March and the study was undertaken in the beginning of April, a difference of two weeks which is about the length of time it takes for the quarantine to start showing results, so it's not surprising that the percentage of infected subjects is somewhat low.
HMMmmm...... "50 to 85 fold larger", eh? That's pretty significant if it's true. It potentially means two things: One, that the fatality rate could be significantly lower than assumed, and Two, that the potential spread of the virus could be much, much worse than assumed to date. So is this good news or bad news? I'd say Yes, it is. Both. But we don't know if it is and neither do the authors, hence the word "suggests".
Also the fact that the subjects were recruited via Facebook makes me suspect that the sample may be skewed in regards to age, with older folks under represented. If the authors mentioned the age factor I missed it.
All in all it's not too bad a study for a preliminary report. It will be interesting to see how it fares in the peer approval process.
|
|
|
|
Post by Johnkenn on Apr 18, 2020 8:35:15 GMT -6
Of course it’s biased. Three of the authors have published opeds previously about how it was overblown and they basically recruited a bunch of karens for the study. And the test they’re using apparently has garbage sensitivity. But that doesn’t mean it’s not good science or a useful data point. Stop...they’re Scientists. At Stanford. |
|
|
|
Post by Johnkenn on Apr 18, 2020 8:36:03 GMT -6
|
|
|
|
Post by svart on Apr 18, 2020 10:39:57 GMT -6
Imagine being so self centered that you want everyone to lose their livelihoods so you can feel a little safer.
|
|
|
|
Post by matt@IAA on Apr 18, 2020 14:49:23 GMT -6
Of course it’s biased. Three of the authors have published opeds previously about how it was overblown and they basically recruited a bunch of karens for the study. And the test they’re using apparently has garbage sensitivity. But that doesn’t mean it’s not good science or a useful data point. Stop...they’re Scientists. At Stanford. Bias isn't bad. Bias exists everywhere, all the time. Svart has a bias for this thing being overblown. You can tell by the selection of articles he chooses to post. John E has a bias to believe negative things about President Trump. I have a bias to read papers because I am a nerd. Weeks ago, John Ionnadis wrote a paper speculating about a 0.05% IFR. I think you might have linked it? And Bendavid and Bhattcharya wrote the WSJ op-ed speculating about a 0.01% IFR. Being a scientist at Stanford doesn't stop making you human, and all humans have bias. These guys believed it was less severe - so much they decided to go do their own study (none of them are epidemiologists!). It's perhaps rare in academia to know in advance, in print, the biases of the person who conducts a study like this (unless you know them / their previous work I suppose). Here was nice, we know exactly what they were looking for. On the other hand, I don't think they got the result they wanted. And I do think their bias shows in how they reported their numbers. For example, they used a demographic adjustment to go from 1.5% (naive prevalence - 50 positives out of 3300 tests) to 2.8%.. and then they adjusted the number for the sensitivity and specificty of their test. That's a bit wonky. And, they didn't really report a thorough 95% confidence interval on their test results. Within the confidence interval of the specificity of the test it is possible all 50 of their positives were false positives. For a nice explanation of how this works out, here: metarabbit.wordpress.com/2020/04/18/did-anyone-in-santa-clara-county-get-covid-19/This works out to a true positive rate of 1.5 with 95% confidence interval ranging from 0.2% to 2.4%. And this doesn't get into the limitations of study design - which they note in their paper, because they are good and honest scientists. |
|
|
|
Post by svart on Apr 18, 2020 16:28:03 GMT -6
|
|
|
|
Post by Ward on Apr 19, 2020 19:19:44 GMT -6
This may rub some people the wrong way . . . another actual expert.
|
|
|
|
Post by svart on Apr 19, 2020 20:06:14 GMT -6
This may rub some people the wrong way . . . another actual expert. Can't wait to see folks rush to condemn what he's saying because they're scared. Trust the scientists! Unless of course he's a scientist who deals with with both the statistics and the diseases and can tell us the truth without the media spin and the political ineptitude. In that case dismiss what he has to say because to accept it now would mean you fell prey to the fear-n-panic headlines. |
|
