International Skeptics Forum

International Skeptics Forum (http://www.internationalskeptics.com/forums/forumindex.php)
-   Science, Mathematics, Medicine, and Technology (http://www.internationalskeptics.com/forums/forumdisplay.php?f=5)
-   -   Continuation The One Covid-19 Science and Medicine Thread Part 4 (http://www.internationalskeptics.com/forums/showthread.php?t=354459)

marting 13th November 2021 04:33 PM

Quote:

Originally Posted by The Atheist (Post 13654747)
Jesus mate, if you're spotting these errors that easily, they should be being stopped long before publication. Have you contacted them?

Always the damned numbers.

I just sent them specific info that hopefully will get read by someone numerically literate. But I don't have a lot of hope. I'm just a random member of the public and no doubt they are inundated by people wondering about stupid things like Luciferace in vaccines.

I'm trying to use their numbers to estimate the population recently infected in Florida which had really high Covid-19 cases/deaths but are now at the lowest levels in the USA. I suspect a rather high percentage of the remaining unvaxxed and not previously infected older people caught it and they will now see little spread. I want to use that to estimate what's going to happen in colder but somewhat more vaaccinated areas of the country.

The Atheist 13th November 2021 06:26 PM

Quote:

Originally Posted by marting (Post 13654776)
... what's going to happen in colder but somewhat more vaaccinated areas of the country.

:dl:

You even got the accent right!

Puppycow 13th November 2021 07:05 PM

Quote:

Originally Posted by marting (Post 13654776)
I'm trying to use their numbers to estimate the population recently infected in Florida which had really high Covid-19 cases/deaths but are now at the lowest levels in the USA. I suspect a rather high percentage of the remaining unvaxxed and not previously infected older people caught it and they will now see little spread. I want to use that to estimate what's going to happen in colder but somewhat more vaaccinated areas of the country.

This is probably an artifact in the data due to the way Florida reports deaths due to Covid:

https://www.news4jax.com/news/local/...recent-deaths/

Now, I'm not claiming they did it for nefarious reasons, but it does mean that you cannot take their most recent death figures at face value because more deaths will almost certainly be reported in the future.

Quote:

Following a report from the Miami Herald that the Florida Department of Health changed the way it reported death data to the CDC, giving the appearance of a pandemic in decline, Gov. Ron DeSantis pushed back Wednesday, calling the story a “totally false partisan narrative.”

According to the Herald, on Aug. 10, just as the surge of new infections fueled by the delta variant hit the state, Florida began to report new deaths to the CDC by the date the person died. Before that, data collected by FDOH and published by the CDC counted deaths by the date they were recorded, a common method used by the majority of states.

The result is that it is harder to see how many new deaths are added each day, as the deaths are spread out over several days of data, as opposed to being reported on a single day.

“If you chart deaths by Florida’s new method, based on date of death, it will generally appear — even during a spike like the present — that deaths are on a recent downslope. That’s because it takes time for deaths to be evaluated and death certificates processed,” according to The Herald. “When those deaths finally are tallied, they are assigned to the actual date of death — creating a spike where there once existed a downslope and moving the downslope forward in time.”

Puppycow 13th November 2021 07:24 PM

I found info for Florida, but you have to keep the above in mind when reading it:

https://www.wptv.com/coronavirus

I'm also seeing some possible discrepancies in these numbers (did deaths increase by 363 or 502 in the last week?)

dann 13th November 2021 07:33 PM

Sweden has been doing the same thing as Florida is doing now throughout the pandemic: Why do COVID-19 deaths in Sweden’s official data always appear to decrease? (Our World in Data, Nov 13, 2020)

Sweden also doesn't test children, and this month they stopped testing people who are vaccinated unless they are ripe for hospitalization, which goes against the guidelines of the ECDC:
Quote:

Att fullvaccinerade svenskar inte längre behöver testa sig för covid vid symtom har väckt kritik. Den nedtrappade testningen går också emot den europeiska smittskyddsmyndigheten ECDC:s riktlinjer.
Sveriges testning går emot Europas riktlinjer (SvD, Nov 12, 2021)

As Trump knew, it makes the numbers look very good, which is the only thing that matters to some politicians.

marting 13th November 2021 07:48 PM

Quote:

Originally Posted by Puppycow (Post 13654835)
This is probably an artifact in the data due to the way Florida reports deaths due to Covid:

I was referring to new daily case rates, not deaths. The latter is indeed reported weeks late and the daily deaths have been very underreported. The positive side is that the numbers are more accurate if you go back at least 3 weeks. I discussed this change here:
http://www.internationalskeptics.com...a#post13586935

The new daily cases are far down, over 90% down from a few months back. Deaths, however, when accurately reported are only down about 75%. There's a long tail in the death distribution from the case instance. Here's a source that is pretty accurate for daily deaths in Florida.

https://newsnodes.com/us_state/FL

Note that about 1/3 of the States in the USA report deaths the same way Florida does so reading the CDC or WorldOMeters recent daily death trends is quite misleading. However, the daily Case numbers are not delayed.

Puppycow 13th November 2021 08:25 PM

3 snow leopards killed by COVID at Lincoln Children’s Zoo in Nebraska

Snow leopards are an endangered species.

Maybe they caught it from a zookeeper?

How SARS-CoV-2 in American deer could alter the course of the global pandemic

Quote:

A recent survey of white-tailed deer in the Northeast and Midwest found that 40% of them had antibodies against SARS-CoV-2.

Now veterinarians at Pennsylvania State University have found active SARS-CoV-2 infections in at least 30% of deer tested across Iowa during 2020. Their study, published online last week, suggests that white-tailed deer could become what's known as a reservoir for SARS-CoV-2. That is, the animals could carry the virus indefinitely and spread it back to humans periodically.
If that's true, then it would seem to follow that you can indeed catch the virus outdoors. If it's spreading among wild deer, it's obviously spreading outdoors.

dann 13th November 2021 09:20 PM

No, it doesn't follow:
Quote:

Sniffing (all deer, all seasons)
Deer sniff each other when they meet. They touch noses or sniff the tarsal glands. Tarsal sniffing may be the way deer recognize family or group members. Deer may be able to determine age and sex. Tarsal sniffing of strangers is often followed by chasing or striking.
Deer Talk (A Field Guide to Whitetail Communication)

It just follows that deer do, due to deer behavior. It may even have been the way they got it from humans - if that's where they got it.

Puppycow 14th November 2021 05:21 PM

Quote:

Originally Posted by dann (Post 13654903)
No, it doesn't follow:



It just follows that deer do, due to deer behavior. It may even have been the way they got it from humans - if that's where they got it.

Fair point. Deer behavior could explain it.

marting 14th November 2021 07:21 PM

On the current CDC esimated covid-19 disease burden page they state this:

1 in 1.9 (95% UI* 1.7 – 2.1) COVID–19 hospitalizations were reported.

This seems pretty high. How can only just over 50% of Covid-19 hospitalizations be reported? Are the USA regulations so bad that they often don't bother to report these? I've not run across this before and it seems pretty unbelievable. Perhaps at the start of Covid-19 some people were hospitalized and they didn't know for sure if they had covid-19 but a lot of time (and hospitalizations) has gone by since then. Hard to believe that number unless hospitaliations for Covid-19 doesn't actually have to be reported. Anyone know?

https://www.cdc.gov/coronavirus/2019...es/burden.html

Chris_Halkides 14th November 2021 08:32 PM

antibody dependent enhancement issue
 
If someone could point me to a good resource for information on this issue as it pertains to vaccines, I would be grateful.

marting 14th November 2021 08:58 PM

Quote:

Originally Posted by Chris_Halkides (Post 13655418)
If someone could point me to a good resource for information on this issue as it pertains to vaccines, I would be grateful.

ADE was a risk factor considered during vaccine dev.

Here's a Nature article that goes into the risks and what was known very early in vaccine dev. I'd start there and follow references as needed. Michael Yeadon was apoplectic over ADE. Don't know if he's calmed down.

https://www.nature.com/articles/s41564-020-00789-5

Chris_Halkides 15th November 2021 04:59 AM

A Pipeline to useful information
 
Quote:

Originally Posted by Chris_Halkides (Post 13655418)
If someone could point me to a good resource for information on this issue as it pertains to vaccines, I would be grateful.

It looks as if In the Pipeline's Derek Lowe has covered it in several entries, such as this one from December of last year. There is also an entry from February of this year. "Antibody-dependent enhancement was specifically tested for in the animal models as these candidates were being developed (re-exposure of vaccinated animals to coronavirus to see how protective the vaccine was). And no cases of more severe disease were seen - I've gone back through the reported preclinical studies, and I don't think I've missed one, and what I'm seeing is not one single case of ADE for any of them. Indeed, as mentioned above, if something like that had shown up, it would have immediately released a bucket of clin-dev and regulatory sand into the gears of the whole project."
EDT
I know just enough about immunology to be dangerous, but I would be at least as concerned about ADE arising from so-called natural immunity. It would see as if there would be many non-neutralizing antibodies created during the course of an infection.

marting 15th November 2021 09:10 AM

Quote:

Originally Posted by Chris_Halkides (Post 13655557)
I know just enough about immunology to be dangerous, but I would be at least as concerned about ADE arising from so-called natural immunity. It would see as if there would be many non-neutralizing antibodies created during the course of an infection.

That was my take as well back when I decided to get Pfizer. I rather like the idea they focused on the spike. Seemed safer and it's panned out to have been the right choice.

Klimax 15th November 2021 09:40 AM

Quote:

Originally Posted by marting (Post 13655729)
That was my take as well back when I decided to get Pfizer. I rather like the idea they focused on the spike. Seemed safer and it's panned out to have been the right choice.

It was obvious choice (Anybody with vaccine that can target subunits of virus did so), the only way to infect a cell (and thus anything attached is likely to reduce ability to bind) and most mutations that would evade induced immunity will make it less binding too.

Chris_Halkides 15th November 2021 10:38 AM

in silico study of ADE
 
There was a third In the Pipeline entry on the possibility of antibody dependent enhancement in August of this year. As part of looking at a pre-print, Derek Lowe wrote something that IMO is applicable more generally: "This is a constant danger with simulations. Readers who have not encountered much molecular modeling are often (and understandably) impressed by the graphics and tables that appear with such work, but if you've been involved with actual experimentation you've seen many, many examples of such hypotheses that turned out to be built on air. Confusing the graphics with reality is a constant danger for all of us."

Planigale 15th November 2021 01:22 PM

Quote:

Originally Posted by marting (Post 13655729)
That was my take as well back when I decided to get Pfizer. I rather like the idea they focused on the spike. Seemed safer and it's panned out to have been the right choice.

All the available vaccines focussed on the spike protein. So not sure where your thinking is going. Pfizer definitely has a shorter duration of protection than the AZ vaccine. The efficacy seems to cross at about 4 months.

All the licensed vaccines seem in absolute terms very effective and very safe. Long term safety and efficacy are hard to determine, as we do not have directly comparable studies over a long enough period to look for this.

Planigale 15th November 2021 01:25 PM

Quote:

Originally Posted by Klimax (Post 13655758)
It was obvious choice (Anybody with vaccine that can target subunits of virus did so), the only way to infect a cell (and thus anything attached is likely to reduce ability to bind) and most mutations that would evade induced immunity will make it less binding too.

Although the evidence is not in favour so far, in theory a whole virus vaccine should be best, as it gives the body lots of sites to 'attack'.

marting 15th November 2021 01:53 PM

Quote:

Originally Posted by Planigale (Post 13655974)
All the available vaccines focussed on the spike protein. So not sure where your thinking is going. Pfizer definitely has a shorter duration of protection than the AZ vaccine. The efficacy seems to cross at about 4 months.

Didn't mean to suggest I picked Pfizer because other vaccines hadn't focussed on the spike. It was just the first that was offered here. Rather, I just liked the vax dev focus on the spike. As for the crossover compared to AZ where it started higher but dropped off more rapidly, it will be interesting to see what the third doses do though it will likely be similar.

Dr.Sid 15th November 2021 04:46 PM

Pfizer hits 50% effectiveness against infection in 6 months, Moderna is just slightly better.
I wouldn't call that good effectiveness. Delta spreads 3 times better than alpha .. so anything under 66% only pushed delta numbers back to alpha numbers. Some delta subvariants are even better. That gives you zero improvement in about 4 months after vaccination.
Sure, it still greatly improves your chances if you get it .. and vaccination + infection gives wider protection. And longer lasting too, as the reinfection rate after year is pretty low.
Some first data also suggest the protection lasts longer after third dose, so there's some hope. But this first wave of vaccination will not stop Covid just yet.

The Atheist 16th November 2021 12:38 AM

Quote:

Originally Posted by Dr.Sid (Post 13656098)
Pfizer hits 50% effectiveness against infection in 6 months, Moderna is just slightly better.
I wouldn't call that good effectiveness. Delta spreads 3 times better than alpha .. so anything under 66% only pushed delta numbers back to alpha numbers. Some delta subvariants are even better. That gives you zero improvement in about 4 months after vaccination.
Sure, it still greatly improves your chances if you get it .. and vaccination + infection gives wider protection. And longer lasting too, as the reinfection rate after year is pretty low.
Some first data also suggest the protection lasts longer after third dose, so there's some hope. But this first wave of vaccination will not stop Covid just yet.

Yet again, the place to watch will be Israel - they're mostly 3 shots now, I gather, and if it breaks out badly there again, we are completely in the soup.

Dr.Sid 16th November 2021 07:01 AM

My favorite page for Czech data (it's actually complete private project !) now includes also reinfection rates. And it's about 2%. That would be plenty to get rid even of Delta .. if everybody got the disease that is.

Here is google-translated link into English: https://www-covdata-cz.translate.goo...&_x_tr_pto=nui

If you click [Czech/age] in the main menu there is also graph for vaccination effectiveness in every age group, for infection, hospitalization and intense care.

The Atheist 16th November 2021 01:39 PM

More good news - Pfizer joins Merck in allowing production of its successful treatment.

https://www.stuff.co.nz/world/300456...s-covid19-pill

Puppycow 16th November 2021 07:42 PM

New cases in Japan are now lower than in New Zealand.

The Atheist 16th November 2021 08:15 PM

Quote:

Originally Posted by Puppycow (Post 13657279)
New cases in Japan are now lower than in New Zealand.

Have no fear, we'll be passing a few countries in the near future. Next target is Aussie and I've picked us to pass them just before xmas, by when we'll be seeing 1000 cases a day.

And it's all down to one particular group of people - gang members.

marting 16th November 2021 08:56 PM

Quote:

Originally Posted by marting (Post 13654708)
One thing about the fringe, is they tend to be more alert to things that make no sense. @Covid19Crusher posted a screenshot of the latest Covid-19 Disease Burden which contained an obvious impossibility. Specifically that, for Covid-19, the total number of estimated infections was lower than the total number of estimated symptomatic cases for ages 0-17.

Here's a link: https://www.cdc.gov/coronavirus/2019...es/burden.html

Infections per 100,000 for 0-17 y/o: 29885
Symptomatics per 100,000 for 0-17 y/o: 30253

I figured the CDC would correct an obvious typo or error in a few days. I was wrong. So I took a closer look and in a few minutes noticed multiple discrepancies which indicate no one bothered to even review the document for sanity before posting it.

The not so bad:
In Table 1 they list numbers of people in each of the age groups together with all ages. The age groups don't add up to the all ages group. But at least they are close. Since these are estimates, perhaps they included estimates of a small group with unknown ages. Problem is some of the columns would require a negative number of unknown ages. But I give them a pass on this. Their models for each age group could be different for their models for all ages and produce small differences so they don't add up.

The truly egregious:
In Table 2 listing a higher rate for infections v. symptomatics, which @Covid19Crusher caught ranks right up their with splendidly wrong. But was it a typo or did they copy the wrong entry? Hard to tell. Here's another glaring error from Table 1's entry of the 0-17 y/o:

Point Number: 25,844,005
Conf. Interval: 18,861,476 – 25,408,407

Wow! The Point number (central estimate) is above the upper limit on the CI. Must be some sort of new math. Grr.

So egregiously wrong data in both Table 1 and Table 2. And it's something that should have been noticed by anyone reviewing it. Thanks a lot for feeding the deniers claiming the CDC's data can't be trusted. Grrrr.

Per TA's suggestion I emailed the CDC about this on the 14th. It's was, um, updated today, Nov 16, 2021 and now looks good. I'm likely one of many that reported it.

Puppycow 16th November 2021 09:09 PM

Quote:

Originally Posted by marting (Post 13657335)
Per TA's suggestion I emailed the CDC about this on the 14th. It's was, um, updated today, Nov 16, 2021 and now looks good. I'm likely one of many that reported it.

Something is still off with the graph. It shows too many deaths in the 0-17 age group. The numbers also add up to 104%. (It should be 0.07%, not 4%)

marting 16th November 2021 09:43 PM

Quote:

Originally Posted by Puppycow (Post 13657341)
Something is still off with the graph. It shows too many deaths in the 0-17 age group. The numbers also add up to 104%. (It should be 0.07%, not 4%)

Good catch. I hadn't looked at the graphs or their linked data tables. They appeared to be trying to restate Table 1 in a simplified form.

All the death numbers are off. I get (to 2 sig.)
0-17: .07
18-49: 6.6
50-65: 17
65+: 76

Pretty sloppy.

Orphia Nay 16th November 2021 11:00 PM

Quote:

Originally Posted by Puppycow (Post 13657279)
New cases in Japan are now lower than in New Zealand.


That's good news!

dann 17th November 2021 10:46 PM

For Japan, but not for New Zealand.

Klimax 18th November 2021 01:07 AM

Quote:

Originally Posted by Planigale (Post 13655974)
All the available vaccines focussed on the spike protein. So not sure where your thinking is going. Pfizer definitely has a shorter duration of protection than the AZ vaccine. The efficacy seems to cross at about 4 months.

All the licensed vaccines seem in absolute terms very effective and very safe. Long term safety and efficacy are hard to determine, as we do not have directly comparable studies over a long enough period to look for this.

Does it? Only some studies show strong decrease in protection by Pfizer, while others don't.

Quote:

Originally Posted by Planigale (Post 13655977)
Although the evidence is not in favour so far, in theory a whole virus vaccine should be best, as it gives the body lots of sites to 'attack'.

Incorrect. Any antibody that targets non-spike parts is a bit wasted, because even if they bind to that protein, they will never prevent infection of cell. (It is race between virus particle infecting cell and antibody-triggered reaction) And non-spike proteins have full mutation space available.

The Atheist 18th November 2021 02:10 AM

Interesting news out of UK regarding the newest Delta+ variant.

It'd be quite helpful if it mutated into a more infectious, but less-harmful bug. Jury still out though.

https://www.theguardian.com/world/20...cause-symptoms

marting 18th November 2021 12:53 PM

This brings into question how efficatious vaccination is for elderly in nursing homes against both transmission and deaths after 9-10 months post vax.

In a skilled nursing home with 70 residents, 67 residents and 22 staff got Covid-19. 8 residents died. 98% of the staff and residents were fully vaccinated but boosters had not yet rolled out. The 8 dead residents had comorbitities in addition to age as is typical in nursing homes.

93% of nearly fully vaccinated is a damned high attack rate showing no significant effect at preventing infection.

A 12% mortality rate in this population is very high and similar to unvaxxed mortality in other nursing homes with frail elderly residents.

Arguably the mortality rate could be a statistical anomoly but the infection rate is beyond anomolous. It's not stated when the residents were vaxxed but likely in Jan/Feb so this is pretty compelling data that the vaccine efficacy almost completely wanes after 10 months in this population.

https://www.newsweek.com/covid-19-sp...deaths-1650473

The Atheist 18th November 2021 04:30 PM

Quote:

Originally Posted by marting (Post 13658495)
This brings into question how efficatious vaccination is for elderly in nursing homes against both transmission and deaths after 9-10 months post vax.

We're walking down that path here too, right now.

Our rest homes were first in line and are now 9 months or so past their second dose. It's already hitting rest homes here and the deaths in the past week have all been fully vaxed oldies. Only half a dozen or so yet, but 100% of 6 is still 100%, and there have only been a small number of cases from two rest homes with infections to date.

Boosters were approved last week, and I imagine they're going pretty fast right now in the 70+ age group.

marting 18th November 2021 10:14 PM

This study seems the most consistent with the high transmission and mortality in the nursing home I posted recently:

Effectiveness of Covid-19 Vaccination Against Risk of Symptomatic Infection, Hospitalization, and Death Up to 9 Months: A Swedish Total-Population Cohort Study

https://papers.ssrn.com/sol3/papers....act_id=3949410

Quote:

In this study, vaccine effectiveness of BNT162b2 against symptomatic infection waned progressively from 92% during the first month, to 47% by month 4-6 and from 7 months and onwards no effectiveness was detected. Effectiveness waned slightly slower for mRNA-1273, whereas effectiveness of ChAdOx1 nCoV-19 was generally lower. Overall, effectiveness was lower and waned faster among men and older individuals. For the outcome of hospitalization or death, effectiveness (any vaccine) waned from 89% during the first month to 42% from month 6 and onwards in the total population. There was notable waning among especially men, older frail individuals, and individuals with comorbidities

Klimax 19th November 2021 12:58 AM

Quote:

Originally Posted by marting (Post 13658775)
This study seems the most consistent with the high transmission and mortality in the nursing home I posted recently:

Effectiveness of Covid-19 Vaccination Against Risk of Symptomatic Infection, Hospitalization, and Death Up to 9 Months: A Swedish Total-Population Cohort Study

https://papers.ssrn.com/sol3/papers....act_id=3949410

Oh, look once again link to that crap paper, where issue of p=0,07 is ignored. Don't you have any better papers on alleged loss of effectivity than that? (Not even speaking about that idiocy of analyzing statistical output when p=0,
21) Confidence intervals are about as useful as saying you are on planet Earth about your location.

Kindly, stop linking to that crap. It just spreads bad information.

jt512 19th November 2021 02:09 AM

Quote:

Originally Posted by Klimax (Post 13658839)
Oh, look once again link to that crap paper, where issue of p=0,07 is ignored. Don't you have any better papers on alleged loss of effectivity than that? (Not even speaking about that idiocy of analyzing statistical output when p=0,
21) Confidence intervals are about as useful as saying you are on planet Earth about your location.

Huh? What exactly about that paper are you finding fault with?

jt512 19th November 2021 02:33 AM

n.m.

marting 19th November 2021 09:46 AM

Quote:

Originally Posted by Klimax (Post 13658839)
Oh, look once again link to that crap paper, where issue of p=0,07 is ignored. Don't you have any better papers on alleged loss of effectivity than that? (Not even speaking about that idiocy of analyzing statistical output when p=0,
21) Confidence intervals are about as useful as saying you are on planet Earth about your location.

Kindly, stop linking to that crap. It just spreads bad information.

Apparently you haven't understood what I wrote in a prior response on this from post #300 so I'll repeat and elaborate.

It's really standard stat. speak when p > .05. The last part does indicate waning after 211 days with a point Ve of 23% but a large CI. For that matter, detecting no protection is never possible. At best one can determine that a benefit or harm is statistically likely. Actually determining no effect can't be done. So what the paper is saying is that the p value is to high to exclude no efficacy in the 95% sense. But it does show some efficacy is likely but with a broad range.

For any population size, as the benefit/harm effect approaches 0, 95% of the time you will get a p value > .05. That's really the definition of the null hypothesis. In simple terms you can't actually determine no effect statistically when there is, in fact, no effect.

The smaller the benefit, the larger population size it takes to expect a p<.05.

Take this from the paper: "From day 211 and onwards no effectiveness could be detected (23%; 95% CI, -2-41, P=0·07).

For example if they had more infections in this group such that the point value was 26% instead of 23% they would have a p<.05 and be able to say the group showed not only a significant decline in efficacy but that it also showed a statistically significant (non zero) efficacy. Since the p value was above .05, they simply gave the point number of 23% and were unable to show a significant efficacy.

The point the paper was making was in support of booster shots and it adds to the evidence boosters are needed.

zooterkin 19th November 2021 10:28 AM

Did you mean to quote my post? I'm not seeing the relevance.


All times are GMT -7. The time now is 11:19 AM.

Powered by vBulletin. Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.
© 2015-22, TribeTech AB. All Rights Reserved.