covid-19

Quote:

---

Originally Posted by RandBlade

No its a closely guarded secret for most countries it seems.

---

Not necessarily. What is a closely guarded secret is the arcane knowledge required to make healthcare systems and governments report accurate data in a timely manner. The Dutch situation is an illustrative example: the officially reported figure for doses administered is half of what the govt. estimates it to be. Many/most places haven't set up automatic reporting yet; health info systems and govt. info systems are extremely fragmented, both within and between countries.

Quote:

---

Originally Posted by RandBlade

The latest available UK data is there are currently 34,783 patients in hospital with Covid 19. Approximately 1 in every 1916 people are currently hospitalised.

Across the different vaccines in the trials quoted that would be 38 people currently hospitalised - and that's only on current data. It doesn't involve any churn over a period of months as real life does.

Age groups may distort that but it's worth noting that while over 80s make up a majority of deaths they are a minority of hospitalisations. The bulk of the hospitalisations are actually 40-65 but the difference is they're more likely to recover while over 80s struggle to do so.

---

Alright, got curious, and checked the statistics, assuming perfect sampling, for a yes/no questions such as hospitalization, AZN data is insufficient to state with 95% confidence that the proportion of vaccinated people is 1/2000 or better (i.e. the same as your actual rate). Close though, but your sample size is too small (very close though, you'd need 5990).

Of course that's excluding any other information you have regarding the actual distribution, you do of course have data on infections and severity of complaints, so it's not really non parametric data. But according to the numbers in that tweet, I can't conclude that AZN is even as good as not being vaccinated, even if you assume perfect sampling.

I will stop nitpicking now, I just like statistics :p

Quote:

---

Originally Posted by wiggin

I am continually depressed how most governments - mine included - have done such an awful job on the rollout. Supply isn't the rate limiting step here and it should be! All this requires is basic coordination - get X people into location Y at Z time, and have an appropriate dose and healthcare worker there to administer it. It's logistics, that's it. And they had months and months to prepare for it. Every country that isn't Israel should be ashamed right now if they aren't pushing against their supply constraints.

I have a friend who now works in biotech and he's livid about the mess. He had a good point: Pfizer delivered an incredibly sophisticated product on a ridiculously short time scale - every single box of doses has GPS tracking, detailed temperature history, and a continuous record of custody and batch production going back to raw materials. I doubt if there has ever been this scale of distribution of something with such restrictive storage conditions on such a short time scale. They know exactly where and how much vaccine there is and have shown that it works. Yet the government simply needs to get the damn thing into people's arms and they can't manage that. It's embarrassing, and we should demand better.

This, possibly more than any other government fuckup, makes me very leery of giving big and complex tasks to the government. If it was limited to specific countries or localities, I'd wave it off as 'bad' government. But given that only one country seems to be doing an even remotely acceptable job, I'd say they're the exception and not the rule.

---

What makes it extra frustrating is my government has good precedent for large vaccination schemes on short notice. Last time it went really well, but should note the logistics were organised by the army then. For some reason, they changed that this time.

I will say there are mitigating circumstances though - it doesn't help that there are numerous different vaccines with different requirements, and for some time (and even now) it's not always clear when which vaccine its available, and for which group. And vaccinating (very) old people comes with complications, they are not always mobile, even those living at home, they may not be sound of mind and need formal approval from family, which can only be done once it's known which exact vaccine is to be used, they are likely to have comorbidities, etc. It's not that simple to just get it in people's arms. I do expect things will go a a lot smoother with 'easier' groups.

I'm still not impressed with my government though.

Quote:

---

Originally Posted by wiggin

I am continually depressed how most governments - mine included - have done such an awful job on the rollout. Supply isn't the rate limiting step here and it should be! All this requires is basic coordination - get X people into location Y at Z time, and have an appropriate dose and healthcare worker there to administer it. It's logistics, that's it. And they had months and months to prepare for it. Every country that isn't Israel should be ashamed right now if they aren't pushing against their supply constraints.

I have a friend who now works in biotech and he's livid about the mess. He had a good point: Pfizer delivered an incredibly sophisticated product on a ridiculously short time scale - every single box of doses has GPS tracking, detailed temperature history, and a continuous record of custody and batch production going back to raw materials. I doubt if there has ever been this scale of distribution of something with such restrictive storage conditions on such a short time scale. They know exactly where and how much vaccine there is and have shown that it works. Yet the government simply needs to get the damn thing into people's arms and they can't manage that. It's embarrassing, and we should demand better.

This, possibly more than any other government fuckup, makes me very leery of giving big and complex tasks to the government. If it was limited to specific countries or localities, I'd wave it off as 'bad' government. But given that only one country seems to be doing an even remotely acceptable job, I'd say they're the exception and not the rule.

---

You want to hear a real crazy example? Care homes in the Netherlands.

First group: caregivers will be invited to get vaccinated at central locations, for others working at care homes I have found no information.

Second group; residents receiving care for chronic conditions that make them living in care necessary, will be vaccinated by the medical staff of the care home.

Third group; residents who do not have any chronic conditions but for whom living at home has become impossible. They will be vaccinated by their GP, and will have to actually go to his office to get the vaccination. Unless they can't go there in which case they will get vaccinated by their GP on a house call.



If this all wasn't n't bad enough for you, then know that there was no coordination foreseen for the vaccination of the three groups.

And some people thought that was smart planning. The outcry once this madness started to be rolled out only made the powers that be reconsider and decide at least group two and three should be vaccinated as one group by the medical staff of their care homes.

Quote:

---

Originally Posted by Flixy

Alright, got curious, and checked the statistics, assuming perfect sampling, for a yes/no questions such as hospitalization, AZN data is insufficient to state with 95% confidence that the proportion of vaccinated people is 1/2000 or better (i.e. the same as your actual rate). Close though, but your sample size is too small (very close though, you'd need 5990).

Of course that's excluding any other information you have regarding the actual distribution, you do of course have data on infections and severity of complaints, so it's not really non parametric data. But according to the numbers in that tweet, I can't conclude that AZN is even as good as not being vaccinated, even if you assume perfect sampling.

I will stop nitpicking now, I just like statistics :p

---

I am not sure if it is possible for me to like a post more than I like this one. There's all sorts of statistical quibbles one could make, but taking a quantitative approach to understanding this data is so refreshing. Kudos.

Quote:

---

Originally Posted by Flixy

What makes it extra frustrating is my government has good precedent for large vaccination schemes on short notice. Last time it went really well, but should note the logistics were organised by the army then. For some reason, they changed that this time.

I will say there are mitigating circumstances though - it doesn't help that there are numerous different vaccines with different requirements, and for some time (and even now) it's not always clear when which vaccine its available, and for which group. And vaccinating (very) old people comes with complications, they are not always mobile, even those living at home, they may not be sound of mind and need formal approval from family, which can only be done once it's known which exact vaccine is to be used, they are likely to have comorbidities, etc. It's not that simple to just get it in people's arms. I do expect things will go a a lot smoother with 'easier' groups.

I'm still not impressed with my government though.

---

Oh, sure, it's not easy by any stretch of the imagination (you may recall some of my earlier posts about my sister's experience rolling it out at her hospital). It requires a lot of coordination, communication, planning, etc. But they had plenty of time in which to figure this out, and clearly some places (okay, one place) are actually managing to do it quickly and efficiently. This exact problem is why government exists. And they are shitting the bed.

The NY Times claims 1.3 million/day are getting vaccinated. That means the percentage of Americans vaccinated should increase by 1% at most every 2.5 days (it's not clear if they're including all Americans in the denominator). And yet the percentage of Americans vaccinated has certainly not increased by over 2% in the past week. What am I missing?

Quote:

---

Originally Posted by Loki

The NY Times claims 1.3 million/day are getting vaccinated. That means the percentage of Americans vaccinated should increase by 1% at most every 2.5 days (it's not clear if they're including all Americans in the denominator). And yet the percentage of Americans vaccinated has certainly not increased by over 2% in the past week. What am I missing?

---

They're counting doses, which will be more than number of people vaccinated (two dose regimen).

I think you're misreading the stats. Are you looking at overall doses or first/second doses?

At least until a single dose vaccine is rolled out then best to look at is total doses, thus disregarding whether its first or second - that way you can make a simple comparison. Otherwise eg as you do a second dose the first goes down one and the second goes up one and you need to make sure you're counting it properly.

On total doses you're looking at 4 days to go up 2%.

24/1 6.60
25/1 6.87
26/1 7.11
27/1 7.45
28/1 7.91
29/1 8.42
30/1 8.94
31/1 9.40
Total change over the week 3.80%

To put that into context as a weekly figure for a selection of countries, 24/1-31/1:

Israel 13.12
UAE 6.34
UK 4.04
USA 3.80
Denmark 0.99
Italy 0.89
Germany 0.75

Messy but frank account of creeping demoralization:

https://nhssugardoc.blogspot.com/202...nes-faith.html

Has been expressed with varying degrees of eloquence by physicians and other HCWs all over the world, but I've mostly encountered it among my friends and acquaintances in the anglosphere. They also describe an even worse situation among junior docs and trainees, esp. in several parts of England. The loss of faith in leadership will have lasting effects on many; the relief of vaccination and an eventual end to the pandemic will do a lot to ameliorate that, but recovering, finding one's feet, catching up—everything will be an uphill battle for months at least, for many people in many parts of our healthcare systems. I don't think most govts have plans for even beginning to deal with the coming burnout epidemic and/or the mass of people who, for whatever reason, have not recovered fully from their bout with covid. I wonder if they have any strategy at all for dealing with the political repercussions. Right-wing parties here are very obviously positioning themselves for an attempt to seize power before the election, with the help of a party of actual honest to God card-carrying racists. So unnecessary.

Twitter Link

Quote:

---

Originally Posted by Aimless

They're counting doses, which will be more than number of people vaccinated (two dose regimen).

---

That makes sense. I'm not sure it explains the sheer difference though. I'll keep an eye on it for the next week. Georgia is at 7.2% today.

Quote:

---

Originally Posted by Loki

That makes sense. I'm not sure it explains the sheer difference though. I'll keep an eye on it for the next week. Georgia is at 7.2% today.

---

Where are you getting that from? I think they're at 9% (by doses).

If you're going to keep an eye using first and second doses rather than total doses you'll need to keep an eye on both figures.

Quote:

---

Originally Posted by Hazir

You want to hear a real crazy example? Care homes in the Netherlands.

First group: caregivers will be invited to get vaccinated at central locations, for others working at care homes I have found no information.

Second group; residents receiving care for chronic conditions that make them living in care necessary, will be vaccinated by the medical staff of the care home.

Third group; residents who do not have any chronic conditions but for whom living at home has become impossible. They will be vaccinated by their GP, and will have to actually go to his office to get the vaccination. Unless they can't go there in which case they will get vaccinated by their GP on a house call.



If this all wasn't n't bad enough for you, then know that there was no coordination foreseen for the vaccination of the three groups.

---

:|

Quote:

---

Originally Posted by wiggin

I haven't been able to find data on doses on hand for much of anyone, have you?

---

Quote:

---

Originally Posted by RandBlade

No its a closely guarded secret for most countries it seems.

But I've seen many reports of places halting vaccinations because they've got no supplies. So either the supplies are being used, or the logistics to get them to the front line aren't there.

---

I actually found the CDC dataset for states where you can figure out how much by subtracting one number from another: https://covid.cdc.gov/covid-data-tracker/#vaccinations

It's not pretty, typically 30-50% of vaccines delivered haven't been administered yet. And I imagine there are more vaccines in gov't hands than have been distributed to states, so the numbers are likely worse than that.

I imagine it's possible to find these numbers in other countries if one digs but I haven't the energy or interest in doing so.

Quote:

---

Originally Posted by RandBlade

Where are you getting that from? I think they're at 9% (by doses).

If you're going to keep an eye using first and second doses rather than total doses you'll need to keep an eye on both figures.

---

I'm looking at first doses. You're right that second doses complicate things, but for now, there aren't that many second doses being given.

The number went up from 7.2% to 7.4% today. I believe the percentage of those with two shots increased by 0.1% (from 1.1% to 1.2%). That's well off the pace of 1% every 2.5 days.

https://www.nytimes.com/interactive/...ine-doses.html

Wig, the above link has data for percentage of vaccines used.

Quote:

---

Originally Posted by wiggin

I haven't been able to find data on doses on hand for much of anyone, have you?

---

Quote:

---

Originally Posted by Loki

I'm looking at first doses. You're right that second doses complicate things, but for now, there aren't that many second doses being given.

The number went up from 7.2% to 7.4% today. I believe the percentage of those with two shots increased by 0.1% (from 1.1% to 1.2%). That's well off the pace of 1% every 2.5 days.

---

The US is trending around 0.4 doses/100 population each day. Given that we probably need on the order of 160 doses/100 population total to get decent herd immunity (or, roughly every adult in the country), we're looking at a very long time before things start looking up. It is true, however, that this rate is creeping up slowly. So it probably won't take 400 days, maybe more like 250 or 300. That's still super slow.

I do think that if the JNJ vaccine is authorized as a single dose formulation, we can speed things up significantly by giving people who are low risk (young, no comorbidities) the lower efficacy and easier to administer vaccine. That might drop that 160 number to something more manageable like 130, potentially lopping off 1-2 months of vaccination time.

Quote:

---

https://www.nytimes.com/interactive/...ine-doses.html

Wig, the above link has data for percentage of vaccines used.

---

They just scraped the data from the CDC website I linked to. I'm more interested in how many doses the US has but has yet to distribute; it might be very few but the information on that has been sketchy at best. In any event, 35% of distributed doses not being administered yet, over a month into the vaccination drive? That's problematic.

Of course, there are complexities here with the lumpiness of supply and projected future supplies; it's hard to know how much of that 35% have been languishing compared to just seeing a rapid scaleup of deliveries. A better metric would be average shelf time of unadministered vaccines - if it's staying steady or decreasing, that's a good sign. If it's creeping up, that means we're falling behind supply. Given how much of a clusterfuck this has been to date, I seriously doubt that they're running near their supply limitations.

Decent herd immunity occurs not just from vaccination but from actually getting the virus.

https://www.cidrap.umn.edu/news-pers...igher-reported

"An estimated 14.3% of the US population had antibodies against COVID-19 by mid-November 2020, suggesting that that the virus has infected vastly more people than reported—but still not enough to come close to the proportion needed for herd immunity, according to a study published today in JAMA Network Open."

That was two months ago so inching to 20%. Add to that the vaccine numbers and that more people will still get COVID I think we are pretty close to the tipping point where infections start going down considerably, herd immunity doesn't need to be absolute to slow spreading. Also since the most at risk will soon be vaccinated the harm of getting COVID for those still potentially vulnerable isn't high as it is very rare for non-old non chronic conditioned folks to require hospitalization.

I for one am looking forward to shaking people's hands, hugging church members, blowing candles from birthday cakes and attending large events. It isn't all going to happen instantly but before the end of the year (hopefully summer) we will be doing what we were doing prior to COVID with a couple of safety items in place.

Quote:

---

Originally Posted by Loki

I'm looking at first doses. You're right that second doses complicate things, but for now, there aren't that many second doses being given.

The number went up from 7.2% to 7.4% today. I believe the percentage of those with two shots increased by 0.1% (from 1.1% to 1.2%). That's well off the pace of 1% every 2.5 days.

https://www.nytimes.com/interactive/...ine-doses.html

Wig, the above link has data for percentage of vaccines used.

---

According to the NYT it is 7.4% for at least one, plus 1.2% for two shots. 7.4 + 1.2 = 8.6

They're reporting that different to Worldometer which shows 1 shot and 2 shot separately, rather than 1 shot being "at least one" it means "only one".

That lags a bit on the data I found, the GA Covid-vaccine website says that 954,786 shots have been administered, which based on the population Google says GA has makes it 9.0% of the population.
https://dph.georgia.gov/covid-vaccine

Quote:

---

Originally Posted by wiggin

I actually found the CDC dataset for states where you can figure out how much by subtracting one number from another: https://covid.cdc.gov/covid-data-tracker/#vaccinations

It's not pretty, typically 30-50% of vaccines delivered haven't been administered yet. And I imagine there are more vaccines in gov't hands than have been distributed to states, so the numbers are likely worse than that.

I imagine it's possible to find these numbers in other countries if one digs but I haven't the energy or interest in doing so.

---

I imagine there's a lot of lag involved there so take those numbers with a grain of salt. The CDC figure for Georgia doesn't currently says 930,517 which lags the 954,786 that I saw last night (which hasn't been updated by this morning) on the GA website.

But also the stats are updated not in real time, but in data dumps, which entails accruing the data and getting it updated. I'm not sure how it works in the USA but it the UK we're seeing a very clear pattern to the data with a 2 day lag for updates. The biggest amount of doses always gets reported on the Sunday (which would match the Friday it seems), then it falls dramatically on the Monday and Tuesday (matching Saturday and Sunday when most GP surgeries are closed), then from Wednesday (matching Monday) it picks back up again. If you also have a 2 day lag in your data then at ~0.4% of population getting a shot per day that's close to 1% of population whose shot will have been administered at any one time but who won't be appearing on the CDC site yet.

The CDC will I'm assuming know how many doses it has delivered to the states in pretty much real time - but it won't know how many injections a hospital or GP (or whatever you call them) etc have injected until that data feeds back to them.

All of the numbers lag. They're obviously snapshots but informative. A two day lag is not particularly substantial and it becomes less substantial with time.

When you're vaccinating at 0.4-0.5% of population per day then a 2 day lag can be pretty significant. 1% of population is nothing to be sniffed at, at this stage it means your doses used in many states will probably be about 10% higher than what it says.

Either way absolutely the key is to tackle both issues simultaneously. You can be limited by distribution or limited by supply - and you should be tackling both not just one of them. The people working on distribution should be operating on a basis of if they can get supply they will use it - and the people working on supply should be operating on a basis of if they can get supply they will and will send it on to be used. Neither should rest on their laurels waiting for the other to 'catch up'.

The key metric is how many days of supply you have on hand. Most places in the US are on the order of 3 weeks. Even if the numbers change by two days (which is not obvious, it depends on how supply numbers are reported), the overall story is the same.

I'm not seeking 3 weeks supply on hand in any states.

Not given the ramp-up of vaccination rates occuring. Given that last week 3.80% of population got a vaccine last week, then 3 weeks would be over 10% of the population or over 34 million stockpiled vaccines. If ramp-up continues as we should expect then 3 weeks would be much more than 34 million vaccines.

California seems to have one of the biggest gaps but even that equates to just about 5% of the population so possibly a bit over a week's doses on hand. Given the potential for a 2 day lag, then a week for delivery to injection isn't ideal but its far from horrendous.

10 million doses done in the UK. :D

Almost 400k here... It's something!

We are following the UK on one thing: the share of "UK" strain has now increased to about 2/3 of new cases.

RB -

The CDC says there are 17.7 million doses distributed but not administered. There's about 1.3-1.4 million being dosed each day. So that's two weeks overall. I was looking at a bunch of states that are closer to 3, but you're right the average isn't 3. Massachusetts is about 18 days, Rhode Island has 23 days, NH has 17. I was focusing on NE because that's where I live, but you're right - a few big states like Texas make the overall numbers better than that.

Of course, that's assuming that 'doses distributed to states' is the complete accounting of doses that Pfizer and Moderna have given the US. I suspect there's more in that denominator but don't know where the information is.

Regardless, there should be days of excess supply, not weeks.

Quote:

---

Originally Posted by Flixy

Almost 400k here... It's something!

We are following the UK on one thing: the share of "UK" strain has now increased to about 2/3 of new cases.

---

Inevitable. :(

A grim few months ahead possibly then. I can't wrap my head around what your Government is playing at sitting on the few vaccines they do have. Its totally inexplicable.

The thing is the attitude to this should be one of "it is a sprint, not a marathon".

Quote:

---

Originally Posted by RandBlade

Inevitable. :(

A grim few months ahead possibly then. I can't wrap my head around what your Government is playing at sitting on the few vaccines they do have. Its totally inexplicable.

The thing is the attitude to this should be one of "it is a sprint, not a marathon".

---

Well, I've already on multiple occasions expressed my displeasure about how they're running things here. Speed is picking up though, then again that's hardly a challenge when you start low.

Numbers are still dropping though, hope we continue that even with the more contagious strain. If we suck at vaccinating, we better not suck at keeping rhe spread in check, I guess..

RIP Captain Sir Tom. 100 years old and a hero to the end.

That's made me tear up. Rest in peace.

Very good news from new data from the Astrazeneca study: https://www.bbc.co.uk/news/health-55910964

A single dose for three months gives 76% protection.

The preprint is one of the messiest things I've ever read.