I previously commented that cases were rising in Florida, but death had not yet risen. About a week ago in an update, I noted they tentatively looked like they had begun to rise. I think it now looks like they have definitely begun to rise. The rate of rise (as a % over previous values) doesn’t not appear to be quite as high as that for cases, but that statement may later modified by inserting “yet”.
This may or may not continue. Needless to say, I do not plan to vacation in Florida this summer. Summer wouldn’t be the time to go to travel to a tropical state anyway, but this July and August is probably a good time to stay at home in Illinois.
Update July 18
Deaths are definitely climbing.
But here’s a ray of hope: the positive/test ratio did drop this week.
I think you’ll have to wait another week until big numbers start coming out of Florida. I expect that the deaths will jump suddenly like has just happened in Texas.
skeptikal,
Please define ‘big numbers’. Predictions without specified values mean nothing. 200? 300? 400? Or maybe 1,000 like NY’s peak rate?
SteveF,
Clearly, I will need to put NY on the graph as this plays out.
Skeptical,
I don’t think they are going to “jump”. But I think they are going to go up quite a bit. The slope for deaths looks smaller than cases. Whatever that means. . .
lucia,
Seeing the delay between the death rate and case rate in Florida and California probably shows, if you needed convincing, how much the infection rate in NY and NJ was underestimated back in March. Thanks again CDC.
DeWitt,
I’ve read lots of possible reasons for the difference in the time lag. That’s yet another one!!
Our (IL) infection rate seems to be going up a bit after having fallen. Not making me happy…. (But right now, it actually might be increase in testing because the positive/case rate is staying very low– near 2-3%) Still… watching.
I have a suggestion as to why there is such a large delay in deaths vs. new cases. This is very simplified in order to illustrate something that might be a factor. It is not meant as a description of what is happening, just an illustration of what might be part of what has been happening.
.
Imagine that we have two “epidemics” which I will call E1 and E2. E1 races through the most vulnerable, then declines as it runs out of fuel. E2 rises much more slowly among the otherwise healthy. The Cse Fatality Rate (CFR) is much higher for the vulnerable than for the otherwise healthy.
At some point, the decline in E1 will be balanced by the increase in E2. At that point, new cases will be at a minimum; that minimum defines the boundary between the E1 dominated time frame and the E2 dominated time frame. Call this boundary t1.
The same consideration applies to deaths; however, the minimum in deaths will occur at a time t2 that is later than t1. At time t1 the deaths will still be dominated by E1, due to the difference in CFR. If there is, say, a factor of 10 difference in CFR, then deaths will not reach a minimum until E2 is producing 10 times the new cases from E1. As a result, the rise in deaths will lag far behind the rise in new cases.
With strong exponential trends, the minima will be narrow and the lag will be small. With linear changes, the minima will be flat and the lag will be large.
IMO, the simplest explanation for a lag is that the time from developing symptoms to death, for those who die, is about 18 days. Early on when tests were in short supply, people didn’t go the the ER and get tested until they were very sick, shortening the lag time.
DeWitt Payne (Comment #187721): “IMO, the simplest explanation for a lag is that the time from developing symptoms to death, for those who die, is about 18 days. Early on when tests were in short supply, people didn’t go the the ER and get tested until they were very sick, shortening the lag time.”
.
That would increase the lag from one week to at most two weeks. The present lag time is much longer. And tests have been plentiful for up to 3 months now.
MikeM,
Now that tests are plentiful, some people are tested before they have symptoms. So the lag could be longer than the time to death after symptoms. That would stretch things past 18 days.
.
I think DeWitt’s notion is probably at least partly true. But it could also be that lag time to death is partly dependent on age. After all, the death rate is too. So yes, that might be part of it also.
.
Other things that might matter: as doctors figure out how to treat people fewer people die, but also, some who do die take longer to die.
.
Beyond that, death is probabalistic. 18 days may be a median. I haven’t done the math to figure out exactly how the death rate would look in a SEIR. (Annan has it in his model. It requires some fiddliness that takes the input of the SEIR.)
.
Right now, I’m glad I’m not in FL…. I’d definitely be holed up at home. Braver (and even plausibly more realistic ones) might do otherwise. But not me!
What, me worry? Ha ha.
.
FL is definitely having their NY moment. We shall see how high the deaths get but the case counts are not encouraging to say the least. FL had a record 15K cases one day, but that was mostly due to one test center dumping 7K tests on a single day with some backlog. FL had a record death count of 122 yesterday as well. The more paranoid behavior change from a week or so ago is showing up in a slow down in case increases.
.
I think there is a great deal of randomness to the hot spots and anyone taking victory laps now for their area might want to wait a year or so to be safe from eating crow.
.
For now, it is paranoid behavior time. I was already pretty low risk, I haven’t seen the inside of a restaurant or store for months. Only groceries, and ummmm, golf. I exercise on trails as well.
Bari Weiss resigns from the NYT, and flames them mightily on the way out, and I should say deservedly.
.
https://www.bariweiss.com/resignation-letter
.
“Instead, a new consensus has emerged in the press, but perhaps especially at this paper: that truth isn’t a process of collective discovery, but an orthodoxy already known to an enlightened few whose job is to inform everyone else.”
.
“My own forays into Wrongthink have made me the subject of constant bullying by colleagues who disagree with my views. They have called me a Nazi and a racist; I have learned to brush off comments about how I’m “writing about the Jews again.â€
.
A hilarious scathing read. Don’t miss it. The usual suspects (Fox) are laughing it up.
For you lower herd immunity theorists:
A New Understanding of Herd Immunity
https://www.theatlantic.com/health/archive/2020/07/herd-immunity-coronavirus/614035/
.
““We just keep running the models, and it keeps coming back at less than 20 percent,†Gomes said. “It’s very striking.â€
Tom,
Thanks. The first half of that article, about non-linear dynamics and heterogeneity, is very good. Then he forgets about that and goes back to the same old same old.
The transition occurs when he brings Tom Britton into the picture.
.
Consider two populations. One consists of hip urbanites sharing an apartment with roommates, working in crowded offices, using elevators and mass transit, taking meals in restaurants, and spending much of their free time in crowded bars, clubs and theaters. The other consists of farmers who spend the day in a tractor or working around the house, take their meals at home, and spend their free time reading or watching TV. R0 must obviously be a lot less in the latter than the former.
.
But the author is almost surely right when he says “Respiratory viruses tend to be more equal-opportunity invaders”. So how can that be? The answer, I think, is schools. They are key to the spread of most respiratory viruses and so are great equalizers as to how the virus spreads in different populations.
.
But the Wuhan virus does not spread via schools. Therefore a huge degree of heterogeneity is to be expected. So I think the under 20% estimate is more likely to be right, but higher in urban areas and smaller in rural areas.
Andrew Sullivan also quits The New York Magazine today, reportedly due to the editors refusing to print his column on rioting and looting a couple weeks ago.
.
I also noticed David Brooks writing less for the NYT and his work appearing at other outlets recently.
.
It should be noted that all of these people are much closer to moderates than conservatives, although with a few right wing views among them. It’s an ideological purge. I don’t really feel that strongly about it since it just exposes what has been there all along and exposes these outlets for what they are, strict purveyors of narrow group think.
.
I can’t think of which job would be worse today, a police officer or a conservative journalist.
Tom Scharf,
“A hilarious scathing read.”
.
Yup. The NYT is a sorry mix of dishonest scumbags and dishonest cowards, and he correctly calls them on it. But I doubt it will make much difference, because the inmates have taken over the asylum.
Lucia,
“Right now, I’m glad I’m not in FL…. I’d definitely be holed up at home. Braver (and even plausibly more realistic ones) might do otherwise. But not me!”
.
Well, my 2 YO granddaughter goes to day care Monday through Friday, and my daughter and son in law (both 31) have some social contact with a few friends. But in reality, where they live is still relatively low risk for them. In a county of 160,000 souls, I think there have been 2,900 confirmed cases, but only 42 deaths… all pretty elderly people, and about half in elderly care facilities.
.
So one in ~55 people have been diagnosed. But between 52% and 70% (~1,500 to ~2,000) of all confirmed cases have been in a smallish Hispanic population of about 20,000; 7% to 10% diagnosed infections! Non-hispanics in the county have a much lower rate of infection. I don’t find this surprising, since the two communities are pretty much socially isolated by language, even though they all live within the same physical areas and go to many of the same stores.
.
Before I drove from Florida to Cape Cod a few weeks ago I was not at all worried about the virus. I won’t be worried when I return in October.
Tom Scharf,
The journalist at the Atlantic (New Understanding of Herd Immunity) seems to understand very little. I am reminded of the journalists who pretend to be “golf betting analysts”, and then confuse club head speed and ball speed… ‘Tony Finau has registered 200+ MPH club head speeds during recent practice sessions!’ I can hardly wait for Finau to pop the inevitable 470 yard drives. Just as I can hardly wait for that journalist’s herd immunity speculations to turn out right. He is blissfully unaware of recent publications that shed light on estimates of herd immunity, and his article shows it.
Steve/Tom,
The article kicking it all off with Lorenz and chaos made me cringe. It just screamed “I don’t really know what I’m talking about but this will make it sound more scientific!”
.
The actual issue of heterogenaity and the possibility of herd immunity at lower % infected is interesting (but we here have discussed that before.)
“I don’t really know what I’m talking about but this will make it sound more scientific!â€
.
Journalism majors don’t study a lot of science… or math. But they do know a lot of gender studies nonsense and social justice rubbish. Yup, nonsense and rubbish… the best substitute for actually knowing something when you write a technical article.
What is striking about the Bari Weiss resignation is she publicly accused them with their own euphemism, a toxic work environment. Normally the immediate response is that this matter will be investigated thoroughly and all matter of bland statements about commitments to equality and diversity. The response from the NYT and the left has pretty much been crickets. It is telling that they don’t even have enough energy to put up a pretense anymore, fear of their own mob I imagine. Some equality is more equal than others.
.
It just isn’t that hard to be professional to your coworkers. You don’t have to like them, but you don’t have to overtly hate them either.
.
The Cultural Revolution, McCarthyism, The Moral Majority, etc. When has a moral panic with a zealous enforcement of a singular set of “correct” values ever ended well?
SteveF, your county has a rate of confirmed tests double the rest of the country.
MikeN,
Yes, and that is driven almost completely by a confirmed infection rate of >7% among a sub-population that represents only ~13% of the county population. I do not personally know anyone who has tested positive. Among the rest, the rate is pretty low.
It’s almost like they were reading this blog: https://fee.org/articles/why-sweden-succeeded-in-flattening-the-curve-and-new-york-failed/
We’ve now had
36 straight days of >20000 new cases
27 straight days of >30000 new cases
21 straight days of >40000 new cases
15 straight days of >50000 new cases
9 straight days of >60000 new cases
I don’t think it is a larger time lag that is keeping the death toll from going into the thousands.
I have been following the discussions here when I have time but have not posted for awhile. I was wondering if the false positive, prevalence, specificity and predictive value for Covid 19 testing has been discussed here.
https://mises.org/wire/false-positive-panic-over-covid-19
Kenneth,
That is a perfectly reasonable analysis. The only missing part is the only part that matters: What are the actual rates of false positives and false negatives? As far as I can tell, the author provides no actual values. So it is impossible to guess how frequent those false test results truly are.
MikeN,
The simplest explanation is that the new cases are dominated by younger people who are far less likely to become very sick or die. The difference in susceptibility across age ranges is enormous… if you could limit those infected to people under age 50, the death rate would be extremely low.
Kenneth Fritsch (Comment #187814): “I was wondering if the false positive, prevalence, specificity and predictive value for Covid 19 testing has been discussed here.”
.
That article is useless.
There is no relation between the false positive and false negative rates. The reference to Bayes’s theorem is mere obfuscation. He effectively admits it:
Rubbish. If that were true, then the positive rate would never be less than 30%. It is often as low as 1% and there is no evidence that those are mostly false positives.
The FDA sent out warnings about a test giving a 3% false positive rate. I don’t know if they pulled it while the manufacturer fixes it. So whatever the acceptable false positive level is, it is much lower than 3%.
I think that false positves are usually not the test per se. They occur from contamination during sample handling.
——–
I have seen reports of a 20% false negative rate. It is not clear if that is the fault of the test, or of sample collection, or if it is just infected people not shedding virus.
MikeM,
Yes. The linked paper is using made up numbers and then flinging around Bayes theorem. The paper also doesn’t distinguish between values that are relevant to diagnosis and those relevant to epidemiology.
If you are a doctor wanting diagnosing a presenting patient/i>, given your test result, you want to know
P(Covid+| patient’s test outcome) = P(test outcome|Covid+ ) * P(Covid+ population )/P(population test outcome) .
If the amount of covid in the test population is very low, this can be quite high relative to the false positive rate. That’s a problem for physicians who want to advise patients. It’s a problem for patients who worry. (But bear in mind: the physician understanding problem is mitigated by the fact that at least some physicians do understand the issue even if they don’t remember the math. They know that they should ask other questions– like do you have a cough. With some diseases, they do follow on tests.)
It’s of course can be a civils-rights issue if the patient is literally going to be quarantined based on nothing other than a test outcome.
BUT, when considering epidemilogy ,you aren’t advising an individual patient. You want to know how many people are sick which is P(Covid+ population|(outcome of ALL tests) ). We could do the math for this postulating different false positive and negative rates and your prior for how many people really are sick., but it’s really nothing like that article says.
In fact: if the tests have 95% sensitivity and 95% specificity, you’re going to get pretty good epidemilogy info, provided you do a decent amount of testing without collecting biased samples. (Most or problems in estimating based on number of cases or + positive has to do with biased sampling!)
I’ve seem as high as 30%. However the article I read specifically linked it to early testing and reported numbers at different amounts of time since infection. The amount of virus in a nasal swab is affected by the amount in your nose. The later is affected by how much virus is in your body and how much you are shedding. If you’ve just been infected, you may not yet have enough in your nose to trigger a positive, but will have a few days later.
So, when people are identified quickly during contact tracing and get tests when they are symptomless, there is a moderate chance of false negative– perhaps 30%. We’d still reduce transmission a lot by catching 70% we otherwise wouldn’t catch. That would greatly reduce spread. But we can’t catch them all.
Is there any point to testing asymptomatic people other than as part of a prophylactic regime of frequent testing for, say, employees of nursing homes?
It seems obviously pointless to get a one-off test if you are both asymptomatic and have no reason to think you have been exposed.
But what if you think you have been exposed? In that case you should isolate yourself for a period of time; 1-2 weeks from exposure, depending on how conservative you want to be. The reason is that you might be about to become sick, in which case you might now be infectious. A positive test does not really change that since by the time you get the test result, you will have symptoms. Neither does a negative test, since it might be positive tomorrow.
So maybe it comes down to the perhaps mythical asymptomatic carrier. But even then, does the test result really change anything?
I suppose that one might argue that a positive test might encourage the subject to take self-quarantine seriously. But any benefit would likely be undone by a negative test convincing the subject that there is no need to self-quarantine.
MikeM
Yep. There are good reasons to testing asymptomatic people. 🙂
They aren’t generally “one off” tests though. They are repeat tests.
.
Testing of asymptomatic people could drastically reduce transmission rates in congregate living facilities (like university dorms and cafeterias) if they tested everyone every day and made people who got positives self isolate. (They could do online classes for 14 days rather than having the entire school do 100% online classes every day for the whole year.)
.
(Note: cost containment might require pooled testing. )
.
Similar benefits could be achieved in meat packing plants — those who test positive could be told to go home for 14 days.
.
Many studios around here are requiring people like ballroom dance instructors who meet multiple people a day to get tested. Some instructors will have false positives and self isolate from tiem to time, but the alternative is having covid not be detected, transmitted to a number of students and propagate. (Students in individual classes usually just have their temperature checked and fill out a form about symptoms. But the high contact person– the teacher– is tested regularly.)
.
Some saliva tests may be quick. I don’t know the false negative rate, but it doesn’t need to be stupendously low to dramatically drop the spread of covid. moving R from 3 to 0.9 in a congregate facilities like dorms could pretty much allow schools to stay open. Some people would be isolated unnecessarily. But since the alternative is to have no school in session….. some people would prefer to go to school knowing they might have a few episodes of online rather than 100% online classes.
I should have added: If they are testing every day in dorms, the fact that you might have a “false negative” the first two days when you are infected (and likely not yet contagious) but are only caught on day 3 when you are starting to shed virus is not a big problem. But you really need very frequent testing.
.
The issues that need to be considered are:
1) Costs of testing. The per test cost needs to be quite low.
2) What action you take when someone has a positive outcome.
3) Cost of the action you take to the person identified, especially to those who will experience false positives.
.
Florida went from 7000 hospitalizations on July 10 to 9100 hospitalizations on July 18.
.
Miami-Dade County ICUs are at 122% capacity and ventilator usage is up 64%.
.
Miami-Dade has a large surge capacity, 457 hospital beds can be converted to ICU beds… so they’re not in any trouble yet.
.
I said earlier that first you fill the hospitals, then you fill the morgues. Well, the hospitals are filling up nicely now… so I’m predicting a very profitable year for casket manufacturers.
skeptical,
Predictions without numbers are meaningless. How many covid deaths will Florida have by October? More or less than New York?
SteveF (Comment #187904)
July 19th, 2020 at 2:45 pm
Not always. Sometimes making predictions with numbers is meaningless… particularly when there are too many changeable things which can (and will) affect the final tally. Remember your prediction of 45,000 deaths by July?…. that wasn’t even close, but getting a prediction right in such a dynamic situation would be more luck than anything else.
.
There are just too many things that nobody knows…. When will Florida go back into lock down? How many aged care facilities will take a direct hit from this virus? Will some miracle cure become available? With so many unknowns, it seems pointless to even make a WAG at the future death count.
.
The only prediction I will make for Florida is that things will get a lot worse before they get better.
skeptical,
Define ‘a lot’.
You won’t? Well, then it seems you have nothing interesting to say.
skeptical,
Duh. That was just as true in March as it is now. But it won’t be as bad as New York City, Spain or Italy. Other countries are seeing a ‘surge’ in new cases. Even New Zealand is still getting the occasional case. But when you have very few cases, you can do contact tracing and quarantine.
SteveF (Comment #187933)
July 20th, 2020 at 5:12 am
.
Again, you want me to quantify something which is a moving target. The situation is constantly changing. Some counties in Florida are already imposing curfews… and I expect more measures to be taken by the authorities as the situation worsens.
.
.
Since you’re so obsessed with numbers, I’ll say something you should find interesting. There are enough hospitalizations in Florida right now to fully support a 7 day averaged death rate of 110 deaths per day over this week (the next 7 days) and 125 deaths per day over the following week. What happens after that depends on what restrictions are put in place during this coming week.
DeWitt Payne (Comment #187936)
July 20th, 2020 at 7:41 am
.
Of course not, Florida has a much lower population density so the virus can’t spread at the same speed as it did in those other places. But the slower transmission just means it takes a while longer to fill the hospitals. Where things will get interesting is if Florida completely fills their hospitals.
skeptical,
“Of course not, Florida has a much lower population density so the virus can’t spread at the same speed as it did in those other places.”
.
Florida and New York have almost the same average population density (356 per sq mile vs 327 per square mile). It is true that more of NY’s population is concentrated in the New York City area, but Florida has multiple metropolitan regions with dense population (Miami, Orlando, Tampa, Jacksonville) all with fairly high densities; most of Florida (like most of NY) is very sparsely populated. Florida also has an older population, so is a bit more susceptible: 42.2 median vs 39 median.
.
Spain has about half the population density of Florida. Italy does have a population density that is higher than Florida: 519 per square mile.
.
There is no doubt that population density makes a difference, but there are multiple factors, which means policies that are suitable for one place not suitable for another. One size does not fit all.
Florida growth in cases is slowing down and is off the fast exponential it had starting in June, not that 10K cases a day is a good thing. Deaths are still increasing and might level off around 100/150 a day. That would be 3x more deaths for 10x the cases. Something changed there.
https://experience.arcgis.com/experience/96dd742462124fa0b38ddedb9b25e429
.
If we assume a 10x multiplier for undetected cases (100K / 21.5M), * 0.5% of the population is getting infected every day *. It won’t be that long until herd immunity.
FWIW, some of the less hard hit European countries are having a resurgence of new cases. Sometimes mild (Austria, Switzerland); sometimes strong (Serbia, Slovakia, Slovenia).
Number of deaths was below 500 at CDC, low for a Tuesday. I was expecting it to go over 1000. We’ll see how the rest of the week goes.
False positive rate has a huge impact.
If 95% are negative and 5% positive, then a 5% false positive rate and a 5% false negative rate will produce
4.75% negatives reported as positives, and 4.75% positives reported as positive.
Half of the positive cases are really negative.
MikeN,
The false positive rate matters. However, I don’t think the false positive rate is as high as 5% for the current set of tests. I’m pretty sure it’s below 3%. Anyway…. the reported %confirmed positive/total tests in Illinois over all is less than 5%– it’s about 3% this week. So whatever tests we are using in Illinois must have a false positive rate less than 5% otherwise, we wouldn’t be seeing the reported rate near 3% in Illinois overall!
.
My understanding is the %confirmed positive/total tests in Chicago is higher than 5%. That means the rate outside Chicago is lower than the state averge of 3%.
.
Since I have no idea what the false positive rate is, I can’t estimate what fraction of “confirmed positives” in illinois might be false positives. It may be quite high. Or not. One thing: it’s pretty well flatlined over the past 6 weeks.
If the positive rate is 3%, doesn’t that make 3% the upper limit on false positive rate? Unless there’s an extremely high false negative rate.
I’ve seen reports claiming the false positive is close to 0, while false negative is around 3%.
MikeN
Yes. Empirically, that would stronlgy suggest 3% is the upper limit on the false positives for the mix of tests being used in Illinois. There are a number of different tests. Some tests many have higher false; some may have lower rates.
.
Reports that the false positive rate is 0% does not contradict the upper limit being 3%. I’ve seen one test the FDA warned had a false positive of 3%. The others I’ve read “very low”…. whatever that means.
.
In situ, where there is always some potential for mishandling and so on, I doubt any test anywhere ever achieves a false positive that is literally 0% if measured by “positive reported by the lab / number uninfected”. But since this one involves RNA strands, it has the potential of being “very low” compared to lots of other tests.
MikeN
I don’t think the magnitude of the false negative rate affects the false positive rate. I also don’t see how it would affect the idea that the reported positive rate is the upper bound on the false positive rate.
.
reported positive rate = (reported +)/(N_all).
False positive= (reported + Incorrectly)/(N_infected)
N_infected is always ≤ N_all.
is always (reported + Incorrectly) ≤ (reported +). 7th grade algebra would allow us to conclude False positive ≤ reported positive rate– at least in the limit of large numbers of tests.
The main difficulty we have in concluding a value for nationwide is that different tests can have different false positive rates and the mix of tests used can be changing over time and may differ in different states.
https://covidtracking.com/data
Some locations have very low positive test rates.
Alaska 1.3%
American Samoa 0.0%, 1037 tests, no positives
Guam 1.8%
Hawaii 1.3%
Montana 1.9%
Northern Mariana Islands 0.35%
Vermont 1.6%, but 0.8% over the last week
Unless the positive test rate has been constant, the false positive rate must be much less than the positive rate.
Lucia makes a good point that there are different tests. But unless the test maker has screwed up, like the CDC did back in February or the one with the 3% false positive rate that the FDA sent out a warning on, it is really not the test that produces false positives. False positives usually result from sample contamination during handling. So they are likely to be a fraction of positive tests and that fraction will vary a lot from place to place.
Lucia, suppose the group being tested is 10% positive, 90% negative, with test results as 8 positive, 2 negative, 80 negative 10 positive.
Am I correct that false positive rate is 10/90 = 11%, and false negative rate is 20%?
I think your algebra is off when you compared denominators.
Assuming my definitions are correct, then with 3% testing positive, suppose the correct results are 50-50 positive negative.
All your 3% is negatives, for a 6% false positive, while all your positives test negative!
I found it interesting to look at how deaths as a percentage of new infections varied as a function of the lag between these two quantities. My death percentage was fairly stable using a lag of 7 days between late March and mid May. Longer (or shorter) lags produced a death percentage that changed more with time. From a purely pragmatic perspective, if you can identify a lag that produces a stable or slightly improving death percentage over the past few months, that lag provides a pragmatic, but not rigorous, way project how increasing infections will translate into deaths over the next week or two – whether or not that lag is physically reasonable
If a given lag stops performing well, then you can speculate why. A priori, one might expect the death percentage to drop as increasing testing identifies more of those at risk of dying from COVID, but only those few who die without any professional care are likely to have been missed since sometime in April. Since then, increasing testing is likely to be identifying milder cases, driving the death percentage down. Treatment is certainly getting better, so the death percentage should be going steadily down. The trickiest variable that is changing (IMO) is the age of those getting infected. If older people remain relatively cautious and younger people are engaging in riskier activities, the death percentage will certainly drop – but it will include some older people infected through essential contacts with people increasingly likely to be infected. The changing age of those being tested and found positive are two important unknowns in Nic’s post about Sweden at Judy’s
In early March, the death rate for a 7 day lag was over 20%, dropped rapidly, stabilized at 6.5% before the end of the month and remained there until mid-May. This hints that 1/3 fewer infections were being detected earlier vs later in March. FWIW, this drop was unrealistically larger for more reasonable lags of 10 or 14 days, suggesting that something besides testing was changing.
Hospitalization rates provide a useful metric between infection and death, but that data isn’t as readily available.
Mike N
I can’t do the algebra because I have no idea what you are claiming the test results are. Scanning across, I though at first, you are suggesting test results on 10 people (8+2) in a sample were 8 positive and 2 negative. But then you seem to give results on some other test of 90 people (80+10) with 80 negative and 10 positive.
.
That’s at least two different tests each with its own result. Can you clarify what you mean?
.
Anyway, you seem to be trying to find the conditional probability a person is infected given a positive test result knowing information about false positive and false negatives. That’s a different question from inferring the bounds of the false positive of an actual test based on population data we are seeing.
Also: I don’t think it’s possible to determine the conditional probability because we don’t have information on the false positive and negative rates. Just making them up can be useful for a thought exercise, but the number coming out of the calculation doesn’t tell us anything about what’s true on the ground in this epidemic.
But.. yes. There must be something wrong with my thinking above, because if the entire population is infected and the false negative rate is 1, then it’s true the ratio of test positives/total will be less than the infection rate. (Specifically, the test postives will be zero, and the true positive rate will be 100%) What I’m getting is
p(positive test outcome) = α + p(positive population) * (1-β-α)
where α is the false positive rate and β the false negative rate. If the test is halfway decent (β + α) <1, then
p(positive test outcome) > α. But if the test is pretty crappy then tge observed test positive rate not the lower bound on the false positive rate.
Still, wrt to this test (and likely any test) the approved tests are definitely thought to have α<3% and β <50%. So it appears (β + α) <0.53 which is less than 1. We don't know what β & α are, but I haven't read numbers to suggest worse than this anywhere. And if (β + α) <1 it appears the the false positive rate of the test must be less than the reported positive rate of any test result on a large population sampled in an unbiased manner.
I was referring to 90% real negatives breaking down 80-10 and the 10% real positives being 8-2. Not trying any new calculations, just checking if I am using definitions correctly, and showing an extreme example where false positive could exceed positive test rate.
MikeN,
I can’t tell if you are using the definitions correctly.
If you test a sample of “N” UNIFECTED people, and get “n” positives, the false positive rate is α = n/N. (Well… in the limit N approaches infinited.)
If you test a sample of “N” infected people and get “n” positives, the false negative rate is (N-n)/N.
That’s it. These are just definitions. There are procedures for figuring them out using “known infected” people and “known uninfected” people. You don’t do a test on a mixed group with some infected people and some not infected people. You do two tests: one using all known positive samples, the other using all negative samples.
So: If you KNOW 10 people are positive, do the test and get 8 positives, your estimate for the false negative rate is 20% because the two who came out negative are false negatives. You would find the error bounds using properties of binomials. That’s one test and is applied ONLY to the 10 known positive people.
Southern hemisphere flu cases (it’s winter down there)
.
Chile last year: 20,949
Chile this year: 1,134
.
Australia last year: 22,047
Australia this year: 85
.
Other countries are seeing similar declines.
https://www.wsj.com/articles/covid-19-measures-have-all-but-wiped-out-the-flu-in-the-southern-hemisphere-11595440682
.
“From Argentina to South Africa to New Zealand, countries in the Southern Hemisphere are reporting far lower numbers of influenza and other seasonal respiratory viral infections this year. In some countries, the flu seems to have all but disappeared, a surprise silver lining that health experts attribute to measures to corral the coronavirus, like mask use and restrictions on air travel.”
Tom Scharf,
Sure, if everyone lived in their basement like Joe Biden, then communicable diseases would become a non-factor. OK, maybe a problem for people who’s jobs require them to actually leave their basement, but nobody else.
.
Funny though how people choose to risk death living outside their basement. I am guessing Lucia will soon be dancing.
SteveF,
I am taking lessons. We also take walks, and do stop to talk to a few of our neighbors (standing some distance apart.)
.
The thing is: life is pretty grim if you can’t socialize at all. I’m not even an extremely social person. Still, most people don’t want to be holed up in their basements alone.
.
I’m hoping the studio comes up with some sort of dance get together. The difficulty is the need to social distance. . . Lightfoot and Pritzker are both threatening stomping down more if ‘new cases’ keep increasing. (Death’s are decreasing. %positives/tests is flat. My theory is that *in the case of Illinois* actual cases are not increase, it really is an increase in *detection*. I don’t think that’s the situation in many other places though. )
.
(Oh. The weekly average “new cases” in FL is the same this week as last week! So the peak may have arrived! We’ll know better in 7-14 days..)
Or maybe people are just not going to the doctor to get a flu test.
Since this started, cancer diagnoses in the U.S. are way down. That is not a good thing.
MikeM
It’s possible people aren’t getting flu test. When I was ill a few weeks ago, I got a Covid test. I didn’t get tested for anything else. (They did do temperature check, pulse etc.)
.
I’m not sure they would have tested me anyway. They’d probably wait for symptoms to be stronger. Right now, the concern is to identify covid so we can isolate people I might have transmitted to. There’s no contact tracing for flu. So it makes sense to just wait and see if you get worse with flu, but you go in for the covid test.
.
That said: I imagine there could be less flu going around. I’m still mystified how I could have gotten whatever it was I got.
Lucia,
“ I’m still mystified how I could have gotten whatever it was I got.â€
.
But you did. That is one of the reasons why I think efforts to indefinitely stop development of herd immunity to covid 19 virus, absent an effective vaccine, are likely to fail. ‘Saintly‘ lefty politicians like Newsom and ‘demonic’ righty politicians like Santis both have rising cases. The fact the MSM declares one a saint and the other the devil’s best friend doesn’t have any effect on the virus. Sweden is just about done with the virus, although at the cost of higher deaths than Florida and California. A truly saintly governor would find a way to reach herd immunity with a minimum of deaths among the elderly, certainly lower than Sweden, and it then wouldn’t matter what the MSM said about the process.
Lucia,
“The weekly average “new cases†in FL is the same this week as last week! So the peak may have arrived! We’ll know better in 7-14 days.”
.
Yes, it looks like the peak in cases was near July 15 -16. Deaths in Florida from covid 19 will likely peak sometime near August 5th, then decline. I guess in the end Florida will do a little better than Sweden (~450 per million versus ~600 per million); much will depend on how many elderly are infected as the number of new cases falls.
Yep. We can’t really call peaks until after they’ve happened. But looking like it could be here.
Hong Kong now has more active cases than they did at their first peak in late March. New cases doubled in about a week. On a per capita basis, the number of cases is small, but that could change. The US is showing signs of peaking in the current hot spots, at least by my eyeball estimate.
DeWitt,
I understand China is playing whackamole with the disease.
.
Whackmole with really strong quarantine when the mole is seen will be the right decision for China if an effective vaccine comes out quickly. It would never have worked in the US.
Trying to lockdown your way out of the virus just leaves the country vulnerable for repeat occurrences, this time by deliberate action.
MikeN,
Perhaps. But honestly, this particular illness isn’t bad enough to use as a bio weapon. I don’t want to get it but it’s not the disease smallpox was back in the day. It’s not The Black Death (which is now curable.)
A hidden dimension of our accelerating ability to create genetics based vaccines that will suppress new diseases is the same ability to modify existing diseases to make them more virulent, deadly, and resistant to existing treatments. It’s almost a feature that by releasing this out in the wild on one’s enemy that it would likely backfire and infect one’s cohorts. A nightmare version of this is constructing a vaccine simultaneously with the new disease and covertly vaccinating one’s own population and then releasing it into the wild. If you want to go full conspiracy theory then covid-19 was just a test case for this evil plot. Have a good night’s sleep, ha ha.
lucia (Comment #188066)
July 23rd, 2020 at 3:39 pm
I would have to disagree with you. Just look at what happened in NYC when it first hit… and that happened in peace-time AND you knew it was coming. Now imagine that being dropped on you in war-time and you didn’t know it was coming.
.
A bio weapon doesn’t have to kill everyone to be effective… it just has to cause disruption and mayhem.
A weapon that caused 20% unemployment, a recession, and forced the enemy indoors for months seems pretty effective. I’ll have to check and see if it’s described in the Mahabharata.
As noted above, it looks like new cases in Florida may have peaked, with total confirmed cases approaching 1.5% of population.
Arizona seems to have peaked two to three weeks ago, also with cases approaching 1.5% of population. A number of other states (AL, SC, GA) seem to be leveling off as they approach 1.5%. On the other hand, Mississippi is at 1.5% with new cases still climbing and Louisiana is over 2% and might not have leveled off yet.
——
The CFR for the USA as a whole seems to have leveled off at 1.3% since the start of July. I calculated that assuming an average 7 day lag between positive test and death.
It looks like new cases might be leveling off at about 70K/day, so 7 day average deaths might stay under 1000/day.
Daily deaths from coronavirus have been over 1000 the last two days.
MikeN,
That’s actually three days (worldometers.info):
July 21 1,165
July 22 1,205
July 23 1,166
And 516 so far today.
Tom Scharf
Better story! Imagine being doctor Weigle forced to develop a typhus vaccine to protect the Nazis. Succeeding and then. . .
https://www.npr.org/2014/07/22/333734201/how-scientists-created-a-typhus-vaccine-in-a-fantastic-laboratory
In the last week, there have been 6132 deaths in the US; that is 876/day. So there is a good chance that will peak at under 1000/day. Data from the covid tracking project. For some reason, Worldmeters has about 8% more total deaths.
Alas… https://open-covid-19.github.io/data/v2/epidemiology.csv is not updating…. at least not for me. It’s stuck with most US data on July 22, some world data on July 23… (This happens from time to time. I don’t know why. Doesn’t seem to be a caching issue. . .)
.
The US’s first wave wave (and 2nd if it occurs) may just always have a series of wavelets. Still, we may be finishing the 2nd of the two “major” wavelets. The parts of the country that haven’t had “waves” are fairly unpopulated.
.
Things to watch for now:
1) Will the decline look fast-is or s_l_o_w relative to the rise during the virus expansion?
2)Will we have a 2nd wave before the vaccine is approved and becomes effectively available?
WRT (1) In a classic SEIR, the rate of increase and decrease in cases look symmetric around the peak. We haven’t been seeing this. I know the lack of symmetry is what we expect if the reason for the peak is not due to herd immunity at “business as usual” behavior. I haven’t modeled what the shape of the case vs. time curve looks like of transmission and so on is heterogeneous. So, the lack of symmetry is consistent with not having true herd immunity, but it may be consistent with other theories.
.
WRT (2): Looks like big-pharma is doing an awesome job. Fingers crossed on getting something approved. Plans to get it pushed out seem to be in place…. but.. well, we’ll see. I’ll want one right away, but I suspect in terms of containing spread— which is important to public health in pandemic, the priority should be to people living in aggregate communities (e.g. college dorms, community housing of any sort), people who must gather in fairly large groups for good reasons (e.g. high school, college, food processing industries), people in public facing jobs (e.g. retail, waitresses, bar tenders, receptionists, teachers), those in health care.
.
My bet is the priority will include “older people”. That’s important to stem deaths (which does matter), but at this point less important to stem spread. Older– especially retired — people who so not fall the categories I indicated as warranting priority can shelter pretty easily. If there is any shortage of vaccine or slowness in distribution, we should be low on the priority. Note: having said that I want to be first in line!!!! Sheltering is super boring.
The politics of getting in line for the vaccine might get pretty ugly. Unfortunately it will likely be racial with some justification.
DeWitt, I’ve been using the CDC numbers. I should start using worldofmeters.
Especially with CDC’s daily count including updates to previous days going back weeks. They had a number over 2000 one day recently.
Tom Scharf,
One thing: Initially, it will need to be voluntary especially if there is a shortage. Given the number of anti-vaxxers and so on, that might alieviate shortages a lot.
.
If the above criteria hold, the “public facing jobs” and some categories like “food processing” rope in enough minorities to not lead to cries that minorities are being deprived of the vaccine. But yes, they might need to prioritize “living in hard hit communities” (by zip code). Plausibly, that should be on my list. I just hadn’t thought of a way to flag it– but living in a particular zip code could work.
Blacks are the hardest hit by the virus and the least likely to say they will get vaccinated, a 20% vaccination gap between Whites/Hispanic (74%) and Blacks (54%). Unclear why this is.
https://www.pewresearch.org/fact-tank/2020/06/04/black-americans-face-higher-covid-19-risks-are-more-hesitant-to-trust-medical-scientists-get-vaccinated/
.
Team Science has some work to do here. I recommend they use different techniques than a mass shaming campaign. Like toilet paper though, I foresee a rush to get vaccinated and the psychology will kick in so everybody will want it.
Tom
Thanks for the link.
If 54% get the vaccine, that will still help a lot. Sure it’s might not be enough for herd immunity, but more than 1/2 the population being immune would be great.
Lucia wrote: “Given the number of anti-vaxxers and so on, that might alieviate shortages a lot.”
.
I’m no anti-vaxxer, but I’m certainly not hurrying to beta test a vaccine that cut every corner possible to get to market! Vaccines do have risks associated with them, some can be extremely severe.
There should be some combination of the worst hit areas and the most vulnerable populations to be vaccinated first. Healthcare workers, essential businesses, old people, and big breakout regions. Oh, and of course rich virtuous liberals, ha ha.
Tom Scharf (Comment #188189)
July 25th, 2020 at 11:06 am
.
“The politics of getting in line for the vaccine might get pretty ugly. Unfortunately it will likely be racial with some justification.”
.
Not sure which direction you’re going with that, but it could go either way.
.
Just imagine if there is a concerted effort to get “urban areas” vaccinated and there’s a safety issue with the vaccine.
Hispanics and Blacks are the hardest hit populations. One could justify them getting priority to a vaccine. One could probably also accomplish the same thing with some less divisive race neutral policies. Given the current political environment I would expect divisive to be a high priority on most politician’s lists.
.
One can refuse the vaccine. There is risk, there is also risk not getting the vaccine unless one’s area can reach herd immunity without one getting the vaccine.
Tom Scharf,
“unless one’s area can reach herd immunity without one getting the vaccine.”
.
I think by the time a vaccine becomes available, herd immunity will be the rule, not the exception, and those identifiable groups who have had high total infections (and unfortunately, deaths) will have extremely low rates of infection by the time a vaccine becomes available. Vulnerable people (elderly, other serious health issues) should have highest priority, but I am guessing politicians will end up having the very highest priority for vaccinations… followed by health care workers, then politically favored groups, the elderly, and everyone else. First and foremost, politicians.
DaveJR,
Maybe you will find this paper of interest: https://www.orau.gov/cdcynergy/web/im/Content/activeinformation/resources/IM_chenvacrisks.pdf
.
The paper seems thoughtful and well written. I’m still going to get the flu vaccine.
SteveF,
Even if herd immunity is reached, I’ll probably get the vaccine. I’ll read reports on safety first, but I’ll probably get it.
All herd immunity means is no-exponential growth. I don’t want to get it. If it’s not eradicated (which it won’t be) I’ll get a vaccine. I’m glad I’m vaccinated against smallpox, polio and so on. If I hadn’t had measles I’d want to have been vaccinated. And so on.
For those of you who are brave and/or paranoid and/or trusting you can sign up for a phase 3 covid vaccine trial here:
https://www.coronaviruspreventionnetwork.org/
Tom,
Some of the studies require you to be studied for 2 years. My main concern is that as part of the study group, I would not be allowed to get a working vaccine that became available before the study was over two years from the start! 🙁
I mean…. there’s a good chance I would only get placebo. I’d remain susceptible for sure and I wouldn’t be able to get a vaccine.
My understanding is there are no placebos in a vaccine trial of this type. It is already known unvaccinated people can get the disease. Some variations are the amount of dose or whether you get two doses (at least that happens in a phase 2 trial). I’m sure your free will is not taken away for future healthcare decisions. I don’t know all the details though. I signed up, I can always just refuse later.
Lucia,
“ All herd immunity means is no-exponential growth.â€
.
Yes, but it also means the incidence drops off exponentially. There is always the possibility of a local “flare-upâ€, especially in places where overall exposure of the population has been low, but at some point the risk of catching the illness becomes very small. The only things that can change the normal course are 1) herd immunity that also depends on changes in behavior, not just lots of resistant people, 2) the virus mutates fast enough that it evades an existing immune response, or 3) immunity doesn’t last long enough for the virus to die out (the pool of susceptibles renews itself constantly). If it is either of the later two, then a vaccine is not likely to be very effective.
.
The Swedes have discussed a couple of different scenarios for the end of the pandemic (without a vaccine). IIRC, their ‘most likely’ estimate is a continuing low level of cases until the illness dies out most everywhere in Sweden, and a thousand or two more deaths.
SteveF (Comment #188216): “immunity doesn’t last long enough for the virus to die out”.
.
Is that not the case with the four common strains of coronavirus? They have not gone extinct.
SteveF,
Yes. But in the context of getting a vaccine: I think it will still be worthwhile (provided it’s pretty safe.)
.
Obviously if the virus mutates in a way that evades existing immunity all bets are off. But in that case, we’ll have herd immunity won’t exist vis-a-vis the new mutation and the vaccine likely won’t work. That’s the situation with any mutation. Covid-19 is, itself a mutation of a coronavirus.
SteveF
I think this assumes the only useful protocol is to take it once and then be done. If the vaccine was safe, I’d happily be vaccinated each year. I don’t think I’d be happy with getting vaccinated every month, but some people do inject medication every day. I injected my cat with insulin twice a day.
.
So I’d consider a vaccine that gives only 1 year immunity effective for the individual. So I’d have one.
.
If the vaccine needs to be taken once a month, likely we won’t wipe out the disease. The only disease we’ve managed to wipe out so far is smallpox. That doesn’t mean measles vaccines are ineffective for individuals.
Studies are saying immunity won’t last because antibody levels drop within three weeks. I would expect this was typical behavior for antibodies.
Mike M,
“Is that not the case with the four common strains of coronavirus? They have not gone extinct.â€
.
Good question. There is an endless supply of naive individuals (from births) that can help allow a virus to remain in the population. My 18 month old granddaughter pretty much constantly had a two or three day cold when she was going to day care, but my wife and I never caught one… no doubt because we had seen the same or similar viruses before. (When she stopped day care for several months, no more colds.) But yes, immunity to any virus has to drop over time, as we know from experience with many illnesses and many vaccines. The severity of the illness on second exposure is very likely to be much less, since some immune memory likely remains for life.
.
I suspect the wide range of severity, and the observation that at least 10 of 11 cases are asymptomatic or nearly so, indicates resistance at least in part due to people having been exposed to other related coronaviruses, which primes their response to covid-19. There are conserved parts of covid-19 which the immune system clearly recognizes from previous exposure to “common cold†coronaviruses. Even a majority of very elderly people (over 85) either do not contract the illness when exposed, or if they do, don’t die from it. Maybe if an elderly person had a coronavirus cold in their 60’s they are resistant, but not if their last exposure was in their 30’s. Remember that only ~15% of common colds are due to coronaviruses.
.
I believe it makes sense to think about covid-19 not as some uniquely dangerous virus, but simply a virus where “the herd†has only a relatively low level of existing resistance. Young children, with very strong immune systems, almost never have serious illness, while older people (like you and me) are at much greater risk, since our immune systems have aged… along with the rest of our systems. 😉 Had we been exposed to covid-19 in grade school, it seems to me we would be at much lower risk of death from the same virus now. The present pandemic is a once-only event. After the pandemic passes, and most people have some resistance, covid-19 should be little different from other respiratory viruses.
Lucia,
The chance of completely eliminating the virus is low, especially since so many carriers are asymptomatic or nearly so. An aggressive immunization program would have to target everyone in the world, be extremely effective, and be conducted quickly, since lots of susceptible individuals are being born every day. Unless the virus mutates very fast, annual immunization is unlikely to be needed. I suspect in the long term, childhood immunization will be enough to keep the toll from the virus low, and even that may not be needed. The transition from now to the ‘long term’ is the difficult part.
SteveF,
For at least some people, they symptoms are worse than a common cold from the other Corona viruses. That’s why I would want a vaccine and would take it even if the protection only lasts a year. I could just get it when I also get my flu shot.
.
I personally don’t want flu. That my getting a shot also reduced the likelihood of an epidemic is a bonus, but my main concern in getting the shot is myself. It will be the same with Corona virus.
.
But yes, there may be good reason to expect the pandemic will only last about a year – – possibly with two waves, possibly with only one. We don’t really know that yet, but it’s in the realm of possibility.
Steve F– Cross posted! 🙂
I agree the chance we will completely eliminate this virus is low. I don’t think it’s going to happen. I don’t think we will succeed in having a sufficiently aggressive worldwide immunization program.
.
Oh… I should add: I might be reluctant to have a vaccine if it happens to be from China. On the one hand, I don’t buy the conspiracy theories that China was somehow doing something on purpose. OTOH: I don’t trust China with something very important unless there is a 2nd very scrupulous external verification process.
.
My view is in their own country they don’t rank risk to individuals highly. Our system can result in screw ups that result in risk to individuals; their often just doesn’t care.
MikeN,
“ Studies are saying immunity won’t last because antibody levels drop within three weeks. I would expect this was typical behavior for antibodies.â€
.
Yes, and those who suggest otherwise are either mis-informed or dishonest.
Lucia,
“My view is in their own country they don’t rank risk to individuals highly.â€
.
The individual is but fodder to the CCP (OTOH, high ranking party members lead a pampered and privileged life…. unless they cross the wrong person). Individuals not in the party are no different from individuals everywhere: most strive to improve their economic lot, and care about their family and friends. The problem is the CCP, which, paraphrasing a well known individual, seems to me nothing more than a group of people running ‘an evil empire’ that willfully does lots of bad things, domestically and internationally.
SteveF,
Yes. My distrust of China is distrust of the CCP. I suspect individual Chinese are just as trustworth as individual Americans, Europeans, Africans and so on.
But it won’t be individual Chinese ultimately vetting any vaccine that comes out of China. It will be the CCP, who I consider “China” the country.
.
So…
My only doubt is how long before he faces discipline from the lefties at Yale who want the response to the covid-19 pandemic to fundamentally change how people are allowed to lead their lives. https://www.newsweek.com/key-defeating-covid-19-already-exists-we-need-start-using-it-opinion-1519535
.
I can almost hear the howls from the NYTs demanding his head.
lucia (Comment #188233): “I agree the chance we will completely eliminate this virus is low. I don’t think it’s going to happen. I don’t think we will succeed in having a sufficiently aggressive worldwide immunization program.”
.
No, the chance is not low. It is precisely zero.
To date, we have succeeded in eliminating *one* human viral disease, although great progress has been made on several others. Those are diseases that are far more serious than the Wuhan virus and for which antibodies last a lifetime. Eliminating them takes decades of sustained effort.
A program to eliminate the Wuhan virus would be a criminally irresponsible waste of resources.
SteveF (Comment #188229): “…The severity of the illness on second exposure is very likely to be much less, since some immune memory likely remains for life. … I believe it makes sense to think about covid-19 not as some uniquely dangerous virus, but simply a virus where “the herd†has only a relatively low level of existing resistance. … The present pandemic is a once-only event.”
.
All of that makes perfect sense.
This says that people get a symptomatic coronavirus infection about once a decade:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7184402/
MIke M,
So maybe a covid-19 immunization once a decade? That is certainly a practical rate… sort of like tetanus. Every year (or worse, more often!) is more of a societal burden.
SteveF,
We should follow the science on the pandemic unless it makes Trump look good.
AFAICT, the designed trials that the press touts as showing that hydroxychloroquine doesn’t work did not include either zinc or antibiotics. They were designed to fail. The Swiss ‘experiment’ reported in the article is pretty solid evidence of effectiveness. Needless to say, this is the first time I’ve heard of it.
It looks like new cases are peaking, well maybe not in CA, so of course the ‘experts’ now say we should lock down again.
Locally we’re running what amounts to a mask effectiveness experiment. Cases have been increasing rapidly lately, so a mask requirement was mandated in the county where I live about a week ago. I do see much more mask use when I’m out. I’m betting that cases will continue to increase at about the same rate.
DeWitt,
“Needless to say, this is the first time I’ve heard of it.”
.
No surprise there. Unfortunately, the left in the USA wants effective treatments to fail, so that the “covid crisis” continues for as long as possible. Or as the despicable and dishonest Rahm Emanuel noted: “Never let a good crisis go to waste” when you want to force your lefty preferences on the public.
Re coronavirus vaccines, I ran across this not encouraging statement as to the common coronavirus infections:
Source: “The common cold: a review of the literature”, Dennis Wat, European Journal of Internal Medicine 15 (2004) 79–88.
Free download, but I neglected to save the link.
The paper he cites is: Bradburne AF, Bynoe ML, Tyrrell DA. Effects of a ‘‘new’’ human respiratory virus in volunteers. Br Med J 1967;3(568):767 – 9.
I have not gone looking for that one.
———-
Tangentially related and possibly encouraging: It seems that the “Russian flu” pandemic of 1889-90 may have been caused by one of the common coronaviruses that is thought to have jumped from cattle to humans at about that time. That pandemic killed a lot of people, mostly old and/or with preexisting conditions. If it was the coronavirus, it then just settled down to being a cause of the common cold.
——-
Taken together, these seem to agree with SteveF’s speculations in Comment #188229.
I’m seeing more mask use, but less social distancing. People used to keep distance at grocery stores, now not so much.
MikeM,
“Taken together, these seem to agree with SteveF’s speculations”
.
There is little new in the world, save for naive people… which is not their fault, since we are all born naive. The lack of perspective in the covid-19 hysteria is not surprising. What is surprising (and more than a little disturbing) is the lack of calm, reasoned voices providing a measure of perspective amid the tumult of the moment.
Changed my mind…. cases may not have peaked in Florida. We’re approaching… but not there.
My county went to a lot of mask use during a period of rapid increase, the cases leveled off within a couple weeks, still not declining. Not clear whether this was casual or not. People are wearing masks, but not staying home. Traffic is as high as ever.
Tom Scharf,
“People are wearing masks, but not staying home. Traffic is as high as ever.”
.
Surprisingly enough, life goes on, in spite of covid-19. That is not a bad thing.
Lucia,
Noise rules. The peak in cases is difficult to pick (see the Florida covid-19 dashboard), but it looks to me like the peak in deaths is very close…. at maybe ~130 per day. But don’t worry, Florida’s peak rate of ~130 is much worse (infinitely worse!) than New York’s peak of ~1,000 deaths per day, because Florida’s governor’s is an evil Republican, and New York’s governor is very close to a (Democrat) saint.
Yeah… but part of the reason detected cases is down is number of tests are down. If I assume tests are drawn randomly from the population, then i get the rate of new infections rose last week. . . (Of course tests are NOT drawn randomly… but still. Anyway, I don’t think I can estimate the effect due to selection bias in who gets tested. )
.
Yes, there is also noise. But my “best” estimate is rate of new infections is still rose rather than fell last week.
.
It’s near the peak though. I’m going to try to look at this over a few weeks … but yeah, noise rules especially for finding peaks.
Lucia,
“Yeah… but part of the reason detected cases is down is number of tests are down.”
.
I think in Florida the number of tests is more a function of symptoms than anything else. If number of tests is down, that suggests the number of symptomatic illnesses is down.
SteveF,
Maybe. I’m just doing calculus…. I’ll explain tomorrow. You and a number of people here will see an important embedded assumption (which everyone will point it can’t be entirely right), but it will explain why I’ve changed my mind. 🙂
.
Bear in mind, what I think most likely isn’t necessarily real. Also: you are right that things are noisy.
Reduced tests should only be a concern if due to a bottleneck in getting the results. A delay in results would mean a delay in positive tests, leading to a misleading drop in positive tests.
MikeM,
It’s not an issue of “concern”. Its merely that if the “rate of infections” is constant, and you reduce number of tests, that tends to result in a lower number of cases detected. This has to hold if the ‘test sample’ is drawn in an unbiased manner from the full population. (It isn’t, but it’s a good first estimate to understand the effect.)
.
Oddly, it’s one of the correct things Trump has observed (in his usually utterly clumsy way) : If you test more, you find more cases. The flip side is if you test less, you find fewer cases. Of course increasing tests is not the only reason you detect more cases. . . (And that’s where Trumps stupid clumsy way of saying things gets him in trouble. )
lucia (Comment #188260): “Its merely that if the “rate of infections†is constant, and you reduce number of tests, that tends to result in a lower number of cases detected. This has to hold if the ‘test sample’ is drawn in an unbiased manner from the full population. (It isn’t, but it’s a good first estimate to understand the effect.)”
.
There is no reason to assume that the rate of infection is constant. That amounts to assuming that we already know the answer. Why are we looking at trends in positive tests? To see if they are going up or down.
.
Samples drawn in an unbiased manner might be reasonable for testing people when there is no reason to expect they have the virus, such as before a medical procedure or because it is a job requirement. But unless such tests are a major source of new cases, that won’t be a big effect.
.
If the *supply* of tests is reduced, then presumably there will be some reduction in positive tests. But there is no reason for that to be true if *demand* for tests has dropped. So it is not an issue unless there is a bottleneck in testing. That may be the case and the bottleneck seems to be in getting the tests processed.
From the abstract of an article in Emerging Infectious Diseases, published by the CDC:
https://wwwnc.cdc.gov/eid/article/26/5/19-0994_article
Published in May, but apparently written last year.
——–
The abstract of the abstract would be something like: We don’t know how to control epidemics.
MikeM
Actually… I don’t. I’ll show the math tomorrow– it will have partial derivatives (holding other things constant.)
That’s what I think.
.
The cause of the reduction (or increase) doesn’t matter unless it changes who is likely to take the test.
.
Wrong. Change in numbers being tested is an issue no matter what the reason for the change.
NPR, ha ha:
.
Florida Case Count Surpasses That Of New York, The Country’s Original Epicenter
https://www.npr.org/sections/coronavirus-live-updates/2020/07/26/895596931/florida-case-count-surpasses-that-of-new-york-the-countrys-original-epicenter
.
“And its death toll is starting to catch up. On Thursday, Florida reported 173 new deaths, its largest increase in a single day.”
.
It’s almost like they are rooting for more deaths. What are those numbers NPR? I guess not important enough to inform the readers, you know, they are “catching up”. They left them out. Here is the data for those discerning individuals:
.
NY: 32,305 deaths
FL: 5,853 deaths
.
The states have equal populations. The WP did a hit job on DeSantis as well this weekend. There’s lots of uncertainty and unknowns in handling this outbreak. The overt bias in the media is ridiculous and callous.
.
I don’t think there are any testing shortages in FL right now, everyone who wants tested is getting tested.
lucia (Comment #188264): “Change in numbers being tested is an issue no matter what the reason for the change.”
.
I can’t imagine why.
I don’t really know what to make of trends in percent positive tested. It seems like a combination of testing ability, the size of the actual outbreak, the size of other potential diseases with covid like symptoms, public awareness, ease of access to testing, etc.
.
I haven’t followed it closely. Perhaps it is a proxy for the size of the outbreak but actual cases and deaths seem like more direct measurements.
MikeM
Yes. I overstated. What’s more accurate if that the change in the number can have an effect even if it’s demand driven. It depends on what drives demand.
FWIW: based on news reports, there do do seem to be some bottle necks in testing. News articles report people waiting in lines leaving the lines because they were too long. There may also be bottlenecks in reporting test results. I’m not sure if the test being administered is logged when it’s administered or only after the test is reported.
With Covid testing now part of every medical surgical procedure in and out patient, it has to be having a positive impact on the number of positives. Anecdotally, the case at my wife’s work was due to the asymptomatic mother testing positive prior to C-section which triggered testing of the entire family none of which were symptomatic.
Of course if there isn’t significant cases in the community the medical testing will also be pushing up the negative results which should decrease the positive ratio. I’d be worried if positives are going up along with the positive rate.
There’s currently a lot of hand wringing in OH about the rise in cases. The positive rate has fallen from June 1st from 9% to the low 6% and deaths from 6% to just under 4%. Deaths could be lag, but the positive rate is trending the good way. Too early to tell for sure but it looks like the current rise peaked around July 17th. FWIW we are sitting at .7% of the population testing positive with fewer than 11% tested.
Andrew P (Comment #188280): “With Covid testing now part of every medical surgical procedure in and out patient, it has to be having a positive impact on the number of positives.”
.
That inspires a Fermi calculation.
A bit of searching reveals that under normal circumstances there are about one million surgeries a week in the U.S. The last week had about 6 million tests, about 5 million in each of the prior two weeks.
So if those getting tested because they are to have surgery are equally likely as others to test positive, then that would up the number of positives by as much 15 to 20 percent. Of course, the number of surgeries being done is still way down; so maybe the effect is no more than 10%. And those awaiting procedures are much less likely to test positive than those getting tested because they think they might be positive.
Lets see, guess that 10% of those sick are getting tested and that the untested but sick would test positive for 10 days. With 65K positive tests per day, that would imply 6.5 million who would test positive on any given day. That is 2% of the population, compared to about 8% of those being tested. So that knocks down the number of positive tests from procedures by another factor of 4.
Bottom line is that people getting tested for medical procedures inflates the number of positive tests by maybe 2-3 percent, uncertain by at least a factor of two in either direction. Not a big deal.
Hi Lucia,
I am not seeing any evidence that the cases in Florida have not already peaked. You must be seeing something I am not.
.
The rate of deaths in Florida looks very close to peaking, although there are two different ways the deaths can be presented which can potentially give slightly different peak dates. Florida (covid-19 dashboard), reports deaths each day, but places those deaths on their graphic on the actual date of death, while Worldometers just takes the total number of deaths reported each day and places them on their graphic that same day. The 7-day trailing average from Worldometers eliminates the weekly reporting cycle, but displaces the peak to ~3 days later than what the true peak was; I think they should use a 7 day centered average, not a trailing average.
.
Taking the 3 day lag into account, the two methods should give very close to the same date for the peak in deaths, but the Worldometer method will show the date of the peak very soon after it happens. It looks like the Florida peak in deaths was July 23 or 24, and the peak in cases about July 14. Time will tell.
https://www.theatlantic.com/ideas/archive/2020/07/scourge-hygiene-theater/614599/
.
“Surface transmission of COVID-19 is not justified at all by the science,†Emanuel Goldman, a microbiology professor at Rutgers New Jersey Medical School, told me. He also emphasized the primacy of airborne person-to-person transmission.
“As he explained to me, as many as 100 people would need to sneeze on the same area of a table to mimic some of their experimental conditions. The studies “stacked the deck to get a result that bears no resemblance to the real world,” Goldman said.”
“COVID-19 has reawakened America’s spirit of misdirected anxiety, inspiring businesses and families to obsess over risk-reduction rituals that make us feel safer but don’t actually do much to reduce risk—even as more dangerous activities are still allowed. This is hygiene theater.”
.
I hope for the benefit and credibility of Team Science this turns out to be not true. 8 months in and we still have NO IDEA how much surface transmission matters, and now people are calling it into question after warning the rubes about this with a religious certainty. Here’s an idea, why doesn’t Team Science go find out the actual facts? The rubes are interested in effective risk reduction.
In the good news department, widespread outbreaks will shorten the vaccine testing cycles by weeks or possible over a month. Apparently they wait for a certain number of trigger events to complete a study. Comparing how many covid cases one group has when the control group has reached a trigger point (like X people testing positive). It does appear the control group may be a placebo.
Tom Scharf,
“This is hygiene theater.”
.
Yes, and the saddest part is that the only known, verified method of reducing transmission (N-95 masks) is specifically not recommended by the clowns posing as experts… “save those for health care professionals” is the usual refrain. It is all political theater.
.
On a lighter note: my nephew and family traveled by air a few days ago, and they encountered a couple at the airport walking around in white hazmat suits with hoods, gloves and booties. Needless to say, they attracted a lot of attention.
SteveF
I posted. It’s not far off….
.
The short version is if the change in (positive/number tests) is increasing, that’s some evidence the peak is not here. Last week, that underlying cases increased.
.
Moreover, I think that is a better estimate than looking at number of cases/ time peaking because of the effect of increase or decrease in rate of tests.)
.
So: I think the evidence it hasn’t yet peaked exists. But I also think we are near it.
.
I wouldn’t really argue about this much, but if I ignore these ideas, I’m pretty sure I conclude wrong things about Illinois where the number of tests has been soaring, the (positive/number of tests pretty flat) but number of cases rising in lock step with number of test. So I think Illinois spread is near peaked even though the number of cases/time is increasing briskly.
.
But I can’t really justyfing using different methods for Illinois and Florida. So I look at the effect of the change in number of tests. Other people dont– and in which case, of course they won’t “see” the evidence, because they don’t look at it. It’s still there.
SteveF
I’d like that dashboard better… but like Sweden, you also need to ignore the last two weeks. (There’s a note on the dashboard.)
Also: if number of tests is changing, I think you need to account for that. That data would be better (when fully reported) but the number of test issue still remains.
Sure. But even after time passes, I would still “correct” for rate of change in tests to get the best estimate of the peak. (Not sure it will matter that much…)
SteveF
Wow! You have to wonder why they were traveling and why they chose by air. Most people can opt to just not travel (or drive. But I guess if you drive you might to … omg… get gas!)
AndrewP
Yes. If both are going up, then the infection is definitely spreading.
Lucia,
“You have to wonder why they were traveling and why they chose by air.”
.
Could be they were making a joke. There are lots of funny hazmat coronavirus memes out there.
Tom, from the beginning CDC listed surface contact as not a primary means of transmission.
Are Naperville schools reopening?
Northern Virginia adopted a hybrid model, then pulled it back and it’s still distance learning only. Can’t really blame the teachers, as majority of parents opted out of hybrid.
MikeN,
I think so. I know Lisle schools are. Parents were discussing their choices on Nextdoor.
.
I’m generally unsympathetic to teacher moaning and groaning as a group. You’d think teaching was the worst job in the world and many don’t seem to realize many of the things they bitch about have analogs in all jobs.
.
But I am sympathetic during this pandemic. OTOH, I have friends working retail who have been public facing since we re-opened and who also worry about their employers staying in business. Some of them are less sympathetic. But honestly, I can see that none of the options are attractive for teacher and many are not attractive for parents or children.
.
Parents are all trying to decide what’s “least bad”. Teachers sometimes have less choice. ( I guess the can quit. But the consequences are pretty severe.)
.
Naturally, I can mostly hole up at home and watch this pann out.
I think there is not a single documented case of a teacher catching the Wuhan virus from a student. There are not all that many cases of children giving it to each other or to family members. There is no basis for closing elementary or middle schools. Distance learning tends to be ineffective for young children, especially if they are from poor families without the resources to compensate. Many parents have to send their kids to schools if they want to be able to feed their kids.
.
Keeping schools closed is an assault on the underprivileged by the privileged.
If you have been curious about the censored press conference by America’s Frontline Doctors, you can watch the video here:
https://www.bitchute.com/video/zr04GsUupOwk/
It is long (45 minutes) and very interesting. I am not much of a video watcher, but I found it worthwhile.
A transcript is here:
https://www.rev.com/blog/transcripts/americas-frontline-doctors-scotus-press-conference-transcript
.
Maybe they are right, maybe not. But that it is censored is very disturbing. What they are saying needs to be heard.
.
IMO, the star of the show was Dr. Joe Ladapo, starting at 33:27.
It occurred to me that maybe I should summarize the key points made by America’s Frontline Doctors.
Many doctors who are actually treating patients have been having great success with hydroxychoroquine.
The treatment protocol matters: dosage (very low), using zinc supplements, and starting early.
The approved studies didn’t use the correct protocol.
The drug is super safe.
The authorities, like Fauci, recommending against it don’t seem to have treated any patients.
There seems to be a bias against the drug, so that fraudulent or inane studies (like the one that used a toxic dose and found that it was bad for the patients) get a lot of attention, but the experience of actual doctors gets ignored.
Mike M.,
I’ve been saying that for quite some time. The studies were either designed by those willfully ignorant of the mode of action of hydroxychloroquine or were purposely designed to fail. Hydroxychloroquine does not act on the virus directly. It facilitates the transfer of zinc into infected cells. It’s zinc that interferes with virus replication.
DeWitt Payne (Comment #188398)
July 29th, 2020 at 8:23 am
I think that they were purposely designed to fail. Big pharma doesn’t want a cheap, readily available, drug taking away the profits they stand to make from a vaccine.
.
The efforts to silence “America’s Frontline Doctors” is extraordinary. Social media outlets are deleting their videos as soon as they pop up and the organizations website has ‘gone dark’.. . their webhost Squarespace hasn’t given any reason for killing their website.
.
The right of free speech is definitely being put to the test.
skeptical,
Well…. anyone who wants to do a big media blitz for anything remotely political and who does not want to risk private-enterprise censorship should spring the ~$15 for a domain name and pay hosting fees. That way they won’t have everything taken down.
.
They can simultaneously put things on twitter, youtube, facebook. But at least have the material online elsewhere.
.
If “America’s Frontline Doctors” had set up their own domain and hosting, Squarespace wouldn’t have had the power to take them down. It looks like the organized enough to pull a group together and have an event in DC. The extra $50 to host their own site should have been well worth it.
lucia (Comment #188404): “should spring the ~$15 for a domain name and pay hosting fees.”
.
Isn’t that what Squarespace is? Real question; I know very little about online publishing.
.
lucia (Comment #188404): “If “America’s Frontline Doctors†had set up their own domain and hosting, Squarespace wouldn’t have had the power to take them down. … The extra $50 to host their own site should have been well worth it.”
.
I don’t understand. Yes, they could do the Roy Spencer thing and be completely independent. But for most people that is untenable.
They could easily avoid being taken down by hosting services that are subject to pressure from ‘progressive’ groups: for example http://zensurfrei.com/
It is not a lot of money…. $250 per year minimal service, $500 per year basic service. They would have to pay someone to develop a site unless they got someone to volunteer; that might add a few thousand dollars. For a bunch of medical doctors, the money is not the problem. Maybe it’s a lack of knowledge.
They’re using the domain name…
.
https://www.americasfrontlinedoctors.com/
.
That looks to me like a domain name that they would have paid for.
As they have a domain, they should be able to change hosts and upload everything somewhere else. The should enquire about take downs first. There are TONS of hosts– (Hostgator, Bluehost, Dreamhost, Godaddy just google “web host”. )
.
If they have any brains and anyone with the slightest amount of Internet ability, they should be able to have things back up quickly.
BTW: By quickly, I mean up in less than 24 hours, or at worst 48 hours. They could probably even ask for a volunteer tech person to do it for them. There are enough potential supporters out there one would likely step forward.
MikeM
Huh? I self host. It might be untenable for an individual doctor. But it’s hardly untenable for a group of doctors who want to get their word out.
In other news, it looks like India is seriously ramping up their distribution of hydroxychloroquine….
.
https://www.ndtv.com/india-news/coronavirus-centre-allocates-4-24-crore-hydroxychloroquine-tablets-to-states-and-union-territories-2270871
.
They don’t make any mention of combining it with zinc, so maybe it won’t be as effective as they’d like it to be.
Internet * doctor’s tech ability = 0.
.
It is surprising how bad most doctors are with tech.
lucia (Comment #188413): “Huh? I self host. It might be untenable for an individual doctor.”
.
OK, but does that protect you from being deplatformed by WordPress?
So I am not sure about the distinction. Is it just that *if* you were deplatformed, you could migrate to a different service relatively easily.
.
People with limited internet savvy and 60 hour a week jobs are going to pick the easiest way of getting their site up. My guess is that they will have learned something when they get back online.
Skeptical:
.
Fuller quote:
.
(1) His claim that he wears the mask so he can visit his mom is clearly false. Mask or no mask, she doesn’t let him visit her and he know sit.
(2) In the same paragraph where he makes it clear he knows the strategy of self isolation exists (because his mom uses it) he claims the mask is the “only defense we have right now”. So: self isolation is also a defense. He doesn’t prefer that defense but clearly it exists.
.
Maybe it’s true that Covid is ravaging him psychologically because the bio on his profile page claims he has a fairly pricey distinguished education. But that paragraph suggests he is currently utterly incapable of critical thinking skills or applying logic.
.
The guy somehow is blind to his option of not going to Napa. If he is not a moron and is really truly worried about catching and subsequently spreading the disease, he shouldn’t have gone to Napa.
MikeM
Yes, it prevents that. I am not hosted by WordPress. As far as I am aware, they have zero power to de-platform me. I use their software– but it’s already released software.
.
The have no control over the database. In the past, I’ve found converters that exist that allow the material in the database to be fished out by other blogging software. There are other things I could do. I’d have to do some work but no, WordPress really can’t do much because they don’t host my site.
.
There are blogs that use WordPress and are hosted on WordPress servers. I know WUWT did that (and may still). I don’t.; ever have.
lucia,
Thanks. I was unaware of that distinction.
lucia,
What you’re saying about self hosting is true. But there’s still the service provider to be reckoned with. I have no idea whether ISPs ever mess with stuff like this / de-platforming people (I suspect not actually), but it’s at least conceivable to me that they could.
Tom Scharf,
Which made a number of the scenes in House where the doctors were supposedly running clinical analyzers, MRI and CT scanners unintentionally funny to me.
Even the radiologists who interpret the data don’t actually run the instruments.
mark,
Yes, at least hypothetically, the host could deplatform.
.
That’s why I think someone posting controversial material should phone to find out the hosts policy on deplatforming. They should probably check their history. But let’s face it: most hosts really, really don’t like to take down. Someone who hosts a lot of porn would probably have a policy to not take you down. 🙂
.
Those posting very controversial material should also look into protection against cyber attacks– merely using cloudflare is likely fine.
.
Hosting on a one of the “blogging software and hosting in one” services (e.g. Squarespace) is much more dangerous vis-a-vis takedown.
DeWitt,
Dentists mostly don’t take your X-rays. Some of this is related to the economics of providing services. The dental technician costs less. So they clean your teeth and take X-rays. The dentist probably can also do those things, but it makes more sense for them to spend their time on the services only they can provide and have someone less expensive do the other things.
.
The same goes for having a receptionist. The dentist could greet each customer at the door, check them in and… But it makes no sense to do that. (Perhaps it would make sense if they lived in some very remote part of the world and were limited to 5 customer a day max. But otherwise.. no. It’s better for them to push through a lot of customers. The main limitation for the dentist is the time he has to devote to doing things only he can do.
.
I’m sure with MD’s there are lots of tests they couldn’t do at all— there are so many specialized ones now. But the fact is, even if they could do it, it doesn’t make sense for them to do it. Better to send you to the lab to draw blood and so on and so on.
>I think there is not a single documented case of a teacher catching the Wuhan virus from a student.
The documented part is hard. How do they know how someone got infected? In a local elementary school, the principal was infected as well as some teachers and staff, one of whom died. I think they had a specific contact trail which is how the asymptomatic principal was detected as the second case. This also was how they could report the kids were not at risk because school had closed the day before.
WUWT left WordPress, but is now looking to move back.
Mike M. (Comment #188174)
July 25th, 2020 at 7:31 am
Average is over 1000/day now and it looks like it still hasn’t peaked.
skeptical,
Prediction is hard. Especially about the future.
lucia (Comment #188507): “Prediction is hard. Especially about the future.”
.
Indeed. But I never actually made a prediction. I merely observed that the 7 day average deaths *might* stay under 1000/day.
MikeM
Fair enough. Also, blog comments tend to be casual. So, it’s not always possible to distinguish between a “might happen” and “a prediction”.
I looked this up for something else, I was pretty surprised.
.
Death rates by state from coronavirus (COVID-19) in the United States as of July 31, 2020
https://www.statista.com/statistics/1109011/coronavirus-covid19-death-rates-us-by-state/
.
Florida is number 20 (?!). No doubt it will climb in the rankings as time goes by but the media coverage is highly distorted, which is not surprising. Illinois death rate is actual double that of Florida for example.
Tom Scharf,
“ Florida is number 20 (?!). No doubt it will climb in the rankings as time goes by but the media coverage is highly distorted, which is not surprising.â€
.
Yes, many of the ‘truly horrible‘ states like Florida, each suffering ‘catastrophic’ rates of covid, are in fact far better off than those ‘heroic‘ states with 3 to 5 times higher rates of death…. but run by Democrats. Talk about Orwellian! The MSM are simply dishonest and utterly shameless, but adamant in their opposition to Republicans who hold any elected office. Their reporting on covid (and every other subject!) reflects that opposition.
lucia (Comment #188507)
July 31st, 2020 at 5:27 pm
Yes it is. The growth in daily deaths in the U.S. over the last week has even surprised me.
SteveF,
The apparent death rate in states like NY and NJ is probably biased high because of the likelihood of a serious under count in cases because of the restrictions on testing at the time. If we had comparably measured rates of infection, FL, TX and CA wouldn’t look nearly as bad. If there were still old fashioned journalists reporting the news, everyone would know this. But journalism has been replaced by what Pielke, Jr. called stealth advocacy. It’s not even all that stealthy any more.
It would have been so easy for Republican-led states to show up the Democrats for the messy, disorganized lot that we are.
In the Tri-State area including NYC, we showed you the scope of the problem. We showed you the path towards a solution.
Florida, Texas and other late arrivals to the party chose not to pay heed.
Thomas Fuller,
So CA has a Republican governor? News to me. /sarc
Thomas
That is such an odd point of view!!!
.
Perhaps NY showed the world some things not to do (e.g. endanger care facilities by State order.) But I really don’t think they showed anyone what to do! One might as well say China showed people what to do. Other examples of how different choices played out preceded NY. These include Italy and so on.
.
I would note NY didn’t pay heed to what China showed them could be done. And despite lots of people dying in Europe, somehow NY didn’t manage to curb their infections and deaths.
.
So if you want to play that game, you really can’t get ahead of this.
Skeptical
I wouldn’t say it surprised me. OTOH: I really thought I had no idea what would happen. This fell in the range of my vague idea.
.
Initially, the infection has sort of localized areas. There was a bulge. Then it started popping up lots of places. I guess that is a pattern that doesn’t surprise me. But I didn’t really have any idea what I thought numbers would be. I’m not stunned that the per capita infection rate is larger now that it’s spread.
.
Sadly, strategies that might work in a geographically contained infection aren’t necessarily designed to prevent spread once it’s spread all over the place. We aren’t seeing the intense death rate per capita of NYC, but we are seeing large value spread over the country.
lucia,
China strictly quarantined Wuhan, internally anyway. But when it was suggested that there should be travel restrictions or quarantines for people leaving the NYC area in March, Cuomo and di Blasio went ballistic. Now, of course, they’re all in favor of travel restrictions for people in the current hot spots.
March 28,2020:
So Cuomo abetted the infection of the rest of the US. Some hero.
I think the infection is winning, and it doesn’t care what color your politics are. One can slow things down with a full lock down, but at great economic cost that is not sustainable in the long term. Those who sustained lock downs are now at the mercy of all the other global hot spots.
.
NYC (and all the other brave and shining lights on the hill who showed everyone the way with their brilliant science and policy as Fuller seems to think) are the biggest messes. High density urban living is the way to go for right thinking people where they can mass up and sneer at the rubes. Unless of course there is a pandemic in which cause this living situation is suboptimal. Sneer at the rubes anyway because they should have learned from the smart people’s stupidity, or something along those lines of thought. Certainty is one’s innate superiority is definitely a sign of … brilliance, love, and tolerance.
.
This pandemic has made every expert and organization look like a fool at some point or other. No vanity has been left untouched. Fuller can continue to look into his mirror and ask it who is the fairest of them all though, he will no doubt keep getting the same answer.
DeWitt
Yep.
Yep. But of course, everyone should all learn from their (self-serving) mistakes. If only Cuome and deBlasio would admit they made them and they should have taken Trump’s advice when he gave it. 🙂
.
Admittedly, Trump has also spewed lots of bad advice long with mystifyingly incomprehensible advise. But he had some spot on suggestions at the start, was roundly criticized, and now those who are adopting those decisions long after they chose not to implement then and, in fact, decried them, are acting as if they were somehow “wise”, their decisions were terrific and that their absolute blunders are somehow excusable while those in the “other” can are not excusable.
.
I’m not surprised mistakes were made. I expect mistakes to have been made. Leaders were acting with very little information. Everyone made mistakes. But the line that somehow NY serves as some sort of good example and all their mistakes were somehow to be excused while others are bad examples because they didn’t learn from NY is just a load of hooey.
.
I’m glad I’m not a mayor, governor or president. This is a hard time. It’s no fun to try to balance risks in a period of uncertainty. But seriously… Cuomo and deBlasio absolutely do not stand out as examples who did better than average– not even correcting for NY being hit early.
Thomas Fuller (Comment #188527): “In the Tri-State area including NYC, we showed you the scope of the problem. We showed you the path towards a solution.”
.
And the “path to the solution” was what? Killing off all the nursing home residents? Because I don’t know what else NYC did that was not done in many places that have had a resurgence,.
skeptikal (Comment #188525): “The growth in daily deaths in the U.S. over the last week has even surprised me.”
.
Indeed. I thought that if the average did get over 1000 a day it would be because new cases had not really peaked. We shall see if the last week or so was growth or a bump.
Brazil and India continue to grow. There is little doubt those places have substantial under counts. I’ll have to check their politics to figure out if this is a bad thing, or if it is simply just a warning to others from very articulated intelligent people.
Tom Scharf (Comment #188533): “I think the infection is winning, and it doesn’t care what color your politics are. One can slow things down with a full lock down, but at great economic cost that is not sustainable in the long term.”
.
Many of the European countries that were so “successful” in containing the virus as well as ones that actually appeared to be successful are now having a resurgence. Some of the East Asian countries also. And Australia, although cases are still very low.
.
The scorched earth campaign against the virus has accomplished little good, possibly none in the long term, while doing massive harm.
Tom Scharf
Clearly, they should have studied NY and learned. . . [/sarc]
.
My guess is given the state of testing back when NY was hit, per centage wise, their case count has a greater undercount than FL. I don’t know what the availibity of testing is in India or Brazil, but likely an undercount.
.
My guess is we are doing a fairly good job of counting in Illinois now but had a large % undercount in March. This is part of the explanation of the lower number of deaths/cases. (It also partly explains a change in lag. Early on, only people with major symptoms tended to be counted. We now catch some people earlier.)
.
What the best strategy will have been will depend partly on when a vaccine becomes available. I doubt NY’s performance will land near the top, especially for a city in a rich country.
Brazil’s testing is in line with most other countries. India is #4 in total tests, but per capita this is way behind. It could be because of high temperatures and less AC, they are not as high-risk.
Japan may now be losing at whackamole.
https://www.bloomberg.com/news/articles/2020-07-31/japan-acted-like-the-virus-had-gone-now-it-s-spread-everywhere?utm_campaign=news&utm_medium=bd&utm_source=applenews
“I don’t know what else NYC did”
Stuffed people into subway cars.
They took my treat it like chicken pox plan for kids and used it on adults.
https://www.dailymail.co.uk/news/article-8583649/At-260-campers-staff-infected-coronavirus-Georgia-sleepaway-camp.html
lucia (Comment #188543): “Japan may now be losing at whackamole.”
.
Thanks for the link.
I’m curious to know if they now have the (probably) more contagious Italian strain there.
.
The authors appear to be innumerate (duh):
Given the underlying logic of the article, there is zero and not-zero, nothing in between.
.
But there is a more reasonable view:
Wise. Sounds like the Swedish model.
.
But of course, there are still the bleatings of educated idiots:
.
No, they can’t reduce the number, just spread them out.
Magic ingredients to super-spreading combined
* Poor ventilation.
* Communal sleep areas.
* Singing/cheering.
Honestly, I don’t know how you would manage to get kids to sleep with masks on. I don’t know how you can get adults to consistently sleep with masks on. Even if you try to wear on to sleep it will probably slip off. You won’t notice because you are asleep!
lucia (Comment #188547): “Magic ingredients to super-spreading combined”.
.
Indeed. And we don’t know the extent to which the kids gave it to each other or caught it from staff and counselors. Presumably, the grade school age kids were not sleeping alone in the cabins.
Lucia,
“Japan may now be losing at whackamole.â€
.
A few places may succeed at whack-a-mole, but realistically, most countries (and most US states) will hit their HIT long before a vaccine becomes available. It is becoming ever more clear that the HIT varies a lot between places.
.
The important difference will be how many elderly die in the process. Evil DeSantis (Florida) will do pretty well relative to other high population places. St Cuomo of New York ought to burn in hell for his idiotic policy choices that killed thousands…. but he will be hailed a hero. We live in interesting and foolish times.