So to conclude from an average CFR of 2% that is a cohort of 1,000 students aged 18-21, 20 will die, is inaccurate. And you could say the same of the cohort was care home residents. So what do you use it for? OK it's useful to track the direction of travel which, fortunately, is downwards, but it can easily can overused.
I assume CFR means "confirmed fatality rate." If so, you can have these for the human race as a whole (all ages, races, and genders) or for some subset of it. There is value even for the overall human rate, because it allows you to compare it to plagues-of-the-past, and see if you can learn anything (like, "we really need to get more aggressive on this one" or "Thank God this one is only X percent, we could have had MERS which is 30%."
Also, if you notice that the CFR as calculated in country X is 10%, and in country Y you get 2%, that tells you it's time to look at what the two countries are doing differently. There might be something you can change to bring X down to a better number.
But this does lead on to another interesting issue about what is immunity. From what (ok- little!) I know, an argument against the much vaunted T-cell immunity is that it's not immunity at all because the mechanism only kicks in post infection. This may be true of current vaccines. I.e. infection is not prevented but the effects are mitigated.
Whilst this is of some interest, I'm not sure how important this distinction is. If a vaccine does not stop me getting infected, but reduced the effect to make the symptoms mild, or even non-existent, that is good enough for me. So if you talked about T-cell protection and avoided the term immunity, there may still be a case for citing it.
As far as immunity--
I am not an immunologist, but I know a little (very little). Immunity is mostly a name for what happens after infection, just as overwhelming firepower with massive destruction and nukes is a name for what happens after a really badly-thought-out invasion. It's a defense. It only gets going once there is a threat. That means at the point of infection. Nobody has an immune system like the one I tease my mother about--one so powerful you can just walk around and kill the viruses in the air around you for six feet on either side.
Immunity has two parts--the innate system, which is basically the OH HELL NO" response that shoves off pirate boarders faster than they can get a foot on the deck of your ship. This is really, really fast but as I understand it, not particularly well-guided. It's a "kill them all, God will know his own" kind of defense. And it works best in children. (Does anybody know, is this the system responsible for allergies when it goes wrong?)
The other part of immunity is the big convoluted and complicated sysem with T cells and B cells and antibodies and all that, which relies on learning an invader's pattern and mounting an attack against it--and then keeping the pattern on file for as long as possible, in case the invader should ever come back again (so an attack response can be mounted far faster, and the human being might have no idea at all there had even been a threat). This is where vaccines work, as well as natural exposure that one survives.
I assume CFR means "confirmed fatality rate." If so, you can have these for the human race as a whole (all ages, races, and genders) or for some subset of it. There is value even for the overall human rate, because it allows you to compare it to plagues-of-the-past, and see if you can learn anything (like, "we really need to get more aggressive on this one" or "Thank God this one is only X percent, we could have had MERS which is 30%."
Case Fatality Rate, the percentage of people diagnosed with a disease who subsequently die of it. And, yes that will vary depending on whether you're considering a global picture for the whole human race or for subsets of it. For covid-19 we know that older people, or those with other health issues, are more likely to die. Access to quality health care will improve your chances of living.
Here's a list of different diseases to compare - so if you catch covid19 (with preliminary CFR estimates of 1-10%) you're at least an order of magnitude more likely to die than if you catch seasonal 'flu. Of course, there's also the question of how easy it is to catch different diseases.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Over 437 thousand new global cases yesterday, the highest total so far, and over 6,700 deaths. Despite clear declines in numbers in Brazil and India. Europe is surging, with close to 200 thousand new cases and 1800 deaths.
The UK saw a record total of new cases, close to 27 thousand, and close to 200 deaths. The brakes may have been applied too late, the second wave surge looks well under way.
The USA saw close to 64 thousand new cases and over 1200 deaths. Both now look to be trending upwards 24 States recorded over a thousand new cases. Immediate forecasts are very grim, particularly for next week.
One in 7 who have covid are unwell for at least 4 weeks, one in 20 people are unwell for at least 8 weeks https://www.bbc.co.uk/news/health-54622059
I’m quite amused that the opening line says old age is a risk factor, yet later in the article it discusses being over 50!
(I was a fit and well 50 year old female, slightly overweight but not obese, when I got covid in March. I was not hospitalised and never had a fever. Yet I still have post-viral syndrome 7 months later, albeit improving.)
I heard on the radio yesterday that the likelihood of post-viral syndrome is greater if one is female, over 50 and obese. Two of the three are out of my control, and yes, I am overweight.
Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
Sorry, false positives, implying that many cases are fictitious. For example, Julia Hartley-Brewer seems to suggest that 90% of positive results are false. However, this seems to confuse different tests, e.g. random ones, and people who feel ill. However, after that, my old brain falters.
The PCR test is highly selective in amplifying genetic sequences unique to this coronavirus. It's almost impossible for the test to give a positive if the coronavirus is not present in the sample - so, for there to be a false positive this would require contamination of the sample during collection and subsequent processing. On the other hand, it's relatively easy to collect a sample incorrectly or at the wrong time such that there's no virus present in a sample from someone who is infected.
Other tests, such as blood tests for the presence of antibodies or the new "results in 15min" tests that are being talked about may not be as selective and could lead to false positives - the work needed to verify these tests including rates of false positive and negative results will be underway, and will need to be completed before these become routine tests.
Of course, the blood antibody test will give a very high false positive rate if it's used to ask the question "do I currently have covid-19?" because it's answering the question "have I been sufficiently exposed to the coronavirus to develop an immune response?"
An interesting point, if false positives are very high across the board, why are cases higher in Manchester than Cornwall?
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Hospitalisations are the reliable stats and they are increasing. I think temperature, new cough and loss of sense of smell are pretty reliable positive indicators but you can have the virus without symptoms or with slightly different ones.
If I get a positive test I’m going into isolation. If I get a negative test and have two out of three of those symptoms I’m going into isolation. Why chance it?
Hospitalisations are the reliable stats and they are increasing. I think temperature, new cough and loss of sense of smell are pretty reliable positive indicators but you can have the virus without symptoms or with slightly different ones.
If I get a positive test I’m going into isolation. If I get a negative test and have two out of three of those symptoms I’m going into isolation. Why chance it?
I may have misunderstood - are you saying you need all three symptoms to suggest you have the virus? I thought it was any one of those three.
There are two measures used to assess the proportion of infected individuals with fatal outcomes. The first is infection fatality ratio (IFR), which estimates this proportion of deaths among all infected individuals. The second is case fatality ratio (CFR), which estimates this proportion of deaths among identified confirmed cases.
This seems largely accepted but if, as you say, Case = Infection, I can't see the point of the distinction. I used to see it because I used to define a Case as Illness + Test, but since the WHO classes a Confirmed Case irrespective of any (or presumably no) symptoms, based solely on a PCR test, I'm not in tune with WHO terminology.
I can see that in the early stages when we were only doing about 20,000 test/week we were only testing ill people, in fact probably only hospitalised people, so the death rate would be an estimate of the CFR and high at that, because everyone would know that there were more cases out there. But as (or maybe if) mass testing goes ahead, does this distinction disappear? Or are people informally using CFR not to mean Confirmed-caseFR (pace WHO definition) but rather Hospital Case FR? Maybe somebody knows.
In general your point is well made.
Quetzcoatl:
I think the skeptics talk a lot about false p
Yes they do and to a limited extent I think this is relevant, but is a dangerous branch to rest too much weight on. The argument often advanced is that we do not have a totally definitive (often called Gold Standard) test against which we can calibrate PCR tests, and in addition there is (so it is said) a lot of variation is test procedures (e.g. maximum number of amplification rounds). Hence I believe there has been no official statement about the FP rate of the PCR test.
To me the main limitations are:
If you are looking for the rate of change, this is not an important factor - just like constants tend to disappear when you differentiate. (OK my maths is slightly more that A-level and maybe it shows.
There is a lot of discussion that the rate of False Negative is a lot higher. There's no point making too much of FP rates if you don't know FN rates.
To me a more valid criticism is that it would be better to give the Case numbers as a percentage of tests done. I'm not sure why this is not thought to be a suitable metric, but it would get rid of the argument that if all you give is the number of cases period, it is entirely possible that the virus is in retreat but you are seeing more cases for no other reason than that more tests are bring done.
Can I give a puff for a good book on the subject?
"Blinded by Corona" by John Ashton.
Before you all roll your eyes thinking I'm plugging a supporter of the Great Barrington Declaration, John Ashton is no such thing. Not in the slightest. And he's a Leftie from Liverpool.
He is hugely qualified, was involved in developing the response for Bahrain which looks to have done pretty well especially given the sensitivity of Shia pilgrimage to sites in Iran. He was amongst other things Regional Director of Public Health for North-West England for 13 years.
Where this opened my eyes was with regards to the importance of Public Health involvement especially at local level, which has been undermined by successive governments. Previously I was really ignorant of what Public Health really is. He is scathing about most aspects of the UK response and highly critical of SAGE. He is obviously a fan of Indie-SAGE. He doesn't get involved in detailed science arguments, mostly it is an account of what happened. He sees lockdown as a blunt instrument, sub-optimal but necessitated by the UK Gov failure to take more targeted and timely approaches. Whilst he is less than totally explicit about this I was left with the impression that he believes zero-Covid is an achievable aim if there is a will. Herd immunity is IHO bollocks.
Well written and deserves to be widely read. Of more interest to sceptics, on the basis of: What's the point of being persuaded of what you already believe.
Your summary sounds pretty similar to what most reputable scientists are saying - you have to have lockdown to get cases under control enough to allow your test, track, trace and isolate system to keep R below 1 without it. Social distancing, mask wearing and hand washing serve to bridge the gap between 3T&I alone can do and imposing more draconian restrictions. I think we could have got to, and stayed, at zero covid in the summer had Johnson not been desperate to re-open and now we're paying the price in lives, health and money.
Immunity has two parts--the innate system, which is basically the OH HELL NO" response that shoves off pirate boarders faster than they can get a foot on the deck of your ship. This is really, really fast but as I understand it, not particularly well-guided. It's a "kill them all, God will know his own" kind of defense. And it works best in children. (Does anybody know, is this the system responsible for allergies when it goes wrong?)
The other part of immunity is the big convoluted and complicated sysem with T cells and B cells and antibodies and all that, which relies on learning an invader's pattern and mounting an attack against it--and then keeping the pattern on file for as long as possible, in case the invader should ever come back again (so an attack response can be mounted far faster, and the human being might have no idea at all there had even been a threat). This is where vaccines work, as well as natural exposure that one survives.
My experience with immunology is limited to a module of my degree that I studied twenty years ago, but I believe this is broadly correct.
<snip>Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
There’s a good piece on this erroneous claim from Reuters - a reputable source, I think.
<snip>Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
There’s a good piece on this erroneous claim from Reuters - a reputable source, I think.
I can however confirm the false negative problem. I doubt it's one in five tests but he could mean that one in five infected people tested test negative.
Hospitals I know from people working there have a category for people they know from symptoms have it but have so far tested negative.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Karl
Symptoms. If I test negative BUT have two out of three best known indicators I’m isolating. Basically it’s a personal response to the risk of a false negative. If I test positive it doesn’t matter what symptoms I have even if none. If I get a false positive and that means I isolate for 14 days for no good reason, that’s a pretty small price to pay.
<snip>Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
There’s a good piece on this erroneous claim from Reuters - a reputable source, I think.
I can however confirm the false negative problem. I doubt it's one in five tests but he could mean that one in five infected people tested test negative.
There are two measures used to assess the proportion of infected individuals with fatal outcomes. The first is infection fatality ratio (IFR), which estimates this proportion of deaths among all infected individuals. The second is case fatality ratio (CFR), which estimates this proportion of deaths among identified confirmed cases.
This seems largely accepted but if, as you say, Case = Infection, I can't see the point of the distinction. I used to see it because I used to define a Case as Illness + Test, but since the WHO classes a Confirmed Case irrespective of any (or presumably no) symptoms, based solely on a PCR test, I'm not in tune with WHO terminology.
I can see that in the early stages when we were only doing about 20,000 test/week we were only testing ill people, in fact probably only hospitalised people, so the death rate would be an estimate of the CFR and high at that, because everyone would know that there were more cases out there. But as (or maybe if) mass testing goes ahead, does this distinction disappear? Or are people informally using CFR not to mean Confirmed-caseFR (pace WHO definition) but rather Hospital Case FR? Maybe somebody knows.
The difference between IFR and CFR lies in the fact that it's basically impossible to test every single human being, and moreover, to do it exactly at the time when that person is infected with the virus rather than before or after. If we could do such a perfect thing, we would give the true IFR rate. But we can't. There are always going to be people who get the virus and die without anyone along the way picking up on the fact that hey, this is a case of COVID. They may misidentify the problem, they may not identify the problem at all (especially if the person dies suddenly in a location where resources are over-stretched and nobody has time/money to investigate the death), or they may assume the cause of death was something else (say someone has severe heart disease and is expected to die soon, but in fact catches COVID and is carried off by the virus before his heart can do him in.) All of these scenarios show why it is hard to go measure IFR directly. Though of course the IFR is what everybody wants, because if we knew it, we would have a much better idea of how dangerous the virus really is.
The CFR is easier to obtain because it comes from testing people known to be infected, that is, known to be cases. Their infection has been confirmed by a lab test. Now you can keep an eye on them to see how their course turns out--whether they die or recover or become "long haulers" or what. So this number is much more obtainable than the IFR.
However, it is going to be skewed, as I mentioned before. People who are confirmed cases are people who had enough privilege or good luck to be tested and probably treated; that means their chances of recovery are higher than those who receive no testing or treating, and so they escape being in the CFR number, pushing the CFR downward. At the same time, people who are confirmed cases and therefore by definition have been tested (and probably treated) have most likely been noticed because they start off sicker than the average person with COVID. That means they have a higher risk of dying, which would push the CFR upward.
Do those two effects cancel out? God knows. We know a skew exists, but we can't be sure just how much or in which direction. Which means we can't just take the CFR and use it as if it were identical to the IFR.
Given that we now know a great many people get COVID and are totally asymptomatic, the CFR is normally going to be higher than the IFR. Which is mildly good news for random human beings sitting at their computers and wondering whether they're going to die in this pandemic. They can look at the CFR for their area/gender/race/whatsit and mentally drop it down just a bit, when they're trying to calculate their odds of making it through.
I'm not sure what else it's good for, though. The historians will someday calculate a true(r) IFR when the pandemic is over, probably in part by comparing the normal expected death rate in a population to what we actually got (the overage is probably due to COVID). But without a time machine, that does us no good now.
The whole situation with numbers counting reminds me of skateboarding. You are never sure exactly where you are in terms of speed, needed power, etc. because those things are constantly changing, but you can ride effectively if you keep an eye on general trends and adjust your efforts where it seems necessary.
<snip>Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
There’s a good piece on this erroneous claim from Reuters - a reputable source, I think.
Basically, the whole "high rate of false positives" fake news comes down to Dominic Raab talking bollocks.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Basically, the whole "high rate of false positives" fake news comes down to Dominic Raab talking bollocks.
Nice!
Not so nice, in fact extremely nasty, is the news from the worldometer front. Another global highest daily case total, over 478 thousand. Trend lines suggest the daily case total is closing in on half a million a day. The total diagnosed cases statistic has topped 42 million, adding another million in a little over two days.
European and USA figures go from bad to worse; over 217 thousand new cases and almost 1800 deaths in Europe, over 74 thousand new cases and almost a thousand deaths in the USA.
Given the trends in both Europe and the USA the plain truth appears to be that unless lockdowns drive new cases down to very low numbers, so that new outbreaks can be quickly contained, the virus will beat attempts to control its spread. Current attempts to balance economic and medical needs by more localised and selective controls seem to be failing. And pressures on health services are ramping up again.
All this needs to be measured against the recognition that very large percentages of people in Europe and the USA have not yet contracted the virus. There are a lot more potential victims out there. For the time being, the virus seems to be winning the battle.
<snip>Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
There’s a good piece on this erroneous claim from Reuters - a reputable source, I think.
I can however confirm the false negative problem. I doubt it's one in five tests but he could mean that one in five infected people tested test negative.
Isn't that exactly what a false negative is?
I was distinguishing between one in five tests being a FN and one in five cases resulting in a FN when tested.
E.g. 100 tests, 10 people actually infected - one in five tests being FN would mean 20 FNs - actually impossible as only 10 people are infected. One in five infections testing as FN means 2 FNs.
<snip>Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
There’s a good piece on this erroneous claim from Reuters - a reputable source, I think.
I can however confirm the false negative problem. I doubt it's one in five tests but he could mean that one in five infected people tested test negative.
Isn't that exactly what a false negative is?
I was distinguishing between one in five tests being a FN and one in five cases resulting in a FN when tested.
E.g. 100 tests, 10 people actually infected - one in five tests being FN would mean 20 FNs - actually impossible as only 10 people are infected. One in five infections testing as FN means 2 FNs.
So it's an important distinction.
Ah, I missed your meaning, thank you.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Two grim statistics as worldometer closes the books on October 23. Firstly the global new daily cases total went up again from yesterday’s record of 479k to almost 490k. Secondly the US new daily cases total was over 81k, the highest in the pandemic so far and the first time it has exceeded 80k.
There will probably be a pause on growth over the weekend simply because less recording is done in some countries because of the weekend break. That’s where the seven day averages can tell us a thing or two. In one week, the global 7 day average has moved from 350k to 409k and the US 7 day average from 56k to 64k. By inspection of the graphs it looks as though the rate of increase is itself accelerating. Over the past week the rate of increase of 7 day averages has been about 15% globally, 13% in USA. It’s harder to get the overall European changes for the comparative period but UK 7 day averages are up 20%, French 7 day averages up over 40%. The daily European total is up from 218k to 231k in one day. We’ll find out whether that’s a spike or a trend in due course.
These rates will almost certainly be reflected in increased hospitalisation numbers in about 2 weeks, increased daily death totals in a about a month. There is now clear evidence that the anticipated Autumn surge has arrived in the Northern Hemisphere and particularly in Europe and the USA, I expect governments in Western Europe to batten down the hatches some more. In the USA it’s going to be State by State.
Meanwhile the President held a rally for senior citizens in Florida, far too many of whom weren’t wearing masks. That’s about as mad as it gets.
I tried to do some analysis to see if the claims that the current deaths from respiratory disease (including COVID) over the last few months are not significantly different from previous years. I would like it if somebody checked these figures and commented as to whether they have any relevance. If they're wrong then it is very difficult to find information, as they are based on UK Government Statistics which break out the deaths due to respiratory diseases. So there has been n o interpretation on my part as to whether the cause of death was respiratory. I have, for 2020 counted COVID with respiratory diseases since I have heard no dispute that it is a respiratory disease.
Unfortunately I can only get figures for this year up to to Oct 8th. For each year from 2010 to 2020 I have totalled the total deaths in the fifteen weeks from start of July to start of Oct where the current year's data ends. The data is taken from spreadsheets provided on the site provided by the UK Office of National Statistics. This provides data on a week by week basis on total deaths and deaths from respiratory diseases, which for 2020 only has COVID as a separate category.
I have adjusted for population growth, taking data from here with a small about of interpolation for the few years not given exactly.
I tried to do some analysis to see if the claims that the current deaths from respiratory disease (including COVID) over the last few months are not significantly different from previous years. I would like it if somebody checked these figures and commented as to whether they have any relevance. If they're wrong then it is very difficult to find information, as they are based on UK Government Statistics which break out the deaths due to respiratory diseases. So there has been n o interpretation on my part as to whether the cause of death was respiratory. I have, for 2020 counted COVID with respiratory diseases since I have heard no dispute that it is a respiratory disease.
For 2020 you only appear to have used the "respiratory disease" row of data. The row below, which is the Covid deaths needs to be included though it's not clear how exactly as some of the "respiratory disease" row may also be in the Covid row below. I've just added figures up for each quarter to simplify things, and noting there may be a small amount of double counting for this year, I get:
Q1 Q2 Q3
Average for 2010-2019 : 22679 16044 12366
Total for 2020 : 21529 63806 12853
Obviously Q4 is meaningless at present as there haven't been enough weeks. But, that's a massive spike in Q2 (over 47 thousand).
Doublethink: I think twitter is becoming a bad joke. The messages from the boards in Private Eye have it right. So I'll pass on that.
Adam Kurcharski is a mathematician and epidemiologist, a Welcome fellow as well, and he publishes a lot of good commentary re Covid data on Twitter.
He has this self-description on his website:
“I'm Adam Kucharski. I am an Associate Professor and Sir Henry Dale Fellow at the London School of Hygiene & Tropical Medicine, where I work on mathematical analysis of infectious disease oubreaks.”
And now the White House Chief of Staff admits the administration is not trying to control COVID 19. Meanwhile, five of VP Prence's staff now have COVID.
Heaven help us.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
The virus can be controlled by social distancing, mask wearing, hygiene. It can be brought under closer control by using isolation and lockdown for a period to get the new case numbers way down. It can be kept under control by testing tracking and tracing the small numbers of new cases. Follow S Korea for example.
But Meadows cannot say those things without contradicting his boss. And admitting that Fauci et al talk more sense than Trump.
For 2020 you only appear to have used the "respiratory disease" row of data.
Well I just checked and I included both. The specific line items were: Respiratory 12,889 and "deaths where COVID is mentioned" 3,423. I've seen press reports that other RD deaths are reduced. Maybe COVID is killing people who otherwise would have died of some other RD. And up to first week of Oct the rise in COVID deaths is modest.
So I agree that the true course of the disease will only become clear as the year progresses, so I don't expect figures up to 9th Oct to be viewed as proving anything. I just wanted to see how possible it is to get data that can be used to see what's really going on rather than what models say might be going on.
Sadly, though, as you suggested, according to a note on the spreadsheet:
Note: Deaths could possibly be counted in both causes presented. If a death had an underlying respiratory cause and a mention of COVID-19 then it would appear in both counts.
It may be that the extent of double counting is only a second order effect - but I don't know. I could plough on, on the basis that since I am trying to find out where "second wave" is an exaggeration for the normal progress of RD, the fact of double counting makes the case stronger.
Of course, I am in no doubt at all about the spike earlier in the year. It's just not what's of interest to me now.
Adam Kurcharski is a mathematician and epidemiologist, a Welcome fellow as well, and he publishes a lot of good commentary re Covid data on Twitter.
OK so obviously someone worth paying attention to. I just wish he didn't use Twitter as my aversion to is is strong.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Because of the medical and economic consequences, those survivors who are suffering from long term COVID-19 disabilities represent another measureable cost of the disease. The death statistics are the numbers which get the attention.
On general grounds, because better remedies and treatments are available compared with the first wave and more testing is being done, we can expect higher new case numbers and lower death rates in the second wave. Such evidence that I have seen re long term sickness effects suggest that it is quite prevalent amongst those whose initial symptoms are mild.
Of course long term post viral effects are known for other viral illnesses. But COVID-19 long term illness is a new cost.
The weekend lull in the statistics still saw over 400k new cases and 4.5k new deaths yesterday. Almost half the new cases and about a third of the new deaths were in Europe. On general grounds the death rates will be expected to rise in countries showing large new case numbers, with a lag factor of about a month. We’ll see by how much more clearly during the next few weeks.
Unfortunately I no longer think that I can get an accurate picture that will be accepted, from the ONS statistics. This is my understanding based on an article from the Cambridge Centre for Evidence Based Medicine, which I expect many on this ship will not accept as a reliable source, as it is broadly more favourable to the sceptical viewpoint.
What this shows is that there has been a large increase in the extent to which COVID gets a mention on the death certificate anywhere, not limited to primary cause of death.
While we found that roughly one in thirteen (7.8%) deaths with COVID-19 on the death certificate did not have the disease as the underlying cause of death, this proportion has risen substantially to 29% (nearly a third) for the last eight weeks of reporting.
In other words at the start and during the main phase of the epidemic, deaths with COVID associated were more likely to be deaths by COVID. But that changed to coincide with the Second Wave.
This doesn't mean to say that it is impossible to disentangle the data, just it's not made easy. The article does have data (from PHE) that shows that now, about 30% of the deaths with COVID mentioned do not have it as the underlying cause.
So looking at the ONS statistics, the line showing RD deaths is based on: "Deaths where the underlying cause was respiratory disease". However, the next line for COVID has a more inclusive criterion: "Deaths where COVID-19 was mentioned on the death certificate". So to make the two lines properly comparable, I should divide the COVID line by 30% so that it becomes (roughly!): "Deaths where the underlying cause was COVID".
I may continue in this basis, but the more I make adjustments like this, which tend to support a conclusion that I am predisposed to believe, the less that is likely to be accepted, because I'm clearly not impartial.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
So far as death rates are concerned excess deaths is I think reckoned to be the most accurate long term measure.
I think that underlines the importance of looking at total deaths without attributing cause (which also has the advantage of not double counting in any circumstance). In the majority of cases there are multiple factors that a death can be attributed to - the question for us is how many of those include covid19 as one of the causes, or as something that has accelerated the death. The "tested positive in the last 28d" measure isn't necessarily that useful, though it is easy to quantify, because some deaths may be unrelated (I could get a positive test then walk out in front of a bus ... but I wouldn't have been killed by covid) and some covid deaths are missed (no test taken, or the death is related to 'long covid'). Likewise a death certificate may not mention covid even if that was a factor (as we get better at testing and identifying symptoms then we should be getting more death certificates mentioning covid compared to March and April when fewer people were tested and doctors issuing death certificates were less aware of exactly what to look out for).
I think that underlines the importance of looking at total deaths without attributing cause (which also has the advantage of not double counting in any circumstance). In the majority of cases there are multiple factors that a death can be attributed to - the question for us is how many of those include covid19 as one of the causes, or as something that has accelerated the death.
I think that last point is an important. There are large number of 'contributory causes' which can be managed for long periods of time and/or otherwise have very little impact on life expectancy as long as they are managed - ask any actuary.
As Barnabas say's excess deaths (picking the time period appropriately) is going to be among the better long term measures of the impact of Covid.
It may be that the extent of double counting is only a second order effect - but I don't know. I could plough on, on the basis that since I am trying to find out where "second wave" is an
Of course, I am in no doubt at all about the spike earlier in the year. It's just not what's of interest to me now.
July-october is vaguely interpeak (although of course some of use were predicting a bounce in may, June, July.).
It's not that surprising it's more or less negligible, the worry is when the graph starts shooting up (like it has for cases in sept/oct and deaths for October) for which you need to act early.
And it doesn't take much to go from mostly fine (lockdown costs exceeding the immediate Corona costs of July), to the bad April and now and the horrendous (no.action)
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
Anteater
Donald Trump has expressed scepticism re death recording in the USA on the grounds that those who certify may get a financial gain out of it if there is a choice between COVID and other pre-existing conditions. His comments fail Fact Checking. And a lot of medics are up in arms.
It may be worth pointing out that there has been concern in the opposite direction. Florida in particular has been criticised for discounting records.
So looking at the ONS statistics, the line showing RD deaths is based on: "Deaths where the underlying cause was respiratory disease".
A note on this; there is going to be a base death rate for respiratory diseases. It will not be possible to determine for an individual case whether that disease or Covid was to blame (in the cases where the individual has both) - However at the level of the whole population a discrepancy between the normal base rate and one including patients with Covid would be detectable, and should one exist it is right that these deaths be attributed to Covid.
How come the UK death toll is being reported as 61 000, when on TV every night, it's shown as 45 000? Presumably, different agencies are being reported. Or is 45k the govt's manicured figure?
I think that underlines the importance of looking at total deaths without attributing cause
I can see the sense of this. If one particular disease is on the scene which kills people, then just look at spikes in deaths and let common sense lead to the conclusion that (at least most of) the excess deaths are due to it. And even though I believe that a lot of deaths will be caused due to the Government/NHS responses to COVID, it is unlikely that so many of them will have already turned up to distort the figures.
We now have the ONS figures for Week 42 (Oct 16) which has a typo implying none in the latest week! I've corrected that in what follows.
So the excess deaths over the 4 weeks up to Oct 16 (week 42) have been:
257 (2.67%) 390 (3.92%) 143 (1.44%) 668 (6.8%)
For last year, the excess deaths in the corresponding weeks were: 2.46% 3.56% 3.23% 4.23%. So it looks like we are getting a rise, and we shall see how serious.
Barnabas62Purgatory Host, 8th Day Host, Epiphanies Host
edited October 2020
I've been having a look at comparative figures for Europe and the USA. The definition of Europe which worldometer uses includes both Western and Eastern European States with a total population of over 750 million, about 2.25 times the population of the USA.
Taking the figures for 27 October to start, the European Totals were as follows.
New Cases, 223k, Deaths, 2757
For the USA, totals were
New Cases, 75k, Deaths 1039
By reference to population size, the second wave of the pandemic in Europe is showing higher figures than the USA and I think that has been the trend in recent days. The USA figures never really showed the drop to much smaller numbers which happened in many countries in Europe during the summer months and there is some evidence that the recent surge in USA cases has started a little bit later in the USA than Europe.
The simple fact is that both have it bad and the surge in European figures is more than offsetting the decline in Brazil and India.
For completeness here are the global figures for 27 October.
New Cases, 459k, Deaths 7023.
Simple arithmetic tells us that together Europe and the USA with a combined population of about 1/7th of the global population, accounted for significantly more than half yesterday's total new cases and a little over half the deaths.
Comments
I assume CFR means "confirmed fatality rate." If so, you can have these for the human race as a whole (all ages, races, and genders) or for some subset of it. There is value even for the overall human rate, because it allows you to compare it to plagues-of-the-past, and see if you can learn anything (like, "we really need to get more aggressive on this one" or "Thank God this one is only X percent, we could have had MERS which is 30%."
Also, if you notice that the CFR as calculated in country X is 10%, and in country Y you get 2%, that tells you it's time to look at what the two countries are doing differently. There might be something you can change to bring X down to a better number.
As far as immunity--
I am not an immunologist, but I know a little (very little). Immunity is mostly a name for what happens after infection, just as overwhelming firepower with massive destruction and nukes is a name for what happens after a really badly-thought-out invasion. It's a defense. It only gets going once there is a threat. That means at the point of infection. Nobody has an immune system like the one I tease my mother about--one so powerful you can just walk around and kill the viruses in the air around you for six feet on either side.
Immunity has two parts--the innate system, which is basically the OH HELL NO" response that shoves off pirate boarders faster than they can get a foot on the deck of your ship. This is really, really fast but as I understand it, not particularly well-guided. It's a "kill them all, God will know his own" kind of defense. And it works best in children. (Does anybody know, is this the system responsible for allergies when it goes wrong?)
The other part of immunity is the big convoluted and complicated sysem with T cells and B cells and antibodies and all that, which relies on learning an invader's pattern and mounting an attack against it--and then keeping the pattern on file for as long as possible, in case the invader should ever come back again (so an attack response can be mounted far faster, and the human being might have no idea at all there had even been a threat). This is where vaccines work, as well as natural exposure that one survives.
Here's a list of different diseases to compare - so if you catch covid19 (with preliminary CFR estimates of 1-10%) you're at least an order of magnitude more likely to die than if you catch seasonal 'flu. Of course, there's also the question of how easy it is to catch different diseases.
The UK saw a record total of new cases, close to 27 thousand, and close to 200 deaths. The brakes may have been applied too late, the second wave surge looks well under way.
The USA saw close to 64 thousand new cases and over 1200 deaths. Both now look to be trending upwards 24 States recorded over a thousand new cases. Immediate forecasts are very grim, particularly for next week.
A horrid day.
I heard on the radio yesterday that the likelihood of post-viral syndrome is greater if one is female, over 50 and obese. Two of the three are out of my control, and yes, I am overweight.
Has anyone heard that the instance of "false positives" is very high in the UK? I'm told by a GP friend that one in 5 tests is a false negative, but someone else I know insists that false positives are high as well and I don't get how that can happen. I keep asking her to send me a link to some reliable information... >rolleyes<
Other tests, such as blood tests for the presence of antibodies or the new "results in 15min" tests that are being talked about may not be as selective and could lead to false positives - the work needed to verify these tests including rates of false positive and negative results will be underway, and will need to be completed before these become routine tests.
Of course, the blood antibody test will give a very high false positive rate if it's used to ask the question "do I currently have covid-19?" because it's answering the question "have I been sufficiently exposed to the coronavirus to develop an immune response?"
If I get a positive test I’m going into isolation. If I get a negative test and have two out of three of those symptoms I’m going into isolation. Why chance it?
I may have misunderstood - are you saying you need all three symptoms to suggest you have the virus? I thought it was any one of those three.
Take The WHO Definition:
This seems largely accepted but if, as you say, Case = Infection, I can't see the point of the distinction. I used to see it because I used to define a Case as Illness + Test, but since the WHO classes a Confirmed Case irrespective of any (or presumably no) symptoms, based solely on a PCR test, I'm not in tune with WHO terminology.
I can see that in the early stages when we were only doing about 20,000 test/week we were only testing ill people, in fact probably only hospitalised people, so the death rate would be an estimate of the CFR and high at that, because everyone would know that there were more cases out there. But as (or maybe if) mass testing goes ahead, does this distinction disappear? Or are people informally using CFR not to mean Confirmed-caseFR (pace WHO definition) but rather Hospital Case FR? Maybe somebody knows.
In general your point is well made.
Quetzcoatl: Yes they do and to a limited extent I think this is relevant, but is a dangerous branch to rest too much weight on. The argument often advanced is that we do not have a totally definitive (often called Gold Standard) test against which we can calibrate PCR tests, and in addition there is (so it is said) a lot of variation is test procedures (e.g. maximum number of amplification rounds). Hence I believe there has been no official statement about the FP rate of the PCR test.
To me the main limitations are:
If you are looking for the rate of change, this is not an important factor - just like constants tend to disappear when you differentiate. (OK my maths is slightly more that A-level and maybe it shows.
There is a lot of discussion that the rate of False Negative is a lot higher. There's no point making too much of FP rates if you don't know FN rates.
To me a more valid criticism is that it would be better to give the Case numbers as a percentage of tests done. I'm not sure why this is not thought to be a suitable metric, but it would get rid of the argument that if all you give is the number of cases period, it is entirely possible that the virus is in retreat but you are seeing more cases for no other reason than that more tests are bring done.
"Blinded by Corona" by John Ashton.
Before you all roll your eyes thinking I'm plugging a supporter of the Great Barrington Declaration, John Ashton is no such thing. Not in the slightest. And he's a Leftie from Liverpool.
He is hugely qualified, was involved in developing the response for Bahrain which looks to have done pretty well especially given the sensitivity of Shia pilgrimage to sites in Iran. He was amongst other things Regional Director of Public Health for North-West England for 13 years.
Where this opened my eyes was with regards to the importance of Public Health involvement especially at local level, which has been undermined by successive governments. Previously I was really ignorant of what Public Health really is. He is scathing about most aspects of the UK response and highly critical of SAGE. He is obviously a fan of Indie-SAGE. He doesn't get involved in detailed science arguments, mostly it is an account of what happened. He sees lockdown as a blunt instrument, sub-optimal but necessitated by the UK Gov failure to take more targeted and timely approaches. Whilst he is less than totally explicit about this I was left with the impression that he believes zero-Covid is an achievable aim if there is a will. Herd immunity is IHO bollocks.
Well written and deserves to be widely read. Of more interest to sceptics, on the basis of: What's the point of being persuaded of what you already believe.
End of puff. I'll see if I get told off.
My experience with immunology is limited to a module of my degree that I studied twenty years ago, but I believe this is broadly correct.
I can however confirm the false negative problem. I doubt it's one in five tests but he could mean that one in five infected people tested test negative.
Hospitals I know from people working there have a category for people they know from symptoms have it but have so far tested negative.
Symptoms. If I test negative BUT have two out of three best known indicators I’m isolating. Basically it’s a personal response to the risk of a false negative. If I test positive it doesn’t matter what symptoms I have even if none. If I get a false positive and that means I isolate for 14 days for no good reason, that’s a pretty small price to pay.
That’s the way I figure things anyway.
Isn't that exactly what a false negative is?
The difference between IFR and CFR lies in the fact that it's basically impossible to test every single human being, and moreover, to do it exactly at the time when that person is infected with the virus rather than before or after. If we could do such a perfect thing, we would give the true IFR rate. But we can't. There are always going to be people who get the virus and die without anyone along the way picking up on the fact that hey, this is a case of COVID. They may misidentify the problem, they may not identify the problem at all (especially if the person dies suddenly in a location where resources are over-stretched and nobody has time/money to investigate the death), or they may assume the cause of death was something else (say someone has severe heart disease and is expected to die soon, but in fact catches COVID and is carried off by the virus before his heart can do him in.) All of these scenarios show why it is hard to go measure IFR directly. Though of course the IFR is what everybody wants, because if we knew it, we would have a much better idea of how dangerous the virus really is.
The CFR is easier to obtain because it comes from testing people known to be infected, that is, known to be cases. Their infection has been confirmed by a lab test. Now you can keep an eye on them to see how their course turns out--whether they die or recover or become "long haulers" or what. So this number is much more obtainable than the IFR.
However, it is going to be skewed, as I mentioned before. People who are confirmed cases are people who had enough privilege or good luck to be tested and probably treated; that means their chances of recovery are higher than those who receive no testing or treating, and so they escape being in the CFR number, pushing the CFR downward. At the same time, people who are confirmed cases and therefore by definition have been tested (and probably treated) have most likely been noticed because they start off sicker than the average person with COVID. That means they have a higher risk of dying, which would push the CFR upward.
Do those two effects cancel out? God knows. We know a skew exists, but we can't be sure just how much or in which direction. Which means we can't just take the CFR and use it as if it were identical to the IFR.
Given that we now know a great many people get COVID and are totally asymptomatic, the CFR is normally going to be higher than the IFR. Which is mildly good news for random human beings sitting at their computers and wondering whether they're going to die in this pandemic. They can look at the CFR for their area/gender/race/whatsit and mentally drop it down just a bit, when they're trying to calculate their odds of making it through.
I'm not sure what else it's good for, though. The historians will someday calculate a true(r) IFR when the pandemic is over, probably in part by comparing the normal expected death rate in a population to what we actually got (the overage is probably due to COVID). But without a time machine, that does us no good now.
The whole situation with numbers counting reminds me of skateboarding. You are never sure exactly where you are in terms of speed, needed power, etc. because those things are constantly changing, but you can ride effectively if you keep an eye on general trends and adjust your efforts where it seems necessary.
Not so nice, in fact extremely nasty, is the news from the worldometer front. Another global highest daily case total, over 478 thousand. Trend lines suggest the daily case total is closing in on half a million a day. The total diagnosed cases statistic has topped 42 million, adding another million in a little over two days.
European and USA figures go from bad to worse; over 217 thousand new cases and almost 1800 deaths in Europe, over 74 thousand new cases and almost a thousand deaths in the USA.
Given the trends in both Europe and the USA the plain truth appears to be that unless lockdowns drive new cases down to very low numbers, so that new outbreaks can be quickly contained, the virus will beat attempts to control its spread. Current attempts to balance economic and medical needs by more localised and selective controls seem to be failing. And pressures on health services are ramping up again.
All this needs to be measured against the recognition that very large percentages of people in Europe and the USA have not yet contracted the virus. There are a lot more potential victims out there. For the time being, the virus seems to be winning the battle.
I was distinguishing between one in five tests being a FN and one in five cases resulting in a FN when tested.
E.g. 100 tests, 10 people actually infected - one in five tests being FN would mean 20 FNs - actually impossible as only 10 people are infected. One in five infections testing as FN means 2 FNs.
So it's an important distinction.
Ah, I missed your meaning, thank you.
There will probably be a pause on growth over the weekend simply because less recording is done in some countries because of the weekend break. That’s where the seven day averages can tell us a thing or two. In one week, the global 7 day average has moved from 350k to 409k and the US 7 day average from 56k to 64k. By inspection of the graphs it looks as though the rate of increase is itself accelerating. Over the past week the rate of increase of 7 day averages has been about 15% globally, 13% in USA. It’s harder to get the overall European changes for the comparative period but UK 7 day averages are up 20%, French 7 day averages up over 40%. The daily European total is up from 218k to 231k in one day. We’ll find out whether that’s a spike or a trend in due course.
These rates will almost certainly be reflected in increased hospitalisation numbers in about 2 weeks, increased daily death totals in a about a month. There is now clear evidence that the anticipated Autumn surge has arrived in the Northern Hemisphere and particularly in Europe and the USA, I expect governments in Western Europe to batten down the hatches some more. In the USA it’s going to be State by State.
Meanwhile the President held a rally for senior citizens in Florida, far too many of whom weren’t wearing masks. That’s about as mad as it gets.
Unfortunately I can only get figures for this year up to to Oct 8th. For each year from 2010 to 2020 I have totalled the total deaths in the fifteen weeks from start of July to start of Oct where the current year's data ends. The data is taken from spreadsheets provided on the site provided by the UK Office of National Statistics. This provides data on a week by week basis on total deaths and deaths from respiratory diseases, which for 2020 only has COVID as a separate category.
I have adjusted for population growth, taking data from here with a small about of interpolation for the few years not given exactly.
The result showing divergence from average is:
Year . . . . Total . . . . Resp . . . . Total . . . . Resp
2010 . . . . 138,281 . . . . 16,065 . . . . 101% . . . . 101%
2011 . . . . 136,719 . . . . 16,395 . . . . 100% . . . . 103%
2012 . . . . 129,284 . . . . 14,531 . . . . 94% . . . . 91%
2013 . . . . 134,808 . . . . 16,191 . . . . 98% . . . . 102%
2014 . . . . 139,896 . . . . 15,231 . . . . 102% . . . . 96%
2015 . . . . 140,100 . . . . 16,116 . . . . 102% . . . . 101%
2016 . . . . 130,808 . . . . 16,385 . . . . 95% . . . . 103%
2017 . . . . 141,304 . . . . 16,198 . . . . 103% . . . . 102%
2018 . . . . 138,884 . . . . 15,527 . . . . 101% . . . . 98%
2019 . . . . 140,017 . . . . 15,891 . . . . 102% . . . . 100%
2020 . . . . 138,096 . . . . 16,312 . . . . 101% . . . . 103%
Average . . . . 137,109 . . . . 15,895
So there is an uptick and it is to be assumed that this will get more pronounced, and I will be monitoring the numbers for my own interest.
I tried (and tried) to space out the table better, that's the best I can do. Nature abhors a vacuum. SoF appears to abhor spaces.
Q1 Q2 Q3
Average for 2010-2019 : 22679 16044 12366
Total for 2020 : 21529 63806 12853
Obviously Q4 is meaningless at present as there haven't been enough weeks. But, that's a massive spike in Q2 (over 47 thousand).
Adam Kurcharski is a mathematician and epidemiologist, a Welcome fellow as well, and he publishes a lot of good commentary re Covid data on Twitter.
He has this self-description on his website:
“I'm Adam Kucharski. I am an Associate Professor and Sir Henry Dale Fellow at the London School of Hygiene & Tropical Medicine, where I work on mathematical analysis of infectious disease oubreaks.”
Heaven help us.
But Meadows cannot say those things without contradicting his boss. And admitting that Fauci et al talk more sense than Trump.
So I agree that the true course of the disease will only become clear as the year progresses, so I don't expect figures up to 9th Oct to be viewed as proving anything. I just wanted to see how possible it is to get data that can be used to see what's really going on rather than what models say might be going on.
Sadly, though, as you suggested, according to a note on the spreadsheet:
It may be that the extent of double counting is only a second order effect - but I don't know. I could plough on, on the basis that since I am trying to find out where "second wave" is an exaggeration for the normal progress of RD, the fact of double counting makes the case stronger.
Of course, I am in no doubt at all about the spike earlier in the year. It's just not what's of interest to me now.
On general grounds, because better remedies and treatments are available compared with the first wave and more testing is being done, we can expect higher new case numbers and lower death rates in the second wave. Such evidence that I have seen re long term sickness effects suggest that it is quite prevalent amongst those whose initial symptoms are mild.
Of course long term post viral effects are known for other viral illnesses. But COVID-19 long term illness is a new cost.
The weekend lull in the statistics still saw over 400k new cases and 4.5k new deaths yesterday. Almost half the new cases and about a third of the new deaths were in Europe. On general grounds the death rates will be expected to rise in countries showing large new case numbers, with a lag factor of about a month. We’ll see by how much more clearly during the next few weeks.
What this shows is that there has been a large increase in the extent to which COVID gets a mention on the death certificate anywhere, not limited to primary cause of death.
In other words at the start and during the main phase of the epidemic, deaths with COVID associated were more likely to be deaths by COVID. But that changed to coincide with the Second Wave.
This doesn't mean to say that it is impossible to disentangle the data, just it's not made easy. The article does have data (from PHE) that shows that now, about 30% of the deaths with COVID mentioned do not have it as the underlying cause.
So looking at the ONS statistics, the line showing RD deaths is based on: "Deaths where the underlying cause was respiratory disease". However, the next line for COVID has a more inclusive criterion: "Deaths where COVID-19 was mentioned on the death certificate". So to make the two lines properly comparable, I should divide the COVID line by 30% so that it becomes (roughly!): "Deaths where the underlying cause was COVID".
I may continue in this basis, but the more I make adjustments like this, which tend to support a conclusion that I am predisposed to believe, the less that is likely to be accepted, because I'm clearly not impartial.
I think that last point is an important. There are large number of 'contributory causes' which can be managed for long periods of time and/or otherwise have very little impact on life expectancy as long as they are managed - ask any actuary.
As Barnabas say's excess deaths (picking the time period appropriately) is going to be among the better long term measures of the impact of Covid.
It's not that surprising it's more or less negligible, the worry is when the graph starts shooting up (like it has for cases in sept/oct and deaths for October) for which you need to act early.
And it doesn't take much to go from mostly fine (lockdown costs exceeding the immediate Corona costs of July), to the bad April and now and the horrendous (no.action)
Donald Trump has expressed scepticism re death recording in the USA on the grounds that those who certify may get a financial gain out of it if there is a choice between COVID and other pre-existing conditions. His comments fail Fact Checking. And a lot of medics are up in arms.
It may be worth pointing out that there has been concern in the opposite direction. Florida in particular has been criticised for discounting records.
A note on this; there is going to be a base death rate for respiratory diseases. It will not be possible to determine for an individual case whether that disease or Covid was to blame (in the cases where the individual has both) - However at the level of the whole population a discrepancy between the normal base rate and one including patients with Covid would be detectable, and should one exist it is right that these deaths be attributed to Covid.
We now have the ONS figures for Week 42 (Oct 16) which has a typo implying none in the latest week! I've corrected that in what follows.
So the excess deaths over the 4 weeks up to Oct 16 (week 42) have been:
257 (2.67%) 390 (3.92%) 143 (1.44%) 668 (6.8%)
For last year, the excess deaths in the corresponding weeks were: 2.46% 3.56% 3.23% 4.23%. So it looks like we are getting a rise, and we shall see how serious.
Taking the figures for 27 October to start, the European Totals were as follows.
New Cases, 223k, Deaths, 2757
For the USA, totals were
New Cases, 75k, Deaths 1039
By reference to population size, the second wave of the pandemic in Europe is showing higher figures than the USA and I think that has been the trend in recent days. The USA figures never really showed the drop to much smaller numbers which happened in many countries in Europe during the summer months and there is some evidence that the recent surge in USA cases has started a little bit later in the USA than Europe.
The simple fact is that both have it bad and the surge in European figures is more than offsetting the decline in Brazil and India.
For completeness here are the global figures for 27 October.
New Cases, 459k, Deaths 7023.
Simple arithmetic tells us that together Europe and the USA with a combined population of about 1/7th of the global population, accounted for significantly more than half yesterday's total new cases and a little over half the deaths.