Sunday, October 04, 2020

Prof Sunetra Gupta

I first came across Prof Gupta as an invited speaker to one of the Royal College of Pathologist seminars, now viewable on-line at

The COVID 19 pandemic: epidemiology*

*Prof Gupta is at Oxford, not Imperial College as the RCPath intro slide incorrectly shows. We all know about the Imperial College modellers.

I thought she was talking sense at the time and that impression has not changed since. I notice that she has been very active recently in the media and I happened to pick up this interview via Faceache:

The Spectator interview

I've pulled this one out from many videos because it answers the question as to what an incompetent Prime Minister says when presented with someone who is telling him that he has done everything wrong. It's near the end so I've clipped out my favourite eight seconds:

Clearly there is no way, ever, that any government is going to say that they have screwed everything up about this pandemic. Completely.

Especially if they have. And honesty is not exactly a hallmark of Boris Johnson.


NB A more competent government in power would probably have done the all wrong things too, but much more effectively. That's a scary thought. I am seriously conflicted about this.


cavenewt said...

"Professor Sunetra Gupta, a leading epidemiologist at the University of Oxford, tells Andrew Neil why a second lockdown would simple delay the spread of coronavirus, and why governments should instead focus on protecting the vulnerable."

Thanks for the video link, Peter. The professor is expressing an opinion that I arrived at several months ago, which is the virus is here to stay much like the flu and colds and other similar things, and rather than trying to prevent it completely, which will only delay the inevitable, we should figure out ways of ameliorating its effects.

For context, I live in the Utah desert in the United States. Donald Trump was diagnosed with Covid-19 2 days ago.

Our political climate is such that I hardly dare post this video on my Facebook page or, gawd forbid, in the local coronavirus group, because then many people would tar me with the anti-masker brush. In the US, masks in particular have become an overt political statement. As a liberal, I must hasten to add that while I privately think some of the recommendations might be overkill in light of current knowledge, I still carefully follow the rules about masking, distancing, etc. That's pretty easy to do anyway because I live in a tiny community of 300 in the middle of the desert, all the houses are pretty far apart, the nearest gas stations and grocery stores are 25 miles away, and I'm mostly retired.

Early on, the lockdowns were supposed to be to "flatten the curve" to give hospitals a chance to gear up for a possible severe epidemic. This seems to have morphed into an effort to prevent every single infection which, as the professor points out in the video, is pretty unrealistic. I wish I could find some reliable information about how Sweden's original approach has worked out for them. The reports are so conflicting, it's hard to know what information to trust.

Peter said...

Hi cave,

Sweden has been pretty successful after their cockup with care homes at the start. Nothing like as bad as Bojo in the UK or the idiot running New York’s care homes (both sides of the political divide there), but they admit they got that aspect wrong in Sweden, and have apologised to the populace. Their current COVID death rate 7 day moving average is zero. The highest value that they have had in the last two months was 4. I use Worldometer for my data and do not read the NewYork Post, though I see their “Sweden has failed” headlines occasionally and think WTF?

I followed Sweden on Worldometer very carefully during the peak and still look occasionally. Now I follow New Zealand via their government website. They are pottering along between zero and 12 new cases per day and are going in to summer. They really might cope until a vaccine is marketed. But let’s say the vaccine is 50% effective, or even 60% effective, in young, health people who don’t need it. Chances of it working in the elderly with immune sencence are pretty low. What will NZ do then? As you say, the virus isn’t going away any time soon.

I have the nightmare that if Corbyn had gotten elected as PM instead of Bojo he would have done a fantastic job and the UK would be another New Zealand. Massive lockdown, minimal COVID fatalities, wide open to the virus and begging Big Pharma for a dud vaccine… No one daring to count lockdown related fatalities.

By the time the vaccine arrives the UK will be pretty close to herd immunity. But we’ll still get compulsorily vaccinated because Bojo is Bojo…


Passthecream said...

Peter I think the Au government has already handed over big piles of kangaroo$$ to one or more of the UK based vaccine projects in lustful anticipation of a magic outcome. We just need to keep the case numbers down below 0 until about the end of next year or the year after. However conservative religious leaders erupted into a furore about the possibility of mandatory vaxxing because, it might contain material derived from embryonic cell lines.

In the state where I live there is no known CV19 for hundreds of miles in any direction except travellers arriving who occasionally erupt whilst in mandatory quarantine. So, it all feels a bit 'looming' atm. We live in an odd bubble moment.

cavenewt said...

Pass—"However conservative religious leaders erupted into a furore..."

You have those too? And here I was thinking it was just us.

Passthecream said...

I think we inherited some of yours in the 1950s up in Queensland but we have our own varieties also.

Second 'erupt' there should have been 'succumb'.

Eric said...

It has been said, e.g. by the Swedish doctor writing guest blogs over at Malcolm's, that the Stockholm region has virtually no new infections, so herd immunity has been achieved. However,
cases are trending up in Sweden, and Tegnell, out of all people, is now considering local and time-limited lockdowns:

The article places new daily infections in Stockholm at 29 in the beginning of September and 81 on Sept. 21.

Average daily infections for the whole country past week were 480 with a growth rate of 25% per week. Deaths were 2. Per capita, both are at about 1,8x the German numbers, where one would not expect herd immunity due to lowish total numbers and deaths.

My goto reference for live data is this article (updated daily):

Regarding excess mortality and dangers of lockdowns, about the fourth graph is labelled "Übersterblichkeit während der Corona-Epidemie". The data are fed from the Economist (link below graph), but I like the graphic presentation of Der Spiegel better.

Quite a few countries had negative excess deaths during the lockdown, probably due to lower infections from other pathogens and less traffic (although even annual traffic deaths are comparatively small).

Lastly, French new admissions to hospitals, to ICU and deaths are trending up significantly, in spite of the fact that partial heard immunity should have been attained.

Eric said...

This article from yesterday says much of the rise is seen in the Stockholm region, but it fails to give numbers for Stockholm:

Eric said...

Has regional two week averages, but not for Sweden. You can go back in time, and it supports the idea that France should be pretty far on their way to herd immunity.

Peter said...

Yes Eric, and admissions to UK hospitals are increasing too, though our data are particularly hard to extract. The virus undoubtedly kills people and there are undoubtedly susceptible people who have not yet been exposed in the UK. Sadly some of them will die of it, as they might of influenza or pneumococcus.

But how much good will suppressing the virus do? Every healthy person will stay a potential carrier. I think we can assume the vaccine will be of limited use in the elderly and that, in all probability, it will merely limit severity of the disease in our grey-zone moderate metabolic syndrome population without stopping them transmitting the virus, so then we are going nowhere.

While I have no problem with the term "herd immunity" (being a vet looking at people) I think I have misused the term. The virus will still be around. Most of us will be okay. Some of us will be ill. Some of us will die. That's viruses for you...


Eric said...

Sorry for the monologue, got some reporting from Sweden in an expat online paper:

923 positives in the week before Sept 29 for Stockholm, slightly less than 1 million inhabitants, so about 95 / (week and 100,000).

Eric said...


I was challenging the notion that through cockups or whatever, some countries have achieved herd immunity, and that we must let it runs its course, as Prof. Gupta advocates. I must say I cannot find reason or rhyme to worldwide numbers, almost as if the virus did not care about past exposures and potential immunity.

How can we have 95 /week & 100,000 in Stockholm unless either immunity is not so great or all precautions were thrown to the wind and the virus was very efficient in seeking out those not yet exposed.

I like the idea of masks acting as an improvised inocculation, keeping one up to date with whatever is circulating while limiting virus dose.

Alternate Future said...

Peter and posters, it would be great if someone could tackle the issue of false positives. It may be outside your scope, but you are brilliant at math, and that is where the answer lies. I've heard a false positive rate of just 1% in a population where 1 in 900 are testing positive (UK figure I believe), means the test is almost meaningless. There are also issues of threshold cycles used being too high, and whether the test is picking up fragments of past infection or current infection. In other words, can we really trust the figures on number of "cases"?

ctviggen said...

There are so many factors involved, it's hard to make sense of it all. For instance, did you know that (supposedly) Sweden has the highest rate of single-occupancy homes in all of Europe? Also, while they didn't "lock down", they did provide some semblance of shutting down. This is one of the better articles I've found:

And I keep seeing the theory that once X percent of people get covid, the virus stops. But the virus is coming back in Brooklyn, which supposedly had a relatively high positive rate. You also can't factor in what people have done in response to covid. I know people who have not gone anywhere -- even to get food -- for MONTHS.

Also, while I find the idea of "protecting the vulnerable" pleasing, the reality of it is difficult. We know a family where the daughter went to a college party, came back with covid, gave it to her parents and her sister. Her sister gave it to her two year old daughter. The two year old daughter gave it to her aunt, who watched the two year old twice a week. The aunt gave it to her relatively elderly boyfriend, with health issues. We're hoping he does not die. And both the daughter who went to college and the two year old were supposedly asymptomatic. How do you protect anyone with a virus where one can transmit it while being asymptomatic?

I'm 56. We know of a dad (55) in my town who spent 12 days in the ICU and 50+ total days in the hospital due to covid. Know of a 55 year old man, dead from covid. Both got them via wakes/funerals, both in NYC.

I have a 75 year old (grand) mother who lives in our house with us. I have no tests I can give my kids. I listen to This Week in Virology (podcast), and they had a guest named Michael Mina, who said $1 spit tests for covid are possible. See this too:

Yet, in the US, we can't get this. My kids are now going to school, one 5 days per week and one two, and we have no idea whether they are positive or not. We use temperature and oxygen sensor on them, but how good are these? The school has a ridiculous amount of protocols, including masks, but with no centralized expert guidance in the US, how do we know these are OK?

I have multiple degrees and read about covid all the time, follow epidemiologists and doctors on Twitter, and have no idea how to adequately protect my mother who lives with us, nor do I know what can be done for groups. Can we have our Thanksgiving in the garage with fans and heaters and masks when possible? I have no idea.

Follow a doctor from the UK. His son got covid and months later still has POTS (Postural Orthostatic Tachycardia Syndrome). On TWiV (This Week in Virology), they have a weekly update from Dr. Griffin, a doctor who has had 1,700+ covid patient interactions. While he did not get covid, his colleague did, who is younger than him. She had no known risk factors/comorbidities according to him, yet has spent several months in the hospital. She's going home, but with oxygen because she still can't breathe.

And I could go on about all the so-called "covid long haulers" I see on Twitter.

In my opinion, we need:
- more, cheaper, faster tests;
- a central authority who can tell us how to actually protect people;
- a centralized response (can't have State 1 doing one thing and State 2 doing something else);
- others, but have given enough of a random diatribe.

Eric said...
This comment has been removed by the author.
Eric said...

@Alternate Future:
False positives are not an issue if dual target testing is used, which is what is done in Germany and most of Europe. Estimates of FP by RKI and Drosten are in the < 0.01% range.

I have made a couple of comments about 60% down and linked a round robin report:

Didn't win me a lot of frieds, there. That blog has been attracting a number of closed minds recently.

cavenewt said...

Total dream: if, as Eric hypothetically suggests, the virus does not care about past exposures and potential immunity; and if Science actually determines that metabolic health is the only way to successfully weather an infection; then, this could turn out to be the event to kick humanity's dietary butt and force us to revert to a more evolutionarily-appropriate diet. Non-believers will pretty much die, natural selection in action.

Of course, that leaves out a whole lot of other important things like overpopulation, sustainable agriculture practices, etc. But one can dream.

Disclaimer: the thought only just occurred to me and has not been thought through to any depth.

Gyan said...

Do we need anything with this virus? Lots of people are producing antibodies on their own. They don't need vaccines. And if >90 percent people have asymptotic infection then doesn't that mean that these people are in fact immune to the virus?

So herd immunity isn't relevant. People are immune right now. +-

JR said...

Sweden functions like a control group for the whole world, so it is of great interest to me (fluent in swedish, have lived there in several occasions and years). This is their own statistics (with <3 to Eric)

Of the 6000 deaths, 250 is below 60 years. 650 below 70 years. Make you math. Their background is two weak influenza seasons, and a combination of immigrants taken care of the old folks in service homes. In the beginning, nodody saw the risks for this, but they corrected the situation.

So, we have 3 weeks worth of extra deaths, mainly old folks. So, my prediction until next summer is, that Sweden shall have 5% less-than-averige deaths for influenza covid etc.

To predict is very risky business, let's see...

Kathryn Rose said...

I'm not sure herd immunity is that easy to achieve. The virus seems to be extremely contagious in some minority of people (who will pass it on to an entire room without really trying), and not very contagious in most others (who might pass it on to a few close contacts but that's about it). If you have a somewhat randomly low proportion of superspreaders early on, or if you focus on limiting the kinds of interactions where superspreading can happen (large groups in enclosed spaces for long periods of time), then you're going to have apparent success in controlling the disease unless you end up with more superspreaders somehow. That isn't herd immunity, it's just a combination of luck and successfully limiting mass infection events. Sweden limited the size of indoor gatherings pretty early on, IIRC. I don't know what their contact tracing is like.

Likewise, we can't really say that the UK is anywhere near herd immunity if, say, three quarters of the cases are coming from superspreader type situations rather than close contact within household groups. There will be lots of people (like everyone in my household) who avoided exposure early on by practising prophylactic social distancing. I gave up public transport at the end of February (for Lent), and started avoiding larger gatherings (including choir rehearsals) a few weeks before lockdown, when most people were still going on about hand-washing (which turns out not to be terribly important in aerosol spreading, though it's generally a good idea anyway of course). I did this because my spouse is at higher risk of severe complications, and because I was at a point in my life and work where it was relatively easy to socially distance in this way (finishing up a PhD). The likelihood that I've been exposed, been totally asymptomatic and also developed an immune response to make me resistant to future infection seems relatively small if superspreading situations are responsible for a lot of cases, because I haven't been in any of those situations.

But the contact tracing the UK was doing in March wasn't "okay, you have covid-19, where did you get it? Let's see if we can find and test everyone who was at that venue on that occasion", it was based on close contact with infected individuals. I don't know if the new fancy-pants NHS app is any better in that respect.

Eric said...

@JR: Thanks for the link. Since I can read Swedish, but barely so, I found an English version:

What did you mean by "<3"?

JR said...

From somewhere with love -you showed great interest into Swedish experience. Well, not such a good joke.
Mfg JR

JR said...

This is even more to the pulse of covid: icu admissions Sweden.

90% of men and over 90% of women, under 60y, survive 30 days of icu. 70-75% of all 60-79y survive as well one month in icu.

I don't know, sounds impressive to me.

Peter said...

Eric, I recall the ITU medic from Lewisham, London, being very cautious about viral load. Once you are infected the virus will multiply so rapidly that whether you get n particles or n X 100 particles, the n particles breed pretty fast to catch up. Not sure what I think about that.

Alternate Future. My wife’s PhD used a huge amount of PCR. It stands Probable Contaminated Reaction… And no, she didn’t use 40 cycles (dunno how many, but in basic research you really don’t want to find stuff which isn’t there).

ctviggen, we have two elderly relatives in a significant at risk category. We have the option of them staying isolated. We have the luxury that our best course would be for the children to get COVID asap, infecting our immediate family and then not having to worry about meeting Ken or Daphne. It’s tough if you cannot isolate specific elderly people. It wouldn’t be for particularly long once the kids were positive. To me there is more of a problem with people of moderate age with DMT2, severe hypertension etc. They will be out at work in society and have to make a choice of how much to isolate themselves. But once their potential circle of acquaintances have had COVID they would be safer. If they work in a shop, the sooner their customers have had COVID the sooner their risk will drop. Slowing the spread extends the risk time, never knowing who is positive. Not slowing the spread increases acute risk. But at some time all of us are going to be exposed, eventually. If people wish to self isolate essentially for ever that is their choice. Mandating we all do this is a problem. Expecting any vaccine to fully eliminate risk is cloud cuckoo land. So we are about managing risk.

cave, I think, ultimately, our metabolic train wreck has to be laid at the feet of the cardiologists and their bizarre lipid hypothesis. If we could kick their butt, that would be a great start.

Gyan, yes. But folks in at risk groups still have tough decisions to make.

JR, yes Sweden is an interesting location to watch. Also New Zealand can’t quite get themselves COVID-19 free, a trickle of positives keeps coming through.

Kathryn, yes, if herd immunity means no one will get ill because too many people are immune, this is unlikely. As I said above, my use of the term herd immunity was incorrect. My guess is the virus will eventually circulate at low levels and highly susceptible individuals will still become very ill or die when they do meet the low level of circulating virus, much as for influenza. This is going to happen (at the risk of predicting the future) but another April sized wave of deaths seems very unlikely…


cavenewt said...

Peter, most of your observations are predicated on the fact that reinfection is not likely to occur. What do you think or know about the reports that antibodies from a previous infection might not confer much protection, much as is the case with the common cold, right?

My understanding is there's never been a vaccine for the cold because it mutates so much. Apparently COVID-19 doesn't. But then we have these weird reports that people are getting reinfected. Could be an artifact of testing or whatever… But it could mean that a vaccine won't be any good.

Mostly I want to see evidence that eating low-carb and/or avoiding PUFAs means COVID-19 won't hurt much. I don't know if anybody's looked at that specifically, though.

Peter said...

cave, obviously there is no evidence base for low carb other than the diabetics with admission BG over 10mmol/l do badly and our older non COVID papers showing elevated double bond index in plasma FFAs at admission predicts ARDS development. It's interesting that many critically ill patients already have high antibody levels on admission to the ITU. This made me cautious about serum therapy, which is an entirely logical (and usually effective) concept but not clearly helpful in this case. So it might come down to T cells and the innate immune system. It also highlights that people going in to the ITU have already made a marked immune response and are suffering from cytokine storm rather than viraemia per se. Hence dexamethasone helps late on. I can think of nothing more idiotic than giving dex early.

The Oxford (2/4000 transverse myelitis, gulp) vaccine stimulates T cell immunity as well as an antibody response.

I also find Japan interesting as very large percentages of their population have seroconverted with no significant increase in hospitalisations, certainly no significant deaths.

They have clearly been exposed and the pre-print only recruited people who were never been ill throughout, anyone with clinical COVID-like signs was excluded. Now 60% of these healthy people are sero positive. They are actually talking about herd immunity (as is Prof Gupta) but their elderly do not develop metabolic syndrome in quite the way that our populations do, so it's hard to say if the elderly are being protected by herd immunity or whether they are being protected by not being diabetic.


Kathryn Rose said...

Thanks Peter. What I'm curious about is whether coronavirus exhibits the same kind of infection patterns as influenza. I've not really heard about superspreaders for 'flu, but is that because 'flu isn't so novel (even if it does mutate a lot), so there is some background herd immunity i.e. lots of people can fight it off? Or is there another mechanism, like most strains of 'flu not being ones where one person might inadvertently infect an entire Tube carriage (or whatever)?

If the basic infection pattern is different -- if individual superspreaders or superspreading events are responsible for a bigger proportion of covid-19 outbreaks -- then sure, we might get to some baseline where a large proportion of the population is immune, but it also might take several outbreaks to get there. My instinct (and it is pretty much instinct, a generalisation from thirty-odd years of choral singing and watching colds/'flu/whatever spread through choirs, usually in the run-up to Christmas or on tours when we're rehearsing more) is that they are different. Some bugs seem to infect only those standing near the first person in the choir to develop a cough; others seem to strike an entire choir, just as the first person is recovering. The mean transmission rate (with a much larger sample than a choir) might look pretty similar, but one bug produces very infectious superspreaders and the other doesn't, so you get these patches of very high transmission even without close contact.

Then there are the longer-term effects. I don't have many people in my direct social circle who think they've had covid-19 -- we are a cautious bunch, and also relatively privileged so able to take more precautions than some. But of those I do know, some people are having some pretty weird symptom relapse issues months after apparent recovery. That isn't the same as death, and of course there are all kinds of weird post-viral illnesses for other viruses, most of which probably get written off as "just one of those things" without necessarily even connecting them to the infection. But it does make me personally more inclined to be cautious, I have enough chronic healthcrap as it is.

My understanding is that Japan has very high mask compliance; I wonder whether that reduces the viral load enough to produce a large number of very mild cases.

I do think we will eventually get to a similar place we are at with 'flu, where there are seasonal waves of covid-19 or something closely related, and a few more vulnerable people die each season. That may be as good as it gets. I'm just not sure we're there yet, or that trying to get there faster is wise. Everyone dies eventually, of course.

karl said...

Please ignore "cases" - they are not counted the same - the PCR test was not designed to be used on the 'well' public - has about a 2.5% false positive - so if you test a million healthy people you get 10,000 cases that are not. The MD's are seeing people that had covid - test positive - no symptoms no fever for months.. So just ignore anything based on case rates. In my book - someone without symptoms or no fever is not a case.

Even the death rates are skewed a bit by the misuse of testing - but if we look at death counts, the virus is doing what viruses do..

The isolation and masking also kills people - there is a 3x rate of psychological interventions - no one wants to count these people - stress kills - war time sees a 3x rate of heart attacks... The economic interruptions in the third world combined with the disruption of TB and malaria programs is likely to kill more than the virus - one estimate - just the TB and malaria program disruptions will cost 4-million - economic deaths are hard to count. When the West gets a bad economy - children in the third world don't eat.

Some good news - I already knew that vit-D effects viral replication - now a controlled study re Covid:

The vit-d level that is call deficient (30) is related to rickets - not immune function - not well studied - but native people have twice that levels with spikes above 100. Some think that weekly dosage might only help with the rickets problem - not immunity..

There were two influenza pandemics in the '50s and '60s that killed a similar number - 1-4 million world wide - cooler heads prevailed at that time.

I know people want to help - want to believe that their mask (that they touch every 90-seconds) matters, but it is a form of fear mongering and is actually hurting a lot of people.