Daily numbers, graphs, analysis thread 8

[Barracker]

Welcome to thread 8 of the daily updates.

Resource links:
Worldometer UK page
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Johns Hopkins Coronavirus Resource Centre
NHS England stats, including breakdown by Hospital Trust
Covidly.com to filter graphs using selected data filters
ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday

Thank you to all contributors for their factual, data driven, and civil discussions.

I was a bit confused by the Spain 5% antibody data. If data were collected at the end of April and if it takes an average of 28 days to build antibodies then these people were infected around the end of March at the latest. Has anyone modeled how this will have changed in the 8 weeks or so since? Will the % have increased much since?

Presumably wider transmission dropped during lockdown but there would have still been transmission within families and in hospitals and care homes as we have found in the UK.

Would updating the serology results tie it in more with the modeling out of Manchester? A part of me hoped that they are true and that's why transmission seems to be low especially in London but I do appreciate the difference between hope and science!

www.manchester.ac.uk/discover/news/over-25-of-the-uk-likely-to-have-had-covid-19-already/

I would also love to see what this statement from the Swedish Ambassador was based on. If anyone knows the study I would love to read it. Again, hope or science?

www.google.com/amp/s/www.aljazeera.com/amp/news/2020/04/sweden-ambassador-stockholm-reach-herd-immunity-200427072044548.html

Also has anyone run the imperial model again using more of the data we have since collected? I would be interested to see how it changes the predictions.

@starfish88

That’s a very good point. The 5% would not be a live figure. Even with a R0 well below 1, numbers would have increased significantly since then. It could potentially be double or triple that.

Possibly ventolin is protective certainly they eem to use asthma meds on covid and a nebuliser. Are steroids an anti inflamatory as well as affecting immune system?

Keep sorry which research is that?

Just to explain, Asthma is an inflammatory condition. The internal surface of the lungs becomes inflamed and so the alveoli close up. There are different pathways though, I can't remember the exact details but it depends on the triggers and the type of immune response.

In basic terms there's allergic and non allergic asthma. But it's usually of no consequence as the drugs work on both ; it's more important to find ten correct drug.

Chronic asthma needs daily therapeutic inhaled steroids, clenil is the most basic, the preventer. If well managed with a preventer, you wouldn't really know you have asthma. Steroids are anti inflammatory, like putting steroid cream on excema.

Ventolin is a bronchialdilator, which is for acute flares to widen the passage ways and support asthmatic exacerbations/attacks. There are combination inhalers with long acting bronchialdilators in them too.

Chronic severe asthma also may need regular courses of oral steroids or even daily. There's also biological therapy.

If it's ventolin they've been using it's sounding like ventolin is helpful to widen the passage ways when patients are breathless?

My asthma is non allergic; chemicals, smoke, cold air and particularly viruses really affect it. Without inhalers a virus would affect my lungs for weeks and months, it's still bad and usually longer than others, they feel inflamed and I have to use the ventolin on top of Fostair to breath and ease huge coughing attacks.

I haven't seen any research defining between asthma types; anecdotally from friends who have allergic asthma, viruses aren't as much of an issue as dust etc.

It also seemed to assume that only government imposed measures had any impact and did not take into account voluntary changes in behaviour, which is why it’s numbers are so wrong for Sweden.

This is true... but back in mid-March it would have been extremely foolhardy to have assumed otherwise. I remember at the time there was a big concern at the time that people wouldn’t comply even if Governments did mandate it!

Clearly the model needs refining and I’m sure it will have been.

Sorry that was long; I didn't understand it all till I got asthma and people often look surprised as it's now well controlled and I don't wheeze!

Btw, a local primary are going to be wearing visors, according to their site.

I just want to add, as an asthmatic I've never felt at risk of dying from this virus. I actually feel apprehensive about all viruses every year, simply as they can last longer than in others, really affect sleep as I need inhalers frequently during the night and am usually exhausted for some time afterwards. It's the cfs and lung damage that worries me with this.

@ShootsFruitAndLeaves I don't think it's helpful to say that the 'vulnerable' as opposed to 'shielding' group aren't at much more risk. That group is far too large and there is no allowance for comorbidities. For example, the group contains all the following:

• someone with co-morbidities of arrhythmia, type 2 diabetes and hypertension.
• an otherwise healthy pregnant woman in the second trimester
• a pregnant woman in the 3rd trimester with moderate uncontrolled asthma that requires a couple of courses of steroids a year and 2 different inhalers
• someone with controlled relatively mild asthma
• someone with existing lung damage and previous heart surgery.

You see, it's a huge group. Some won't be at more risk of dying, or even much more risk of severe illness and slow recovery. Some will be at a substantial increased risk of death and certainly an increased risk of lengthy illness. Add into that disparity of age, sex, ethnicity etc and it's nowhere near as straightforward as 'moderate increased risk'.

But anyway, my point was that teachers these days aren't necessarily healthy and people posting 30% clearly aren't using statistics, they are using their own experience in their schools.
Back to the covid data!

Age, sex and ethnicity aren't inputs into whether you are considered vulnerable. However it's certain that older people not in the vulnerable group are at greater risk than the youngest people in the vulnerable group. That doesn't mean individual young people who count as vulnerable aren't, but on the whole they won't be nearly as vulnerable as say 65 yo men without underlying conditions

The vulnerable group is a selection of conditions the government thought likely to be worth being wary of. It's not a statement that they ARE at higher risk - the point is more 'just in case'.

That's not to say it should be ignored entirely, it's just that there are hundreds of times more men in their 60s dying than there are pregnant women (5 in total as I understand it)

I mean as a starting point it seems logical to try and keep pregnant women out of hospital if only to avoid damage to the child, but that wasn't really the point I was making, which was really that there is a risk 200x-300x greater for someone who is male & 69 than someone who is female and 20 and yet the government imposed this on people with 70+ without making it clear that there is no such cut-off and then relatively speaking anyone in their 60s is going to be hugely more at risk than younger people. It's this risk gradient which is above all others in importance. Not ethnicity. Not underlying conditions. Age.

Yes, of course they aren't on the official list. But when the government tells schools to 'make their own risk assessments' about vulnerable people who should 'work from home if possible' and you need to staff your school and choose who you are going to ask to come in, it would be foolish to leave out those factors. Same way that hospital trusts are being told to consider black staff as higher risk. Since the recent data showing that 55% of pregnant women admitted to hospital with covid were BAME, I'd hope that when choosing who is coming back into the classroom an SLT would look at those stats and take that into consideration even though it isn't in the government's (frankly pathetic) guidance.

The concern wrt in schools are for teachers and other staff, together with older family members,
not pupils - who have almost no risk

however, everyone in any kind of work outside the home will have a higher risk
For young people it is a trivial one; for workers in their early 60s it is not.

I would also love to see what this statement from the Swedish Ambassador was based on.

I am not 100% sure but there are a few preliminary non peer reviewed studies that appear to suggest you can potentially achieve herd immunity at much lower levels than 60-80% normally quoted by scientists. I assume this hypothesis feeds into the Swedish strategy.

The usual high herd immunity threshold applies to vaccinations. However, if the virus spreads naturally, once it infects potential superspreaders (e.g. medical personnel, hospitality workers, socially active individuals, etc) it may "run out of fuel" and start slowing down naturally.

Here is one of the research papers that suggests that you may be able to achieve natural herd immunity at approximately 20% infection level.

www.medrxiv.org/content/10.1101/2020.04.27.20081893v2.full.pdf

The usual caveats apply - it is a pre-print study that requires further review and analysis (as do most C19 related publications that seem to come out every day).

"I was a bit confused by the Spain 5% antibody data ....
Would updating the serology results tie it in more with the modeling out of Manchester?"

The % has likely increased since then, maybe to the higher single figures
However, with lockdown and flattened curves, it is most unlikely to be anywhere near the Manchester

• that was based on theory and has been panned by reputable scientists

I would be v interested in updates from earlier studies, e.g. the ongoing Uni Bonn study of the Gangelt epicentre in Germany,
to see how much the 15% infection rate there has changed

There are too many people desperate to clutch at theories - not backed by tests - from any publicity-hungry chancer with a PhD
Here, we should stay with data and reliable science

The Imperial model may still be relevant in UK planning;
however, it never was on the continent, where lockdown was imposed as reaction to Italy's carnage and warnings from our own virologists and epidemiologists

Since then, we have been relaxing lockdown mostly relying on mass testing and the curently low level of new infections

Even in the UK, I'd expect it was the CMO's worst case back-of-the-envelope-sums that convinced the PM & Cabinet, not Imperial

One aspect of Imperial's modelling that is concerning for after lockdown is that flu-based models like theirs and IHME have cases dropping to near-zero in summer due to seasonality

Current falls are believed here to be due to SD memasures;
some of the public hope for another 40C heatwave, but so far scientists in the field here are relying on SD, until significant seasonality can be proven

It's not exactly BAME is it.

It's very strongly 'black'

www.npeu.ox.ac.uk/downloads/files/ukoss/annual-reports/UKOSS%20COVID-19%20Paper%20pre-print%20draft%2011-05-20.pdf

3.7% of the control group (women giving birth three years ago) were black Vs. 21% of the hospitalised cohort with covid-19. That's 6x higher

For Asian women it was 24% vs. 11%. That's more than double, but a large part of that will be geography in that BAME populations live mostly in cities hit by covid-19 more than the rest of the country.

For Chinese women it was 6 out of 400, Vs 7 out of 700, not a sign of increased risk.

There was, as expected, a straightforward link between age and risk.

Being South Asian might be a small risk factor, but it is clearly not a risk factor even as big as sex, let alone age. In other words, given that 6 or 7 years of age will double your risk generally, it doesn't really make sense to treat South Asian ethnicity as a big risk factor, given that if you adjust out the increased exposure due to employment, geography, possible cultural differences, etc., any remaining risk size is not going to be that large.

The evidence for black people is more disturbing as it does seem to be coming in from multiple sources. It is unfortunate that 'BAME' is being treated as a monolithic group. We should likely be talking about black people.

This might be a dumb question, but looking at the link below, mortality exceeding expected levels (spring 2018) is still high in weeks 12-15 for some countries, but not others (attributable to course of influenza outbreak across Europe) - how does this make comparisons between countries reliable, especially if some publications are using a mixture of 2 year averages and 5 year averages, e.g. The Economist?

"Weekly all-cause mortality for all ages exceeding the threshold level (+2 z-scores) during the 2017/18 winter season by country, based on the EuroMOMO algorithm."

www.clinicalmicrobiologyandinfection.com/article/S1198-743X(19)30058-8/fulltext

Also, in one of the links from last night, I notice that under Table 1, it says; Notes: *Some of the FT’s published P-scores look unreliable as the data and graphics do not match;

www.inet.ox.ac.uk/files/6-May-20-Muellbauer-Aron-Excess-mortality-in-England-vs.-Europe-and-the-COVID-pandemic.pdf

I agree with quarantine. This isn't just about deaths. It's encouraging that asthma death risk is low; I know though that lungs that are quick to inflame and get secondary infections are at risk of further damage.

Re the 30% teacher thing; I'm not aware of that but I do know in primary schools the majority of staff are women.

In our school the vulnerable wfh include asthma, MS, type one diabetes (from very young), type two with obesity and pregnancy. One who is expect shielding as she had cancer.

The only other thing that I've wondered besides pregnancy is that autoimmune diseases are more common in women of which ms and type one are. Asthma can be too.

There was some research linked here I think about COVID infections that lingered and that data was interesting as it's a very different aspect of the issue. I didn't save it unfortunately.

Also, we mustn't minimise the effects of stress on pregnancy outcomes. The risk is very low, yes. We don't know if there's any impact of infection on the baby long term. The evidence of stress on maternal mental and foetal physical health is well known however.

Loss of taste and smell is finally added to the list of symptoms that should trigger self isolation

I know several people working in frontline areas who had this symptom over the last few weeks that, on advice from senior managers, continued working with colleagues and vulnerable people. They didn't even qualify to get tested at that point.

Yes five there have been cases where people (not uk afaik but i suppose could happen here) have been shot, fallen or RTa.
So cause of death is stab wound but they also tested positive for CV and so it's mentioned on DC. I guess it's the uncertainty that having CV meant they didn't survive the stab wound whereas they might otherwise have? It's so complex.

RTAs, shootings and all but uncontroversial falls (i.e. old ladies falling over at home etc.) during covid-19 won't have been registered yet. This will take several months.

So they won't appear in death statistics yet, whether they tested + for covid-19 or not.

I'm really just thinking out loud here, but it occurred to me that most children currently in school are those of key workers, many of whom are in touch with the infected; therefore we might have expected to see large numbers of cases arising from this if the dangers were as great as some claim

I'm not aware that any such large clusters have arisen - surely the unions would have told us if so - and wonder if this is perhaps some reassurance around more children returning to school?

I know I'm a bit late but the Brazilian president has already had it, he held a rally and was shaking hands with people whilst symptomatic.

My brother lives there & barely leaves the house because there's no social distancing/lock down at all. It's going to be awful there but the official statistics don't show the reality (their numbers are even more fudged than ours).

Same across the rest of Europe, where several countries e.g. Germany, have had children of keyworkers in school

There have been NO clusters of child deaths
and I'd expect headlines too if there had been any clusters of teachers or other adult deaths related to schools