Daily numbers, graphs, analysis thread 16

[BigChocFrenzy]

Welcome to thread 16 of the daily updates

Resource links:

Uk dashboard deaths, cases, hospitals, tests - 4 nations, English regions & LAs
MSAO Map of English cases
[[https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/909430/Contain_framework_lower_tier_local_authority__14_August_2020.pdf
Slides & data UK govt pressers
UK added daily by PHE & DHSC
R estimates UK & English regions
PHE Surveillance report infections & watchlists every Thursday
ONS England infection surveillance reports
ONS UK death stats released each Tuesday
ECDC rolling 14-day incidence EEA & UK
Daily ECDC country detail UK
WHO dashboard
Worldometer UK page
Plot FT graphs compare countries deaths, cases, raw / million pop
Covidly.com world summary & graphs
Plot COVID Graphs Our World in Data test positivity etc

We welcome factual, data driven, and civil discussions from all contributors 📈 📉 📊 👍

Decisions about restrictions here seem to be a bit more nuanced now, rather than "lockdown". So shops may be allowed to stay open but people told not to mix households. Of course there have been so many changing messages you have to wonder who is still listening. I know I gave up some time ago - just watch my local figures (when I can) and decide what I'll risk myself.

boys3 you are missing those who test positive and still roam about infecting people, there are known to be some where I live. If I knew the names they would be reported to the police and their employers.

@MRex

It's worth noting that while 1.5% of those tested are using positive, it's more like 0.05% of the country testing positive in the randomised swabs. New infections per day gives part of the picture, but people currently infected gives risk of spread. That's estimated to be 28,300 in the UK right now, or one in every 1900 people.

I was just thinking how we can work out the number of undiagnosed cases. Say anyone is infectious enough to get a positive swab result for just 5 days (before day 5 of symptoms is recommended). At a rolling average of 1200 identified cases per day, 6000 will have been identified. That leaves 22,300 unidentified and potentially roaming around infectious, whether asymptomatic or selfishly untested with symptoms. A tiny proportion of people, but they are what's keeping this going. Some may be false positive swabs, but while we know there is spread it's not worth removing them from the figures.

Would be interesting to get thoughts on tweaking these assumptions - anyone?

.... Mrex It seems reasonable

Both our different means of calculation agree on a tiny number of cases compared to the number at peak,
which together with the much lower age profile of cases explains the very low deaths in the UK and other European countries

Measures to combat infections need to take into account the far lesser danger atm
This doesn't mean exponential growth can't take off again if we get too careless for weeks, hence the need for continuing SD,
but not the draconian measures in the early days of dealing with a "Novel" virus

imo, other European countries need to at least lower the maximum numbers at assemblies to the UK limits.
Allowing sometimes hundreds of people in weddings etc is reckless

Both 7-day and 14-day case incidences are useful to give a fuller picture,
just as the 4-day and 7-day R-numbers do (the RKI calculates both R-numbers , but I don't know which time periods PHE uses)

We don't have a good idea of how many people were infected back in March,but it was thought to be 1-5 % (I calculated this would be between 800,000 and 3.7million-ish from the figures suggested). This was the level of infection that gave rise to 1000+ deaths per day.

We were told at the time that there were about 1000 infected for every 1 death (presumably prevalence rather than incidence).

The deaths have been readjusted so that deaths over 28 days from diagnosis don't officially count. However, there is a suggestion that this underestimates the true figure by 15-25% (and there is a complicated medical question about how you would make that judgement medically, for each individual patient).

If we have 30,000 approx infections in the community then 10-15 deaths would seem to be in a similar ballpark.

So it would seem that the virus hasn't actually become that much less dangerous, it's just that lockdown has knocked back the numbers, what do you think?

The problem is that if infections are seeded all the way through the community, the potential for exponential growth remains if we all cram back into offices, buses,schools.

And then after about, um a month or so, we could get back to where we were before?What would make a difference would be everyone not going to the pub or having parties or eating out...

7-day incidence rates Per 100,000 here:
www.statista.com/statistics/1139048/coronavirus-case-rates-in-the-past-7-days-in-europe-by-country/
Show that most countries are above 20 now. However UK is 16.

I am really not sure about using death rates now to show prevalence of Covid19 - we have more effective treatments, people are getting seen sooner, and we don’t know yet whether there will be less risk or death than in March or April. Positivity rate and amount of testing is more useful? As is the contact tracing - was it from a known contact - or not? If not known and many cases ‘popping up’ spread out with no obvious link - that shows seeding and community transmission and likely to mean many more ‘unknown’ cases.

The same with schools, as children not usually tested / asymptomatic - then they may have been under counted.

I think prevalence is something we more retrospectively see doing random samples to see who has antibodies?

"I am really not sure about using death rates now to show prevalence of Covid19 - we have more effective treatments"

Improved treatements may have cut deaths to ½ or even ⅓ of what they were, but currently UK deaths are about 1% of the peak 1,445 deaths
Remember, we have no effective anti-virals for COVID, no magic bullet, even Dexamethasone only reduced deaths by up to about 30% and only in certain types of patients

it really is about the massive reduction in the number of cases

ONS data shows that the real daily deaths are somewhere between the 28-day and 60-day cutoffs
i.e. the 28-day cutoff under-estimates, but the 60-day over-estimates

We need to use all data to estimate the number of cases, but the ONS surveillance reports look to be the most reliable indicator that daily cases are very low indeed, compared to March-April

Infections are so low, that the ONS samples are too low to show accurate trends any more.

Hence, the ONS COVID-19 Infection Survey will be expanded from regularly testing 28,000 people per fortnight in England to 150,000 by October
and eventually to 400,000 people in England, also adding the other 3 UK nations.

This also seems useful in risk assessment (ECDC quotes)

In EU/EEA countries and the UK where a recent increase in cases has been reported:

• High or Very High The risk of further escalation of COVID-19 is high in countries that have also had an increase in hospitalisations, providing a strong indication that there is a genuine increase in transmission occurring - must introduce more measures
• High or Very High No increase in hospitalisations but having seen an increase in test positivity* (if testing capacity is sufficient and intensity has remained stable) - must introduce more measures
• The risk of further escalation of COVID-19 is moderate-to-high for those countries reporting no increase in hospitalisations or test positivity (if testing capacity is sufficient and intensity has remained stable). The countries that have multiple physical distancing measures in place should conduct local risk assessments to better understand the groups or settings driving the increase in cases.

However it does not say how much increase or any thresholds.

"prevalence is something we more retrospectively see doing random samples to see who has antibodies?"

The surveillance reports show the % with antibodies reducing quite quickly over time in the various regions

It is more difficult to find the % with T cells and in any case, the research indicates that T cells reduce the severity of infection, but do not stop someone passing on infection

@BigChocFrenzy I just meant that I’m not sure using death rates to indicate prevalence is so accurate now as we are not really sure all the reasons why they ‘appear’ to be lower at present - (but may well rise). Not sure I explained myself very well. That is not to dampen the seriousness at all of covid19.

German Traffic Lights

This was the most recent set of criteria, but it may get updated as more knowledge is gained about the virus

Stricter national containment measures come with each colour level

Currently green

Yellow - any 2 of 3:
. R >1.1 for 3 days
. 7-day incidence of 20 / 100,000 cases
. >15% of ICU beds occupied by COVID patients

Red - any 2 of 3:
. R >1.3 for 3 days
. 7-day incidence of 30 / 100,000 cases
. >25% of ICU beds occupied by COVID patients

red on its own to lockdown any of the individual 401 administrative districts:

7-day incidence of 50 / 100,000
Bavaria chose a stricter limit of 30 /100,000

@BigChocFrenzy yes good point too, antibodies are a slippery thing to measure! Another unknown really. Although they did show very high rates of antibodies in Mumbai and other areas so we could see that there was a very high prevalence in the slum areas.

@BigChocFrenzy your last post was EXTREMELY useful (sorry to be in capitals) thank you very much! That is exactly the kind of numbers tied to restrictions that I was looking for. Very, very useful.

[quote IceCreamSummer20]@BigChocFrenzy I just meant that I’m not sure using death rates to indicate prevalence is so accurate now as we are not really sure all the reasons why they ‘appear’ to be lower at present - (but may well rise). Not sure I explained myself very well. That is not to dampen the seriousness at all of covid19.[/quote]
.....
Death rates are fundamental, because scientists say there is not yet evidence that the virus variants are significantly weaker
and studies show that improved treatments can maybe reduce deaths by ½ or ⅓, but not by 99%

Also, the management of the crisis is driven by the number of deaths or potential deaths
Since deaths have been so low for a couple of months or more, there are only limited SD measures that are appropriate, when balanced vs the economic impact

Agreed and the long term ‘long covid’ poor health also suffered by people who will then have to be off work, can’t look after their children or other effects.

I'm still looking forward to the first covid sewage report. They started the monitoring on 3rd August, so surely there will be a report in early September? I think it will throw in some surprises about added of greater or lower infection rates and become very important as a metric.

I actually quite like that the surveillance report is explicit that they also use judgement about local capabilities and other measures. If you know a hospital is struggling from other factors, or a new testing facility is in place

Whoops, to finish my point...

... or if 90% was linked to a particular outbreak / overseas travel, then the reaction should vary. Not every consideration can be written in traffic lights.

@MRex

I'm still looking forward to the first covid sewage report. They started the monitoring on 3rd August, so surely there will be a report in early September? I think it will throw in some surprises about added of greater or lower infection rates and become very important as a metric.

I actually quite like that the surveillance report is explicit that they also use judgement about local capabilities and other measures. If you know a hospital is struggling from other factors, or a new testing facility is in place

I'm still looking forward to the first covid sewage report. I know what you mean but this is sentence I never expected to read

Hello! Canterbury (ex) calling in from Italy. I was with you all in the deep dark lockdown days and still follow your thread.

Just to elucidate the reasoning behind the 6pm-6am thing- as masks are still generally used in all shops and commercial places, but weren't being worn (because they didn't have to be) in the evenings at the summer nightspots (which are HUGE- we are talking at times of open air nightclubs with 1000+ people) the new mandatory mask hours were introduced and nightclubs closed.

The % of people testing positive coming in from , respectively, Malta, Croatia, Greece and Spain is high - we are envisaging a total ban on travel between the first two possibly sooner rather than later.

My own region had 11 new cases, 6 from Croatia, 3 from Malta, 2 relatives of two people returning from Croatia.

Haha. Yeah, it's funny what becomes interesting in life. I spent a brief period working with the water industry, so maybe that helps my fascination levels and also has me wondering if certain estimated cases will be logged to England or Wales as the sewage crosses the border down the hill. I really liked the scientific accuracy implied by the research, while by and large the UK water companies are very careful and responsible, so I'm hopeful it'll cut through some of the false positive / swab volume issues.

Were results of the vitamin D studies published? Does anyone know if they are tracking if the only people getting ill now are those who are still vitamin D deficient?

Texas did some fancy maths to correlate vitamin D deficiencies with deaths: www.news-medical.net/news/20200702/More-evidence-on-vitamin-D-deficiency-and-death-rates-from-COVID-19.aspx. As they note though, correlation is not necessarily causation. It was odd to me that they didn't seem to cover a hypothesis that the same factors may cause vitamin D deficiencies and covid spread, i.e. weather. I don't believe it's been fully peer reviewed yet. The Lancet has an article that basically says "the studies aren't up to much and patient deficiencies would be hard to prove, but it's worth taking vitamin D anyway so promote it": www.thelancet.com/journals/landia/article/PIIS2213-8587(20)30268-0/fulltext.

Hi ex Canterbury! Nice to see you again!

I'm still wondering how Murraythemonk from the old Worried** about Coronavirus threads is....

@meercat23 @MRex what have our lives become 😂