Daily numbers, graphs, analysis thread 11

[BigChocFrenzy]

Welcome to thread 11 of the daily updates

Resource links:

Slides & data UK govt pressers
NHS England stats including breakdown by Hospital Trust
ONS UK statistics for CV related deaths outside hospitals, released weekly each Tuesday
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Worldometer UK page
Covidly.com to filter graphs using selected data filters ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday
Plot COVID Graphs Our World in Data

We welcome factual, data driven, and civil discussions from all contributors 💐

Rapid tests are being used quite widely on entry into various countries. E.g., for Malaysia if you fail then you go to hospital, and if you pass you do a 14 day quarantine.

They will filter out 'some' of the infected people, which is better than none, but if you consider the case of a country with 'zero corona' and afterwards no internal controls then if you allow people to come in with rapid tests alone, then it follows that if there are many international arrivals, then some will get through and infect your population. This doesn't necessarily mean 'a second wave', but it does make it difficult to go 'zero corona'.

It is balancing the risk of local outbreaks vs the loss of tourist income

which is a risk some S. European countries are also having to take - without any tests afaik

Just relying on only accepting visitors from certain countries

• and assuming that not too many have the time / money to dodge this by coming via another country

The risk is reasonable for countries with an efficient test, track & trace system for new outbreaks

Once countries have COVID under control,
the knowledge that public health has accumulated about what works to stop COVID spreading
gives a good chance of localising and squashing new outbreaks

https://www.newstatesman.com/world/2020/07/data-shows-economic-recovery-requires-more-reopening-shops

https://opportunityinsights.org/wp-content/uploads/2020/05/tracker_paper.pdf

shows that there is virtually no difference in spending patterns in US states that have reopened versus states which have not.

In turn, the authors suggest, this would indicate that health concerns are a significant driver of spending decisions
– not government orders to shut down high street establishments like salons and shops.

Gilberto Montibeller, a professor at Loughborough University who studies risk, told the New Statesman that
any policy intended to restart the economy must take into account how risky people perceive certain activities to be
– regardless of whether those activities are truly likely to transmit coronavirus.

“Covid-19 is a classic example of a high perceived risk.
Any policy has to deal with both perceived and actual risk and understand why people are feeling afraid.

“You cannot open the economy and expect activity to resume at steady levels, because there are health concerns.
The two issues are not as isolated as policymakers assume.”

For anyone who is interested in local data, the Govt stats site
coronavirus-staging.data.gov.uk

now has deaths/cases/hospital admissions by area over time.

@PatriciaHolm thank you - is this just pillar 1?

Also useful to see the 4 test pillars combined on 1 graph

Reportedly they plan to add UTLAs & LTLAs for deaths & hospitals later,
but may not be able to do so for tests for some time

Patricia I can only see it by the 4 nations on that Beta site, not yet LAs

That's only 4 nations for tests
The deaths/cases/hospitals have English regions too, but not for the other 3 nations and not for LAs

Cases are there by local authority area. They are pillar one and pillar 2 now.

So we are sure # cases = positives from Pillar1 + Pillar 2 ?

Then mainly what we are missing there is the total # tests per LA, to calculate % positives

@BigChocFrenzy

So we are sure # cases = positives from Pillar1 + Pillar 2 ?

Then mainly what we are missing there is the total # tests per LA, to calculate % positives

Yes, they are, it says so in the notes. And yes; the data on weekly positive rates is in the ONS spreadsheet, but that is only released weekly.

Several people on Twitter are calculating daily case rates per 100,000 population - @rp131 has been doing some, and @cricketwyvern

Royal Society DELVE: Nosocomial (Hospital Acquired) COVID infection

rs-delve.github.io/reports/2020/07/06/nosocomial-scoping-report.html

•	Using publicly available data, we estimate that <span class="italic">at least 10% (95% confidence interval: 4-15%) of all COVID-19 infections in England were among patient-facing healthcare workers and resident-facing social care workers during the period from 26</span><span class="italic">th</span><span class="italic"> April to 7</span><span class="italic">th</span><span class="italic"> June 2020.</span>

An estimated further 1% of infections in this period were acquired by inpatients in hospital,
with an additional 6% of all infections among care home residents.
This does not consider secondary cases arising from hospital acquired infections elsewhere.

•	Although surveillance systems and large-scale hospital-based studies have recently been set up, there remain gaps in availability of surveillance data on hospital-acquired infections, particularly of healthcare workers (including agency staff) and in nursing homes; 

important questions remain unanswered, including about the impact on Black, Asian and minority ethnic (BAME) health and social care workers.

•	At present, there is no single organisation with clear oversight of hospital-related infection surveillance, monitoring and response. 

Data obtained from surveillance, monitoring and outbreak investigations need to feed into epidemiological and modelling research, including to evaluate interventions. 

There remain significant opportunities to link epidemiological and phylogenetic datasets to improve the ability to identify and respond to changing epidemic dynamics.

It does appear data provided on testing has finally improved, and I’m quite impressed with the new UK gov site.

We finally have daily data for Pillar 2 on total tests actually processed, which helps make some sense of the data. But according to the notes Pillar 2 data for number of people tested (as opposed to tests processed ) will only be made available once a week as part of the weekly test and trace reporting. So positivity rates overall can only be calculated weekly. Overall rate from the latest test and trace report was 1.7%.

@AprilLady

It does appear data provided on testing has finally improved, and I’m quite impressed with the new UK gov site.

We finally have daily data for Pillar 2 on total tests actually processed, which helps make some sense of the data. But according to the notes Pillar 2 data for number of people tested (as opposed to tests processed ) will only be made available once a week as part of the weekly test and trace reporting. So positivity rates overall can only be calculated weekly. Overall rate from the latest test and trace report was 1.7%.

The Government have just said today they have given up (!) trying to give daily figures for people tested. I'm assuming they are going to continue the weekly number for test and trace.....!

A lot of data for them to check & publish - and they do seem overwhelmed sometimes

Weekly figures tested would be ok, if we get positives, total # people tested etc

(Chief Exec NHS England) Sir Simon Stevens:

https://www.hsj.co.uk/workforce/stevens-nhs-must-gear-up-for-huge-vaccination-programmes/7027972.article

"The health service needs to prepare for its largest-ever flu immunisation season,
and potentially to run a covid-19 vaccination programme in the autumn/winter "

https://www.dailymail.co.uk/news/article-8491351/NHS-chief-Sir-Simon-Stevens-says-hospitals-preparing-second-wave-Covid-19.html

'NHS chiefs feared in March that their hospitals would be overwhelmed with ...
two million coronaviruss^ patients,
660,000 of whom could have needed intensive care.

Scenes of patients in corridors in hospitals in northern Italy petrified officials in Britain who threw all their efforts into efforts to protect the NHS.

In reality far fewer Covid-19 patients needed hospital care
- a total of 128,737 people have been admitted to wards with the disease so far,
with the majority recovering.'

@BigChocFrenzy

Royal Society DELVE: Nosocomial (Hospital Acquired) COVID infection

rs-delve.github.io/reports/2020/07/06/nosocomial-scoping-report.html

•	Using publicly available data, we estimate that <span class="italic">at least 10% (95% confidence interval: 4-15%) of all COVID-19 infections in England were among patient-facing healthcare workers and resident-facing social care workers during the period from 26</span><span class="italic">th</span><span class="italic"> April to 7</span><span class="italic">th</span><span class="italic"> June 2020.</span>

An estimated further 1% of infections in this period were acquired by inpatients in hospital,
with an additional 6% of all infections among care home residents.
This does not consider secondary cases arising from hospital acquired infections elsewhere.

•	Although surveillance systems and large-scale hospital-based studies have recently been set up, there remain gaps in availability of surveillance data on hospital-acquired infections, particularly of healthcare workers (including agency staff) and in nursing homes; 

important questions remain unanswered, including about the impact on Black, Asian and minority ethnic (BAME) health and social care workers.

•	At present, there is no single organisation with clear oversight of hospital-related infection surveillance, monitoring and response. 

Data obtained from surveillance, monitoring and outbreak investigations need to feed into epidemiological and modelling research, including to evaluate interventions. 

There remain significant opportunities to link epidemiological and phylogenetic datasets to improve the ability to identify and respond to changing epidemic dynamics.

Wow. I am so surprised by that... Only 1% of cases hospital acquired? I thought it would be much higher than that

Patricia, yes I think they have to include it in the test and trace report, as how else do they come to a target number of people each week need to trace! As I read their notes, they do seem to be committing to weekly reporting of the figures.

"The health service needs to prepare for its largest-ever flu immunisation season,
and potentially to run a covid-19 vaccination programme in the autumn/winter "

When do they make the 2020 flu vaccine? I know they do a sort of a guess of what strains will be around on the winter and sometimes the vaccine is better than others.

How much depends on an accurate guess this winter? A bad flu vaccine combined with Covid could overwhelm health systems... or does the risk overlap so much the additional impact might be negligible?

Wow. I am so surprised by that... Only 1% of cases hospital acquired? I thought it would be much higher than that

Yes - They say "at least" - and go on to say -

• "Patient infections. While the identities (and thus socio-demographics and comorbidities) of individuals testing positive are often captured, the time and place of onset (and thus possible transmission source) is not documented in a consistent or coordinated manner."

So I think it's more a case of really not having the data to be able to put a better figure on it. The paper seems to me to be more aimed at highlighting this knowledge gap and emphasising the various ways transmission will occur in hospitals than really trying to establish a number.

@AprilLady

Patricia, yes I think they have to include it in the test and trace report, as how else do they come to a target number of people each week need to trace! As I read their notes, they do seem to be committing to weekly reporting of the figures.

You don't need numbers of people tested, you just need the number of positives and their details for test and trace. Though if you have the latter I would imagine you have the former (who knows atm though..)

@PatriciaHolm

Wow. I am so surprised by that... Only 1% of cases hospital acquired? I thought it would be much higher than that

Yes - They say "at least" - and go on to say -

• "Patient infections. While the identities (and thus socio-demographics and comorbidities) of individuals testing positive are often captured, the time and place of onset (and thus possible transmission source) is not documented in a consistent or coordinated manner."

So I think it's more a case of really not having the data to be able to put a better figure on it. The paper seems to me to be more aimed at highlighting this knowledge gap and emphasising the various ways transmission will occur in hospitals than really trying to establish a number.

Ah thank you. Yes 1% seems too low to be accurate.

@Nihiloxica

*"The health service needs to prepare for its largest-ever flu immunisation season, and potentially to run a covid-19 vaccination programme in the autumn/winter "*

When do they make the 2020 flu vaccine? I know they do a sort of a guess of what strains will be around on the winter and sometimes the vaccine is better than others.

How much depends on an accurate guess this winter? A bad flu vaccine combined with Covid could overwhelm health systems... or does the risk overlap so much the additional impact might be negligible?

www.ema.europa.eu/en/news/update-eu-recommendations-20202021-seasonal-flu-vaccine-composition

This suggests initial strain recommendations were April 1, updated mid May, and applications to change should be made by June 15.

86% of flu deaths were in the over 65s last flu season. I would imagine there is quite some overlap in susceptibility.