Thread by @StroudCorona

Here's our weekly #Covid19 data update for #Stroud and #Gloucestershire.
We now know that the past few weeks have been badly affected by false negative PCR results from the Immensa lab... but have done our best to account for this.
Thread. Read in full: neighbournetworks.uk/2021/10/16/16o…

1/ Here's the clearest chart showing impact of the Immensa testing lab fiasco. This shows the number of people to test positive in the past 7 days. The shaded are covers 8th Sept - 11th October, but data since (to 15th) is still an undercount.

2/ It's worth underlining that 696 people testing positive in a week in Stroud district is highest ever number by a distance *despite* data from 11th and 12th Oct being a massive undercount.
People being re-tested from a whole week into a few days 'over-corrects'... but still

3/ Above chart isn't adjusted for population (because specimen date data is known to be incomplete for recent days, so isn't included in dashboard rates data).
Adjusting for pop'n, we still don't have good data for the county - but you can see the divergence from England rate

4/ From this chart we get an idea which of the districts might have been worst affected by the Immensa-ly dodgy lab results. Forest of Dean maybe sent more tests to a lab in Wales, Tewks - nearest Wolverhampton - has a notably lower rate

https://twitter.com/StroudCorona/status/1449504317757661196?s=20

5/ So far, so annoying in data terms - but obvs much more of a problem in terms of people who were LFD positive/clearly symptomatic being ridiculously told not to isolate on the basis of a false negative PCR from a broken private testing lab. Let's look at some more robust data..

6/ Here's our standard comparison chart. You can see the testing data that's affected by Immensa scandal, but also hospital numbers unaffected by it.
Admissions have been slowly rising - 72 people in week to 10 Oct, similar to 73 in week to 25 Sept, and 76 to week 11th Feb

7/ We may see further hospital admission increases (it takes time for people to get sick) but... it's clear many are quickly discharged.

Number of beds occupied by Covid19 patients is ~flat: 34 on 12 Oct, up from 26 on 5 Oct but similar to 33 on 28 Sep

https://twitter.com/StroudCorona/status/1449506358546554883?s=20

8/ Across the South West, the hospital admission picture is worse - a 27% increase over the week, and unlike across #Gloucestershire, that's on top of a previous increase rather than bounce back after a dip

https://twitter.com/COVID19actuary/status/1448309345104678913

9/ We can maybe get a better idea of true prevalence in the South West (and hence a guide to #Stroud district/#Gloucestershire) through the ONS infection survey. This estimates 2% of people in the South West had infections in the week to 9 Oct.

10/ While the rate and increase are relatively bad for SW compared to all regions of England, they aren't totally out of kilter: East of England, West Mids, North West and Yorkshire and Humber have similar increasing trends and rates (~1.6%-2%)

https://twitter.com/StroudCorona/status/1449507761478307844?s=20

11/ Worth noting that the ONS estimates vary much more by age than by region: 8.1% of those in school years 7-11 may have tested positive in the week ending 9 Oct - vs 3.1% age 2 to school year 6, and 0.6% 70+ for example

https://twitter.com/john_actuary/status/1448971799480487938

12/ Anyway, age breakdown data for Gloucestershire is basically junk thanks to Immensa (wishing I'd waiting rather than wasting the time working out how to make that plot work!)... let's close on that impact... positivity for #Gloucestershire.
Spot the lab issue!

13/ Read more on our website neighbournetworks.uk/2021/10/16/16o…. (includes a little national and international context not covered by this thread)

P.S 14/ please see our page all about the false negative PCR / "I tested positive repeatedly on lateral flows and had symptoms but because of the broken Immensa lab and terrible Test and Trace guidance was told to stop isolating" scandal. More to come!

https://twitter.com/StroudCorona/status/1448377735613714434?s=20

PPS 15/ should've included this for national context

https://twitter.com/BristOliver/status/1449393034790150145?t=6Qkqu7RnbZYqjuKY0shM3A&s=19

PPPS 16/ This thread has had a lot more attention than we normally get and gained us some new followers, so please read some backstory if you're interested

https://twitter.com/StroudCorona/status/1449811907217694725

PPPPPP-Pick up a Penguin PPS /17

Read more and see some more insightful and well visualised data in this thread from one of the people who spotted this problem earliest and helped us understand WTH was going on locally

https://twitter.com/ArtySmokesPS/status/1449699271142424576?t=q4Wq2UqhEW7iuvnin23VKA&s=19