On Friday evenings I'm going to release a thread like this on English covid data that I think is useful / reliable. If you don't want to follow the rest of my ranting, the # will be #ahcveng
• Cases: pillar 1
• Care home outbreaks
• Hospital deaths & admissions
• Sundry
1. English pillar 1 (clinical need & NHS) cases and % positive
1a: why not pillar 2? Because I regard the data as corrupted by door-to-door testing of asymptomatic cases in hotspot areas, and testing of children / students with mild symptoms which will 🔼% positive while 🔽 usefulness (more non-infectious high Ct "weak" positives).
2. Care home outbreaks
Compiled from weekly surveillance reports (recording protocol was changed in week 30 report)
This section will be different every week. This week:
4.1 Spanish '2nd wave': cases (dark green), deaths (red) and the localised outbreak cases in Madrid & Catalunya (blue)
4.2 Compare the English and Spanish death curves above to the projections in a fascinating paper by @mgmgomes1 team, in July. They projected 2nd waves; time will tell.
This is normally extremely useful data, but was always likely to be thrown out by an abnormal no. of calls for 0-18 year olds before schools went back, and 18-21 year olds before uni / college. >70 (RH axis) is the vulnerable group.
This is Toby Young (@toadmeister)'s response to my piece in the @spectator, which he very fairly follows by a frontline account from an in-house senior doctor in London.
I am still a lockdown sceptic, and Toby sets out very well the principled objections to lockdown that I still believe too. 2/n
I still think the 1st lockdown went on too long, with harms likely to have outweighed the benefits. TikTok punting videos while people missed basic healthcare is revolting.
I don't think the 2nd lockdown was well justified: the arguments that convince me now didn't then. 3/n
ONS: Incidence now up to 35k per day, 0.9% overall; huge variance regionally and by age group, with young getting infected *much* faster than old (good news), and this is backed up by Zoe (KCL/CSS) data: (
My experience of contact tracing.
---
My daughter was sent home from school to isolate Tuesday morning, following a positive test by her teacher (who had been isolating since Friday).
She rapidly developed (moderate) fever and kept falling asleep (fine now), so was tested /1
This was 2.30pm Tuesday.
Result came through 10pm Thursday; within half an hour we had phoned all contacts (piano teacher, school, houseguests, other direct contacts).
Contact tracers phoned us 17 hours later (Friday afternoon); we gave them all the same information. /2
I asked what the point of contacting people we had already contacted; they confessed "basically... not much".
Since then they have repeatedly been contacting us (in isolation) by phone and text to confirm our details, even though I've told them we're isolating. /3
Note sharp age distinction in ONS / REACT1; if you favour the "GB Declaration" approach (see below) you would want the difference between young and old incidence and trajectory as great as possible
The headline is that during the most recent week (18 to 24 September) there were around 8,400 new infections per day not including those living in institutional settings, *down* from 9,600 previous week.
We also had the Imperial "REACT1" interim incidence report this week which also pointed to a slowdown, and the KCL Zoe app (tracking symptomatic cases) points in the same direction