Read on Twitter
Post this lockdown we are going to again be faced with a lobbyist demand for careless rushed opening based this time on the idea that IFR will have fallen because those most at risk have been vaccinated. This sounds sensible but is another recipe for disaster /1
Core problem is that over half those who were hospitalised were under 65 & about 60% of those who went into ICU. The 6k cases a day peak in January exhausted ICU capacity, forcing cancellation of routine surgery & turning operating theatres into surge ICU. /2 ht @1987Andrewk
161 under 65s had died to December, stats on ICU admission to that date suggest that without such access for 385 who needed it 2 to 3 times as many would have died. 2768 needed hospitalisation from those ages, presumably without access many (50%) would also have died. /3
This outcomes by age for cases from when testing was not constrained, July to November allows us to approximate demand if a premature opening up meant large numbers in under 65 age groups became infected. 76k hospitalisation, 7700 ICU, 2,000 plus deaths (next tweet for calc) /4
Assuming rapid 80% infection of the segments of the under 65s populations but no infection of those over 65s we need 76k hospital beds, 7,600 ICUs & even with those get 2,000 deaths. But as we only have 11k bed and 350 surge ICUs deaths can exceed 20k due to denial of care /5
That's a 'not going to be that bad' worst case. However it doesn't factor in that every bed and ICU being used for Covid would mean a lot of deaths due to other causes because those people can't access a hospital bed or ICU. There is no need to do a precise calculation here /6
The earliest significant risk taking by this timetable is July as vaccinating everyone over 55 reduces these totals by at least 1/3 & presuming some level of protection against infection R at that point should be reducing. But still need under 100 cases a day /7
But really the hospital situation only becomes anyway realistic for a back to normal once you have everyone over 19 vaccinated. The coping level is pretty much the hoped for herd immunity level. gsd
All approximations above can be argued about which will push number up or down 10 or 30%. However the outcomes of a 'everyone over 65 is vaccinated so its grand' projection are bad enough that enhanced precision wouldn't add anything. The idea is a non runner /8
The main takeaway from this thread is that any schemes to experiment with opening up on the basis of vulnerable groups having been vaccinated are high risk & so should only happen if cases are very low AND there is agreement that as soon as they start to rise things shut down /9
I'm not the only one fearing this outcome and projecting terrible consequences. And it has to be said Germany has about 6 times our ICUs per head of population https://twitter.com/newschambers/status/1352941419934191616?s=20
We are have a couple of unfortunate examples of what happens when ICU fails for technical reasons that confirm why itâs essential capacity is maintained - Egypt: Entire ICU ward dies after oxygen supply fails â Middle East Monitor https://www.middleeastmonitor.com/20210104-egypt-entire-icu-ward-dies-after-oxygen-supply-fails/
The cabinet appear to be aware of that risk but don't want it communicated clearly to the public going by the accidental brief release & then withdrawal of slides from a modelling presentation to the sub committee 
https://twitter.com/andrewflood/status/1362804489485512706?s=20
https://twitter.com/andrewflood/status/1362804489485512706?s=20
SAGE in Britain have modeled the same danger of additional waves in Britain if lockdown is relaxed early there before almost everyone is vaccinated https://twitter.com/paulwaugh/status/1363895910145658885
Ronan Glynn laid out the premature opening due to vaccination trap is next above at tonight NPHET press conference. Otherwise tonight was quite upbeat - we do have a clear path out now if we can avoid the last pitfalls
NPHET advise to government letter of Feb 18th carried a similar if much more low key warning to that in this thread. I hope its not so low key that the usual 'open it up' pundits will not once more get away with a 'nobody warned us' if things go wrong https://twitter.com/andrewflood/status/1364986580134465543?s=20
The HPSC two week report gives the breakdown for hospitalization & ICU admissions for cases in the previous 14 days. This underestimates severe consequences as particularly for ICU admission may take place a while after the case is detected but gives some idea https://twitter.com/andrewflood/status/1366335329997848576
This ISAG visualization of whatâs actually happened is useful as if there was no ICU the second ICU bar also becomes a deaths bar as does a fraction ( Iâve no idea how much) if the hospital bar. Interested in what that fraction might be & the additional survivor outcomes
New HPSC 14 day table out so I did similar calculations for it and as part of that did a more precise estimate of the actual age structure of the population in the 2nd tweet. https://twitter.com/andrewflood/status/1371435418160365568?s=20
New 14 day HPSC table this morning allowing doing the calculation again. Virtually same result of up to 130k hospitalised if a new out of control wave was allowed to happen after everyone over 65 vaccinated (but 6% of under 65s are D1 vaccinated so -8k) https://twitter.com/andrewflood/status/1372490020880326657?s=20
How much will vaccination slow spread? Israeli data suggests Pfizer 94% effective which for B117 that means R goes under 1 if *fully open* at 83% vaccination. Lower levels of vaccination allow only small R increases
R 1.6 Vc is 40%
R 1.2 Vc is 18% https://twitter.com/andrewflood/status/1369991212405522432?s=20
R 1.6 Vc is 40%
R 1.2 Vc is 18% https://twitter.com/andrewflood/status/1369991212405522432?s=20
Not factored in is those who have protection because they survived Covid19 - Danish study suggests thats about 80% protection against symptomatic infection & working backwards from our observed CFR against expected IFR maybe 500k have that 80% protection https://twitter.com/andrewflood/status/1369609146446778370?s=20
That 80% isn't great, it would put immunity through infection for B117 at a level requiring 97% of the population to have survived infection
Vcâ=â(1â 1/4.5â)/80%=97% https://twitter.com/GidMK/status/1372351899593109504?s=20
Vcâ=â(1â 1/4.5â)/80%=97% https://twitter.com/GidMK/status/1372351899593109504?s=20
I've had a few objections from people who thinks its not possible for lots of people to get infected fast. We have an example of exactly that happening here in Ireland, in January Belmullet very rapidly went to 15%+ of the populating being infected from zero most weeks before
The 94% protection against infection finding in Israel was great but this study from the US finds a much lower 80% protection - similar to the protection obtained through survival. So that moves herd immunity into the high 90%s https://pubmed.ncbi.nlm.nih.gov/33704435/
NPHET provided some modelling today that confirms much of what this thread has been about, including these 4 slides that have a strong resemblance to the accidentally published 'vulnerable period ahead' ones I linked to a few tweets up thread
Running the same calculation again from latest HPSC outcome table for 14 day period to 19-4-21 doesn't yet show vaccination reducing the total number of hospitalisations that could be expected if we allow a 4th wave. There is probably a lag of 4 weeks in this measure though /
I've taken to updating NPHETS 'vulnerable period ahead' graph of March 31st with the latest cases each day (the dots are a reasonable approximation using photoshop). There is an issue that vaccine administration needs to be caught up on in the next 10 days for it to hold/
The original of the 'vulnerable period ahead' slide is on the right, NPHET Mar31 - although its more about the threshold level for vaccination to start significantly helping case reduction its also the rollout point around which high risk should have got at least dose 1 /
If we get through the next 5 weeks below the line then the threat of the health service being overwhelmed recedes considerably - it doesn't vanish but we would need to allow a hell of a 4th wave from that point on via premature opening of indoor hospitality or similar /
Plague Tapes 91 looks at plans being floated by politicians & lobbyists to reopen pubs long before everyone is vaccinated. Give it a rest lads, no one wants to see a last minute 4th surge for the sake of drinking outdoors another week or two. https://radioactiveinternational.org/gotw-pt91/
Most recent HPSC outcome table appears to now show start of vaccination protection for older age groups appearing in reduced hospitalisation ratio. 75-84 and 85+ age groups with Covid19 in last 14 day period hospitalised at about half the previous percentage /
Seychelles being forced to reintroduce some measures even at a 62% vaccination rate after opening too much too soon - with over 1% currently infected this is similar two the levels feared here if things went wrong https://www.bloomberg.com/news/articles/2021-05-04/world-s-most-vaccinated-nation-reintroduces-curbs-as-cases-surge
Calculations off latest HPSC 14 day outcomes table has worst case unchecked pandemic hospitalisations down to 68% of the pre-vaccination period - still 170k but this doesn't account for reduced risk of infection of eldest groups so arguably 135k closer /
