Thread by @andrewflood

Todays 917 #COVID19Ireland hospital cases are 140% last Saturday but 2nd day of a row of falling numbers. I don't want to prematurely call that a peak, it will rise tomorrow, but it is very good news indeed which I'll explain in detail /1

Looking at last years hospital graph makes it clear we are on a very different trajectory. Today had 1153 in hospital and was the first 1/3 of several days of rapidly rising hospital cases as a consequence of xmas intergenerational mixing - we are not seeing that /2

We have seen a period of rapidly rising new hospital cases but they are no longer doubling despite cases more than doubling in the week 6 days back, 1007 new hospital cases this week is 136% previous rather than 200%+ /3

There were 140 new hospital cases this morning, of which 73 (52%) had a positive test prior to admission (remember to hard to access PCR testing now). This day last year there were 139 /4

So similar numbers are going to hospital but today there were 124 discharges, last year only 60. Vaccination has mean hospital stays are on average shorter this year (7 days) which is why despite similar numbers of hospitalisations its possible for number in hospital to fall /5

Last week we had more cases than in the entire month on January last year so the other thing vaccines are doing is keeping people out of hospital. New hospital cases this week are 0.89% cases 6 days back - it also appears Omicron is helpfully sending a smaller proportion in /6

Add todays hospital cases as black dot onto my tracker of the NPHET hospital cases Omicron scenarios you can see the temptation to take an optimistic read of this as an earlier & lower than expected peak. Here is where I really wish we had enough testing to track cases still /7

First off as I said up thread expect hospital numbers to go up tomorrow, they do almost every Sunday & Monday because there are fewer discharges at the weekend while admissions continue at almost normal levels /8

The key indicators tomorrow will be the New Hospital Cases as a percent of the previous weeks 136% today so a drop is good, a rise not so good. And the number in hospital as a percent of last Sunday, todays that 140%, again a drop is good, a rise not so. /9

My suspicion is now we are seeing a short term peak as Omicron pushes out Delta completely with its lower rate of hospitalisation but as long as cases are rising we will then see a rise in hospital admissions and hospital numbers resume. /10

So as before cases matter in predicting hospital numbers but now we have no real idea of what case numbers are. NPHET think they may be double whats counted, ie rather that 142k in the week we may have had 280k. And remember they think infections are twice cases /11

We may still be on course for a hospital peak over 1,500 towards the end of this month as shown in the NPHET models. This linked thread explains what I'm doing in the tracker up thread but I'll continue updating in the next days to 'see' where we are /12

https://twitter.com/andrewflood/status/1478019441137008645?s=20

The big unknown is if boosters will be enough to stop a major surge in over 65s. In most waves it takes a while for the virus to get to older groups - so far that hasn’t happened but still could & this age data only runs to 4 days back

https://twitter.com/andrewflood/status/1479164787288383495

83 #Covid19Ireland ICU is 98% last Saturday
ICU this day last year was 107
/10

45 new ICU this week is 110% previous - less frequently severe does not equal 'mild' many of these will be Omicron
/11

New ICU this week is 0.06% cases, 4.64% new hospital cases - this is where the less frequently severe effect of Omicron appears to be showing up as but in part this is people getting infected who would have dodged Delta for now but whose outcome immunity holds up for Omicron /12

Under testing makes cases a bit meaningless as predictors but
26,122 cases is 112% last Saturday
144880 cases this week is 136% previous week
2,362,211 population boostered (12+ 56.5%) 47.2% all pop (+37,733 since yesterday) /13

Another way of looking at this where the waves are synchronised roughly to where NPHET said a variant was dominant - very nicely captures how initially Omicron speed of transmission advantage made it look worst yet but how that now seems to be changing

https://twitter.com/SeanMK1/status/1479760346189545474?s=20

Tracking the actual ICU numbers against scenarios suggests the vaccines have held up against very severe Omicron outcomes more than expected

https://twitter.com/andrewflood/status/1479823977518444545