Thread by @andrewflood

Common Sunday increase in #Covid19Ireland hospital cases due to delayed discharges at weekend with 984 in Hospital 137% last Sunday - that rate has slowed & this day last year there were 1,285 in hospital. Discharges typically catch up Tuesdays so tomorrow may see over 1,000 /1

Today last year there were a similar number of new hospital cases (134) but we were in the middle of a rapid escalation of occupancy due to xmas intergenerational mixing - so far this year is different because of higher discharge rate, probably vaccine driven /2

Again due to vaccination the proportion of cases going to hospital is a tiny fraction of last year. New hospital cases are 0.83% cases in week to 6 days previous - just as well as there were 123k cases detected that week & probably as many again that could not access testing /3

In contrast to same days last year the number of cases going to hospital in a week has slowed. 1022 new hospital cases this week is 126% previous week. These hospital cases would mostly be from infections that took place from 21st Dec to 27th Dec so now include Xmas & Stephens/4

Plotting todays hospital cases as a black dot on my NPHET Omicron tracker we can see that in the last week we shifted from being on the lower end of the worst 3rd to slightly above the middle curves which peak under 1500. I suspect that may turn out too optimistic though /5

I say that because the north has a significant increase in cases in older people late this week. The south has not yet released age data which is as recent but presuming its similar that will increase hospitalisations & length of stay next weekend so may push us well over 1500/6

The second key factor looking at last years hospital numbers is that flattening that takes place in a few days is because of the New Year lockdown. Obviously that doesn't apply this year although there will presumably be some impact of a return to normal mixing /7

This is the big difficulty in knowing what will happen - PCR testing has essentially broken down with very large numbers unable to get timely tests & many giving up. So daily case numbers tell us little apart from how much testing was done (and 52% will be positive) /8

As long as infections increase then hospitalisations will increase 2 weeks later. Cases no longer give an accurate picture of whether infections are increasing or decreasing & won't until positivity comes back down again. We won't know where hospital peak will be until after /9

Thankfully these trends are continuing into ICU occupancy with 83 ICU is 95% last Sunday however deaths in ICU remain the major reason why we don't see a rise week on week /10

The contrast with this day last year is very striking, now was the middle of a consistent period where ICU numbers rose every single day, there were 121 there today last year 146% of the current occupancy /11

44 new ICU admissions this week is 102% previous - new ICU admissions per week have hardly risen at all in the Omicron period which has seen infections at the very least doubling every week /12

New ICU in week is 0.05% cases in week 10 days earlier, 4.52% new hospital cases in week 3 days earlier. Both of these are fractions of the proportions in October before the boosting program reached out but have also quickly halved in the Omicron period /13

Adding todays ICU occupancy as black dot onto my NPHET Omicron ICU demand tracker its now very clear that the guess at effectiveness of boosters & the less frequent severity of Omicron were pessimistic & we are seeing much better outcomes than expected. Preferable to last Jan/14

What exactly the percentage cases changes mean right now is anyone's guess & it will be interesting to see positivity figures when released tomorrow but
21,384 cases today is 125% last Sunday
149,193 cases this week is 132% previous