The other Omicron mild or wild experiment.

Well Hong Kong did hold things back till Omicron came along.
Vaccination was a bit more than 60% fully vax 2 doses at start of wave.
Though unvaccinated skewed to older population, maybe only 25% over 80's.
Boosters for everyone a bit over 10%, minimal for older.

Then the vaccine being initially offered being Sinovac, western vaccines only later.
Even Singapore required Sinovaxed to get western vax as their booster for valid vaccination certificate.

Antivax activists also capitalized on Sinovac reputation unfortunately with variety of messaging.
Plus politics, imagine you swam the pearl river risking death to escape , and ...

True during wave vaccination rate got much better and better vaccines.

Approx. numbers... What I can find YMMV
Hong Kong 7.15 million population , 9,361 Deaths, pretty much all but 200 in the Omicron wave.

The control measures were pretty drastic too.
One person in your building gets positive test you all have to get tests and are isolated etc. 

https://inews.co.uk/news/science/how-hong-kong-went-from-zero-covid-success-story-to-the-worlds-worst-omicron-wave-1517171

To reinforce the fact that covid is now significantly impacting our death rates, the plot below shows numbers I ran a few days ago comparing our raw reported covid death numbers with 'excess deaths' in NZ this year (death stats ex Stats NZ).

Still early days.  March this year the first month with large numbers of covid deaths and the last month for which monthly death stats have been reported.  Hence only month where a meaningful comparison is available.   That nevertheless does indicate:

• Apparent close alignment between NZ 'excess deaths' and reported raw covid death numbers in March
• After two years of being one of the few countries with negative 'excess deaths', the situation has now reversed.  Now ~10% 'excess' deaths in March and could rise to ~15% for April (?) based on the reported covid death numbers
• Our World in Data shows a peak weekly excess of 19% in Mar for NZ, equal to the maximum peak experienced in Finland at any stage during the pandemic, and higher than the max peaks in Denmark, HK, Taiwan & Japan.  Only exceeded in AU during their January peak earlier this year (26%)

Unfortunately will be another month before the monthly death stats for April are released.   Will be interesting to see how the covid vs 'excess' deaths trend unfolds over the next few months

Tech notes!

'Excess' deaths shown here = total reported deaths per month minus expected. 

The expected monthly values were calculated from reported (Stats NZ) monthly deaths for the period 2010 - 2019 using a linear forecast function.  Deaths per year show a reasonably strong linear trend over that period (R^2 = ~0.9).

Possibly slightly overestimates the 'expected' values though due to reduced recent net population growth?  If so, excess deaths may be slightly underestimated.

Data for 2020 & 2021 excluded due to Covid impacts.

Omicron deaths estimate for May extrapolated from 1-15 May data.

Re: #2914666: DS248: NZ covid deaths more than double the rate in AU as a proportion of reported cases (2.0 deaths per 1000 cases vs 0.8 in AU)

@Batman: possibly due to the way deaths are classified? 

Possibly, but also possibly not.  As I understand it NZ now operates per WHO methodology for death numbers for consistency with other (conforming!) countries.

But ...

Perhaps the ICU numbers below are part of the story? 

Covid patients in ICU (based on ICU and case averages for the 7 days to 17 May):   

• AU = 4.96 per million population, 2.45 per 1000 cases
• NZ = 1.64 per million population, 1.05 per 1000 cases

That is, AU has 3x as many in ICU per capita and 2.3x as many per 1000 cases

Plus AU an earlier adopter of the new antivirals?  Be interesting to know the real relative use of these in AU and NZ.

Better medical care?

Below - a telling plot?  Well plots plural (late decision to add NSW plot).  Especially for anyone over 70!
(plotted a couple of weeks ago so not fully up to date)

Plot below compares ICU:Hospitalisation and Deaths:Hospitalisation ratios.

Of course we are well aware of the strong age bias in Covid death rates.

But to be honest I was not aware of the severity of ICU triaging of over-70's in NZ.  The very groups with much higher than average covid death rates.

Time for our news organisations to dig deeper into our current Covid death rates instead of accepting the spin that we are doing fine?

Might explain why our ICU capacity does not appear to have been stretched, despite initial concerns?

Raw ICU and death counts ex NZ MOH website. Data is for all NZ Covid ICU and hospital admissions to 24 Apr 22 but is predominantly Omicron related.

Of course there are 'similar' trends elsewhere including in AU.  Unfortunately hard to find directly comparable data.

From what I could tell from very limited NSW data, the trend was similar except 60-79 groups appeared not so heavily triaged.  Or should that be culled?

Had to extract the NSW data from weekly pdf reports, so the data I had only covered two weeks (plot below to same vertical scale).

Indeed, re-looking at this plot, there is much less triaging (culling?) of the 60-69, and 70-79 age groups, or even the 80-89 age group - and their deaths:hospt. ratios are also significantly lower.

Eg. for 70-79  (for some reason an age group of personal interest 😀)

• NZ ICU:hospt. ratio    =   1.6%
• NSW ICU:hospt. ratio = 15.7% (albeit based on only 2 weeks data)

That is a factor of 10 difference!  

And not just 70-79.  Their ICU:hospt. ratios are considerably higher across all age groups except 90+

Time to start asking hard questions.

DS248:

 

 

Raw ICU and death counts ex NZ MOH website. Data is for all NZ Covid ICU and hospital admissions to 24 Apr 22 but is predominantly Omicron related.

That chart is a real shocker!

DS248:

 

Re: #2914666: DS248: NZ covid deaths more than double the rate in AU as a proportion of reported cases (2.0 deaths per 1000 cases vs 0.8 in AU)

 

@Batman: possibly due to the way deaths are classified? 

 

Possibly, but also possibly not.  As I understand it NZ now operates per WHO methodology for death numbers for consistency with other (conforming!) countries.

 

But ...

 

Perhaps the ICU numbers below are part of the story? 

 

Covid patients in ICU (based on ICU and case averages for the 7 days to 17 May):   

 

• AU = 4.96 per million population, 2.45 per 1000 cases
• NZ = 1.64 per million population, 1.05 per 1000 cases

That is, AU has 3x as many in ICU per capita and 2.3x as many per 1000 cases

 

Plus AU an earlier adopter of the new antivirals?  Be interesting to know the real relative use of these in AU and NZ.

 

Better medical care?

 

 

 

Below - a telling plot?  Well plots plural (late decision to add NSW plot).  Especially for anyone over 70!
(plotted a couple of weeks ago so not fully up to date)

 

Plot below compares ICU:Hospitalisation and Deaths:Hospitalisation ratios.

 

Of course we are well aware of the strong age bias in Covid death rates.

 

But to be honest I was not aware of the severity of ICU triaging of over-70's in NZ.  The very groups with much higher than average covid death rates.

 

Time for our news organisations to dig deeper into our current Covid death rates instead of accepting the spin that we are doing fine?

 

Might explain why our ICU capacity does not appear to have been stretched, despite initial concerns?

 

 

 

 

Raw ICU and death counts ex NZ MOH website. Data is for all NZ Covid ICU and hospital admissions to 24 Apr 22 but is predominantly Omicron related.

 

 

 

Of course there are 'similar' trends elsewhere including in AU.  Unfortunately hard to find directly comparable data.

 

From what I could tell from very limited NSW data, the trend was similar except 60-79 groups appeared not so heavily triaged.  Or should that be culled?

 

Had to extract the NSW data from weekly pdf reports, so the data I had only covered two weeks (plot below to same vertical scale).

 

 

 

 

Indeed, re-looking at this plot, there is much less triaging (culling?) of the 60-69, and 70-79 age groups, or even the 80-89 age group - and their deaths:hospt. ratios are also significantly lower.

 

Eg. for 70-79  (for some reason an age group of personal interest 😀)

 

• NZ ICU:hospt. ratio    =   1.6%
• NSW ICU:hospt. ratio = 15.7% (albeit based on only 2 weeks data)

That is a factor of 10 difference!  

 

And not just 70-79.  Their ICU:hospt. ratios are considerably higher across all age groups except 90+

 

Time to start asking hard questions.

 

 

 

 

 

 

 

 

 

deserves an article in the herald or a 15 min slot in 6pm news

There was a decent article on this at the Beginning of April in Stuff, 

"Professor Tony Blakely, an epidemiologist at the University of Melbourne, described the way Covid affects the elderly as a “giant ski jump” of death.
The ski jump refers to the very steep gradient of Covid's fatality risk with age, with the risk roughly doubling for every extra five to eight years of age.

........

Experts say the low ICU numbers actually give a clue as to why the case fatality rate in the 90+ population is so high.

Many of the 88 per cent of deaths not admitted to ICU are likely to be in aged care facilities, where residents are already receiving palliative care for other ailments, Blakely says.

Roughly a third of the 30,000 deaths each year in New Zealand occured in aged care facilities already, even before Covid.

https://www.stuff.co.nz/national/explained/300555559/covid19-why-elderly-are-dying-with-omicron-at-disproportionately-high-rates

wellygary:

 

There was a decent article on this at the Beginning of April in Stuff, 

 

"Professor Tony Blakely, an epidemiologist at the University of Melbourne, described the way Covid affects the elderly as a “giant ski jump” of death.
The ski jump refers to the very steep gradient of Covid's fatality risk with age, with the risk roughly doubling for every extra five to eight years of age.

 

........

 

Experts say the low ICU numbers actually give a clue as to why the case fatality rate in the 90+ population is so high.

 

Many of the 88 per cent of deaths not admitted to ICU are likely to be in aged care facilities, where residents are already receiving palliative care for other ailments, Blakely says.

 

Roughly a third of the 30,000 deaths each year in New Zealand occured in aged care facilities already, even before Covid.

 

https://www.stuff.co.nz/national/explained/300555559/covid19-why-elderly-are-dying-with-omicron-at-disproportionately-high-rates

 

Doesn't explain the wide gulf in ICU admissions between NZ and NSW for people under 90

Nor the (resulting?) large discrepancy between the death rates (either per hospitalised case, or per notified case)

Revision/correction of my earlier plot in #2915803  (2nd to top post on this page).

First up apologies - found an error in the forecast formula due to not properly anchoring the year field 🙁!

However, have also changed the forecast basis.  To reduce scatter, now extrapolate based on year totals, then scale by 2010-2019 monthly averages (rather than extrapolating separately for each month). 

Resulting excess deaths are higher than reported C19 deaths for March but need to see data for another couple of months to assess correlation & scatter between the two curves.  The C19 deaths not looking disproportionate relative to estimated excess deaths at this stage though (& there are reasons why the excess deaths may be higher than the reported C19 deaths, if that turns out to be the case).

Possibly could slightly improve the estimated deaths by smoothing the monthly averages.  However, the expected values are but forecasts so uncertainty will remain regardless.

Press release:

 

With New Zealand expecting to see Omicron cases rise during the winter, the Orange setting remains appropriate for managing this stage of the outbreak, COVID-19 Response Minister Chris Hipkins said today.

 

“While daily cases numbers have flattened nationally, they are again beginning to increase in the Northern region and hospitalisation rates have also increased slightly over the past month. In addition, our latest COVID-19 modelling indicates that under current conditions, there is a likelihood of a secondary wave of cases appearing,” Chris Hipkins said.

 

“The COVID-19 Protection Framework has effectively managed the Omicron outbreak at Orange. By following public health advice to remain at that setting, we can maintain some protections while ensuring businesses can continue to operate.

 

“I urge everyone to continue to be cautious and think about the health of others, especially those who are immunocompromised or at higher risk of long-term health impacts from infection. Please also get boosted, if you haven’t already.

 

“Self-isolation, vaccination and mask wearing continue to be our main defences against COVID-19. People will continue to be required to wear a face mask in many indoor settings.

 

“It is encouraging to see case numbers remain steady, but there are other factors at play that tell us to remain cautious and not yet move to Green. These include the arrival of new strains of colds and flus, which will add to the workloads of our already busy hospitals.

 

“The next review of the traffic light settings will be in late June,” Chris Hipkins said.

 

DS248:Doesn't explain the wide gulf in ICU admissions between NZ and NSW for people under 90

Nor the (resulting?) large discrepancy between the death rates (either per hospitalised case, or per notified case)

the question of the under 80s is one very easy one

did those who die die in ICU or die without ICU?

Batman:

 

DS248:Doesn't explain the wide gulf in ICU admissions between NZ and NSW for people under 90

 

Nor the (resulting?) large discrepancy between the death rates (either per hospitalised case, or per notified case)

 

 

the question of the under 80s is one very easy one

 

did those who die die in ICU or die without ICU? ...

Why the question?   The plot I posted answers that, at least for 60+ age groups.

Indeed, if look at just the last month (black plot below), then the answer for all 40+ age groups is overwhelmingly "without ICU".  That is, far more people died than were ever admitted to ICU.

==

The issues here are:

• Use of ICU in AU much higher than in NZ (factor of ~10 higher in NSW)
• Higher death rate in NZ relative to AU, whether measured per capita, per case, or per hospitalised case.
• And the fact that the powers that be here basically sweep it under the carpet.  I have not seen any evidence of urgency in dealing with the situation.

AU puts more covid cases into ICU and has a lower death rate.  Related?   

==

Plots below are for NZ (month to 24 May) and NSW (4 wks to 7 May, the latest hospital admission data I located for NSW).  Not fully consistent time periods but a ~50% overlap.  In both cases, 7-day average case numbers have been reasonably constant over the last ~5+ weeks.

Note that the ICU:Hospitalisation ratios for mid-age groups are now lower than in my earlier plot, which was based on all hospitalisations since 16 Aug 2021 (ie. now worse than earlier in the Omicron phases).

NZ data from https://www.health.govt.nz/covid-19-novel-coronavirus/covid-19-data-and-statistics/covid-19-case-demographics

NSW data from https://www.health.nsw.gov.au/Infectious/covid-19/Pages/weekly-reports.aspx

Note that the above plots are based on people admitted to hospital and ICU, not the numbers in hospital and ICU each day, as reported in our daily news and media updates.  The latter figures depend on both the numbers admitted and the average length of stay, which for ICU appears to be quite a lot longer in NZ.  Unsurprising given the small numbers admitted to ICU here; ie. the condition of those admitted likely considerably more deteriorated before they are admitted, or possibly only the most at risk (in younger age groups) admitted to ICU 

It may be worth looking in to general ICU utilisation?  It may be that we're running pretty low and that they're triaging Covid 19 patients to allow space for other emergency requirements?

DS248:

 

Batman:

 

the question of the under 80s is one very easy one

 

did those who die die in ICU or die without ICU? ...

 

 

 

Why the question?   The plot I posted answers that, at least for 60+ age groups.

 

Indeed, if look at just the last month (black plot below), then the answer for all 40+ age groups is overwhelmingly "without ICU".  That is, far more people died than were ever admitted to ICU.

 

==

 

while i don't disagree with your statement, i wasn't able to see that from your graphs.

i thought your graphs plotted ratios of x:y and y:z according to the axes labels

Benoire:

 

It may be worth looking in to general ICU utilisation?  It may be that we're running pretty low and that they're triaging Covid 19 patients to allow space for other emergency requirements?

 

RNZ has this diagram but it is based on capacity data from Oct 2021.  It is also clear from the ANZICS link that the capacity estimate is based on a mis mash of data. Also that actual capacity (spare capacity?) varies by the day depending on staffing levels.

Another indicator is perhaps that in March this year there were up to 33 Covid patients in ICU (possibly ICU + HDU?) some days.

As of yesterday (reported today) there were 11 in ICU.  Average for the last two weeks = 10.

From the RNZ article: https://www.rnz.co.nz/news/covid-19/450874/covid-19-data-visualisations-nz-in-numbers

Batman:

 

DS248:

 

...

 

Why the question?   The plot I posted answers that, at least for 60+ age groups.

 

... 

 

while i don't disagree with your statement, i wasn't able to see that from your graphs.

 

i thought your graphs plotted ratios of x:y and y:z according to the axes labels

 

No, the plots show x/z (= ICU admissions/hospital admissions) and y/z (deaths/hospital admissions), both plotted against age group.  Or in the shorthand notation used in the legend 'ICU : Hospt ratio' and 'Deaths : Hospt ratio'.

That is, there is a common denominator (= 'hospital admissions').

In effect one line is the number of ICU admissions and the other the number of deaths, both normalised by the number admitted to hospital.

The reason for normalising the plot is to aid comparison of relative risk between age groups and with NSW.

===

In case still not obvious, in plot version below the brown line is 'Number deceased' and the light blue line, 'Number admitted to ICU'.

Hopefully it is now clear that for people in the 40+ age groups, the numbers of deaths are far larger than numbers the admitted to ICU (ie. deaths and ICU admissions during the month 25 Apr - 24 May 2022).

Hence the large majority that died had clearly never been in ICU