Thread by @MoriartyLab

Continuing on the topic of hospitalization and ICU admission reporting across the country....

Yesterday I was talking about under-reporting/delayed reporting of hospitalization and ICU admission data in ON and AB during the 2nd wave, compared to QC.

Today: Omicron reporting

🧵

https://twitter.com/MoriartyLab/status/1476745626784309249

So, based on the thread yesterday, you can see that most of Canada likely has a similar Omicron burden--certainly everywhere except maybe PEI and NL are similar. Probably minimum 2% of the Canadian population infected from Nov 15-yesterday.

I'll get to the hospitalization reporting oddities shortly, but first I want to emphasize 2 things that maybe weren't clear in yesterday's thread.

1. Just because most of Canada is about the same as Quebec for Omicron burden, I am NOT saying QC is doing well, by any stretch. What I'm saying is that the ROC needs to get its head out of its collective ass and recognize that we're NOT different than QC (read NOT better).

2. I want people to understand just how horrifyingly fast exponential spread is. In yesterday's thread I explained why a MINIMUM of 2% of the Canadian population has likely been infected from Nov 15-Dec 30.

Here's an illustration of how fast this wave is moving.

a) Take est. current Omicron doubling time from Dec 30 @COVIDSciOntario dashboard (5.4 days, probably slow b/c testing is maxxed out).

b) Estimate % Canadian pop that will be infected in coming month.

100% by Feb 1.

@COVIDSciOntario It's unlikely we'll reach 100%. I hope. For example, yesterday's @COVIDSciOntario dashboard estimates double-vaxxed people have about 20% protection from any infection.

And there may be further slowing, but not based on current restrictions in most Canadian regions.

@COVIDSciOntario The South African Omicron wave just peaked yesterday. Their wave was fairly advanced when variant was 1st identified. If it slows at same rate that it grew, it may be done by end Feb.

If Canada is similar, we might see peak in mid-late Jan followed by slowing till mid-late Mar.

@COVIDSciOntario We don't really know how this wave will play out in Canada--so many variables are different than South Africa, UK and other regions. But it's clear that a lot of us will be infected very, very quickly.

@COVIDSciOntario The @GovCanHealth epidemiology update will be updated daily at 9 a.m. starting Jan 4, and it's critical to watch the hospitalization, ICU admission and death data for the country there.

health-infobase.canada.ca/covid-19/epide…

@COVIDSciOntario @GovCanHealth To understand speed and completeness of nationally reported data during Omicron, I've plotted cumulative hospitalizations, ICU admissions and deaths in each age group from Nov 15-Dec 30.

Shown here: All ages, 0-19 and 20-29.

Comparisons: Canada to Quebec; also Ontario all ages

@COVIDSciOntario @GovCanHealth Here are the comparisons for 40-59, 60-79 and 80+.

I can't do these comparisons for Ontario because Ontario does not report cumulative hospitalizations and ICU admissions by age groups in its daily epi reports.

@COVIDSciOntario @GovCanHealth It's important to break this down by age group, where possible, because most cases until recently were in people 40 and younger, although this is changing fast. So you might not see signs of exponential increase in older age groups, or not as much, or not until more recently.

@COVIDSciOntario @GovCanHealth These are CUMULATIVE severe outcomes (hospitalizations, ICU admissions, deaths) from Nov 15-Dec 30.

Hospitalizations typically lag cases by about 1 week, ICU by about 2 weeks, deaths by about 4 weeks.

@COVIDSciOntario @GovCanHealth We should certainly be starting to see visibly exponential growth in hospitalizations by mid-late Dec on these graphs. This is visible as a steeper and steeper slope each week. A steady increase means linear (non-exponential growth).

@COVIDSciOntario @GovCanHealth You can readily see evidence of exponential growth of hospitalizations in QC after Dec 10 for the ALL AGES graph (middle row: left), but you don't see it at all in the ON ALL AGES graph (3rd row), and it's not starting till after Dec 24 in the Canada ALL AGES graph (top left).

@COVIDSciOntario @GovCanHealth The Ontario ALL AGES data are also weird because you'd expect ICU admissions to be ~2.5X higher than deaths (like QC), but they're less than 50% higher.

ICU admissions are 4X higher than deaths in QC, but they're only 1.3X higher for ON.

Canadian ALL AGES ICU admissions are about 1.5X higher than deaths, likely reflecting a big impact of the Ontario data, but also possibly reflecting slow/incomplete ICU reporting from other provinces outside Quebec.

When we look at different age groups, you can see clear signs of exponential growth in ICU admissions in people younger than 40 in Quebec, although it's not quite as obvious in older age groups yet.

The data I pulled from @INSPQ only go to Dec 29, so the Dec 30 QC point is low.

@INSPQ You can see both ICU admissions and hospitalizations speeding up in people older than 40 speeding up markedly in QC, but less so in the ROC until the last week.

Note: people 80+ typically can't survive ICU itself, so often aren't treated there, in case you're wondering.

@INSPQ Assuming that most of Canada has a similar COVID burden as Quebec, AND seeing the difference in reporting of key hospitalization and ICU admission indicators in and outside QC, it's obvious that QC is providing the most complete/reliable data the rest of us should be watching.

Finally, I've updated my spreadsheet that estimates how many deaths, ICU admissions and hospitalizations we're likely to see if 100% of unvaccinated people and ~80% of double-vaxxed people are infected.

You can check your PHU/region at link here:
docs.google.com/spreadsheets/d…

This spreadsheet now incorporates the current vaccine protection estimates posted daily by @COVIDSciOntario at the dashboard here:
covid19-sciencetable.ca/ontario-dashbo…

It's based on current rates of double vaccination in every age group and Canadian province and public health unit/region for which I could find age-specific vaccination data.

The sources and methods are listed at the bottom of the Google sheet, for those interested.

The age-specific infection fatality, hospitalization and ICU admission rates I use for these estimates are OECD averages for the original (wild-type) SARS-CoV-2 strain. All Omicron severity estimates to date have found that Omicron is more severe/similar to original strain.

This spreadsheet is updated daily with the most recent estimates for vaccine protection against severe outcomes and all infections, based on @COVIDSciOntario updates, so it reflects our best current Canadian estimates.

@COVIDSciOntario Finally, the spreadsheet also shows how many influenza deaths would be expected annually in each Canadian region, based on the average 2010-2019 annual mortality rate/100,000 people from the U.S. CDC.

It also shows expected number of deaths caused by vaccines (<1 in Canada FYI).

@COVIDSciOntario As of Dec 30,

🔵COVID-19 vaccines have saved ~474K Canadian lives
🔵If 100% of unvaxxed people and 80% of vaxxed people are infected, Canada will likely still see 340K NEW hospitalizations, 56K ICU admissions, 71K deaths.

Expected flu deaths: 4K
Expected vaccination deaths: <1

@COVIDSciOntario It is also fairly likely that QC will report all/most severe outcomes, unless its healthcare system collapses, and that hospitalizations, ICU admissions and deaths will be under-detected/under-reported by 50% outside Quebec, based on trends throughout the epidemic.

For those who find the spreadsheet overwhelming, here are the estimates for different provinces:

ALBERTA: 36,000 NEW hospitalizations, 6,500 ICU admissions, 6,500 deaths

BRITISH COLUMBIA: 49,000 NEW hospitalizations, 8,600 ICU admissions, 9,700 deaths

MANITOBA: 10,000 NEW hospitalizations, 1,500 ICU admissions, 1,800 deaths

NEW BRUNSWICK: 9,000 NEW hospitalizations, 1,500 ICU admissions, 2,100 deaths

NEWFOUNDLAND LABRADOR: 5,000 NEW hospitalizations, 700 ICU admissions, 1,100 deaths

NORTHWEST TERRITORIES: 30 NEW hospitalizations, 60 ICU admissions, 30 deaths

NOVA SCOTIA: 9,400 NEW hospitalizations, 1,500 ICU admissions, 2,100 deaths

NUNAVUT: 200 NEW hospitalizations, 30 ICU admissions, 30 deaths

ONTARIO: 137,000 NEW hospitalizations, 23,000 ICU admissions, 31,000 deaths

PRINCE EDWARD ISLAND: 1,300 NEW hospitalizations, 160 ICU admissions, 260 deaths

QUEBEC: 74,000 NEW hospitalizations, 11,000 ICU admissions, 14,000 deaths

SASKATCHEWAN: 10,000 NEW hospitalizations, 1,700 ICU admissions, 1,800 deaths

YUKON: 300 NEW hospitalizations, 60 ICU admissions, 40 deaths

Eliminate all non-essential indoor contacts with people outside your household, wear the best masks you can get, get vaccinated/boosted.

With that, I won't bombard you anymore tonight, will go have New Year's Eve dinner with my long-suffering spouse, and try to stop yelling at the top of my lungs on Twitter.

Love to all of you.

May 2022 be better for all of us.

And may Canadian leaders finally lead.