Thread by @MoriartyLab

Here's an illustration of reporting differences between Ontario and Quebec that mask magnitude of ON deaths in current wave.

First, note that by the end of January, ON's new daily deaths will likely exceed QC's new daily deaths during 1st wave.

But we won't "see" them.

Why?
🧵

This graph shows the expected deaths in each province, based on IHME and ICL estimates of true case numbers.

My method for estimating expected deaths using these case numbers is described in the thread linked here.

https://twitter.com/MoriartyLab/status/1482786120417095682?s=20

Yes, the estimates take into account previous C19 deaths in each region, age-specific vax rates, reduced Omicron severity, changing protection vs infection and severe outcomes from boosters. It's all in the linked thread, as are estimates of total future deaths we may see.

For Canada as a whole this method works exceptionally well to predict the 7 day average actual reported deaths we're seeing daily.

See yesterday's update for Canada here:

https://twitter.com/MoriartyLab/status/1483644969244139521?s=20

In the graph for Canada, you see see 3 things:

1. Expected actual deaths
2. Expected reported deaths
3. Reported deaths

Expected reported is the number of actual expected deaths divided by the global under-detection/under-reporting rate for COVID-19 deaths in Canada.

In Canada, about 60% of actual COVID-19 deaths are detected/reported.

The expected new daily deaths are shown with the green dashed line.

60% of those deaths is the black dashed line.

Pink is the actual REPORTED deaths. It matches the black line almost perfectly.

Right now, I'm watching what expected and reported deaths look like in QC and ON.

QC typically quickly reports all its COVID-19 deaths.

You can see that here. The light blue dashed line for expected QC deaths perfectly lines up with reported QC deaths (solid darker blue line).

Ontario should be reporting very similar new daily deaths as Quebec right now (compare dashed light blue and dashed orange lines).

But it's not. The deaths Ontario is actually reporting are shown with the red solid line, and the gap is growing.

ON typically is quite slow reporting COVID-19 deaths. They often still show up 6 months after they happen. There are still ON C19 deaths from Wave 1 showing up in Cdn national mortality database that were not reported, or perhaps so late they're buried in deaths for later waves.

Ontario is actually 3rd best in the country for complete/timely death reporting, after Quebec and Manitoba. The other provinces are worse (Atlantic provinces may not be quite as bad, but it's harder to know because death waves 1 and 2 were fairly small.

Again, over and over and over:

What you see in QC is almost certainly currently happening in your province or will happen soon, even if you're not seeing it in reported death numbers.

Expected death waves may be somewhat smaller SK and west b/c so many died in last 6 months.

I'll do similar estimates for other provinces where we have infection number estimates from IHME, but maybe not until the weekend when I have a bit more time.

Finally, it's important to note that expected QC deaths may peak in coming week, but ON deaths will keep rising.

Death estimates for ON, QC are expected to be wobbly over next 2 weeks because of testing problems Dec 18 onward, but so far QC and Canada data still fit closely.

And I'm not sure if "peak" is the right word for expected Quebec deaths in the next week or so. It's probably more accurate to describe it as a plateau.

Finally, for those asking, yes, I can do these estimates for BC, Alberta, Saskatchewan, Manitoba and Nova Scotia. I'll be trying to do them one by one in the evenings after regular work, but may not finish till this weekend. Once it's done, I'll post them all daily.