Thread by @MoriartyLab

Update on excess death story:

So far, journalists @ericandrewgee of @globeandmail @arthurwhiteLP of @leaderpost have on record confirmation that ON and SK do NOT include probable #COVID19 deaths in reported C19 deaths.

This is counter to guidance of @GovCanHealth and @WHO.

By contrast, QC DOES report probable
C19 deaths in their reported numbers.

This goes at least some of the way toward explaining why excess deaths in ON and SK are greater than reported C19 deaths from March-Nov 14, 2020, and why this is not true in QC.

Today's @ericandrewgee story is below. @arthurwhiteLP story is going to press now.

Looking foward to hearing how BC, AB, MB report C19 deaths.

Looking forward even more to learning why ON, SK, possibly others don't follow @GovCanHealth, @WHO guidance.

https://twitter.com/picardonhealth/status/1379051118295064581?s=20

P.S. thank god for journalists.....

P.P.S. this practice of not including probable cases is common in countries that have had excess deaths during the #COVID19 epidemic greater than reported C19 deaths. There's no reason to think Canada is immune from this type of data massage.

Another common practice in other countries (cough USA) is NOT reporting C19 as cause of death if C19 and another condition are listed.

@GovCanHealth guidance is explicit about NOT doing this.

Would like to have every province on record about whether they follow this guidance.

Here's the @GovCanHealth national #COVID19 case definition; canada.ca/en/public-heal…

From this you can see that if a case clinically compatible with #COVID19 is NOT accompanied by a specific laboratory test, it is considered a probable if IT IS ALSO epi-linked to a lab-confirmed case in a close contact or if it occurred in the setting of a lab-confirmed outbreak.

From these descriptions you can see how some cases and deaths might not fall under probable cases (example when testing/contract tracing have collapsed, which has happened in multiple provinces).

That's if the province even reports the probable cases and deaths.

People need to keep in mind that if test % positivity is more than 2 or 3%, we're likely missing a lot of cases, hence many epi-links, plus likely quite a few deaths.

Also crucial to keep in mind that #COVID19, like many infections, often presents atypically in older adults, and may not be associated with a lot of symptoms initially.

Do you remember how fast C19 spread in LTC? It's partly because many residents didn't look symptomatic.

We caught a lot of C19 cases in LTC because we started testing all residents during an outbreak to catch more cases, or at least this happened in ON--I'm not absolutely sure it did in other provinces. We wouldn't have known about these cases if we hadn't tested...

Now apply this to older individuals in community, where families or doctors might not have recognized C19, and where not all household members might have been tested, especially the frail.

@LauraCRosella and colleagues reported a lot of excess deaths during the ON first wave that occurred in people's homes, not LTC or hospitals.

I know I'm raising this issue over and over and over again, to the point of being obnoxious, but I am so worried about this issue.

I'm worried because in the third wave, it's all about deaths in community, not LTC, and I don't think we're equipped to monitor for this.

I think Canadians think that 70-80% of all our C19 deaths happened in LTC, which is double the average of OECD countries, still.

I think we think there hasn't been a lot of death outside LTC.

I think we have a VERY serious blind spot about this.

I also suspect that some provinces might be vaccinating with greater urgency if they didn't think most of our C19 deaths so far occurred in LTC.

Do you think that some decision makers are hesitating to rein in the 3rd wave b/c they think LTC is protected by vaccination?

Anyway, enough.

I feel like I'm yelling into the void, and I'm grateful that some journalists are covering this topic now and asking the questions that need to be asked.

I could be totally wrong. But I'm quite afraid that I'm not, which is why I keep insisting about this.

Wait. One more thing. Here are excess deaths in various age groups, reported C19 deaths and expected/predicted C19 deaths based on provincial seroprevalance data and current province-specific CFRs.

cont'd below

In many cases where there's a discrepancy b/wn excess and reported C19 deaths, predicted/expected C19 deaths based on seroprevalence match excess deaths better than reported C19 deaths.

https://twitter.com/MoriartyLab/status/1378391231051530245?s=20

@brycehoye @DoubleEmMartin @bkives @__m_pereira any chance you can find out how Manitoba reports C19 cases and deaths? Do they include probable cases/deaths? I spoke with @withfilesfrom about this today. He is starting to inquire.

My hunch is that MB may report in line w/ @GovCanHealth @WHO guidance, which would account for province's higher reported death rate, but it's possible it's even higher than reported. It would be good to know this before the next batch of StatsCan excess death data drop (soon).

To clarify, we're not seeing much of a gap in excess and reported C19 deaths in Manitoba, at least up till Nov 14, which is the last date when we have solid-ish excess death data for most provinces/regions.

@JenLeeCBC @charlesrusnell @carolyndunncbc any chance you can get Alberta on the record explaining whether reported C19 cases and deaths include probable cases/deaths, and whether deaths listing C19 on the certificate are always reported, even if there's a second condition?

@AdrianaBarton @MichelleGhsoub @j_mcelroy @bethanylindsay @AndreaWoo @PennyDaflos any chance you can get BC on record stating if reported C19 cases/deaths include probable cases/deaths & whether death certs listing C19 are always reported, even if there's a second condition?

Finally, another piece of info I'd be interested in. The federal government reports cases & deaths of First Nations people living on reserves, as well as federal inmates, I believe.

Do we know if all provinces include these populations in their reported cases/deaths?

It seems like a really dumb question to ask, but total deaths in Canada (used to calculate excess deaths) are reported by province, and deaths/excess deaths in First Nations and federal penitentiaries are not reported separately.

In provinces where there have been significant numbers of cases and deaths in these populations, it's possible that deaths might show up as excess deaths for a province, but not reported C19 deaths, hence potentially explaining part of the gap between excess and reported deaths.

@Kanhehsiio reports on #COVID19 cases and deaths in First Nations communities. Here's her most recent update:

https://twitter.com/Kanhehsiio/status/1377633333690175488?s=20

It would be good to have provinces on record indicating whether cases and deaths in First Nations communities are included in the numbers they report publicly.

Finally, there's also the possibility that C19 death reporting is just very delayed, whereas death data used for StatsCan excess death calculations is reported by date of death, not date of reporting.

In some provinces, you can see excess deaths rising before reported C19 deaths. That lag could be the reporting delay, and it may mean that reported provincial C19 deaths may eventually catch up, but at this point, that's not looking so likely.

For those who want to look at the data all of this refers to, here's a link to most recent update. Data start on this slide and go down for another 15 or so slides.

https://twitter.com/MoriartyLab/status/1378391098360524800?s=20

Here's the original thread with some more textual explanations. The Apr 3 data include excess death data after excess deaths due to toxic drugs for those 65 years and older. This isn't an age group where toxic drug deaths are prominent.

https://twitter.com/MoriartyLab/status/1372009067766542338?s=20

Signing off for the night. I really, really want to see provinces on the record about their reporting. And I'd really, really like to know how many probable cases and deaths are excluded from reporting. We can't interpret excess death data w/out this information.

It's not like this is an outlandish request. @GovCanHealth and @WHO guidance is to include probable cases and deaths in reporting.

Let's see which provinces actually comply with this guidance, and how many cases and deaths have been left out.

Apparently not signing off. I was reminded of an excellent piece on this by @adamsmiller in June of last year. According to @ericandrewgee reporting today, it looks like QC may have changed its reporting to include probables sometime after June?
cbc.ca/news/health/ca…