"Not protected yet" people - what happened to them? And, where are their COVID-19 deaths counted: among vaccinated or among non-vaccinated?
In its Cases following Vaccination table (Table 2), PHAC uses label "Not protected yet" to categorize people who received the vaccine within 14 days of reporting COVID-19 deaths...
This newsletter is in the response to the above raised questions, which were asked during our last week’s “Open Canada Data” seminar.
Remember, we already talk a lot about various “algorithmic biases” that have been embeded in various Government of Canada reports publishing statistics related to C19 pandemic and C19 vaccines, and even submited a"Concern" related to that to the Office of the Auditor General of Canada? It appears that another “algorithmic bias” could have been used to skew the data towards a particular (possibly, a particularly desired) outcome, by excluding C19 deaths of "Not protected yet" people from the consideration. Here we’ll look into how large this bias is.
Surprisingly, up till now in all graphs that we generate from official Canada Data by “Independently Verified Information Machine” on www.IVIM.ca (such used in our communications with Chief Science Advisor of Canada), we simply ignored cases of people categorized by PHAC as "Not protected yet". However this data is available in our Google Spreadsheet and here will look into it closer.
Image below shows these data, with new column added (in blue) in Death statistics, which computes "Not yet protected" / Total metric. The image also shows the plot of weekly observed percentages of “fully-vaccinated” among cases since August, which is when PHAC started publishing these data.
What do we see:
In August-October, the percenage of “not yet protected” people among COVID-19 deaths was of quite large: 5-7%, which is of the same order of magnitude as the percentage of those who are categorized as “Partially vaccinated” and “Fully vaccinated”.
The question remained however:
Whom does the “not yet protected” category include? Does is include only those within 14 days from receiving 1st dose only, or - as in Sweden, and possibly in UK - also those within 14 days from receiving 2nd dose?
Much analysis on the issue of “imiscategorisation of vaccinated deaths soon after injection as unvaccinated” was done using UK and Swedish data
Here is the summary of their main findings. Check orinal articles for more details.
Analysis of Swedish Public Health Agency data
A December 2021 article casted doubt on the vaccine efficacy implied by the UK’s official mortality statistics as they related to vaccination status, raising miscategorisation of vaccinated deaths soon after injection as unvaccinated as a possible significant factor.
Despite repeated FOI requests by several parties, no UK government agency has still released sufficiently granular data broken down into the necessary categories to permit any meaningful analysis of the extent (if any) of this miscategorisation issue. Finally, FOI request to the Swedish Public Health Agency has been successful in obtaining such data and showed the following.
It shows that individuals dying within 2 weeks of vaccination have been classed and counted as unvaccinated.
Of particlar interest and concern was the fact that such classfication was applied to the 14 day period after the second as well as the first dose.
The authors state that “the correct categorization turns the vaccine efficacy calculation totally on its head, suggesting a significantly increased risk of death in the vaccinated compared to the unvaccinated, rather than the vice-versa conclusion the authorities had originally touted”, however also noting that there is no age-breakdown in the data. And we do know that age is very large confounding factor.
The main take-away from their analysis is not that the vaccines are or are not efficacious (vs death), but rather that there has been a systematic miscategorization error which (1) seems likely to have been deliberate and (2) resulted in an extremely misleading picture of what the data suggests.
And they asked:
“How many other countries have played similar tricks with their data?”
and conclude that
Such incidents – which now appear all too common in many countries – are likely to shatter the public’s trust in the institutions upon which we are supposed to rely.
Additional Analysis of Canada Data required.
We plan to extend the analysis presented below to include the effect of excluding "Not protected yet" people on calculation of the Odds (such as those shown in image below and discussed last month).
The main problem in conducting such analysis however is that the definitions of “people with various number of doses” used in different PHAC reports are different.
This is how they are defined in [PHAC0] COVID-19 vaccination in Canada (https://health-infobase.canada.ca/covid-19/vaccination-coverage)
Cumulative number of people who have received at least 1 dose
Cumulative number of partially vaccinated people
Cumulative number of fully vaccinated people
Cumulative number of fully vaccinated people with an additional dose
And this is how they are defined in [PHAC1] COVID-19 Daily Epidemiology Update (https://health-infobase.canada.ca/covid-19/epidemiological-summary-covid-19-cases.html):
Unvaccinated cases: include those who were unvaccinated at the time of their episode date.
Cases not yet protected from vaccination include those whose episode date occurred less than 14 days after their first dose of the vaccine.
Partially vaccinated cases include those whose episode date occurred 14 days or more after their first vaccine dose or less than 14 days after their second dose of the vaccine.
Fully vaccinated cases include those whose episode date occurred 14 days or more after receipt of a second dose in a two-dose series or 14 days or more after receipt of one dose of a one-dose vaccine series,, and, if an additional (i.e., third or booster) dose was received, 0 to <14 days after receipt of the additional dose.
Fully vaccinated cases with an additional dose include those whose episode date occurred 14 days or more following the receipt of at least one additional dose (e.g., third or booster) of a COVID-19 vaccine product, after being fully vaccinated.
So, perhars the best way to match one category to another is what they have done in Sweden - using Freedom of Information (FOI) act, which in Canada is called "Access To Information” (ATI) act, to request such information. Anyone wants to do it?..