For those that do not know, I track Alberta’s statistics. I am meticulous. I do so because they have the best Canadian data available, and yet it is deeply flawed. Thus far, I have identified dozens of errors in the way they add keep track of the virus. Today, they fixed some of them.
Thank you. I have been proof over the Alberta data as well. Alberta is using he statistics manipulatively to give the general impression covid is much more dangerous than it is, that the vaccine is more effective than it is, and that he vaccine is much less dangerous than it is.
Some notes:
- being overweight or obese is universally recognized as being very strongly correlated to severe covid outcomes and deaths. Alberta does not recognize obesity as a comorbidity. If you added obesity as a comorbidity to Alberta’s data, my guess is that you would find covid hospitalizations and deaths would occur almost-never in people with no comorbidities (inclusive of obesity), especially in younger people.
- Alberta’s adverse event reporting per dose he last time I checked discloses adverse events and serious adverse events following vaccination in similar proportions (relative to the kinds of adverse events). Curiously, Alberta reports lower case rates of serious adverse events (per dose) than the rest of Canada. I do not believe this is a coincidence.
There are many others.
THe most egregious behaviours are the misleading claims, and the general impressions given that the disease is of equal deadliness to every person in Alberta.
If a pharmaceutical company were making such statements, they would be violating many provisions of the federal Food and Drugs Act which prohibits advertising that:
- give a general impression that a disease is more harmful than it is, and that a drug designed to treat or prevent it safe without identifying real risks.
- obscures or misstates the medical risk/benefit of the drug.
- does not differentiate between the risk/benefit by demographic group or patient status.
Because all of these things get in the way of achievement of the goal of informed consent (which was universally demanded by principles of medical ethics - until covid vaccines were authorized on an emergency basis).
Pharmaceutical companies have received multimillion dollar fines for such behaviours. What the Alberts government is doing by definition in this federal statute is in fact contrary to he Food and Drugs Act by definition inasmuch as such behaviours are considered advertising, and the drug need not be offered for sale. Advertising of this nature by a party engaging in Distribution of the drug is treated the same as a sale.
There is so much more to say as the propaganda is wide and deep. And that is before doctors are threatened by their governing bodies for disagreeing with the narrative.
We need more analysis. Less misleading statements. Less lying with statistics. Less outright lying.
Agreed. It becomes really hard to trust the data on Ads especially is not being manipulated when the people in charge of collecting it are also threatening/suing doctors that do not agree with the narrative.
I applaud your data-driven approach. Even if an interpretation warrants reexamination, we should always be searching for the objective truth in the actual data. Thank you for doing so.
It is true that looking at total outcomes dating back to Jan 1 provides little context in terms of the percentage of the population by vaccine status. I'm glad you could help those unfamiliar with this information understand what is and isn't clear from those specific data. We should not look at those numbers and assume all those deaths happened among a tiny number of unvaccinated people.
However, you neglect 2 critical points in your post:
1) Inclusion of negative outcomes within 2 weeks of vaccination is not a misrepresentation - vaccines were designed to provide strong protection, but only reliably after 2 weeks. We can't blame COVID for the death of someone already imminently dying of cancer. We can't say that exercise doesn't make us stronger just because we don't see results after the first workout. And we can't blame the vaccine for not working the instant it is delivered, in some cases after a person is already infected. It isn't a cure, it is a preventative tool that takes some time to work. If anything, the reduced cases and bad outcomes seen for some diagnosed at day 11 or 12 or 13 after vaccination is actually a bias against vaccines - because some of those people would have derived partial benefit as their immune system responded to the vaccine, but their positive outcomes are lumped in with those people who had not been vaccinated at all.
2) You also should be driving readers to the published statistics on the same page which relate to current outcomes (under Vaccine Outcomes. NOT Severe Outcomes.). As we agree, the Jan 1 numbers provide little context. But AHS publishes detailed stats that relate to cases in the past 120 days (when a majority of people had at least some vaccine protection), demonstrating disproportionate risk of bad outcomes among the unvaccinated - normalized by the number people actually vaccinated at the time. Moreover, even your own screenshots show *current* data in the bottom half. Current (confirmed) active cases, new cases (from this week), and current hospitalizations. Given the sharp rise in hospitalizations over raw case numbers for the unvaccinated, it becomes very clear that they are being hospitalized at a much greater rate, especially since now 75% of the entire population are partially vaccinated, and 70% are fully vaccinated. (Another detail this data cannot tell us: why exactly fewer confirmed unvaccinated cases lead to greater hospitalization rates. Is it because the unvaccinated get much sicker - or because the unvaccinated tend not to get tested when they are sick, so there are fewer cases confirmed by testing out of a larger-than-average pool of actual COVID cases, leading to hospitalizations only *appearing* to happen more often, due to greatly undercounting COVID infections among the unvaccinated.)
Actually, to be strictly fair, we should look at vaccination rates from 2+ weeks ago, which are 73% (partial) and 68% (full) respectively. So those current numbers show that 69% of current hospitalizations are among the unvaccinated (or got their shot less than 2 weeks ago), while they themselves are only 27% of the population. Meanwhile only 25% of current hospitalizations are from the 68% of the population which were fully vaccinated at least 2 weeks ago.
I agree that the opacity with some of this data is troubling, and I would really like to understand more about what goes on underneath the hood. The statistics you reference are updated retroactively with new information, and therefore are understood not to match the daily summary update. Robson Fletcher at CBC has a page on COVID statistics that explains this in more detail, and why it is done. It isn't meant to obfuscate, but to offer mostly-reliable and timely updates daily (in the summary) and then provide a record of fully-researched and confirmed information - which takes longer to produce - to be available for posterity and long-term analysis.
There is a lot to unpack here, so forgive me if I do not fully articulate my rebuttal. Unfortunately, a perfect response that will satisfy you would require a lot more than I can provide here with time limitations.
For point (1), I am not neglecting that the vaccines are meant to provide protection. Indeed, I have point out repeatedly that I believe the vaccines have relatively impressive medium term benefits. You are absolutely correct that we cannot, without reservation, attribute a vaccine death in the first 14 days to the vaccine. I would like to make a couple points here:
1. We do that regularly with the unvaccinated. The most high profile case was a 14 year old boy that died in Alberta. The media said that he died of the virus. His sister said that he had terminal brain cancer. He was critically ill and happened to get tested a few days before he died, but the virus did not contribute in any real way. If he did not have cancer, he would not have died. If he did not catch the virus, he would have still died.
2. I am not being a hypocrite by speaking about the above case; rather, the reason I am exceptionally specific about the first 14 days not being lumped in with the unvaccinated is because the unvaccinated are less likely to catch the virus in that period. You can see this with the eye test by looking at the upside down-V shape of the curve directly after vaccination. This can be verified statistically, and I have verified it statistically. There are also examples of it in the literature (a dutch study found negative 40-110% VE in the first 14 days) and UK data showed 1.8 million cases in the vaccination vs 1 million in the unvaccinated at the beginning of their vaccine campaign (around 50% of the population was vaccinated at the cut off date). The Indian vaccine also shows negative VE for the first dose -- in short, it takes until day 80ish of their trials until the cumulative risk from the vaccinated group is equal to the unvaccinated group.
3. It is hard to ignore the healthy vaccine effect that has been present in vaccination literature for a long time. In short, people are less likely to get vaccinated when they are feeling sick. So the sharp rise a few days after vaccination in case counts is quite astonishing.
In short, I absolutely agree with you we need to be very careful about what we call a covid death. But we have not been. So I am forced to analyze their data with imperfect information.
As for (2), yes, severe outcomes is a typo which I missed! I think I linked the right page though. The current data is wrong. It has been wrong every single day for as long as I have checked. New cases record about 0.88 to 7.84% more cases each day than there actually are, most of these cases are attributed to the unvaccinated population. One day in October, they added 118 phantom cases to the unvaccinated and 87 less the fully vaccinated than actually occurred. This happens to some degree every single day and it is not simple miscounting because it is always directed in one direction.
Simply put, I am highly skeptical of the data on active cases and current hospitalizations. But I do track that data. I do not think it is because the vaccinated population is more likely to get tested, largely because many unvaccinated must be tested for work on a regular basis. Even during this time, the proportion of current hospitalizations has fallen consistently for months. I will posit some theories though...
1. In British Columbia and other provinces, those in a LTC home are not counted as hospitalizations (since, technically, they are not). I am not 100% certain about Alberta, but I would imagine they follow similar trends. I would posit that someone in a LTC home is more likely to be vaccinated than a person of similar age that is not in one.
2. Some of these hospitalizations are data artifacts. Again, we are not working in discrete time but continuous time. The number of people getting vaccinated has been growing for a long time; thus, if someone entered a hospital a month ago, the proportion of unvaccinated would be higher.
3. People with extremely fragile health may be less likely to be vaccinated. I know this does not square with the public health campaign, but it is unlikely any decent doctor would give a patient undergoing chemotherapy a vaccination just to check a box.
I appreciate that the data is not meant to obfuscate, but it does. Some of the suggestions I made would help make it more transparent and less confusing for people that are using this data as a hammer against the unvaccinated. For example, the last 120 days data, as I said, is problematic as they are changing the denominator after every vaccination. They have gone up disproportionate to what they should have. It is simply not informative and potentially destructive. A better way of doing things would be to change the rate calculation to person years lived. It would go a long way towards reducing bias. Case counts should be in person years lived as well instead of since Jan 1.
Well, hopefully that offers some insight into my line of thought. Unfortunately, when the government is forcing many people to take an effective but risky (safe, the way it is used by regulators, being a function of risk) medicine, and people are using skewed statistics that are produced for the sake of timely updates to attack those peoples livelihoods, we do not have time for posterity.
Thank you. I have been proof over the Alberta data as well. Alberta is using he statistics manipulatively to give the general impression covid is much more dangerous than it is, that the vaccine is more effective than it is, and that he vaccine is much less dangerous than it is.
Some notes:
- being overweight or obese is universally recognized as being very strongly correlated to severe covid outcomes and deaths. Alberta does not recognize obesity as a comorbidity. If you added obesity as a comorbidity to Alberta’s data, my guess is that you would find covid hospitalizations and deaths would occur almost-never in people with no comorbidities (inclusive of obesity), especially in younger people.
- Alberta’s adverse event reporting per dose he last time I checked discloses adverse events and serious adverse events following vaccination in similar proportions (relative to the kinds of adverse events). Curiously, Alberta reports lower case rates of serious adverse events (per dose) than the rest of Canada. I do not believe this is a coincidence.
There are many others.
THe most egregious behaviours are the misleading claims, and the general impressions given that the disease is of equal deadliness to every person in Alberta.
If a pharmaceutical company were making such statements, they would be violating many provisions of the federal Food and Drugs Act which prohibits advertising that:
- give a general impression that a disease is more harmful than it is, and that a drug designed to treat or prevent it safe without identifying real risks.
- obscures or misstates the medical risk/benefit of the drug.
- does not differentiate between the risk/benefit by demographic group or patient status.
Because all of these things get in the way of achievement of the goal of informed consent (which was universally demanded by principles of medical ethics - until covid vaccines were authorized on an emergency basis).
Pharmaceutical companies have received multimillion dollar fines for such behaviours. What the Alberts government is doing by definition in this federal statute is in fact contrary to he Food and Drugs Act by definition inasmuch as such behaviours are considered advertising, and the drug need not be offered for sale. Advertising of this nature by a party engaging in Distribution of the drug is treated the same as a sale.
There is so much more to say as the propaganda is wide and deep. And that is before doctors are threatened by their governing bodies for disagreeing with the narrative.
We need more analysis. Less misleading statements. Less lying with statistics. Less outright lying.
Keep up the great work!
Agreed. It becomes really hard to trust the data on Ads especially is not being manipulated when the people in charge of collecting it are also threatening/suing doctors that do not agree with the narrative.
Apologies for typos. iPads confound me at times
I applaud your data-driven approach. Even if an interpretation warrants reexamination, we should always be searching for the objective truth in the actual data. Thank you for doing so.
It is true that looking at total outcomes dating back to Jan 1 provides little context in terms of the percentage of the population by vaccine status. I'm glad you could help those unfamiliar with this information understand what is and isn't clear from those specific data. We should not look at those numbers and assume all those deaths happened among a tiny number of unvaccinated people.
However, you neglect 2 critical points in your post:
1) Inclusion of negative outcomes within 2 weeks of vaccination is not a misrepresentation - vaccines were designed to provide strong protection, but only reliably after 2 weeks. We can't blame COVID for the death of someone already imminently dying of cancer. We can't say that exercise doesn't make us stronger just because we don't see results after the first workout. And we can't blame the vaccine for not working the instant it is delivered, in some cases after a person is already infected. It isn't a cure, it is a preventative tool that takes some time to work. If anything, the reduced cases and bad outcomes seen for some diagnosed at day 11 or 12 or 13 after vaccination is actually a bias against vaccines - because some of those people would have derived partial benefit as their immune system responded to the vaccine, but their positive outcomes are lumped in with those people who had not been vaccinated at all.
2) You also should be driving readers to the published statistics on the same page which relate to current outcomes (under Vaccine Outcomes. NOT Severe Outcomes.). As we agree, the Jan 1 numbers provide little context. But AHS publishes detailed stats that relate to cases in the past 120 days (when a majority of people had at least some vaccine protection), demonstrating disproportionate risk of bad outcomes among the unvaccinated - normalized by the number people actually vaccinated at the time. Moreover, even your own screenshots show *current* data in the bottom half. Current (confirmed) active cases, new cases (from this week), and current hospitalizations. Given the sharp rise in hospitalizations over raw case numbers for the unvaccinated, it becomes very clear that they are being hospitalized at a much greater rate, especially since now 75% of the entire population are partially vaccinated, and 70% are fully vaccinated. (Another detail this data cannot tell us: why exactly fewer confirmed unvaccinated cases lead to greater hospitalization rates. Is it because the unvaccinated get much sicker - or because the unvaccinated tend not to get tested when they are sick, so there are fewer cases confirmed by testing out of a larger-than-average pool of actual COVID cases, leading to hospitalizations only *appearing* to happen more often, due to greatly undercounting COVID infections among the unvaccinated.)
Actually, to be strictly fair, we should look at vaccination rates from 2+ weeks ago, which are 73% (partial) and 68% (full) respectively. So those current numbers show that 69% of current hospitalizations are among the unvaccinated (or got their shot less than 2 weeks ago), while they themselves are only 27% of the population. Meanwhile only 25% of current hospitalizations are from the 68% of the population which were fully vaccinated at least 2 weeks ago.
I agree that the opacity with some of this data is troubling, and I would really like to understand more about what goes on underneath the hood. The statistics you reference are updated retroactively with new information, and therefore are understood not to match the daily summary update. Robson Fletcher at CBC has a page on COVID statistics that explains this in more detail, and why it is done. It isn't meant to obfuscate, but to offer mostly-reliable and timely updates daily (in the summary) and then provide a record of fully-researched and confirmed information - which takes longer to produce - to be available for posterity and long-term analysis.
Hey and thanks for your thoughtful comments,
There is a lot to unpack here, so forgive me if I do not fully articulate my rebuttal. Unfortunately, a perfect response that will satisfy you would require a lot more than I can provide here with time limitations.
For point (1), I am not neglecting that the vaccines are meant to provide protection. Indeed, I have point out repeatedly that I believe the vaccines have relatively impressive medium term benefits. You are absolutely correct that we cannot, without reservation, attribute a vaccine death in the first 14 days to the vaccine. I would like to make a couple points here:
1. We do that regularly with the unvaccinated. The most high profile case was a 14 year old boy that died in Alberta. The media said that he died of the virus. His sister said that he had terminal brain cancer. He was critically ill and happened to get tested a few days before he died, but the virus did not contribute in any real way. If he did not have cancer, he would not have died. If he did not catch the virus, he would have still died.
2. I am not being a hypocrite by speaking about the above case; rather, the reason I am exceptionally specific about the first 14 days not being lumped in with the unvaccinated is because the unvaccinated are less likely to catch the virus in that period. You can see this with the eye test by looking at the upside down-V shape of the curve directly after vaccination. This can be verified statistically, and I have verified it statistically. There are also examples of it in the literature (a dutch study found negative 40-110% VE in the first 14 days) and UK data showed 1.8 million cases in the vaccination vs 1 million in the unvaccinated at the beginning of their vaccine campaign (around 50% of the population was vaccinated at the cut off date). The Indian vaccine also shows negative VE for the first dose -- in short, it takes until day 80ish of their trials until the cumulative risk from the vaccinated group is equal to the unvaccinated group.
3. It is hard to ignore the healthy vaccine effect that has been present in vaccination literature for a long time. In short, people are less likely to get vaccinated when they are feeling sick. So the sharp rise a few days after vaccination in case counts is quite astonishing.
In short, I absolutely agree with you we need to be very careful about what we call a covid death. But we have not been. So I am forced to analyze their data with imperfect information.
As for (2), yes, severe outcomes is a typo which I missed! I think I linked the right page though. The current data is wrong. It has been wrong every single day for as long as I have checked. New cases record about 0.88 to 7.84% more cases each day than there actually are, most of these cases are attributed to the unvaccinated population. One day in October, they added 118 phantom cases to the unvaccinated and 87 less the fully vaccinated than actually occurred. This happens to some degree every single day and it is not simple miscounting because it is always directed in one direction.
Simply put, I am highly skeptical of the data on active cases and current hospitalizations. But I do track that data. I do not think it is because the vaccinated population is more likely to get tested, largely because many unvaccinated must be tested for work on a regular basis. Even during this time, the proportion of current hospitalizations has fallen consistently for months. I will posit some theories though...
1. In British Columbia and other provinces, those in a LTC home are not counted as hospitalizations (since, technically, they are not). I am not 100% certain about Alberta, but I would imagine they follow similar trends. I would posit that someone in a LTC home is more likely to be vaccinated than a person of similar age that is not in one.
2. Some of these hospitalizations are data artifacts. Again, we are not working in discrete time but continuous time. The number of people getting vaccinated has been growing for a long time; thus, if someone entered a hospital a month ago, the proportion of unvaccinated would be higher.
3. People with extremely fragile health may be less likely to be vaccinated. I know this does not square with the public health campaign, but it is unlikely any decent doctor would give a patient undergoing chemotherapy a vaccination just to check a box.
I appreciate that the data is not meant to obfuscate, but it does. Some of the suggestions I made would help make it more transparent and less confusing for people that are using this data as a hammer against the unvaccinated. For example, the last 120 days data, as I said, is problematic as they are changing the denominator after every vaccination. They have gone up disproportionate to what they should have. It is simply not informative and potentially destructive. A better way of doing things would be to change the rate calculation to person years lived. It would go a long way towards reducing bias. Case counts should be in person years lived as well instead of since Jan 1.
Well, hopefully that offers some insight into my line of thought. Unfortunately, when the government is forcing many people to take an effective but risky (safe, the way it is used by regulators, being a function of risk) medicine, and people are using skewed statistics that are produced for the sake of timely updates to attack those peoples livelihoods, we do not have time for posterity.