Public Health has estimated Delta's CFR to be 0.096 %.

They need to include vaccination status of the cases, hospitalizations and deaths. Canada is one of the most widely vaccinated countries in the world, and we can see from Israel data and UK data that the vaccinated are particularly vulnerable to the Delta variant.

This is incorrect. The UK study "Effectiveness of Covid-19 Vaccines against the B.1.617.2 (Delta) Variant" August 12, 2021
N Engl J Med 2021; 385:585-594, shows 88% and 74.5% vaccine effectiveness for Pfizer and AstraZeneca respectively. Only slightly lower than the vaccine trials, as was expected as trials are a highly controlled environment.

The Israeli study has been widely misinterpreted in the press, owing in part to some terrible actions by Israel's own health officials before the data was analyzed. After analysis the Israeli study "Correlation of SARS-CoV-2 Breakthrough Infections to Time-from-vaccine; Preliminary Study" (preprint), concludes only that there may be decreased effectiveness of vaccines, but more data is needed. The reason for the inconclusiveness is that both confounding behavioral aspects (i.e. greater risks taken by vaccinated) and that initial cases (also Pfizer and AstraZeneca trials) were with early virus strains whereas current cases are with the Delta strain, meaning the increased virulence of Delta may be to blame. The Israeli study is not apples to apples. However, the conclusion we can draw from this study is that because of Delta's virulence, boosters may be needed, which is exactly what Israel is doing for 50+ y.o.

Agreed the fact that this Canadian study did not have vaccination status of the participants, and that Delta wasn't specifically sampled, are glaring holes, however. I wouldn't put too much these Canadian results.

This is at the stage of calculating CFR where we were with the initial wild SARS-Cov2 in February 2021. That is, we have no idea how many cases there actually are, because most people aren't getting tested.

Testing was extremely common prior to mass vaccination, but post-vaccination, most people no longer bother to get tested as soon as they have symptoms, most institutions (colleges, healthcare, etc.) stopped requiring weekly testing for those with proof of vaccination, most institutions requiring immediate tests for participation (sports, travel, etc.) stopped requiring it for those with proof of vaccination, and the CDC, among other public health bodies, announced they would stop collecting stats for asymptomatic cases in the vaccinated.

That means the denominator for the CFR for VOC-202012/1 ("Delta") is completely unknown, just as it was in February 2020 for the wild virus, when it was estimated at 3.5% or higher, at least an order of magnitude higher than it turned out to be.

Unfortunately, this study doesn't yet allow us to compare apples to apples.

Data from the UK shows an almost complete decoupling between cases and hospitalizations/deaths - this won’t pass peer review because it’s complete and utter nonsense, par for the course for Fisman and Tuite who have been almost completely wrong with their modelling the entire pandemic.

That's due to vaccination. Virulence is measured against those without seroprevalence.

Proof?

proof of what? That vaccination has lowered the CFR?

Correct. Where is the proof, that the effect you postulate is caused by vaccination? It is an idea, yes, but where is the proof? Without real data, this idea is just an idea.

Dear authors, I do not get the point: In your raw data (table 1) the percentage of people dying from Corona Delta is 0.7%. All other variations cause 0.9% deaths for infected people. So, how can the risk to die from Delta be higher than for other variants? Where can we see how the "adjustment for age, sex, comorbidities, health unit, and temporal trend of the raw data works? Here in Germany people go wild because of this study, but I can not comprehend it. Thank you very much!

you are forgetting that vaccination lowers CFR. The more of a population with seroprevalence, the lower the mortality rate. This doesn't mean delta is less virulent, it means there is more population protection.

This may be a logical conclusion, but too "proof" delta to be more dangerous this assumption is weak, as long as it can not be supported by real data. And exactly this is missing in the paper. So, the main conclusion is just a best guess.

The confounder may be vaccinations and vaccination breakthroughs. Canada has had high vaccinations during the study timeframe. And was particularly high during the period Delta became dominant. Since Delta deaths were skewed towards the end of the period it's probable that deaths were missed in their 2 week end window. Also, while vaccines are quite effective, infections do occur. Unfortunately, the authors did not have individual vaccination records of the cases. So the key to the higher virulence numbers may be their modeling of vaccination impact. It would be good if the paper had some tables and details of that model and how their results followed. I don't find their results unreasonable given that.

They are unreasonable in the extreme looking at far better UK and India data. Almost an order of magnitude worse on IFR vs Alpha on older strains.

R0 also badly abused. An airborne virus with an average R0 in the 7+ range? Do you have any idea how explosive spread is under that scenario? These SSS data fit papers with terrible data control (Delta isn't genome sequenced directly).

Just awful. At least spend a few minutes comparing global data before using their academic heft to get something this terrible published.

Shameful.

Where does it makes conclusions on R0?

Thank you or your contribution! I still do not understand, how this works. If vaccination is interfering the ration between deaths from Delta and all the other variants, what is the study based on? Hard numbers shown are just those in table 1. How can a "lower death rate" cause a "higher death risk"? This for me only makes sense, if the time between infection and death is much longer for Delta. But there are no data presented supporting this thesis. I`m really confused here, especially because the German press and also the first "experts" forecast the end of the world because of this study. You don't find their results unreasonable, but did you find a proof for the conclusion "Delta ist more dangerous" in the paper? I didn`t..

The Delta cases are younger and with fewer comorbidities, so they should have a lower death rate if Delta were the same as Alpha. However, the Delta cases have the same death rate as Alpha cases (not significantly different). Therefore, Delta appears stronger. You can't see it because it's math, but this is a verbal explanation. Another way of saying it is that the model compares death rates of each age & comorbidity group and they appear higher for Delta. I do hope someone repeats the analysis because the ages and comorbities don't appear different enough intuitively to cause that much of a higher death rate. Table 3 is especially non-intuitive because the comorbidities in the Delta group is actually higher, with a lower crude death rate, and the ages are fairly similar. Hmmmmm.

But all of your explanations are not stated in the paper. This means you explain data that are not shown with math that is not shown. Additionally you say "delta is worse than alpha because the victims are younger, even if in total less people are dying from delta". Correct?

Delta is infecting those who are not vaccinated at a much higher rate -- this tends to be those less at risk (i.e. younger). The older/vulnerable populations are vaccinated at higher rates and hence are infected with Delta in lower numbers and also experience less severe infection. This is why the CFR is lower -- younger population infected + less severity if infected with vaccination.

OK, slowly I get, what's meant: Given the case, vaccination hepls to avoid CFR, and delta cases show the same fatality as alpha cases do, the aggressivity of delty is relatively higher. Correct? Finally got that. That means, the authors are not talking about just viruses, but viruses in different environments defined by vaccination status and effectivity. It's not about math, it's just about the vaccination state and its (assumed) effectivity. So what's still missing is the proof for the postulation about the effectivety of vaccination. In Germany we are told, delta is an escape mutant, so vaccination is not as effective here as against alpha.