Jeffrey Morris won't show us the correct analysis of the Czech Republic data

He tries to gaslight people into believing my analysis is flawed. Where is HIS analysis?

Executive summary

UPenn Professor Jeffrey Morris sent me a DM referencing a long thread on X designed to debunk my analysis of the Czech Republic record level data.

In this article, I’ll point out the flaws in his post.

But here are the key points:

1. There are differences between the unvaccinated vs. vaccinated because those who opt to be vaccinated generally are healthier. This has been demonstrated time and time again for example here. So of course the vaccinated die less. If vaccination really reduced all-cause mortality, me and my misinformation spreader friends and followers would all be lining up to get our shots, wouldn’t we? It seems like Morris has never heard about selection bias.

2. The key point of the Czech data is it shows a 30% 1-year higher mortality difference for those who got Moderna. This is an age-standardized mortality rate (ASMR). The 95% confidence intervals are tiny. See the analysis here.

3. The most important point is that Morris never gives us a more plausible explanation for the higher mortality in the Moderna group. He provides a hand-waving “it was comorbidities” without addressing our extensive work showing it wasn’t. In particular, we show below that in Jan 2021, a month with virtually zero comorbidities, Moderna was deadlier than in any other month.
   
   And in the example he cited, the absolute differences between the groups were minor (it was a small percentage of a small percentage).
   
   And the 30% difference wasn’t due to Pfizer being better at protecting against COVID deaths because the mortality differences were there even in months of non-COVID.

The key point is Morris acknowledges the mortality differences, but he lacks a plausible explanation for it if the vaccines are equally safe. He only provides hand-waving arguments.

I wasn’t convinced.

And I keep asking him to present his analysis of the Czech data which he refuses to do.

The Morris post attempting to debunk the Czech Republic data

Here is where he admits I was right (something I know must be very deeply painful for him):

Not sure how he gets that Moderna is "killing many people" and needs to be "removed from the market — Prof. Jeffrey Morris

OK, I’ll explain it again.

The ASMR death rate from one year from the time of the shot was 30% higher for Moderna. That’s huge.

Is there an explanation other than a higher kill rate?

1. It cannot be explained by worse protection from COVID because COVID itself is on the order of a 20% mortality rate increase so it’s impossible, even if Pfizer was a perfect vaccine, for Moderna to have a 30% higher ASMR.

2. We have the same higher mortality if we only examine deaths happening in LOW COVID months. So clearly, unless vaccines reduce non-COVID ACM (which nobody has ever claimed otherwise we’d all be taking the shots), the ONLY possibility left is that Moderna INCREASED all-cause mortality from baseline.

In summary, the difference wasn’t due to comorbidities because:

1. They are too small (in absolute terms) to have such an impact (a fraction of a fraction).

2. There is no evidence of #1 causing a significant net mortality difference.

3. We showed from the data when the comorbidities were higher for Pfizer, Moderna was even worse which completely trashes the argument. The reality is the comorbidities were just noise which is why we can show the effect when comorbidities were reversed.

4. It happened in all age groups. Every single age group without fail.

5. It even happened when we excluded everyone vaccinated in the first 3 months of 2021 (we’d have extended it out to the first 6 months but we’re left with too little data so it’s really noisy).

6. There was no systemic or systematic bias.

Morris cites this paper. Let me tell you what the paper showed about Moderna. You’re going to love it.

In his argument, Morris cites this paper, “Effectiveness of COVID-19 Vaccination in Preventing All-Cause Mortality among Adults during the Third Wave of the Epidemic in Hungary: Nationwide Retrospective Cohort Study.”

When you look at the raw data (that’s before they make all the signals disappear), you see major differences between brands. Guess which vaccine is at rock bottom and KILLED THE MOST? Yup. Moderna. More “bad luck?”

Look at the data yourself

Click the image to see the data yourself. In the chart below, the M/P ratio is Moderna/Pfizer mortality ratio for every 5 year age group from 15 years old to over 100+ year old (where the effect nearly vanishes):

It’s easy to play with the parameters in the pivot table to restrict the month of vaccination to exclude the first 3 months of 2021 so we avoid the time when the sickest got their shots. Here are the results. Note the part in red. This is the kind of stuff that Jeffrey Morris never points out to you.

vax_5.xlsx from the analysis directory in the github repo shows that for shot 1 given in the last 9 months of 2921, Moderna was still a huge mortality disaster vs. Pfizer.

Making claims I never claimed

I told you the comparison with the unvaccinated is misleading and problematic. This is precisely why I focused the analysis on the brand comparison because that is relatively free on confounding.

Look the comorbidity data yourself

It’s all in my github analysis directory: '

from comorbidy.xlsx.

Note that in Jan and Feb 2021, both vaccines were given to people with less than 1% comorbidities. Therefore, looking at the MR in Jan and Feb, we can see what happens in the virtual absence of comorbidities.

To visualize that, let’s look at the Triangle plot of Moderna/Pfizer MRR by month of shot (y-axis) and month of death (x-axis):

Mortality triangle plot. X axis is month of death, y-axis is month of shot. Value in each cell is age standardized relative mortality rate. >1 means Moderna is deadlier. In all lines, when there is sufficient data to be above the noise, (from July onwards the numbers are not statistically significant - see below) Moderna is more deadly. Source

The MRR is through the roof in Jan at 2.32, and about normal 1.3 in February.

So the high mortality for Moderna was not caused by comorbidities. It was there in the near complete absence of comorbidities!!

Morris’ summary argument

He wrote:

And as I’ve repeated said, you cannot get any reliable causal estimate of vaccine effects from simple deaths and vaccine status data even if you also have age, since you do not have the necessary information to even begin to try to adjust for confounders and other sources of bias making the vaccine groups inherently different. So you cheat puffing saying “where’s your analysis” is meaningless — these data no matter how good you say they are, are not fit for purpose for estimating vaccine effects And if you are going to keep trying to make causal statements from these data you will continue to make invalid arguments, especially if you blatantly ignore key features of the data like unvaccinated death rates and available confounders

There was no systemic or systematic bias to the distribution by brand in the Czech Republic. There will be differences between the cohorts that will happen by random chance, but these differences are negligible. The only factor Morris brings up is comorbidities, but the data just doesn’t support that since in months of tiny comorbidities (fraction of 1%), the mortality differences were highest of all the months measured. The data doesn’t just doesn’t fit his hand-waving argument.

Summary

Morris acknowledged the consistently higher mortality for those who got Moderna.

What he never does is provide us with a viable explanation for this effect.

He expects us to believe his hand-waving argument that it must be the comorbidities but that can’t be because the effect was there when virtually no one vaccinated had comorbidities.

I am not persuaded by Morris’ arguments. He uses hand-waving arguments not backed up by calculations (you know… numbers).

So what produced the 30% mortality difference between the brands? And why was higher Moderna mortality observed in Hungary as well? Coincidence?

The Czech Republic data is the single best record-level dataset on vaccine mortality available today. Why isn’t anyone analyzing it to prove the vaccines are safe???

If there is a better dataset, Morris should produce it. If there isn’t, then Morris needs to analyze this dataset and prove to the world it shows the vaccines are safe. Where is his analysis of the data? All I get is excuses and attacks when I ask.

And for that matter, why is it that no epidemiologist in the world is looking at this data and showing us that it clearly shows the vaccines are safe? They should be rushing in to analyze the data in order to show that people like me are spreading misinformation. But there is nothing but the Sounds of Silence.