Did clearly healthy people die from the mRNA vaccines?

There are hundreds of healthy people in foreign VAERS death reports

Summary:

• There are over 500 death reports in foreign VAERS where the writeup includes a keyword which suggests the patient was healthy

• Some of these are false positives, but only a small percentage

• There are over 100 death reports where the symptom onset was on the same day of the vaccination

• There are about 30 death reports for people under 30 years of age

• Of the 500+ reports, nearly 190 reports were made by healthcare professionals. Of these, 140 reports were made by physicians.

• Even if you combine all these conditions to make it more stringent, there are still 8 death reports with a) same day symptom onset b) <= 30 years of age and c) reported by a healthcare professional. That is 8 more than what the CDC claims

In the previous article, I pointed out how doing just a little bit of legwork (i.e. reading the actual text writeups) can prevent the vaccine pushers from making elementary errors.

There is another example of “elementary error”, and this is the idea that everyone who died right after taking the vaccine already had some underlying condition.

For example, that’s the gist of this tweet:

Link to tweet

But if you actually start reading the report, you will see that this is just plainly wrong.

Previously healthy people who died after taking the vaccine

So I did a simple text search on the foreign VAERS death writeups for words like “healthy” and “good health” and “fit” and “shape” (in good shape)

Now this can also include some false positives - for example if someone writes “the patient was not very healthy” it will be counted in this query.

But you can process this information further and also manually evaluate them to flag problematic reports.

Why was this never done?

I took the list of reports returned by the SQL query above, and turned it into a searchable index using Algolia.

Click here to see the search demo

As you can see, there are over 500 death reports in foreign VAERS which have one of the keywords in the writeup.

Also, there is an important update in this search demo. The title of the result now includes the sentence which contains the relevant keyword (like healthy, fit etc) so you can read at-a-glance the reason why that particular report was selected by the SQL query.

If there are multiple such sentences containing one of these keywords, the Python script selects the last one. There is no specific reason for this choice, it just made it a bit easier for me to write the code :-)

False positive example

The very first search result is in fact a false positive.

But it is also clear that the three reports which follow right after that are for healthy people.

Same day symptom onset

Of the 500+ reports, over 100 had same day symptom onset.

Now you DO need to read the reports to check if there are any false positives, but the main point I am making here is that this can be done manually by a small team without a lot of effort.

Are we supposed to believe ALL of these deaths were NOT because of the vaccine?

Deaths of under 30

What about age?

Weren’t most people anyway very old and close to their death?

There were 28 deaths for people under 30 years of age.

Reported by health care professionals

There are also some suggestions that these are somehow “fake” reports.

But there are 189 reports by healthcare professionals, and 140 of them are from actual physicians

These are all reports which are derived from the original list of 500.

Combining all the conditions

What if we combine ALL these conditions - same day symptom onset, filed by a healthcare professional and age <= 30?

And there are STILL 8 such reports.

Are we supposed to believe not ONE of these was because of the vaccine?

How to improve this analysis using Machine Learning

I wrote this in a previous article about VAERS reports filed for legal purposes:

One big advantage of understanding how Machine Learning works with Natural Language Processing is that you can go well beyond basic keyword search and start using word vectors etc. for more advanced analysis. By combining ML models and human review, it should also be possible to decide whether or not litigation is actually a source of VAERS over-reporting.

Link

It is possible to do the same thing to decide whether or not the patient was healthy.

While the HISTORY field in VAERS is usually populated when the patient is unhealthy, an empty HISTORY field does not automatically imply the patient was healthy.

You do need to read the full writeup to see if you can assess this - and in fact, quite often there isn’t sufficient information in a VAERS report to evaluate this. But there are also some VAERS reports (as I have shown) where the reporter quite clearly states that the patient was healthy.

Incorporating ML techniques will

a) likely identify all VAERS death reports where the patient was healthy before taking the vaccine. The 500 reports (minus the false positives) that I have presented here is not likely to be a tight lower bound.

b) also remove the false positives

In other words, this is another example of using data science techniques to improve VAERS analysis. When will the data science community start paying attention to vaccine injuries?