The hidden underreporting factor in VAERS
CDC says followup reports are not added to public facing VAERS. Would it make a difference to VAERS analysis if they were added?
Around Sep 2022, I first learnt that CDC actually deletes VAERS reports every week.
That did not make a lot of sense, so I tried to get a better picture of what was going on.
First, I realized that some VAERS death reports were being deleted
Of course, people who were discussing this had already mentioned that the deleted reports were duplicate reports.
But you had to take the CDC’s word for it given
a) VAERS does not provide the public the VAERS_ID of the original report
b) there wasn’t any easy way to actually find out the original report’s ID
I did a bit more analysis, and noticed that the deleted report often had more information than the original report:
And then I tried to quantify this, and wrote an article where I demonstrated that for a large majority of reports, the “retained” report made the vaccine look much better than it was.
I actually wanted to leave it at that, because unless you have a foolproof way of matching the deleted report to the original report (I haven’t found one yet), you cannot really make any major claims about the impact of these deletions.
Someone asked me if the deletions could be explained simply by the fact that the deleted reports were all follow ups to original reports, and the CDC simply chose to keep the first report.
And this was correct of course.
I realized I DID have a piece of information (the RECVDATE field) to validate this, and I had also managed to identify duplicates for nearly 50% of the deletions, so I was actually able to verify this is true. There were some exceptions, but they were very rare.
And I thought that must be the end of the analysis, given that
a) CDC simply chooses to delete ALL follow up reports
b) to the extent that I can observe this as an outsider, this policy may be idiotic (given that follow up reports are not going to be filed unless the condition became worse), but not nefarious. Especially if it is consistently implemented.
But I had missed an important point in my very first analysis about deleted death reports. Actually, I had posed it as a question, but I had not revisited my own question to complete the analysis!
See, there are some reports where the original report says DIED is unknown, but the followup report clearly states the patient died. When you think about it, it is very obvious that this happened because when the first report was filed, the patient was still alive, and the followup report was to update the VAERS database that they had died.
Here is a clear example:
There are actually more such reports where the DIED status went from ‘Unknown’ to ‘Yes’ and I will be writing about them in the next article of this series.
And this has some important implications
If you look at the Wonder tool to see how many people died, you are clearly undercounting deaths. There are not so many deleted reports that this would dramatically change the results (at least that’s what I think till now), but this is clearly an under-reporting!
As the founder of OpenVAERS keeps saying, VAERS is clearly two sets of books
It is my view that if you move from looking at only the deleted death reports, to also taking into account serious adverse events, and then also taking into account the full narrative text of both the original and the followup report, the sum total analysis you can potentially do will end up being drastically different.
Is this a hidden under-reporting factor?
I don’t know about the under-reporting factor of 41 (to me, it seems too high for COVID19 given that nearly the whole world was waiting for a vaccine so we could get back to normal life), but it is definitely not possible for VAERS to be “over-reported” as some pro-vaccine advocates say.
In my opinion, the deleted reports alone create a substantial under-reporting factor. Especially in terms of quality, even if not in terms of quantity.
It is not hard to see why.
No one is going to bother sending in a followup report to the less serious adverse events (“wrong age group!” “changed to a different vaccine!” “vaccine thawed!"). On the other hand, the more serious the adverse event, the more likely there is going to be a follow up.
In other words, even if the total number of deleted reports (which is actually a HUGE number - it has now crossed 30K reports) is a small fraction of the total number of VAERS reports, it is still a large fraction of the reports which actually need to be taken seriously. And from that perspective, it creates a substantial under-reporting factor.
I will try and quantify this in the next few articles.