See — look at that. Even made the subtitle more or less follow the tune. This here writing stuff is a hoot.
Alright. So a few weeks ago I posted1 an observation that the VAERS2 data not only includes reports of possible adverse effects tied to the COVID vaccines (and vaccines in general), but also apparently includes vaccine breakthrough cases and deaths. In fact there is apparently a legal reporting requirement for both the vaccine associated adverse events AND breakthrough cases3. This means a few things:
It is not correct to blindly count the number of COVID entries in VAERS and state they are exclusively injuries or deaths caused by the COVID vaccines4. There in that sense is over-reporting of vaccine injury in VAERS.
As we’ll see (in a followup post) the breakthrough numbers that are in VAERS are quite low. This implies there could be significant under-reporting in VAERS5 6.
That there is over-reporting AND under-reporting of vaccine associated fatalities in VAERS does not mean it is on its face a correct or incorrect indicator of such. Like any data, it must be understood and interpreted. An officer in a patrol car on the side of the road sees many cars go by above a speed limit, and misses many when distracted by that donut box in the passenger seat, or parked elsewhere. This does not mean there are always or never dangerous speeders on a particular stretch of road. You can try get an estimate of the right number knowing what the officer is seeing, and accounting for, for example, the amount of time they are watching or not watching.
To start to address #3, I’m going to try to get some handle on #1 and #2. What I’ve done personally so far for #1 is below, what I’ll try to do to tackle #2 is mostly in a followup that substack size limits push into another post.
Trying to understand the extent of over reporting in VAERS involves scrutinizing the data itself, and trying to identify entries that are not likely to have a causal relationship to the vaccine in question. In most cases some medical expertise is needed. I am not a medical doctor, so do not entirely have that expertise. I can do some simple checks though, one of which is how I stumbled on what I’m going to look at related to #2. This lack of expertise by the way also is why I so far really have focused on the “DIED=Y” entries. Despite not being a MD, I do have some confidence that I have a handle on what that indicator biologically means.
If I walk through the entries in the latest set of data that I have (data up to 12/10/2021), and count unique ID’s:
There are 1,823,613 unique VAERS ID’s in the entire dataset.
If I remove the “NonDomestic” data — reports from outside the United States, but still must be reported in the VAERS system due to the drugs being manufactured in the US, or from US companies7 — There are 1,440,237 entries left. These are in principle the number reported of US domestic adverse events for all vaccines over the last 30 years.
If I then select for ‘COVID19’ in vaccine type, I get 688,136 entries. I should note here that at this point, by far COVID19 dominates any other vaccine in the VAERS system. The runner up is ‘VARZOS’, or the Varicella Zoster (shingles) vaccine at 96,616 entries. COVID19 though is exclusive to 2021, Varicella summing up over the course of the last 15 years.
Selecting on fatalities fortunately whittles this down to a much smaller number, though not necessarily small: 9262. Next runner up here BTW is the polio vaccine, at 829 fatalities summed over the entire 3 decades of VAERS, but very much weighted towards the earlier years. Hmm. Learning some of that history now added to the too long “to read up on” list…
I was able to identify a few duplicate entries by looking for entries that had the same age, date of death, state, and some of the symptom text. There were 64 of these, where it looks like the same person in each case received a new entry in the system with a new ID, but some of the information was corrected8. That brings us down to 9198.
Finally, the subject of my previous post on this, I constructed a simple filter to try to identify the covid vaccine breakthrough entries from the severe vaccine event entries. This differentiates that number into 5812 severe event fatality entries, and 3322 breakthrough fatality entries. These aren’t hard numbers by the way — there is likely some mis-attribution in either direction with my simple filter, and given it ultimately depends on the words the reporter did or did not include in their submission. Visually scanning what I have on each side though looks pretty close.
Now, I said earlier that even though I’m not an MD, I’m pretty confident I have a handle on what condition a person who’s died is in. But one thing I do not have here, and can’t say I have a lot of expertise in, is whether the death recorded in this data was in fact directly tied to the application of the vaccine that triggered its inclusion in this data sample.9 In most cases it appears obvious, with the entry including a text description of the events, and a time difference within a few days.
There are several entries that extend out many days and weeks after the vaccine was applied. The breakthrough events cover the vast majority of them. You can see a plausible looking pattern if you then plot this data over time:
This now is the US domestic VAERS covid vaccine associated fatality entries by day, with the entries I select with my breakthrough filter identified in orange. Blue is then in principle the otherwise severe vaccine events. Dates are date of death according to VAERS. Blue subjectively follows the burst of vaccinations early on, followed by the slower vaccination rate that’s frustrated single metric focused officials to the point of fascism10.
Some further spot checking I’ve done so far to try to get a handle to what extent the remaining (blue) entries are tied to the vaccines:
I opened up the online “Report an Adverse Event” form11 to see what that looked like — the warning at the bottom of the form seemed pretty clear not to take submitting something into it lightly:
There are 38 fatality entries for children under the age of 18 so far12. Most of these have text descriptions of the case. I read through each of them. 3 of those deaths were from suicide. Most of the rest were cardiac arrests or strokes, or with “flu-like” symptoms following in death. In my non medical professional opinion, they really read to be vaccine caused. The footnote above is a link to a substack post displaying all of them. Regarding the suicides though:
There are 27 entries among all the COVID19 fatalities with “suicide” in the description text13 (not just children, all ages)— a few of these state that the reporter didn’t believe they were related to the vaccine, but most of them have detailed text descriptions of the suicide being a result of personality changes, disorientation following vaccination, tinnitus apparently driving to suicide…. Out of thousands that at least couple dozen had severe events that drove the person to suicide seems plausible.
Right now after reading through the text descriptions of several hundred entries, my subjective assessment of an upper limit on “not tied to the vaccine” over-report is in the 10% range. As in ~90% look to have something to do with the vaccine. So my sense is the blue in the above plot is a pretty good indicator of the direct vaccine injury. Still wading through these, but also really having to ration out reading through them. Very hard, as in heartbreaking, to read.
But. OK… to shift gears a bit. I think I have some confidence that most of the blue in the plot above is fatalities from vaccine injury. Then the question is to what extent are those a count of the actual number, or is there an inefficiency in getting the data into this database in the first place. One obvious deterrent I could see, is the disclaimer in the form I include above — it could discourage a valid report from the more liticaphobic14. There are also many sociological forces these days I would expect to cause hesitation in putting one’s name on a report in here. There is a legal requirement to report too though, that ought to encourage reporting.
But here is why the orange in the plot above is interesting. Because in the orange in the plot above is a potential sample of data that can be used to try to calibrate the VAERS fatality numbers in some way. There are other sources of public data that one may compare against to see how well they match up to what is in this part of the dataset. As observed in the earlier post, the overall number of breakthrough fatalities in VAERS seems surprisingly low compared to other reports provided by several states. Could one use this to get some sense of an under-report rate?
But this waits till the next post, as substack is once again unhappy with my verbosity…. I will leave this post though with a quote from the VAERS data guide15:
"Underreporting" is one of the main limitations of passive surveillance systems, including VAERS. The term, underreporting refers to the fact that VAERS receives reports for only a small fraction of actual adverse events. The degree of underreporting varies widely. As an example, a great many of the millions of vaccinations administered each year by injection cause soreness, but relatively few of these episodes lead to a VAERS report. Physicians and patients understand that minor side effects of vaccinations often include this kind of discomfort, as well as low fevers. On the other hand, more serious and unexpected medical events are probably more likely to be reported than minor ones, especially when they occur soon after vaccination, even if they may be coincidental and related to other causes.
https://vaers.hhs.gov/index.html
https://vaers.hhs.gov/faq.html — I’ll repeat from the earlier post: (emphasis theirs) “The reporting requirements for COVID-19 vaccines are the same for those authorized under emergency use or fully approved. Healthcare providers who administer COVID-19 vaccines are required by law to report to VAERS the following after vaccination:”, then at the bottom of their list, which of course also includes adverse events from the vaccine: “Cases of COVID-19 that result in hospitalization or death”
https://vaers.hhs.gov/data.html — “Disclaimer” in shaded box.
In fact there is another under-report source, or maybe more correctly and under-select source, in that an entry in VAERS may be present, but a piece of one of the selection columns was not entered. There are for example many entries with a “NUMDAYS”, or days since vaccination column, thats zero, but in fact in the comments there is information indicating the true span of time (usually within a week or two). Extracting this in an automated way though is not so trivial.
And of course there is this: https://www.skirsch.com/covid/Deaths.pdf. They find a VAERS under-report factor of 41x in a few ways. Or a 2.5% efficiency of getting a case reported in VAERS.
https://wonder.cdc.gov/wonder/help/vaers/VAERS%20Advisory%20Guide.htm: “VAERS occasionally receives case reports from US manufacturers that were reported to their foreign subsidiaries. Under FDA regulations, if a manufacturer is notified of a foreign case report that describes an event that is both serious and unexpected (in other words, it does not appear in the product labeling), they are required to submit it to VAERS. It is important to realize that these case reports are of variable data quality and completeness, due to the many differences in country reporting practices and surveillance system quality.”
Actual patient identities are of course obscured in what is publicly available, or identifying duplicate entries would be pretty easy. Its interesting these exist though — since the actual forms include that personal patient ID info. Somebody on the VAERS end ought to be able to trivially identify duplicates and somehow flag those. Makes one wonder to what extent effort is directed at really maintaining this.
Probably this is where I ref the VAERS disclaimer that is often the only pro vaccine objection to looking at this data: https://vaers.hhs.gov/data.html. — namely: “VAERS reports alone cannot be used to determine if a vaccine caused or contributed to an adverse event or illness.” Definitely true. Note the word “alone”. That same document also says: “The strengths of VAERS are that it is national in scope and can quickly provide an early warning of a safety problem with a vaccine. As part of CDC and FDA’s multi-system approach to post-licensure vaccine safety monitoring, VAERS is designed to rapidly detect unusual or unexpected patterns of adverse events, also known as “safety signals.””
www.whitehouse.gov
https://vaers.hhs.gov/index.html, linked in blue menu at top. Please don’t play with this and get in trouble though…
There BTW actually appear to be more than these 38, but do not have the ‘AGE_YRS’ field populated. I select by the age in that field, so miss entries that instead have an age listed in one of the comment fields.
I was disappointed that I didn’t just make this up. https://medical-dictionary.thefreedictionary.com/liticaphobia#:~:text=Psychology%20Fear%20of%20lawsuits.
https://vaers.hhs.gov/data/dataguide.html