My powerpoint above is probably the quickest overview of the various ways the data is obfuscated. However, if I could summarize my overall observations:
1) There are twice as many dead victims in VAERS (now since dead) because only initial reports are made public, even though follow-up data is still collected, like death.
2) VAERS does not publish all legitimate reports received.
3) VAERS delete legitimate reports (after publication) aka not duplicate and not fake/false.
Terrific work as always. Wondering if you're familiar with machine learning to analyze VAERS reports? A great substack on this - might be a nice collaboration
Yes I am. Gary Hawkins and I have been working on it for a very long time. I was familiar with it working with billing and EMR systems for the past 25yrs. Watching the evolution SOAP notes, charts, OCR to EMR, then to diagnosing from the encounter data. I made a niche for myself in my career linking EMR's to billing modules with the NCD's (National Coverage Determination) codes to get the physicians paid! If the right ICD code wasn't coming out of the EMR it would bounce back to the physician for a a proper code. Or at least a suggestion of a payable ICD code for services rendered. lol It's actually this exact spat I had with openvaers.com where they more than doubled all the myocarditis in those red boxes. In the end they were doing word hunting, not MLN and diagnosing. It resulted in a ton a false positives results they were then calling myocarditis as and example. See here: https://www.vaersaware.com/post/openvaers-com-the-myocarditis-stats-the-full-story
Gotcha. I started looking at peer reviewed literature on VAERS and it looks thin, particularly for papers that explore the SOAP field. Seems that a paper on methodology is needed for amateurs like myself to improve their signal search/ interrogation of VAERS
I think it was the University of Kentucky that spent a lot of time in this area attempting to extract/diagnose from physician SOAP/EMR. At one point the MSO/IPA I was in charge of had 11 different EMR's we had to dive into at once! One being from Stanford Hospital. Back in the day Stanford use to use IDX which eventually rolled into GE Centricity. But now like most University Hospitals they mostly use EPIC. I hate EPIC in the sense that is locked down, no more customization is allowed even at the hospital level. As a hospital admin you need to go to Oklahoma for a week and get your EPIC cert to comeback and be allowed to make any customizations. Kind of like air traffic controllers and their training...
Beyond having a data viz interactive dashboard for VAERS, the other just as important feature is to automate all the algorithms to "ethically cleanse" and even up-code events where appropriate. Would this kind of pharmacovigilance evolution be important to the medical freedom movement? I think so. God Bless you doctor!
Albert- were any of those lots tested from my batch I can message you the lot numbers if you don’t have them
No hun none from your batches 028L… or 011J… ;)
AB,
Did you get my correspondence? If so, please respond.
Yes, I responded on your original comment: https://www.vaersaware.com/my-powerpoint
My powerpoint above is probably the quickest overview of the various ways the data is obfuscated. However, if I could summarize my overall observations:
1) There are twice as many dead victims in VAERS (now since dead) because only initial reports are made public, even though follow-up data is still collected, like death.
2) VAERS does not publish all legitimate reports received.
3) VAERS delete legitimate reports (after publication) aka not duplicate and not fake/false.
4) VAERS throttles (purposeful delay) publishing reports.
5) VAERS strips out data elements that were there on initial submission like age and state location data.
I have compelling evidence on every assertion I've made.
I got that one. I sent you a personal correspondence this morning.
checking now...
Terrific work as always. Wondering if you're familiar with machine learning to analyze VAERS reports? A great substack on this - might be a nice collaboration
Yes I am. Gary Hawkins and I have been working on it for a very long time. I was familiar with it working with billing and EMR systems for the past 25yrs. Watching the evolution SOAP notes, charts, OCR to EMR, then to diagnosing from the encounter data. I made a niche for myself in my career linking EMR's to billing modules with the NCD's (National Coverage Determination) codes to get the physicians paid! If the right ICD code wasn't coming out of the EMR it would bounce back to the physician for a a proper code. Or at least a suggestion of a payable ICD code for services rendered. lol It's actually this exact spat I had with openvaers.com where they more than doubled all the myocarditis in those red boxes. In the end they were doing word hunting, not MLN and diagnosing. It resulted in a ton a false positives results they were then calling myocarditis as and example. See here: https://www.vaersaware.com/post/openvaers-com-the-myocarditis-stats-the-full-story
Gotcha. I started looking at peer reviewed literature on VAERS and it looks thin, particularly for papers that explore the SOAP field. Seems that a paper on methodology is needed for amateurs like myself to improve their signal search/ interrogation of VAERS
Hawk is good substack to follow as well as the young man from India.. https://vaccinedatascience.substack.com/p/cdc-wonder-vaers-search-ui-is-comically
I think it was the University of Kentucky that spent a lot of time in this area attempting to extract/diagnose from physician SOAP/EMR. At one point the MSO/IPA I was in charge of had 11 different EMR's we had to dive into at once! One being from Stanford Hospital. Back in the day Stanford use to use IDX which eventually rolled into GE Centricity. But now like most University Hospitals they mostly use EPIC. I hate EPIC in the sense that is locked down, no more customization is allowed even at the hospital level. As a hospital admin you need to go to Oklahoma for a week and get your EPIC cert to comeback and be allowed to make any customizations. Kind of like air traffic controllers and their training...
Beyond having a data viz interactive dashboard for VAERS, the other just as important feature is to automate all the algorithms to "ethically cleanse" and even up-code events where appropriate. Would this kind of pharmacovigilance evolution be important to the medical freedom movement? I think so. God Bless you doctor!