FWIW I have started wading through the csv files themselves — so far expect analyses of this will by definition come to benign conclusions. As self reported data, it by definition cannot contain debilitating severe events. (also if fields in the CSV are any indication, possible there is no option to enter that in the app — could be a built in bias)
VAERS or tying external data to these (which CDC can in principle do, and technically possible with the existing anonymization, but won’t do unless compelled) is going to be required to normalize/compare to VAERS I’m guessing.
The big time streamed event data could be interesting, but takes work & head scratching.
FWIW I have started wading through the csv files themselves — so far expect analyses of this will by definition come to benign conclusions. As self reported data, it by definition cannot contain debilitating severe events. (also if fields in the CSV are any indication, possible there is no option to enter that in the app — could be a built in bias)
VAERS or tying external data to these (which CDC can in principle do, and technically possible with the existing anonymization, but won’t do unless compelled) is going to be required to normalize/compare to VAERS I’m guessing.
The big time streamed event data could be interesting, but takes work & head scratching.