Here’s a an editorial written by my good friend James Lyons-Weiler entitled, “If Vaccine Adverse Events Tracking Systems Do Not Support Causal Inference, then “Pharmacovigilance” Does Not Exist.”
Suffice it to say that it is a big mistake to take statements from the CDC about what you can and cannot do with VAERS data at face value.
See also the Bradford-Hill causality criteria.
The linked site has been pulled. Please can you share the paper to James' discussion directly?