confounders can potentially explain all the discrepancies in vaxxed and unvaxxed data. for instance prevalence may just be from differences in behavior. vaxxed are more scared of the disease so they will take precautions more often limiting exposure than most unvaxxed. then the differences in deaths can be from differences in quality of treatment. vaxxed are more affluent on average so have access to better hospitals. better hospitals means less likely to die and early measures that keep patients out of the icu in the first place. there may also be a bias that creeps in when doctors treat patients. a lot of doctors are probably not very fond of anti-vaxxers and see them as a problem so these feelings mat creep in when they're treating patients. maybe on some animal instinct level want the unvaxxed to die. until these confounders are accounted for all the stats that show discrepancies, just show discrepancies and nothing more.