>>>Here's my issue with that (and this is not directed towards you)...the definition of "effective". I'm not challenging the 39 and 42 numbers - just want the context is.
>>>If it's 39% effective against acquiring enough viral load that I don't test positive, I'm not necessarily scared. Since the vaccine still provides much higher protection against serious illness, that's what's more important to me.
>
>For vaccines, "efficacy" means risk reduction measured by controlled study versus an untreated control group. So Pfizer's 95% efficacy doesn't mean that for every 100 exposed vaccinated people, only 5 get infected; it means that after vaccination you have 1/20 of the control group risk of infection.

bit of nit-picking:
Those efficacy percentages are operationally defined as positive PCR test, which was main criterion set forth by FDA. You are totally correct to underline the importance of the group the study was made in: Any new vaccine testing today will have to work against already mutated strains, practically a heavy bias to overcome. IMO a better test would be a triple blind including "best of breed" of current vaccines to level playing field against stronger antagonist.

At least over here further specification like "efficacy against hospital admission" is typical usage here as well as "efficacy against death with/by Covid" - also measured only inside study group and "back then strain".

At least efficiency can be guessed at by comparing monthly age bracket CFR or Covid percentage of total death # in correlation with vaccination level of age group for that time period if infection rate is high and institutions delete vaccination information like CDC did May 1. Even worse, DBs like Global.Health have gender and ethnicity field in data dict, but no list of vaccination tuples of (Vaccine / date). Yuck...