As of August 9, net breakthrough deaths stood at 8/million and net breakthrough hospitalizations stood at 35/million, up 900% and 733%, respectively, since April 20.
Current numbers: https://www.cdc.gov/vaccines/covid-19/health-departments/bre...
Snapshots: https://archive.today/https://www.cdc.gov/vaccines/covid-19/...
Putting my Bayesian hat on, a reasonable conclusion is that if the breakthrough hospitalization/death rates had improved, they would not be holding them back on the eve of the approval of the vaccine.
Also, interestingly, CDC stopped reporting breakthrough infection rates that do not result in hospitalization/death earlier.
The page says:
> Vaccine breakthrough cases are expected. COVID-19 vaccines are effective and are a critical tool to bring the pandemic under control. However, no vaccines are 100% effective at preventing illness in vaccinated people. There will be a small percentage of fully vaccinated people who still get sick, are hospitalized, or die from COVID-19.
First, I must applaud them for the reasonable argument there. Now do Covid19.
Second, it is not the absolute percentage that matters. It is the trend. Frankly, a vaccine where 1/100,000 who take it die of the disease it is supposed to protect against sounds a little far away from "vaccines are 95% effective".
Also, deaths lag.