Read on Twitter
1) Thread addendum to my previous vaccine study analysis-
Looking at "final" Pf*zer, pre-EUA data...Looks like they dropped the efficacy population from 44,000 to 36,000...and there's this little gem "Some participants became infected in between taking the first and second
Looking at "final" Pf*zer, pre-EUA data...Looks like they dropped the efficacy population from 44,000 to 36,000...and there's this little gem "Some participants became infected in between taking the first and second
2) ..doses, highlighting the need to get the second dose (efficacy after just the first dose was only 52%). If you take both doses, it's very likely you'll be protected from COVID-19, at the very least in the short term."
3) For those not in the Clinical Research world, this is a big no-no...you just dropped treatment failures in an overwhelmingly skewed manner, whether intentional or not. They dropped anyone who got sick before the second shot from the "efficacy evaluation"
4) ...What needs to be assessed is the ITT population...It's called Intent To Treat, as this captures those who get the disease anytime after the study begins...you know, how things work when you don't live in CandyLand. Pretty pathetic...
5) That story about designed to be approved (in Forbes) was being kind!
Also, they claimed no new "severe" covid cases from the last interim analysis...although the interim stated "covid cases" and this analysis claims "severe covid cases"
Also, they claimed no new "severe" covid cases from the last interim analysis...although the interim stated "covid cases" and this analysis claims "severe covid cases"
6) ...maybe I'm giving too much credit, but if those two evaluation points are equal, it's a statistical impossibility.
