I have just been reading an excellent blog on statistics with the most recent articles on statistical aspects of vaccine efficacy trials. I am not going to repeat the full analysis and invite you to read it if you are into statistics. It is very well written and explains difficult ideas very well.
If I follow the author’s argument, essentially the Astra Zeneca trial is an example of how not to carry out a highly sensitive vaccine efficacy trial.
While I do not think the author of the blog actually says that there is something wrong with the AZ vaccine, but he presents a very compelling story on why the AZ claims are not supported well by the evidence.
It would have deserved just a note in statistical textbooks, where it not for the fact that the consequences are important for our health. How much protection does the first shot give? Is it OK to delay the second shot? It might all be OK, but sloppy evidence undermines the message.
