[pepperberg]

I have one question regarding the current Pfizer data (it would be great if anyone with experience in statistics could help out):

UK healthcare professionals received this information called "REG 174 INFORMATION FOR UK HEALTHCARE PROFESSIONALS" [1] It states that the "efficacy of COVID-19 mRNA Vaccine BNT162b2 [for 75 years of age and older] was [...] 100% (two-sided 95% confidence interval of -13.1% to 100.0%)"

What does this mean? With my very minimal university statistics knowledge I would say that they are 95% sure that the efficacy lies between -13.1% and 100.0% for age >=75. And that means: We simply do not know (probably due to lack of data in this cohort). Where am I making an error?

Thank you in advance!

[1] https://assets.publishing.service.gov.uk/government/uploads/...

[spuz]

If you look at the formula for determining vaccine efficacy, you'll see it's determined by comparing the risk of being infected while unvaccinated vs the risk of being infected while vaccinated. If the risk of being infected while vaccinated is higher than the risk while unvaccinated then you will get a negative efficacy.

I believe you are right in saying we simply don't know what the efficacy is for those over 75 as a result of this study. Given that the range extends to 100% that tells us that no one over 75 who received the vaccine also tested positive for the virus. However the lower bound tells us that even if the vaccine had increased the chance of being infected by 13% we still would have expected 0 over 75 to be infected by the time they were tested simply because the probability of being infected is already so low.

https://en.m.wikipedia.org/wiki/Vaccine_efficacy

[pepperberg]

Thank you for your explanation!