2 ms·

Regarding the Cleveland Clinic link, a random one I picked out to verify, the confidence is actually very low, or more specifically, the statistic most relevant

by lucb1e 5y ago

Regarding the Cleveland Clinic link, a random one I picked out to verify, the confidence is actually very low, or more specifically, the statistic most relevant to your quote has a 95% confidence interval of zero and infinity: > vaccination was associated with a significantly lower risk [...] but not among those previously infected (HR 0.313, 95% CI 0 to Infinity) I'm no mathematician or statistician, but that sounds like the conclusion should have been that no reliable recommendation can be formulated without more data. Regarding your chosen quote specifically, there the sentence continues and says > vaccines can be safely prioritized to those who have not been infected before. The keyword being "prioritized"; not that they don't recommend vaccinations if you already lived through COVID-19, which (to me) your comment is suggestive of. Don't know about the rest, I don't have time to check them all out now. Hopefully someone else has more time to properly look at this because your suggestion goes against what most countries are currently implementing so that's interesting.

hammock 5y ago
>95% confidence interval of zero and infinity The reason the CI is like that has to do with divide by zero, not a low sample size. There were ZERO reinfections among previously infected people, regardless of their vaccination status. As the study says: >COVID-19 did not occur in anyone over the five months of the study among 2579 individuals previously infected with COVID-19, including 1359 who did not take the vaccine. The conclusions of the researchers are not wrong or mistaken - they accurately represent the data. As for the "prioritized" comment, that's just the researchers being politically correct. As you point out, indeed these conclusions go against most countries' policies today. It's a tragedy that natural immunity is completely suppressed from the policymaking conversation.

by lucb1e 5y ago

hammock 5y ago

>95% confidence interval of zero and infinity The reason the CI is like that has to do with divide by zero, not a low sample size. There were ZERO reinfections among previously infected people, regardless of their vaccination status. As the study says: >COVID-19 did not occur in anyone over the five months of the study among 2579 individuals previously infected with COVID-19, including 1359 who did not take the vaccine. The conclusions of the researchers are not wrong or mistaken - they accurately represent the data. As for the "prioritized" comment, that's just the researchers being politically correct. As you point out, indeed these conclusions go against most countries' policies today. It's a tragedy that natural immunity is completely suppressed from the policymaking conversation.