Read on Twitter
This appears to be an interesting study and would be consistent with other studies showing a 4x-5x reduction in risk of (re)infection after previous exposure (i.e. there's immunity, but not complete and reinfections are not "rare"). But where's the data? https://www.gov.uk/government/news/past-covid-19-infection-provides-some-immunity-but-people-may-still-carry-and-transmit-virus
Specifically, I'd like to see numbers for infections in the group without existing antibodies. It's quite unlike the UK to push stuff out without the actual study being available, so does anybody know if it's available somewhere?
Okay, paper is here: https://cdn.prgloo.com/media/download/d44829149b77403eb23c5f93a0a6aad8. Some key summary points:
Reinfections in 0.67% of cohort
New infections in 2.91% of cohort
These numbers are consistent with a ~4x reduction in risk, which was also observed in this study: https://www.medrxiv.org/content/10.1101/2020.11.18.20234369v1
Reinfections in 0.67% of cohort
New infections in 2.91% of cohort
These numbers are consistent with a ~4x reduction in risk, which was also observed in this study: https://www.medrxiv.org/content/10.1101/2020.11.18.20234369v1
Symptomatic infections were ~2x lower in reinfections (34%) vs new infections (78%). In the reinfections, ~90% had antibodies upon reinfection.
These data have been used to conclude that reinfections are rare - because they're only seen in ~1%. That's not correct, since we need to compare to the new infection rate, which in this study was ~3%. So reinfections are _lower_ than new infections, but not "rare".
Similarly, it's often said that reinfections are mild. Based on this study, there was a 2-fold reduction in symptomatic disease upon reinfection vs new infection, so while it's true to say that reinfections have symptoms less often, in ~1/3 of cases they still lead to symptoms.
