Thread by @DearPandemic

1/ Q: Has almost everyone been infected with COVID by now?
A: Recent estimates suggest around 58% of the population in the US and over 70% in England have been previously infected, with BIG increases during the Omicron wave.

dearpandemic.org/numbers-infect…

2/ A recent @CDCgov MMWR report estimated seroprevalence from a convenience sample of blood collected for medical tests.

➡️Roughly 58% of Americans showed evidence of previous SARS-CoV-2 infection in their blood by Feb '22.

cdc.gov/mmwr/volumes/7…?

3/ ➡️ During the Omicron wave from December 2021-February 2022, this estimate increased from 33% to 58%.
➡️ Rates vary a lot by age, ranging from 33.2% for those over age 65 to 75% for those under age 18.

cdc.gov/mmwr/volumes/7…

4/ ⚠️Convenience samples have limitations-people having blood tests for medical reasons may be systematically different from the overall population. Its hard to know which way this bias goes for COVID, but it gives us data Nerds serious pause.

#poptwitter #epitwitter

5/ 🇬🇧 The UK has been notable during the pandemic for having an AMAZING data resource that overcomes some of these challenges. The @ONS Coronavirus Infection Study has been PCR testing hundreds of thousands of UK citizens every 2 weeks since April 2020.

ons.gov.uk/surveys/inform…

6/ What makes data Nerds *swoon* is the population representative sampling. People are tested regularly whether or not they seek out medical care. This type of sampling should give much more accurate estimates of the true infection levels.

7/ The @ONS study estimated:

➡️70.7% of people in England had been infected from Apr 27, '20 to Feb 11, '22 (NOT including the first wave or recent Omicron cases).

➡️These #s increased dramatically during the #Omicron wave—going from roughly 40% to 70%.

Plot from @chrischirp

8/ ⬇️ BOTTOM LINE:

We can now confidently say that the majority of the population in the US & UK has had a COVID-19 infection, w/ huge increases due to Omicron.

9/ The UK estimates are considered more reliable. The US estimates likely capture the trend (huge increases during Omicron) and a broad “ballpark” idea of the overall infection prevalence, but the precise number should be interpreted more cautiously.

10/ Since re-infections are happening more often w/ Omicron, the presence of antibodies or a previous PCR test does not imply protection from future infection. So we can’t really use these data to say we are getting close to “herd immunity,” (even if we could define such thing!).

11/ ⭐ Vaccinations (including boosters!) remain the best way to minimize the harm of encountering this increasingly hard to avoid virus.

12/ Full thread from @chrischirp:

https://twitter.com/chrischirp/status/1517509865714995201

13/ Link to @ONS data:

ons.gov.uk/peoplepopulati…

14/ Nerdy Girl @drjenndowd 's love letter to the @ONS infection study in @bmj_latest:
bmj.com/content/376/bm…

@drjenndowd @ONS @bmj_latest 15/ And some updated to April 2022 estimates for England:

https://twitter.com/mroliverbarnes/status/1521765133067984897?s=20&t=5LGs67oOE18_urV6mAsUIQ