India has already begun to implement random sampling of people who display flu-like symptoms but don’t have any history of travel to outbreak zones 13/https://economictimes.indiatimes.com/news/politics-and-nation/icmr-intensifies-random-testing-to-check-for-covid-19/articleshow/74643047.cms?from=mdr …
-
Show this thread
-
While a large scale stratified random sample may be costly & impractical, in Holland they are taking 10,000 samples from all blood donations in the past week (shared by
@PeterLugtig) https://nos.nl/artikel/2327585-bloedbanken-gaan-opbouw-immuniteit-coronavirus-onderzoeken.html … 14/1 reply 6 retweets 10 likesShow this thread -
You could also imagine beginning with a few small representative cities across the US, as several people including
@cdsamii have suggested. 15/1 reply 1 retweet 3 likesShow this thread -
But the current approach, no matter how large, will always provide biased estimates of prevalence & morbidity/mortality because, as
@deaneckles pointed out, computing standard confidence intervals from big, biased data does not yield valid inference. https://statistics.fas.harvard.edu/files/statistics-2/files/statistical_paradises_and_paradoxes.pdf … 16/2 replies 1 retweet 11 likesShow this thread -
Sinan Aral Retweeted Yashar Ali 🐘
Also... There are some reports that CA has been doing random testing and that this surveillance informed their policy:https://twitter.com/yashar/status/1241525616219787264?s=19 …
Sinan Aral added,
1 reply 1 retweet 6 likesShow this thread -
CA is using testing as "community surveillance"... "broad random tests of otherwise young, healthy people, which allowed health officials to improve their use of scarce medical resources." https://m.arcamax.com/currentnews/newsheadlines/s-2341501 …pic.twitter.com/KbQWjzEJbO
4 replies 7 retweets 17 likesShow this thread -
Health Nerd Retweeted Gregg Gonsalves
Worthwhile thread here:https://twitter.com/gregggonsalves/status/1241793670409596928?s=20 …
Health Nerd added,
Gregg GonsalvesVerified account @gregggonsalvesWant "better data"--well perhaps we should take those non-existent tests and do a random sample of the entire US for a few weeks and then decide to act? As@mlipsitch an actual real infectious disease epidemiologist has said we have more than enough information to act now. 6/Show this thread1 reply 0 retweets 0 likes -
I appreciate the sentiment and, frankly, agree, but fear we're right back where we left it... To reiterate: not arguing we take our limited tests and "do a random sample of the United States." Instead asking if smart sampling of marginal cases or w/ antibody tests makes sense.
3 replies 0 retweets 0 likes -
Is there a fast antibody test yet? Last I heard it was in the works but some time away
2 replies 0 retweets 0 likes -
The FDA in the US updated it's guidance to include serological tests last week. China has conducted millions of them and it looks like an FDA approved antibody test is about 6 wks away (not great)...https://techcrunch.com/2020/03/20/scanwell-aims-to-launch-at-home-15-minute-coronavirus-test-but-it-still-needs-fda-approval/ …
2 replies 0 retweets 0 likes
Fair enough. Don't think anyone's arguing against targeted testing, hopefully the serology will come online soon
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.