The better testing/tracking you have, the less social distancing you need to achieve equivalent health outcomes. You can focus social distancing efforts more precisely at people who are sick or at immediate risk of being sick and others can get back to normal life a bit more.
-
Show this thread
-
Replying to @NateSilver538 @kylemathews1 reply 0 retweets 1 like
-
Absolutely. Especially epidemiological testing (the population) rather than medical testing (the suspicious cases). If you have sings or symptoms, operate under the belief that you're infected.
1 reply 2 retweets 5 likes -
Replying to @oliverbeige @_StevenFan and
Here's the task for Nate and Nate fans: How do you test for a rapidly evolving exponential distribution under signal lag, with limited testing resources? This is the question everyone should try to answer.
1 reply 2 retweets 8 likes -
I feel like a greedy algorithm that just looks at people adjacent (home/work) to the infected would be the best, cant prove it, but it seems like most bang for buck. Someone more mathematically inclined could say what it means if adjacents are under infected.
1 reply 0 retweets 1 like -
You mean as an early warning system? The problem is that you have lots and lots of asymptomatic transmission, so every case you detect is already a small sample of infected population with time lag. Thermometers at every checkpoint are a much cheaper option to do dragnet search.
1 reply 1 retweet 2 likes -
This might be super naive but just testing within the communities a known infected person has been in contact with. Move from medical testing to epidemiologic testing moving outward from the symptomatic folk.
1 reply 0 retweets 1 like -
You always have to assume extremely limited supply in (diagnostic) test equipment. Things like thermometers are luckily ubiquitous.
2 replies 0 retweets 1 like -
Replying to @oliverbeige @_StevenFan and
My intuition for highly contagious pre-/no-symptom transmission is that testing people *randomly* then extrapolating total population would be most accurate. That may not be best for treatment, though.
1 reply 0 retweets 3 likes -
Replying to @least_nathan @_StevenFan and
My understanding of the epi sampling protocol is that it takes everyone who was in contact with an early observed case and extrapolates transmission patterns from there.
1 reply 0 retweets 2 likes
this is what I meant in my first post, thanks for putting it clearly and a name to it
-
-
Replying to @_StevenFan @least_nathan and
There's a methodology paper in there somewhere. Fast-tracked to Econometrica.
1 reply 0 retweets 0 likes - End of conversation
New conversation -
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.


"Nasty little Buddhist"
Seeking via neuroscience and psychology informed dharma.