An argument for *random* COVID-19 testing. Tesing is biased toward the sick & ill. So we don't know how much COVID-19 there is or how dangerous it is. This uncertainty => over and/or under reaction. We can't manage what we don't measure respresentatively. https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-as-the-coronavirus-pandemic-takes-hold-we-are-making-decisions-without-reliable-data/ …
-
Show this thread
-
Replying to @sinanaral
I dislike this argument very much. Decision-making is important, but there is no feasible, equitable, or realistic way to randomly test people for
#COVID19 during an outbreak1 reply 0 retweets 1 like -
Replying to @GidMK @sinanaral
If nothing else, I can imagine the panic if you suddenly approached a truly random sample and asked them all to get tested, while denying people with symptoms that avenue
1 reply 0 retweets 1 like -
Replying to @GidMK
Totally random sampling makes no sense as we want probability sampling to favor finding positive cases. But the current approach doesn't give us statistical knowledge. I'm not an epidemiologist (don't know the field also see ethics thread) but current test frame teaches us little
1 reply 0 retweets 4 likes -
Replying to @sinanaral
I don't think that's true - it teaches us a lot. What it doesn't give us is perfect information, but I think the reality is that we simply aren't going to have perfect information for some time yet
1 reply 0 retweets 0 likes -
Replying to @GidMK
What we care most about is prevalence and risk (morbidity/mortality). Testing people as they drive up to a tent indeed teaches us little about prevalence or risk, not to mention asymptomatic shedding and transmission. The errors are so large, we can't plan.
1 reply 0 retweets 2 likes -
Replying to @sinanaral
I disagree wholeheartedly. With an infinite supply of testing and a perfectly happy population, perhaps, but in the real world we care most about identifying as many cases as possible in as short a space of time with limited testing facilities
2 replies 0 retweets 0 likes -
Replying to @GidMK
Wow, yes, we completely disagree. Learning about the prevalence of cases in regional & demographic subpopulations will help us find more true positives. Assuming testing ramps up over time, the tests later are better deployed with knowledge of where to deploy them.
2 replies 0 retweets 0 likes -
Replying to @sinanaral
Ofc testing will ramp up over time, but so will cases. If there's some hypothetical future where we are on top of the epidemic and have extra tests to run, sure. For now, it's not a good suggestion imo
1 reply 0 retweets 1 like -
Replying to @GidMK
So then you're in favor of testing anyone who drives up to a testing facility?
1 reply 0 retweets 0 likes
No of course not. Targeted testing is extremely important, using pre-test probability as a guide. If there's no chance someone has COVID-19, and we have limited resources, test those who are most likely to be sick
-
-
Replying to @GidMK
Then you are in favor of probability sampling and we just disagree on the minutiae of the sampling frame and procedure.
1 reply 0 retweets 0 likes -
Replying to @sinanaral
I think it's impossible to avoid that the first priority right now is convenience rather than sampling perfection
1 reply 0 retweets 0 likes - Show replies
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.