New paper @nataliexdean @datcummings using fraction of #COVID19/SARS-CoV-2 tests that are positive to estimate incidence.
Key challenge is an assumption that they lay out - quote below.
I wrote why this is problematic:
medrxiv.org/content/10.110…
Thread
Key challenge is an assumption that they lay out - quote below.
I wrote why this is problematic:
medrxiv.org/content/10.110…
Thread
I'll add 3 additional thoughts to thread above.
- There is increasing outreach to get people w/ mild symptoms to get tested b/c it's needed for contact tracing to be effective ()
This increases pool of tested w/ lower prev due to allergies, colds, etc.
- There is increasing outreach to get people w/ mild symptoms to get tested b/c it's needed for contact tracing to be effective ()
This increases pool of tested w/ lower prev due to allergies, colds, etc.
- Re-agent shortages have caused test capacity to show huge fluctuations over 1-2 wk time scales.
- Many univ. labs have ramped up (are ramping up) capacity to do asymptomatic screening such as SNF staff. These add 100s-1000s of tests/day of even lower prev pop.
- Many univ. labs have ramped up (are ramping up) capacity to do asymptomatic screening such as SNF staff. These add 100s-1000s of tests/day of even lower prev pop.
Disentangling testing capacity/criteria from infection is super tough. Authors come to same conclusion I did in thread that we need extra data w/ each test to interpret it:
I'd welcome comments from authors of paper!
