I’m afraid I think it is worse than “not perfect” given the sheer number of non symptomatic screenings
The problem stems from the way they have set up counting of people tested, splicing together people from two distinct and very different cohorts.
I think it worse than before
The problem stems from the way they have set up counting of people tested, splicing together people from two distinct and very different cohorts.
I think it worse than before
https://twitter.com/_johnbye/status/1316852121199943680
1/ Routine screening of people with no symptoms and 2/ testing symptomatic are very different pools of people.
They should never have been mixed up.
I completely understand that when the datasets were set up originally repeat screenings were probably not in their minds.
They should never have been mixed up.
I completely understand that when the datasets were set up originally repeat screenings were probably not in their minds.
But that is no reason to continue to embed a mess.
Especially as one group (repeat screening) could probably be filtered through a pooled testing pathway to filter out the negatives in an early triage.
So there is a FINANCIAL and also efficiency/ resources benefit.
Especially as one group (repeat screening) could probably be filtered through a pooled testing pathway to filter out the negatives in an early triage.
So there is a FINANCIAL and also efficiency/ resources benefit.
But first you you have to disentangle to two strands so the base point data on positivity/negativity and unclears can be transparent and the pooled testing methodology can be validated at scale.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
