Thread by @barrysmyth

Slovakia's drive to test its entire population over 2 weekends got off to a good start on Saturday with 2.5m tested and 1% positivity. Those 25k positives are now in quarantine. A voluntary programme with opt-outs required to quarantine for 10 days.
(1/n)
theguardian.com/world/2020/nov…

Next weekend the second half will be completed and presumably another ~25k positive cases will be found. So approx. 50k people will be in quarantine for 10-14 days. The remaining >5.4m people will presumably be free to go about their business with limited restrictions. (2/n)

There will be false positives among the 50k positives, possibly a fair few of them, but the alternative is that everyone goes into lockdown so this seems like a reasonable trade-off. There will also be false negatives circulating but there shouldn't be too many of them. (3/n)

Could this work in Ireland? No way our testing maxes out at just over 100k per week. How could we possibly scale up to 2.5m tests per week? No way? (4/n)

It shouldn't take 5m tests to implement, because of pool testing (washingtonpost.com/health/what-is…). This is an approach to testing that combines the swabs of groups of say 20 people & analyses the group/pool as one. If the group is positive then the individuals need to be analysed. (5/n)

If the pool is negative then the individuals must be clear and so no further testing/analysis is needed. If prevalence is low then most pools will be negative so you get to test 20 people at a time, mostly by doing one single analysis of the combined pool sample. (6/n)

E.g. imagine 1,000 people and 10 are positive. We swab everyone and general 50 pools of 20. So that's 50 analyses. In the worst case, 10 pools show up as positive so that means a further 10x20 analyses. So we test 1,000 people with 250 analyses. But we can do better than this ...

There are limitations to pool testing. E.g. it might not show up people with a low viral load by diluting their samples with others. On the other hand, such individuals represent lower transmission risks, so perhaps that a risk worth taking.

IWe can estimate the optimal pool size for a given prevalence (sciencedirect.com/science/articl…) for even better gains. E.g if 1% of people are positive then the best pool size turns out to be around 10. That means that 5m people can be swabbed and tested as 500k groups/pools. (7/n)

Assuming 5,000 (1%) of pools come back as positive, then 50k (5000 pools x 10 people per pool) extra analyses will be needed. Thus, testing 5m people turns into 550k tests, give or take. (8/n)

In Ireland, we can do 100-120k analyses per week so testing the entire population in two weeks means we need at least 2x more. Might that be possible if we pulled out all of the stops? Doing it in L5 would mean minimising spread as testing was going on. (9/n)

By December we'd be done with 50k people in quarantine, but in the clear for Christmas, and with very little virus circulating by then, bar new imports. Too late to plan for now? If so, might it be worth considering for Jan? (10/n)

Practical for Slovakia why not Ireland? What do they know that we don't? Or what do we know that they don't? Discuss?

Slovakia didn't test the under 10's so this brought the numbers down a bit. For us, it would mean testing about 4m people I guess which would require about 440k tests instead of 550k. Slovakia pressed 20k medical staff into service over the 2 weekends of the programme.

What’s wrong with the above analysis? Is it just that we cannot hope to coordinate so twice as many tests as we have been doing? Remember its not a 2x scale up for good, just for 2 weeks? Is it too costly? Consider the economic cost of L5 now and the one we will need in Jan/Feb?

Of course even if population-level testing worked it doesn’t mean cases won’t climb again. Maybe it takes a little longer but it underscores the fact that we need to learn to live with the virus and we are not doing that yet. Lockdown is not living with the virus.

Absolutely. It’s really not a fix unless we really learn to live with the virus. But it might get cases low enough to reset/reimagine test & trace: pool-test border counties & congregate settings, including schools, & implement backward tracing to locate sources of transmission.