According to Crisanti, the director of the virology lab of U Padua, as little as 10% of #COVID2019 carriers show any symptoms at all. He sampled repeatedly the entire 3k+ population of Vo ', one of the initial clusters.
He managed to virtually eliminate the disease in Vo', by testing not only the symptomatic patients to confirm they had the disease, but all of their contacts in the preceding days. These were then tested even if they did not show symptoms. And in fact 9/10 of positives did not
show any symptoms. It is these asymptomatic carriers that were doing the bulk of the spreading, at least after the initial measures were put in place. This means social isolation alone is not effective, or at least not efficient. He +- eliminated the disease in Vo' by :
1} confirming the disease in symptomatic walk-ins 2) testing all possible contacts, REGARDLESS OF THE PRESENCE OF SYMPTOMS 3) securely quaranteeing all positives. Now the only people in Vo' with Covid19 are asymptomatic carriers in isolation.
I believe this is similar to the South Korean approach? Maybe also the German? In any case this would imply that the US need a gazillion tests. Assuning the US had 30k known symptomatic infections, and each known infection required testing 100 contacts, 3,000,000 tests are needed
*would be needed for one full screen. And the screen will certainly need to be repeated over time as previously unknown asymptomatic chains would become symptomatic and would need to be tracked down.
The implication is that current measures of social isolation alone will slow down the spread of Covid 19, as they have in Italy, but are unlikely to contain it, as they have in South Korea and China, and hopefully Germany. Only sufficient testing capacity can do that.
If the disease had been contained early (through testing of all contacts of symptomatic patients found positive) the number of tests needed would have been less.But as the number of active cases is now unlikely to be less than 10k, testing would need to increase by
... several orders of magnitude. (Note: the number of contacts to test is a pure guess. Perhaps there will be many shared contacts between different chains, so the multiplier will be in the 10-20 tests per symptomatic carrier?)
Likely we are talking millions of tests, perhaps tens of millions. And the later the tests come online, the more will have to be run. The implication is that the social isolation is not only about flattening the curve to spare the ventilators, but also about pushing it out enough
to allow time for the testing to scale up and quarantine the asymptomatic carriers. Without them around, those infections need never happen, and (it seems to me) the social isolation can be relaxed if not yet entirely eliminated, as in South Korea.
Each day that the sun sets off the coast of Hawaii, the US has produced 54bn dollars in GDP. If social isolation is responsible for 10% of that, the cost of needlessly prolonging social isolatuon is $5bn a day, 150bn dollars a month. I think that should buy a fair bit of tests.
Sources: only in Italian! This is why I am writing this, I am worried that without a WP out, the info might otherwise take a few days to spread outside of Italy. sanitainformazione.it/salute/scovare…
The analysis is basically a translation of this letter to the Corriere Fiorentino from a professor emeritus of internal medicine. Google translate should work pretty well. google.com/amp/s/corriere…
@albertobisin ma si sa se si progetta di applicare questa procedura solo in Veneto o Lombardia o si progetta di estenderla in tutta Italia? Non vedo come il costo possa essere superiore ai benefici di evitare che quello che é successo al Nord si ripeta nel resto di Italia.
Another important point is that testd are more valuable the earlier they are available. If there are a 10,000 active infections to backtrack, a full screen would require perhaps one million tests. If in one week those cases had become 30,000, then 3mn tests.
So a test a week earlier is -say- three times more valuable then a week later. We could imagine the CDC publishing a specification document for a standard test to be run. When a test is needed, it is brought to the lab that can produce the result quickest.
Each lab receives a bounty which decreases through time. How much can we afford to pay? Well if 10 million tests are needed, then we can pay about $500 per test, PER DAY that deploying the tests shortens the lockdown. So if using the tests shortened isolation by two weeks,
We should be willing to pay up to $15,000 per test. Hower in one week, with 100 million tests needed, we would only be willing to pay $1,500. And this is just money. The thousands of lives saved would be worth immeasurably more! And kids could go back to school!
Test everyone. What do I mean by everyone?
(everyone that's been in contact with a known positive case)
Note: this means social isolation is more important! Not less! It suggests we will need to test our way out of this, and the slower the spread the less tests we will need!
Note the 10% of symptomatic carriers is taken from this interviews. I couldn't find any working paper with the actual data. tgcom24.mediaset.it/2020/video/cor…
Important! As @guido2020 pointed out, the study found 50-75% asymptomatics! 90% was the Drs best guess of what fraction of infected patients never make it to a hospital (which in the US and Italy, basically would mean you don't get tested, as of now? Right?)
This is basically in line with the 60% figure @NAChristakis quoted a few days ago. The difference number is important for some policy decisions, and I apologize for the inaccuracy, what I think you should take from the thread is that we need tests like the air we breathe.
