So far we've been trying to figure out the proportion of the population that had already been infected, hence the focus on serological studies, but do we have good estimates of the number of new infections that are still occurring every day? 1/n

I ask because this seems key to determine whether "test and trace" is a viable strategy. We need to know that, as much as how many contacts people who are infected typically had while infectious and how much overlap there is between them, to figure this out. 2/n

It seems to me that, in principle, we should have enough data to come up with rough estimates, but has anybody tried? If we need to test and/or quarantine 250,000 people every day, I have no idea what the actual figure is, then "test and trace" may not be realistic. 3/n

So far the main takeaway from serological studies has been that too few people have been infected, which makes sense if you're thinking about herd immunity, but if you're thinking about "test and trace" the right conclusion might be that too many are still infected every day. 4/n

Of course, as long as R is below 1, we could just stay locked down until the number of new infections that occur every day is small enough for "test and trace" to be doable, but depending on how long that is, it may not be realistic for economic and political reasons. 5/n

I think "test and trace" makes a lot of sense if you do it early enough in the process, like they did in South Korea, but it may not be realistic in a lot of places where the epidemic was allowed to spread too much. It seems like an important question that should be studied. 6/6

Here's a back-of-the-envelope calculation to show it's not a purely theoretical concern. Today there was ~25k new cases in the US. If you only detect 1/10, the real # is ~250k. If for each you trace 20 contacts, that's ~5 million ppl minus the overlap you have to test every day.

Again, I have no idea how many people you'd actually have to test every day, but it could easily add up very quickly. It seems to me that, if only for planning, we need to estimate this, but I don't see anyone trying. Repeating "test and trace" like a mantra is not enough.

Here is a website @Roflsaurus16 just told me about. It says that we need to scale up the number of workers doing contact tracing. In theory, it makes sense, but how realistic is this if you'd need to trace 2.5 million people a day or whatever? https://testandtrace.com 

All I'm saying is that, if you're advocating the "test and trace" strategy, you can't just say very general things like that. You have to go into the details of how it would be implemented in practice and show that it can realistically be done, because it's not obvious.