Sunday, April 26, 2020

Half-Baked Ideas

As usual, I feel irritation when people make arguments without thinking them through.  The two half-baked ideas I’m thinking about today are (1) We cannot reopen the economy safely without wide-spread testing and (2) App infection tracing will allow us to reopen safely.

First, I could dispute that reopening safely soon is possible in any set of circumstances.  But I don’t want to go that route.  Instead, let’s look at what testing or app infection tracing get us.

I’ll start with testing.  We have several options:

  1. Tested recently and you have the virus
  2. Tested recently and you do not have the virus
  3. Tested for antibodies and you probably have recovered
  4. Tested for antibodies and you probably never had the virus
  5. You haven’t been tested recently and you are a complete unknown
Who do we allow to go back to work?  Remember, these great medical thinkers want us to be safe, so they don’t want workers spreading the virus.  Tested #3 is the only really safe worker.  Tested #2 is good for at least a day after testing; but thereafter, could have picked up the virus.  Tested #4 is a person who hasn’t had the virus and enough time has passed to develop enough antibodies for protection, but they might also still have an infection, or they might get one tomorrow.

In short, only tested #3 going back to work is the safe approach.  But that means we only allow you to go back to work if you’ve had the virus.  Numbers today show only 1 million are in that category, and most of those are presumed from virus (not antibody) testing and recovery.  Actual antibody testing has been limited to sample sizes of about 1,000 people in several areas.

So, let’s assume we can scale up antibody testing massively in a short period (probably a poor assumption).  If the current sampling is correct, maybe 2% of the population has had the virus and recovered.  It’s almost certainly a much smaller percentage since the sampling was in high-case areas.  But let’s make an even more ridiculous assumption, 5% of the working population has had the virus and recovered.  People want to keep social distancing and masks, so mitigation is expected to continue to slow the spread.

Can our economy run on 5% of it’s workforce?  Can it run on even 20% of it’s workforce?  My assertion would be no.  Too many without paychecks and too many businesses without sufficient employees.  We would have a crash.  So the ‘must do testing first’ is a half-baked idea.

Now let’s look at app tracing of infected people.  Let’s make the ridiculous assumption that 75% of the population old enough to be out on their own with phones has signed up, downloaded the app and keeps their status up to date.

Let’s further assume that we have the 5 test cases we looked at in the previous discussion.  Everyone keeps their test status up to date in their app.  However, in this situation, we just have slowly growing testing.  It’s not yet widespread testing.

So what is the safety benefit from this app?  Supposedly, it’s going to tell you (the government) who has come in contact with a contagious individual.  Maybe the app will also tell you (the user) when you’ve been exposed.  So far, no safety benefit yet.  If you’ve been exposed, you might get priority for testing.  Some might assume that contagious and exposed people will be ineligible for working and would be quarantined and isolated.  But if the infected are quarantined or isolated, there would be no exposures for the app to detect and trace.  Hence, one assumes there is neither effective mandatory government quarantining or voluntary isolation.  Again, no safety benefit.

Suppose the primary safety benefit or use is to identify individuals ineligible to work.  And the app tracking is all post exposure after someone is tested and updates their status.  That is, any contacts in the last 14 days has to self isolate for another 14 days.  So, nobody gets to work unless the app gives them a clean bill of health for the last 14 days; i.e., no exposure during that period.

Ok.  Now suppose the app became available and the 200 million workers in the US all downloaded it on day 1.  The 1 million cases update their status.  The 6 thousand tested for antibodies do the same.  There is no historical contact data, so everyone else shows clean, irrespective of their actual status.  Let’s assume that the government is smart and prevents reopening for 14 days until contact histories are available.  With current mitigation, you might have 1.5 million now possibly exposed.  So everyone else can go back to work safely?

Well, unless you take into account the 75% of infected who were asymptomatic or had symptoms comparable to a cold or flu.  The app simply cannot find those people and in turn, the people that they came in contact with.  The app would give government the ability to do contact tracing for those hospitalized.  But with 75% asymptomatic, that’s a pretty useless capability.  It also assumes a huge contact tracing operation.

So app tracing might supplement testing, if enough people sign up.  But by itself, all it would do is allow widespread reopening with a lot of contagious, asymptomatic workers.  Workers that would be spreading the infection.

So, I consider app tracing/tracking to be another half-baked idea.

No comments:

Post a Comment