COVID-19 May Be Slowing: Now What

In a short humor piece entitled "The Dog Who Bit People", James Thurber tells about his travails with a dog of his mother's named Muggs who (as the title explains) bit people.

One morning when Muggs bit me slightly, more or less in passing, I reached down and grabbed his short, stumpy tail and hoisted him into the air. It was a foolhardy thing to do, and the last time I saw my mother, about six months ago, she said she didn't know what possessed me. I don't either, except that I was pretty mad. As long as I held the dog off the floor by his tail he couldn't get at me, but he twisted and jerked so, snarling all the time, that I realized I couldn't hold him that way very long.

I've been thinking of late about that image of Thurber holding the snarling Muggs in the air, as people talk optimistically about how "the curve is bending" in regards to the COVID-19 pandemic. The curve is indeed bending in much of the world, and indeed perhaps here in Ohio where the number of new cases and the number of new hospitalizations each day have been pretty flat for the last week.

The problem is that the curve is bending under current conditions, and those conditions involve much of the world being under a stay at home order. It's not exactly a surprise that after a few weeks of most people staying in their homes and only coming out to do a few essential activities (and social distancing when they are going about essential activities) the spread of the virus would have significantly slowed. However, that slowing is dependent upon the staying at home and social distancing that we are all doing. The virus is not like an army that will realize it is defeated and fold up its tents and go home. It's still out there, moving through hosts at a slower rate than before, but biding its time until more hosts become available. The spread of a virus is fairly mathematical. How fast it spreads is determined by how easily passed from one host to another it is balanced against how many contacts the people with the virus have. We've reduced the number of contacts, and so the spread has slowed. But it's still just as contagious as before. If we go back to our offices and restaurants and schools and theaters, the transmission rate will pick up again and we'll be back where we started.

So here we are, like Thurber holding Muggs in the air. Under present conditions, the virus's ability to bite us is reduced. But we know we can't maintain those circumstances indefinitely.

There are, of course, a couple of means of relief. Some viruses spread less well in typical summer weather than they do in typical winter weather. That's why "seasonal flu" is, well, seasonal. Thus far there's not necessarily a lot of evidence that this is the case with the coronavirus, but it may be. In that case, we might get a break over the summer (while the virus went off to ravage the southern hemisphere) only to see a second wave in the fall. That would at least buy us time to improve our production of tests and to keep working on a vaccine.

The other way out is through "herd immunity". Right now, almost every human the virus encounters is a potential host able to give shelter to the virus and pass it on to other humans (though some humans will feel sick in the process and others will not.) However, if it's the case (and we think it is) that someone who has been infected with the coronavirus then develops antibodies capable of fighting off the virus and is thus immune to the virus for a long period of time, then after a sufficient portion of the population has had the virus and developed antibodies, the spread of the virus will slow. Herd immunity works a bit like social distancing, but without the necessity of staying home. Just as with social distancing, herd immunity slows the spread by cutting down the number of contacts through which the virus can be passed on. If one contagious person comes in contact with a dozen different people, but all of those people are already immune to the virus because they've already had it, then the virus doesn't find another congenial host and after fighting its last stand against its current host's immune system, the population of viruses in that person dies off and becomes a dead end for the virus population. Herd immunity is what keeps diseases that used to be common like chicken pox from spreading much at all anymore.

The problem with herd immunity is that a lot of people need to have the disease before it develops (unless you are able to take a shortcut by developing a vaccine and vaccinating everyone.) So far we know of 400,000 cases of COVID-19 in the US. Those, however, are just the number of people who have been tested and have turned out to actually have the disease. The number of people who have actually had the disease must be much larger. The question is: how much larger? If we look at the data on the COVID-19 tests that we've been giving to people over the last week, we find that on average 20% of the tests are positive. However, the people who are being given tests right now are not a random sample of the population. Given the limited supply of tests, we're mostly testing only people who either show symptoms for the disease or who have been exposed to someone known to have had it. This means we should take that 20% number as a very high upper bound. We can't imagine that more than 20% of the population has had the virus, and indeed we need to assume that it's a good deal less because we're only testing the people who are most likely to have it. If for each person who has tested positive, there are ten more who do have the virus but haven't been tested, that would give us about 4 million infections. That doesn't seem like a totally crazy number. If I had to commit to some sort of a guess as to the number of actual Americans who have had COVID-19, I'd put it somewhere between twice and half that: 8-2 million people who have had it. That's between 2.4% and 0.6% of the US population. And that's not nearly enough to produce herd immunity.

So even if it's true that our number of new cases per day is slowing, and thus that after a lag of a week or two the number of deaths per day will cease to grow and start to shrink (which yes, means that today's nearly 2,000 deaths is not the worst day we'll see, because the deaths we're seeing today are a result of the smaller number of people who got sick a week or two ago) we can't imitate Thurber. His solution to the problem of the furious dog he was holding by the tail was to heave it out the door and slam the door after it. The problem was that he forgot the bad door was open, and so Muggs ran around, came up the back stairs, and was ready to vent his fury on Thurber until the writer saved himself by climbing onto the mantelpiece and sheltering in place. To play out this now rather stretched analogy, we can't just come out of our current defensive posture which is holding off the virus, because if we do so we'll just see it spike again. And yet we also can't continue to all stay at home. At the slowing rate the virus is spreading we would still be staying at home a year from now and we'd no longer have an economy.

If it sounds like I've just said we can't come out of isolation and we can't stay in isolation, you'd be right. Are there other options? What countries such as South Korea have done is to test lots and lots of people, and to have the people who test positive remain in isolation until they recover. This would allow the people who do not test positive as actively having the disease (which would include both people who have never contracted it and those who did but have recovered) to go back to somewhat normal behavior. To pull this off, however, we'd need to be able to move up from testing an average of 140,000 people a day (as we have over the last week) and instead test millions. If we concentrated those tests on the people most likely to have been exposed to the virus, we could probably start to identify and quarantine the people who are actually contagious and thus be able to start to lift restrictions without seeing a spike in infections. How fast we can get to that testing volume (and whether we can successfully get the people who are contagious to stay quarantined) remains to be seen. But it's important to be clear that just seeing the new case or new hospitalization curve "bend" does not mean that we've beaten the virus and we can all go back to normal. Unless it turns out that the virus is weather dependent, seeing the curve peak and decline under current conditions does not mean that the virus is beaten for good, it just means that it's beaten so long as we maintain current conditions. If we change the conditions, we are quite likely to see spread pick up again.