The Epicenter and What It Can Tell Us About Prevalence

We seem to recall some science-y/medical types in New York City grumping the other day about the results of a community survey of coronavirus antibody prevalence in Los Angeles. They didn’t like the 1.5-2% prevalence that was found, for some reason.

We guess they all shut up after New York did their own survey and found that the antibody prevalence in New York City was about 20%. We haven’t seen that they have done any more surveys, again for some reason.

And when we say for some reason, we mostly think it’s because these results are counter to the current prevailing narrative.

That’s too bad, because this is very important information. The New York Times has reported that some researchers believe that when the first case of official “coronavirus” was reported in New York City, there were already 10,000 infections there.

They also report that most of the infections in the United States can be traced, by examining viral genetics, to New York City, even in some places on the West Coast. All of Louisiana’s infections come from New York.

To the math!

New York City (proper, not including suburbs) has a population of 9 million. Twenty percent of that is 1.8 million. The NYC department of health has a good COVID-19 information website. They report, as of May 11, 2020, that there were 178,766 reported cases of COVID-19 diagnosed since tracking began, and 44,812 hospitalizations attributed to COVID-19. It is important to state here, that these are not all at once! Current infections and hospitalizations are a fraction of this total. Reported deaths, as not understood by some New York politicians, lag behind cases and hospitalizations, both because it takes time for some people to die after they become sick, and because it takes a while to process the paperwork and make the reports afterwards.

Anyway, doing the math, if there were 1.8 million exposures in New York City significant enough to generate antibodies, 180,000 of these caused enough symptoms to trigger testing and about 2.5% of the total (all the cases not just the tested ones) got hospitalized. New York City reports both “confirmed” and “probable” deaths, by which they mean that all these people are all the way dead, but they are noting their COVID-19 status. Either way it works out to roughly 1% (14,753 versus 19,931) of the 1.8 million.

That’s…not great. Good news/bad news, of course, is that 90% or so of these deaths are in the elderly or infirm and most people have mild or no symptoms, but that’s still not a great number. For perspective, if everyone in the United States got this virus, and the proportions stayed the same, that would be 330 million cases, about 33 million would get tested, about 9 million would end up hospitalized and about 3 million would die.

Luckily, that’s not a possible occurrence, because, even if immunity is only partial or not permanent, it will occur and significant slow transmission of a disease, and also, even in a doomsday scenario, eventually all the people who were going to die will have done so.

What percent of a population can a virus infect? That is, if a population does nothing to mitigate risk and just runs around sliming each other?

That turns out to be a difficult question to answer. This particular virus, again, has that good news/bad new thing going on. It is almost always mild which is good, but then people can run around infecting others, which is bad. Even if it kills less than 1% of the people it infects, if it infects a large number, then those fatalities can add up.

As with all viruses, the spread contains the seeds of its own destruction as well. It basically runs out of victims. When this would happen for the COVID-19 infection if we did nothing about it is anybody’s guess.

Estimates of what would happen if it were business as usual in the US (that is, ignoring respiratory viral threats) are all over the place. Twenty percent of the population infected? Forty? Eighty? And pure case numbers don’t tell the whole story either. Sixty million colds a year happen in the US but that is more annoying than anything else. If this virus spread unchecked would it be closer to the cold scenario, or would it be more like a serious strain of flu, and maintain its current one percent fatality rate?

These are important questions, because viruses are a constant issue in human life. We need to know all the information about prevalence and transmission rates we can uncover, in order to improve our response, and to design meaningful interventions, while discarding those that don’t work.