Most people who live in Ohio are probably aware by now that we're experiencing another new surge of COVID cases. If you hadn't seen yet, or if you don't live in Ohio, now you know: Ohio is experiencing another new surge of COVID cases.
But you might not know in what parts of Ohio most of those new cases are happening.
I live in Cleveland. I'm certainly very concerned and troubled by new spikes in COVID cases anywhere in Ohio or anywhere else for that matter, because I don't want people to get sick and die, but if I want to assess the current level of risk for myself or people around me, I want to know what's happening here in Cuyahoga county. So let's take a look at what's been happening in my county and elsewhere in the state.
I'm going to focus on new hospitalizations. New case counts are important, but are highly dependent on how many tests are being run and on who is being tested. To me, changes in hospitalization numbers are a better indication of how the spread of the disease is changing. (If you're interested, you can see county case data on the state government's COVID dashboard.)
Here's a graph of the hospital admissions in Cuyahoga county by day of the pandemic (all results in this post use the data made publicly available by the Ohio Department of Health):
It's important to note that the numbers from the most recent days are incomplete and will increase somewhat as time goes on. I do see a hint we might be starting to tick up after a long declining period. But there's certainly been no major spike here. In July we were briefly getting around 30 or more new admissions per day. Then the numbers steadily declined to the point where for much of September we only averaged around 2-3 admissions per day. That's quite an improvement!
What happened in July? As far as I can tell, the most important thing that happened in July was a mask mandate went into effect - first county specific mask mandates, but not long after, a statewide mandate. I suspect, though, that adherence to the mandate is not as good in some parts of the state as others.
Ohio has three major cities, Cleveland, Columbus, and Cincinnati, that are in Cuyahoga, Franklin, and Hamilton counties, respectively. Wherever I go around Cleveland, people seem to be generally good about wearing masks now. I suspect that tends to be more true in large cities in general and less true in other parts of the state. Here's a graph showing changes in the new hospitalization numbers (per capita) for the three largest Ohio counties vs. the rest of the state:
(Note: this uses the extrapolation method I detailed in this post to get more accurate estimates for the recent days that have incomplete data.)
We can see that in the big cities, after the mask mandates went into effect there was a big decline and although there may be a recent uptick, the numbers have remained at a level that is close to the lowest since March. In the rest of the state combined, the recent spike has pushed the numbers to the highest level at any point during the pandemic.
I will say that although I think masks clearly help, they aren't the whole story - there are also many other factors that basically can be summarized as how seriously are people taking the pandemic and how seriously are people taking efforts to slow the spread.
Cuyahoga, by the way, has had the most dramatic improvement. From March through July in aggregate, out of 88 counties in Ohio we had the very worst per capita hospitalization rate. August? 53rd out of 88 (first being best). September, 23rd out of 88. And October thus far? 14th out of 88. So I think we should appreciate the efforts we've all made that have saved a lot of people from sickness and death.
Anyway, it occurred to me that this divergence between the big cities and the rest of the state could very well be related to how politicized the pandemic has become. We have a Republican president and we have a right wing media who have pounded into the heads of their followers messages like: COVID is just like the flu, masks don't work and/or might be dangerous and/or are "unmanly," stay-at-home orders were government tyranny, we should just open everything up and let "herd immunity" take over, etc., etc., etc. And it's very clear that as a result of this, Trump supporters are much less likely than the rest of the country to view COVID as continuing to be a serious threat that we have to continue to take serious measures to address, even as the deaths continue to pile up. So I decided to take a look at whether data support the hypothesis that this polarization is affecting COVID spread in Ohio.
Here's a graph of the per capita hospitalization numbers by county from August through today (August being the cutoff point because mask mandates went into effect in July) plotted against the share of the 2016 presidential election vote received by Donald Trump:
Indeed, we see that the more a county voted for Trump in 2016, the more likely it is that that county has had a bad COVID outbreak since the statewide mask mandate went into effect. I was actually surprised at the strength of the apparent effect - the two most pro-Trump counties are the two counties with the worst outbreaks!
Now, certainly, this is suggestive evidence but not proof that Trump support leads to worse COVID outbreaks. There are other potential explanations and I'm open to hearing them.
One alternate explanation that someone raised when I made a similar point recently, and that I don't buy, is that the reason outbreaks are now worse in rural areas is because the worse outbreaks that previously occurred in cities caused a population immunity effect that has subsequently slowed the spread in cities.
This explanation doesn't make sense, because even the hardest hit places in Ohio had nowhere near the fraction of the population become infected as somewhere like New York City. In New York City, nearly 3 out of every 1000 residents has died of COVID, whereas in Cuyahoga county, about 2 out of every 1000 residents has been hospitalized, to give an idea of the magnitude of the difference. Even in New York City it is questionable whether there has been enough population immunity to have a significant effect in slowing disease spread. In Ohio the idea is completely far-fetched.
But let's humor that explanation and address it with data. If cities in Ohio are now doing better because their previous worse outbreaks have caused a protective effect, we would expect that there would be an inverse relationship between how hard hit a county was in the earlier part of the pandemic and how hard hit it has been recently. Is there such a relationship?
No, there is not. There is no relationship whatsoever between hospitalization rate from March through July and hospitalization rate from August through the present. (There is a measure called R-squared of how strong a linear relationship is that goes from 0 (no correlation) to 1 (perfect correlation). The R-squared of this graph is literally 0.00. If you're curious, the R-squared of the previous graph on correlation with Trump support was 0.10, with a p-value of 0.002 - so the claim is not that Trump support guarantees a bad outbreak, but rather significantly increases the chances of one.)
So that takes us back to the explanation that rural counties are now tending to be hit worse because people there aren't taking the pandemic as seriously, which is probably related to Trump support.
This analysis is specific to Ohio, but I know that there exist data showing similar effects at both the county level and state level when looking all around the country.
So I guess the message is: the virus doesn't care whether you believe in it. In fact, the virus (if we are anthropomorphizing the virus) prefers it if you don't believe in it, because that makes it easier for the virus to infect you and people around you. If people don't take the pandemic seriously, these problems will only continue to get worse.
But the other message is, be concerned, but don't panic, about rising COVID numbers in Ohio. We collectively know so much more about how to stay safe than we did in March. If you take that knowledge into account and take measures to stay safe, you're much more likely to stay safe than if you ignore that knowledge. If you and all the people who live around you do that, it is possible to keep the numbers down. So keep wearing masks in public, and keep avoiding dangerous gatherings of people, especially crowded indoor gatherings. We are all in this together and the data suggest that in places where people believe that we are all in this together and act accordingly, the results will be better.
No comments:
Post a Comment