One country is on the down slope of a wave and get kudos for an effective strategy. Some country on an upslope get trashed. And it changes every few months.
We won’t know which countries handled it well for another 5-7 years once analyses can be done that take into account the full impact of strategies.
Then add on top their stats aren’t even comparable due to different levels of testing, different categorization, etc.
Just looking at raw stats says Africa did really well with Covid even without the vaccine. But if you don’t test, you can’t find cases. No cases, no covid deaths.
> Anyone interested in a real world case study on the extraordinary effectiveness of face masks in suppressing the spread of the coronavirus need look no farther than Prague. In this capital city of 1.3 million, among the first in Europe to decree mandatory universal mask-wearing almost four months ago, life has now returned to normal.
Most of Europe in general was presented as having beat COVID before the late summer surge.
No matter what happens, people say it proves their preconceived beliefs correct. "This country is doing better? Well, that's simply because they did [THING I LIKE]. Now it's doing worse? It must be because they're not doing [OTHER THING I LIKE]."
[1] https://www.usatoday.com/story/opinion/2020/07/14/how-czech-...
https://www.businessinsider.com/delta-variant-made-herd-immu...
Africa is doing surprising well with COVID-19, considering expectations. Reasons may include:
* Not kidding themselves like UK and US did that they were "world leaders", and had this in the bag. (1)
* Recent and ongoing experience of combatting other diseases, e.g. HIV, TB, Ebola.
1) "In October 2019, an extensive report two years in the making placed the US and the UK first and second respectively in a global ranking of countries’ pandemic preparedness." https://www.huffingtonpost.co.uk/entry/uk-coronavirus-pandem...
https://www.wfae.org/health/2021-09-30/novant-says-9-of-10-c...
I’d argue it’s all about how Covid deaths are counted and how widespread testing is.
Having worked in healthcare attributing a cause of death is not easy in many cases. There are underlying factors and comorbidities.
In Singapore pneumonia deaths were typically 4,000 per year. That went down dramatically with covid. Why? Because they died of covid not pneumonia.
That might indicate that the mask mandate was ineffective and the evolution was due to something else, or it might indicate people didn't abide by the mandate as soon as it came in.
Top points if politicians noticed this difference off the bat and create the mask mandate to engineer the illusion that it was their policy that worked, even though it would've happened anyways -- aka action bias.
In any case, I think we can all agree people need to stop (1) using different random X-day moving averages and (2) give us the data.
edit: there's an entire literature on methodologies (1) Regression Discontinuity Design (RDD) and (2) Difference-in-Difference (DiD) that's designed to address this problem in analysis.
Confounding trends might exist for various reasons like: labs want people to have a day or two off, so process less on the weekends or some days of the week in spreading that weekend out.
People might interact in different risk settings over the course of a week. (E.G. For some higher risk on the job and shut in on the weekends. For others time to go out clubbing on the weekends and shut in jobs during the week).
So the 7 day running average just says: we can't disentangle to get at the true raw numbers, this is the best we can do to get useful data out of the data-feed.
PS: The only way to get truly useful data would be sufficient testing capacity to issue fully randomized tests on everyone, in a mandatory way, on a regular basis. A 'well designed survey' would probably still have biases but might approach that. To the best of my knowledge (public news media), currently most tests are probably focused on populations perceived to be at higher risk, who show visible possible symptoms, or who are obviously ill.
Or at the very least show the 7-period SD so we can see if the confidence intervals overlap.
You can absolutely use averages to hide data points and data quality.
PS: I disagree with your PS -- if we're being wildly theoretical and impractical about this we should just create a second Earth where everyone wears mask and a third Earth where no-one wears masks and compare all the Earths....
And none of what you're saying addresses the core problem I mentioned, which is that the divergence in the series starts before the policy date.
If I wanted to do the opposite I would pick cities that are inland with temperate climate during that fall, with the air dry and most people inside poorly ventilated buildings.
Transmitting and date rates are determined by multiple factors, many that are known. Risk groups, ability and convenience to do social distancing, population density, ventilation, weather, population movement patterns and more. In this case we got Kansas city vs St. Louis, through the article is trying to making an argument about the world in general.
Graphs like this are basically pointless except for one thing. They imply that there might be a benefit of masks, and thus while its not evidence of effectiveness its is not evidence of the opposite. It is what researcher use as support for doing further studies that later accounts for known factors. If however one uses it to jump ahead, someone else can create a similar graph in a different locations and time to create the opposite conclusion.
(Small note about the time period). It is worth looking at the period between April and August. The two lines swap places which each other. St. Louis started out with a higher rate than Kansas city, and end up with a lower rate afterward. This is a pretty common pattern and follow the old "even out the curve" concept.
https://twitter.com/pbump/status/1466261627074920452
Basically, the more urban and politically blue the county is, the more success they had against COVID, despite the fact that urban locations with higher density and more shared spaces have an intrinsic disadvantage in stopping the spread of disease.
By politicizing public health measures, they got outcomes split by political culture.
Try this: shift the blue curve about 15 days to the left. We may then interpret masks acted in lowering the infection peak, since masks were introduced at an earlier infection stage. In other words, if St Louis reinstated masks at the same infection count as Missouri it might have seen a higher infection peak.
But then, that’s not a result of masks being mandated, is it?
Regardless, if people are willing to voluntarily wear masks before they are required, like you claim, then one thing is clear: Masks do not need to be mandatory for people to wear them. Imagine if they dropped the speed limit from 70 to 50. How many people would start driving 50 a week early to "prepare"? Not many. If people are so resistant to an idea that it has to be mandated for them to do it, then they aren't going to suddenly do it before it is mandatory. And if they aren't resistant to it, then it won't take a mandate.
Also, you can’t fake the response from an impending mandate without a mandate. Most people fill out their taxes early, but if taxes were optional then nobody would be filling them out. The test would simply be to look up when a mandate was announced and then when it took effect.
At the end of May, the places that would institute a mask mandate two months later fell below the other places. Then they stayed below the other places. Attributing that to the mandates is bordering on absurdity. We aren't talking about a week. We're talking about two months. That's long before people are preparing for a mandate that nobody is even considering yet.
You might as well say, "Places that will eventually have mask mandates have higher infection rates two months prior to the mandates."
Look at the underlying data for each mandate in the state and things become much more obvious.
The labeling is weird. The reinstatement arrow should point SSW not NW. That says that during the surge, those counties* had a lower death rate.
Presumably the counties that ever had a mask mandate had higher compliance even during periods when the mandate was lifted, but the article doesn’t discuss any of that, and it’s only a presumption.
* I think the blue line should be labeled “death rate in the four counties”.
It points to the the point on the blue line at July 26. Why should it point SSW? Does SSW even intersect with the blue line at any point?
Unless that chart is a horribly misleading presentation of the data (quite possible!), it's evidence against mask mandates having any discernible impact.
To me, the story here is not the data itself; it's another example of elected officials hiding data that doesn't agree with their position. Is the headline the best interpretation of the data? Not at all. But it's one more data set showing that the areas that are most closely following relevant recommendations are seeing better outcomes than areas that aren't.