Interesting Study on Covid and Lockdowns by Johns Hopkins

[CitizenBBN]

Basically conclude that lockdowns had almost no impact on Covid mortality rates. They did find that bar closures helped, but that was about the only thing they found that made a real difference.

They did find that the outdoor lockdowns some states had, like closing golf courses, were actually counterproductive, which I think any sane person would see right away. You WANT people outside during these things, something we found during the Spanish Flu epidemic. The viruses spread extremely poorly outdoors if at all, so get out of the enclosed spaces. Going to golf courses etc. was actually a great idea.

https://www.foxnews.com/us/lockdowns-reduced-covid-19-mortality-by-2-study-finds-lockdowns-should-be-rejected-out-of-hand

This was a review of various other studies. I'm curious if they took into account the response of the health systems of the countries and whether it prevented them from being as overwhelmed. That was the general idea of lockdowns and sheltering, etc. This article didn't address that and I don't have time to read their full study review right now.

That to me is the central question. Did it change the curve of hospitalizations significantly? If it did then you would think that would impact death rates, but if it didn't, or if that is not actually strongly associated with death rates, then we have some very interesting conclusions.

Especially given the economic and other impacts of such a shutdown.

[PedroDaGr8]

I try to avoid this sub but I saw the title and knew exactly which study it would be talking about and had to step in.

This study has a NUMBER of significant issues which throw the results into question. As usual, the media coverage of science is abysmal. In this case, Fox News isn't questioning this nearly as well as they should because it fits their agenda. They aren't the only ones who do this and it doesn't make it OK in ANY form, I have similar problems with the fear mongering around variants (and have been vocal about such). Were this paper robust and even moderately legitimate then I would have no problem with the results and questioning lockdowns effectiveness for COVID. In reality, it is appearing far from it, smelling more and more like a propaganda piece to fit a specific narrative.

For example, let's look at the stated mortality effect. The -0.2% effect on mortality is almost entirely dominated by a SINGLE study.


Excluding Chisadza et al. (2021) from the precision-weighted average changes the average to -3.5%

Looking deeper, it becomes clear that the authors basically reinterpreted the results of the Chisadza paper and only took the data that showed that lockdowns don't cause an impact. Otherwise, the impact would have been even greater than 3.5%.

Other issues with this study (which should have been obvious to any media person):

The paper has undergone ZERO peer-review. The paper was published in "Studies in Applied Economics", a journal which primarily publishes undergrad research and is run by one of the authors. This is VERY unusual in the sciences.
The authors are all economic policy experts having little to no training in public health, epidemiology, etc. While this is sometimes OK, taking it in light of the previous, something is awry.
This study is a meta-analysis, the lowest in rigor form of analysis. It is viewed as such because the authors can easily generate what ever result they want by including or excluding specific papers. In this case, the authors excluded a huge number of studies settling on a very small subset of their original batch of papers. Some of the criteria used to eliminate papers were rather arbitrary. For example, "We exclude papers which analyze the effect of early lockdowns in contrast to later lockdowns." This is a puzzling decision which hints at excluding papers which don't fit their goal.
By design it essentially ignored all countries which implemented significant lockdown measures like New Zealand, Australia, and Taiwan. The one paper included which covered NZ only covers through July 2020 and indicated the lockdowns had no impact. This is bewildering as lockdowns were the primary method of control in NZ and are well known to have kept the death toll exceptionally low.
The paper intentionally ignores a number of other factors like the fact that healthcare capacity limits were seldom exceeded, that there are other outcomes besides death , etc.


To be clear, none of this GUARANTEES that the results of this paper aren't valid but it sure appears suspicious.

[CitizenBBN]

I was suspect of the methodology because the article did not discuss having to account for the impact to the health system of not locking down. that is the primary justification, to "flatten the curve" so that the health system can treat everyone effectively.

I had no idea if it was due to typically poor journalism (don't get me started on what I think of them, regardless of political views), or a poor study.

Now, if Fox or anyone is reporting an undergrad non-reviewed paper as a "study from Johns Hopkins", that's beyond shoddy in and of itself. Not surprising but really bad.

that's like reporting one of my college term papers as an "economic analysis by Washington University in St. Louis". lol.

I imagine that's why they didn't push it as a big lead story, but honestly it shouldn't even be reported.

I do think it will be interesting to see the research on this in the coming years as the data is sorted through. I have a feeling that the conclusion will be to have much more targeting in our approach rather than just a complete shutdown. I particularly want to see some international comparisons where more variables can be better controlled, and maybe state to state analysis in the US.

[CitizenBBN]

IMO a good example of targeting would be something like closing bars but not retail stores (due to density), and absolutely not shutting down anything outdoors. that one I never understood. Of course that all has to hold up legally as well.

[VirginiaCat]

IMO, there are way to many variables involved to make a positive or negative statement on this subject without a lot of wiggle room.

How do you prove a negative. Since we had/have lockdowns how do we know how much more widespread the virus/infection rate/mortality rate be without them?

I have similar thoughts on the Vaccine issue with worry about unvaccinated. From all I can read and find, a person vaccinated has just as high of an infection rate as an unvaccinated person. The difference appears to be only in possible severity of infection although that data is not perfectly clear. This seems to me to indicate a person who is vaccinated can spread at the same rate as an unvaccinated person but the only ones in any real danger are those that choose to remain unvaccinated. And in my libertarian mind, if they are willing to take that risk, so be it. That is on them. So lets get rid of the damn masks.