Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Coronavirus Update: Regional Analysis
I’ve done further analysis of the Coronavirus death data for the US. Specifically, I calculated the seven-day moving average of reported deaths for each state (plus DC), from the beginning of the outbreak in early March through December 10. My goal was to determine whether there are any regional patterns that might provide insight into the effectiveness, or ineffectiveness, of various policy responses.
It is difficult to present such data. I can easily make a chart showing all 50 states (plus DC), but such a chart would be too confusing and busy, with 51 overlapping lines.
As you’ll see in the chart below — and have seen previously if you’ve read my prior posts — the US experienced a first wave between March and May, peaking in mid-to-late April; then a smaller second wave in the summer, peaking around August 1; and is now experiencing a third wave that continues to mount.
Sometimes it’s helpful to just sort the data, look at it, and see if anything jumps out at you. To do so, I picked three dates, and ranked the states by their average death rate on those dates:
- April 20, which was around the peak of the first wave
- August 1, which was around the peak of the second wave
- December 10, which was the last date reported in my data set
Here is what I found. The reported “rate” is the seven-day average of deaths per million for the indicated state and date.
Notice anything?
The first wave is highest in the corridor from NYC to Boston, followed by other areas of early outbreak (like New Orleans and Detroit). The second wave is highest in the hot states. The third wave is highest in the cold states.
So, as I apparently have too much time on my hands, I prepared a chart grouping the states as follows:
- NYC to Boston: New York, New Jersey, Connecticut, Rhode Island, Massachusetts
- Hot States: Florida, Georgia, Alabama, Mississippi, Louisiana, Texas, New Mexico, Arizona, Nevada
- Cold Stats: Michigan, Illinois, Wisconsin, Minnesota, North and South Dakota, Iowa, Nebraska, Kansas
Here is the result:
The US as a whole is in orange. You can see the three waves.
NYC to Boston is in green. You can see the huge initial spike, very low rates over the summer, and a recent modest increase in the third wave.
The Hot States are in red. You can see that the second wave in the summer is concentrated in these states.
The Cold States are in blue. You can see a modest first wave, low rates over the summer, and a very significant spike since November 1.
The rest of the states are in yellow. You can see that there is less variability in these states, with a modest increase in the third wave.
I hope that this proves helpful. It is not conclusive, but it tends to indicate that various policy differences between the states are not particularly helpful. After the first big outbreak in the major cities of the East Coast, the weather seems to be a significant factor.
It would be nice if we had a news source that presented this sort of relatively simple, dispassionate analysis. It is annoying to have to answer these questions on my own, by going to the original data.
ChiCom delenda est.
Published in Healthcare
Glad we’ve got you, Jerry.
Thank you.
Great post. Glad you have too much time on your hands.
I’d recommend removing the “ChiCom delenda est” part. Our ChiCom Overlords most likely will not appreciate the sentiment when they take over in January.
Tim
Coconino County is the cold county in Arizona, with the County Seat in Flagstaff sitting at 7,000 feet. The week ending December 5th was the highest on record. The statistics come in a few days late, so I anticipate that the week ending December 12th will also be well higher than the current number shown. The Navajo Nation is on a 57 hour lock-down on the weekends again, from Friday nights to Monday mornings.
They prefer to lead the nightly news with “stunning numbers of new cases”. So much scarier than some actual analysis.
It is like when people go indoors, where transmission is most common, there is more spread.
What is the over/under on the percent of Americans that get a Covid19 vaccine this year? I put it at 30%.
I hate that we keep referring to “waves.” There are no waves. It’s a seasonal virus, and the US is a big place that spans a few different climate zones. But that’s a bit harder to sensationalize. Imagine if we talked about daily deaths from all causes, or if we did this every cold/flu season. This national hysteria is a mental illness, but I’m not sure that people are still buying it in quite the same numbers as before.
I have a theory about the recent ‘surge’: flu season
Hospitals especially ICU are always busy in December.
In California the governor said 85% ICU usage is code red.
In December ICU usage is normally 90% or higher because it’s flu season!
If you set the PCR cycle at 35 or above, everyone will test positive for covid
There was a test in Austria or Germany where a glass of coke tested positive for covid.
In Los Angeles, half the covid cases and deaths are Hispanic. How many of these patients are undocumented?
The regions that border Mexico have more ICU patients than the northern part of CA and AZ.
I remember years ago Heather Mac Donald asking why are lines for the Emergency Room so long in CA, AZ, NM?
The answer is obvious
This generally comports with a theory I’ve had for years, based on admittedly subjective observation. The “flu season” used to be pretty well limited to the winter months, when people tended to stay indoors and in close proximity to other people, coupled with limited air exchange. The “flu season” nowadays can go on pretty much all year. It seems to follow the widespread use of air conditioning. People are just as likely to stay inside during the hot months, especially in hot weather, and particularly areas with high humidity.
This doesn’t explain everything of course. The Northeast spike was undoubtedly caused to some extent by Cuomo’s & friends’ imbecilic project of population control and herd immunity. People do more outdoor recreation in hot but dry areas, particularly in spring and fall.
It’s a credible theory which also confirms my bias.
97% of viral transmission occurs indoors.
The first one’s not true. Nationally, the percent of tests that are positive for Covid was – last I checked at end of November – was about 10 or 11 percent. And my understanding is that most PCR tests are at cycles over 35.
The second one is true, but it was for a new rapid response test developed in Austria.
You may explain but is it deaths per million, cases, positive test, or population?
Jerry presents hard stats on COVID-19 deaths; you respond with inherently soft stats on COVID-19 “cases”. Why?
I grow irritated by scare headlines on “case” numbers. Hospitalizations and deaths provide a more reality based snapshot of how things are going.
Even the term, case, has become muddied. Pre-COVID, the medical community uses that term to indicate someone who is receiving medical treatment. So a positive test, where you go home and wait it out for 14 days, isn’t really a case.
While we’re talking about information, I’d like to know what’s going on in hospitals. The information is generally vague with little hard data.
We are getting numbers on actual hospitalizations and occupied ICU beds.
As someone else pointed out, perhaps some comparisons in previous years, seasonally, would be helpful.
Another statistic I’d like to see is length of stays.
Percentage wise, how many people have a stay of no more than 3 days? How about 4 days? Run that up to, say 2 weeks. After that, we’ll call it long term.
Instead we get stories of nurses (it’s usually nurses, not doctors) of how “terrified” they are (I see headlines using that very word).
We’re also being told there are staff shortages. That the real choke point isn’t number of beds, but enough staff to service those beds. So give us some numbers.
The more raw data we have, the more we’re able to sort through the emoting we’ve been seeing.
To know what these piles of numbers actually mean to us humans–what the impact of covid is–we’d need to see the change in life expectancy for the population, and also for those who are paying the highest cost in human happiness.
CDC have that data. Have they published it? If so, could you provide a link?
If they have that data they should publish it. But why is the CDC involving itself in human happiness? It’s not that it’s not important, but part of the problem the CDC has as an institution is it doesn’t stay in its lane.
So you have people who advocate that CDC should involve itself in traffic deaths, because someone referred to traffic deaths as being into epidemic proportions.
Or that racism is a disease, so the CDC should be involved.
If anything, the CDC should be careful to stick to its knitting.
As for measuring human happiness, that’s a pretty hazy endeavor. But there are disciplines that attempt it. The CDC should cede that endeavor to those disciplines.
If you mean as to the collective us, yes that would be helpful. As to what it means to individuals, that is a different story.
I’m afraid you completely misunderstood what I meant when I used the term “happiness”. We will both be “happier” if we leave it there, rather than attempting to understand each other’s thoughts. We have a bad connection.
I am so sorry to disappoint you that my answer did not come up to your standards. /Sarc off/
There are just so darn many variables involved with this virus. And, of course, there is at least one “conclusion” and one “solution,” it seems for each variable. I’m just going to wait until it all plays out before offering my own. In my opinion, the virus is going to do what it’s going to do. I’m 82, so just a disinterested spectator. But it does help keep our experts occupied writing analyses and advice.
Flu season is not all year- it has typically peaked in February for many years, but is significantly active December-March. Clearly not all year. Almost every year February is the month the hospital administrators are scouring the floors for possible discharges to free up beds. If you check the CDC the peak month for the flu is Feb about 50% of the time with Dec, Jan, and March equally splitting the rest of the peak times.
Discussing COVID-19 “cases“ instead of deaths has been described as a means of political deception; so I thought I’d give you an opportunity to provide a different explanation.
Cases can be useful in understanding trends if the data is used carefully and normalized for the amount of testing and used in conjunction with positive rates.
Thank you for the clarification. I literally lack sufficient time to address that level of analysis. Sorry.