COVID-19 Compilation: Part 26 - September 6-12, 2020
On March 14th, I began posting daily updates to my facebook page, regarding the status of the COVID-19 pandemic. I have used data from worldometer as a primary source for daily and cumulative data, with frequent supplementation from various national and local sites.
This blog post compiles the period of September 6th through September 12th.
see https://www.worldometers.info/coronavirus/ for more"All the News that's Fit to Confuse"
Consider:
- On June 25th, the Washington post reported that CDC director Robert Redfield, in a conference call with reporters, said that the number of people in the USA likely to have been infected with COVID-19 was likely 10 times as high as the 2.4 million cases that had been confirmed at the time (the USA is now closing in on 6.5 total confirmed cases).
One of the common readings of those remarks by many readers was that the actual case fatality rate (CFR) for COVID-19, which was then about 5% overall in the USA, was actually 0.5% or less. - An article in The New York times published on August17th was titled, "What if 'Herd Immunity' Is Closer Than Scientists Thought?"
Key passage:
"To achieve so-called herd immunity — the point at which the virus can no longer spread widely because there are not enough vulnerable humans — scientists have suggested that perhaps 70 percent of a given population must be immune, through vaccination or because they survived the infection.
"Now some researchers are wrestling with a hopeful possibility. In interviews with The New York Times, more than a dozen scientists said that the threshold is likely to be much lower: just 50 percent, perhaps even less. If that’s true, then it may be possible to turn back the coronavirus more quickly than once thought."
On August 17th, the USA had about 5.6 million total cases. 'Herd Immunity' at 50% of the total population would have been about 165 million cases - On August 29th, the New York Times published an article titled, "Your Coronavirus Test Is Positive. Maybe It Shouldn’t Be." It contained this passage: "In three sets of testing data that include cycle thresholds, compiled by officials in Massachusetts, New York and Nevada, up to 90 percent of people testing positive carried barely any virus, a review by The Times found."
The USA was averaging about 40,000 reported cases per day at the time; the suggestion to many readers was that the actual caseload was closer to 4,000. - On September 4th, the Wall Street Journal published an article titled, "Covid-19 Positivity Rate Might Not Mean What You Think it Does"
Key passage: "in the U.S., there are no standards for how states report the figure; it’s sometimes calculated using duplicative or irrelevant data; and the meaning of the result is often misconstrued."
Let's add it up:
If we look at #1, we might conclude that the USA has had close to 65 million people with COVID-19 to date.
If we look at #2 and some other published sources, we know that some sources have speculated that herd immunity could be as low as 20% of the population.
65 million is about 20% of the USA population.
But... #3 suggests that our actual cases are far lower than the 6.5 million total for the USA, because our tests are too sensitive, and therefore we cannot be close to "herd immunity." Item #3 also gets in the way of the CFR conclusions from item #1 (there's an answer for that, too, though I'm not going to get to it in this post).
And #4 suggests that, without national standards, the numbers don't mean the same thing from place to place.
The bottom line could be summed up as, "trust none of what you hear and less of what you see."
But we'll continue to try.
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 27,319,919 | 236,542 | 892,756 | 4,207 |
| USA | 6,462,315 | 32,153 | 193,248 | 430 |
India had 91,723 new cases today, marking the 5th consecutive day on which India has set a new world record for most cases by one country in a day. India passed Brazil as the country with the 2nd most cases in the world. India is still 87th in per capita cases, between Denmark and Germany.
While standard disclaimers about lower Sunday numbers apply, today's case count in the USA was the lowest since June 21st. Missouri tallied 32 deaths, a new record.
SEPTEMBER 7
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 27,517,874 | 197,955 | 896,604 | 3,848 |
| USA | 6,487,722 | 25,407 | 193,534 | 286 |
SEPTEMBER 8
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 27,759,349 | 241,475 | 901,081 | 4,477 |
| USA | 6,516,241 | 28,519 | 194,030 | 496 |
Numbers here are still low due to yesterday's federal holiday, as many states are a day behind in their updated.
SEPTEMBER 9
"‘Worst case scenarios’ at Sturgis rally could link event to 266,000 coronavirus cases, study says"
So blared a headline in yesterday's Washington Post.
The study was put together by economists at IZA, San Diego State and University of Colorado, Denver.
My reaction, in a word: BULLSHIT.
3 days ago, I posted about various headlines that contradicted each other and that might lead to confusion among readers.
This study is in another category. It is fatally flawed in many ways that are staggeringly obvious. Yet it has been treated credulously nearly across the board.
Let's just look at the article in the Washington Post. The first paragraph is: "The Sturgis Motorcycle Rally led to significant spread of the novel coronavirus in the event’s home state of South Dakota and in other parts of the United States, a team of researchers said in a newly released study that is disputed by state officials."
You have to read seven paragraphs before getting to this: "...state epidemiologist Josh Clayton and Secretary of Health Kim Malsam-Rysdon noted that the paper has not been peer-reviewed." In paragraph eight, we get this: "'What I have to say at this point is the results do not align with what we know for the impacts of the rally among attendees in the state of South Dakota,' Clayton said."
Finally, in paragraph TEN, the Post article shows hard data: "The South Dakota Department of Health has reported 124 cases linked to the Sturgis rally through contact tracing. A Washington Post survey of health departments found an additional 204 rally-linked cases in 20 states. One death has been reported — that of a Minnesota man in his 60s, who had underlying conditions and was hospitalized with the virus after returning from Sturgis."
Finally, in paragraph TEN, the Post article shows hard data: "The South Dakota Department of Health has reported 124 cases linked to the Sturgis rally through contact tracing. A Washington Post survey of health departments found an additional 204 rally-linked cases in 20 states. One death has been reported — that of a Minnesota man in his 60s, who had underlying conditions and was hospitalized with the virus after returning from Sturgis."
That is, the post has 328 cases, and the study has... 266,796. You may be excused for wondering, here, how the analysts at the Post, who have been pretty good so far, could possibly have missed 99.88% of all the cases linked to the Sturgis rally.
Or how a rally that that held about 0.14% of the total population of the United States could somehow have been responsible for 19% of all national cases during a one month period, a month during which overall cases in the United States fell by 25%.
Or how these conclusions can be based on "anonymized cell phone data" correlated to case rate changes in impacted areas.
Well... THEY CAN'T.
Let's be clear here: there was a spike in cases due to Sturgis, in South Dakota and in other states. It will likely end up being more than the few hundred that the Washington Post has found so far, but it will not be in the hundreds of thousands.
I posted the South Dakota positivity rate on August 29th, which happened to be the day the spike hit its peak in South Dakota (it's at https://mattorel.blogspot.com/2020/08/covid-19-compilation-part-24-august-23.html#August29, for anyone who wants to check it out). For the week ending that day, South Dakota had a positivity rate just over 20%, and had 425 total cases on the 29th. That's a lot of cases for a state with a population under 900,000, but it's nowhere near the case rates that several other states have had recently. For that week, South Dakota -- which had been around 8% positivity prior to Sturgis -- was 3rd in the country for positivity, behind South Carolina and Iowa.
And here is just one of the many places where the study goes off the rails. Quoting from the study:
"For the state of South Dakota as a whole we find that the Sturgis event increase COVID-19 cases by 3.6 to 3.9 cases per 1,000 4 population as of September 2nd 2020. This represents an increase of over 35 percent relative to the 9.7 cases per 1,000 population in South Dakota on July 31, 2020." The authors put "over 35 percent" in italics, as if it were a Very. Big. Deal.
"For the state of South Dakota as a whole we find that the Sturgis event increase COVID-19 cases by 3.6 to 3.9 cases per 1,000 4 population as of September 2nd 2020. This represents an increase of over 35 percent relative to the 9.7 cases per 1,000 population in South Dakota on July 31, 2020." The authors put "over 35 percent" in italics, as if it were a Very. Big. Deal.
It. Was. Not.
Did they bother, for example, to look at the extreme spikes in cases in Iowa during the same period?
Did they note that South Dakota has had a low case rate overall?
Did they note that there was no spike in deaths in South Dakota?
Did they not notice that many states -- mostly red states -- had seen enormous case spikes in July, much bigger than the 35% South Dakota case spike in August? The United States, as a country, was up 126% in July.
Did they not notice that many states -- mostly red states -- had seen enormous case spikes in July, much bigger than the 35% South Dakota case spike in August? The United States, as a country, was up 126% in July.
Did they not account for a myriad of other reasons why cases might spike in given locations, and why anonymized cell phone data correlations is a pretty awful way to measure cases?
Did they not account for the demographics of the attendees? Check out this footnote, which indicates they tried... sort of: "We also explored whether the policy findings could be explained by state political preferences, but found that, in
the main, results were not sensitive to controlling for whether the majority of the state’s voters supported the
Democratic presidential candidate in 2016."
Did fail to notice that cases went down in the USA by a half million during the study period? Do they propose that cases would have been down by 750,000 had it not been for Sturgis?
Did they not even notice that they reported South Dakota's population at 858,000 -- which is 3% less than the actual population of the state -- or that the state's population during the event was around 1.3 million?
If I wanted a case study in confirmation bias, this might be it.
As a study of "The Contagion Externality of a Superspreading Event," it isn't close.
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 28,041,784 | 282,435 | 907,703 | 6,222 |
| USA | 6,551,469 | 35,228 | 195,238 | 1,208 |
The USA death total is the highest so far in September, but likely also reflects a reporting backlog over the Labor Day holiday.
SEPTEMBER 10
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 28,338,351 | 296,567 | 913,290 | 5,987 |
| USA | 6,590,703 | 39,234 | 196,327 | 1,089 |
There were 96,760 new case in India, a new single day record for any country.
France had 9,843 new cases, a new record there.
Israel had a record 4,429 new cases today. Over the past 4 days, Israel has averaged 3,721 new cases per day; before then Israel had never had as many as 3,000 cases on a single day. Deaths are also beginning to spike in Israel; the 23 deaths recorded today were the 2nd highest of the pandemic. Israel's coronavirus cabinet announced a 2 week lockdown to commence on September 18th -- erev Rosh Hashanah, and extending through Yom Kippur.
SEPTEMBER 11
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 28,649,043 | 310,692 | 919,097 | 5,807 |
| USA | 6,637,319 | 46,616 | 197,421 | 1,094 |
The worldwide case count today is a new single day record.
Another look at the CDC Excess Deaths from COVID-19 numbers:
As of this week's updates, the CDC estimates that between 192,767 and 252,307 excess deaths have occurred in the United States as of August 29th. As the CDC notes on its site: "Data are incomplete because of the lag in time between when the death occurred and when the death certificate is completed, submitted to NCHS and processed for reporting purposes. This delay can range from 1 week to 8 weeks or more, depending on the jurisdiction and cause of death." As a result, numbers after about August 8th are incomplete; we're likely to see about 20,000-25,000 deaths added for the last 3 weeks in August.
The attached graphs show the overall excess deaths as well as the totals at worldometer, both cumulatively and on a week-by-week basis.
Due in part to some outrageously misleading headlines lately, there are some who believe that the "official" death counts in the USA are higher than real number. As should be clear from the graphs, that's not true; the "official" death count in the USA (to the extent that worldometer may be used as an "official" count) is less than the CDC excess death count, and that the discrepancy grows whenever there is a spike in US deaths.
 |
| CDC Excess Deaths in the USA |
SEPTEMBER 12
COVID-19 has returned to France in a big way. There were 10,561 new cases in France today, the first time France has surpassed 10,000 cases in one day. That's the per capita equivalent of about 54,000 cases in the United States. Per a report on France24, "Despite this resurgence in cases, [French Prime Minister Jean] Castex did not announce any major new restrictions, saying France's aim is to "avoid a general lockdown" and succeed in living with the virus through social distancing, mask-wearing and ramped-up testing."The Independent reported, "French health health authorities reported that 772 clusters were being investigated, an increase of 86 in the past 24 hours."
Deaths have not yet risen in line with cases, but the Prime Minister noted "a substantial increase in the number of hospitalized people."
Per worldometer data, 15 countries/entities of 5 million or more people had at least 100 cases per million population today, including 6 in South America, 5 in the Middle East, 2 in North America and 2 in Europe. The attached table lists the 15 countries.
 |
| Highest per capita case rates, September 12, 2020 |
The largest per capita number today -- by a wide margin -- is in Israel, with 452.1 cases per million population (I wrote a blurb on Israel in Thursday's update). To put things in perspective: 19 USA states have surpassed that per capita rate at least once, including North Dakota today.
| Today's counts | ||||
|---|---|---|---|---|
| Location | Total Cases | Cases Today | Total Deaths | Deaths Today |
| World | 28,931,434 | 282,931 | 924,080 | 4,983 |
| USA | 6,676,601 | 39,282 | 198,128 | 707 |
| This week's counts | ||||
|---|---|---|---|---|
| Location | Cases | Change | Deaths | Change |
| World | 1,848,057 | -2.2% | 35,531 | -5.2% |
| USA | 246,439 | -16.8% | 5,310 | -11.3% |
Comments