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 4-15-2020: The Other Guys
I have a new COVID-19 update for you, based as usual on the Johns Hopkins data through April 15 (here). As you probably know, I’ve been tracking and graphing reported deaths, and sometimes cases, in the US and the Western European countries. Western Europe is defined as everything west of the old Iron Curtain (other than micro-states, and not including Greece or Iceland).
I call this post “The Other Guys” because I’m going to include some reporting on the smaller countries, which I’ve generally combined into my “Other Western Europe” category.
Tiny Belgium has recently emerged as one of the hardest-hit countries, surpassing Italy in reported COVID-19 deaths per 100,000 as of Tuesday, April 14. If present trends continue, Belgium will surpass the hardest-hit country, Spain, tomorrow (April 17).
In fact, Worldometer (here) shows Belgium higher than Spain at this very moment, though this is based on current-day reporting that might change. (One more caveat — per Worldometer, the micro-states of San Marino and Andorra are actually the hardest-hit. I have not been tracking them separately, due to extremely small size.)
You can click the graphs to enlarge.
Here are total reported COVID-19 deaths per 100,000 population, in linear scale, for the larger countries plus Belgium:
You can see that the US and Germany remain at quite low levels compared to the others, and you can clearly see a quite rapid rise in France and an even more rapid rise in Belgium. Part of the recent increase in Belgium was a pair or reporting delay, detailed below.
Here’s the graph for the smaller countries, leaving in Italy as a point of reference. It is the same information as the prior graph, reported deaths per 100,000, and is on the same scale:
You can see the rapid rise in Belgium, with the Netherlands being the second hardest-hit of the “other guys,” followed by Switzerland. The next graph is the same data in logarithmic scale, which would allow you to observe any sudden changes in rates:
I don’t really see any sudden changes here, though different countries have progressed at somewhat different rates.
Here are the counts of daily reported deaths per 100,000 for the larger countries, excluding France for the moment:
You can see that our hopes that the peak was April 10, initially claimed by IHME, have not proven correct. April 15 was the US peak for reported deaths, and the day before is the second highest (thus far). My own prediction (here), posted on April 5 based on data through April 4, was that tomorrow (April 16) would be the peak. We’ll see.
In the US trend line, you can notice a dip around April 12-13, and a prior dip around April 5-6. This will be more apparent in the rate of growth graph later on. There is a recurring tendency for US reported deaths to be lower on Sunday and Monday, rebounding on Tuesday and Wednesday.
The graph above also shows a recent upturn in Spain, and shows that the UK is at levels comparable to Italy at the height of its suffering about two-three weeks ago. Germany has a strange trend line, with a correction on April 11 that actually caused a negative number of reported cases for that day.
Here is the graph of daily reported deaths for France, Belgium, and the Netherlands, including Italy and Spain as reference points. Note that the scale of this graph changes from the prior graph, due to Belgium’s very high figures on a couple of days.
The two spikes in Belgium are a data anomaly, but the extremely high level of reported deaths remains accurate even if this anomaly is corrected. Per Worldometer:
- On April 7, Belgium reported 241 previously unreported deaths in retirement homes. This is the first spike in the graph above. This error was 2.08 deaths per 100,000, the equivalent of an error of about 6,900 deaths in the US. Even without this error, Belgium reported 1.40 deaths per 100,000 on April 7, a very high figure.
- On April 10, Belgium reported 171 previously unreported deaths in elderly care centers (the deaths occurred in March). This error was 1.48 per 100,000, and even without this error, Belgium reported 2.80 deaths per million on April 10, which is still the highest daily total that I’ve seen in any country.
France also had similar corrections on April 2 and 4, explaining the first spike in the data for France.
Next are my usual graphs for the daily percentage growth in total reported deaths. Note that the reporting period for this graph is different, focusing on March 25 to April 15. The first shows daily percentage growth for the US, with Italy and Spain for comparison:
This is the graph where you can really see the weekend effect in the US trend line (darker blue). You can see a decrease on March 29-30, with a spike the next day; a decrease for April 5-6, with a spike the next day; and a decrease for April 12-13, with a spike the next day. This is a pattern of a lower reported death figure on Sunday and Monday, rebounding on Tuesday, which is understandable and probably due to reporting being slower over the weekend. The only real relevance to our ongoing analysis is that we shouldn’t jump to conclusions based on favorable figures on a Sunday or Monday.
Here is the graph of the three-day moving average in total reported deaths, by country:
Here you can see the generally good news, which is the decrease in the growth rate virtually everywhere. There is a slight upward curve in the US trend line, probably as a result of the Sunday-Monday reporting phenomenon discussed earlier. In fact, notice that in this graph, you see a stepwise reduction in the US trend line, which corresponds to this same effect.
Germany has a notable upward turn, also due to very low figures over the weekend. Thankfully, Spain has finally reached the very low level of daily growth as Italy, around 3%.
Finally, I can’t resist patting myself on the back. As noted above, I made a prediction back on April 5. It has been a little bit high, but actually better than the IHME prediction made the same day. Here is the graph:
The real test is still to come and will depend on whether my prediction is accurate about the peak. I predicted tomorrow, so we’ll see.
ChiCom delenda est.
Published in Healthcare
I have noticed that ihme hasn’t been updated since the 13th. When they do, it will probably be a major revision.
I should add an update note here.
I rely on the Johns Hopkins data. It has generally been quite close to the Worldometer data. For example, total US deaths as of April 12 and 13 were: 22,020 and 23,539 per Johns Hopkins; 22,105 and 23,640 per Worldometer.
There was a major divergence on April 14, when Worldometer reported 6,185 daily deaths, while the Johns Hopkins data had just 2,303. The Worldometer site noted that New York City reported an additional 3,778 deaths classified as “probable” that had occurred since March 11.
At the end of my data reported in the OP, April 15, Johns Hopkins showed 28,326 deaths in the US, while Worldometer showed 32,443 — a difference of 4,117.
The Johns Hopkins figure just released for April 16 is 32,917 reported deaths in the US, an increase of 4,591 from the yesterday — and a new daily high by over 2,000. I suspect that some or all of the NYC deaths have been added to the Johns Hopkins number. However, the Johns Hopkins number remains significantly lower than Worldometer, which is at 34,580 at the moment.
I’ll let you know if I find an explanation at the Johns Hopkins site.
Good Stuff. May I link to the graphs for discussion elsewhere? And, by the way, I am too lazy to look it up BUT if you know of the top of your head, any idea how John Hopkins is handling the data revisions caused by the “new definition” of a COVID death as well as the seeming flood of old data be sent in to the CDC in light of new data?
I suspect, especially in the last few days, something else than the old trendline is pushing of daily death counts on WorldoMeter (among others).
You may link or steal my graphs at will. That goes for everybody.
I do not know how Johns Hopkins is handling the data revisions. I spent a couple of minutes looking around at their “read me” pages, but didn’t find an explanation. It may be there, and I just didn’t look hard enough.
It will be strange if my April 16 prediction for the peak turns out to be correct because of a data reporting issue.
Belgium had the equivalent yesterday of 8800 deaths in the US and has a total equal to 150000 US deaths.
God help them.
Finally some good news!
From the Global Times:
From our good friends at “the Center for Disease Control and Prevention (CDC) of Tongzhou district in Beijing clarified that farts, normally, do not constitute another transmission route of COVID-19, unless someone takes a good and rather close sniff of gas from a pantless patient.”
“The Tongzhou authority drew a conclusion at the bottom of the article stating that it’s safe to say that farts will not transmit the novel coronavirus provided that pants are being worn.”
“But if the infected patient were not wearing their pants, and they released a large amount of gas, and someone else happened to be in the vicinity and took a close and hard sniff at the gas, anything could be possible, the article said. “
Got that – flatulence (a fart) is not normally a transmission route of COVID-19 if proper pants and underwear are worn. But never fear I’m sure my ever helpful Mayor Eric Garcetti is looking into a “flatulence detail” for the LAPD right as I write this, so there might be only a minor inconvenience of a licensed police officer checking you if proper pants and underwear are worn by you. I’m so relieved! We are protected!
That reminds me of a bizarre Chinese martial arts story in which one master’s secret weapon was a secret diet which produced appallingly foul flatulence which he would emit in a fight..
While we’re talking about other communities, here’s a community in the US which is suffering greatly:
Probably ranging farther afield after food, too.