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 Through 3-28-2020
I have an update on COVID-19 reported cases and deaths for the US and Western Europe, through yesterday (March 28). These graphs focus on population-adjusted figures and growth rates.
Italy and Spain remain the hardest-hit countries on a per capita basis, by a wide margin. Spain actually surpassed Italy in reported cases per million yesterday, though Italy remains highest in reported deaths per million. On a per capita basis, the US has only 4-5% as many deaths reported as Italy and Spain.
As before, my data source is Johns Hopkins. Details in the comments.
I start with reported deaths per million, by country, for the US and the major Western European nations.
The line for the US (medium blue) and Germany (orange) are difficult to distinguish at this scale, at the bottom of the graph.
Here are the deaths per million figures for the smaller Western European countries. I’ve left Italy in the graph for comparison, so the scale is the same as the first graph above.
I know that it’s a bit hard to differentiate these at this scale. None of the other countries are comparable to Italy or Spain, at least not yet. The Netherlands, Belgium, Switzerland, and Luxembourg are similar to France at present, while the others (Portugal, Sweden, Austria, Denmark, Norway, and Ireland) are significantly lower, around the level of the US, Germany, and the UK.
The next graph shows the daily growth rate in reported COVID-19 deaths, for the US, Italy, Spain, and France. (I left out Germany and the UK because inclusion made the graph too difficult to read.) This figure varies fairly widely from day to day, so I have smoothed the data by calculating a 4-day moving average.
You can see a general downward trend for Italy (light blue) and Spain (green), which is very good news. France’s figure is more variable. The high early rate in the US was when there were very low numbers of reported deaths (12 or less).
The daily growth rate in reported deaths, in the US, has been fairly steady at around 30% for the past 10 days or so. This is approximately exponential growth, but remember that we expect this to decline, as has occurred in other countries. The difficult question is when.
The next graph is transitional between reported deaths and total reported cases. This graph shows the case fatality rate, meaning the number of reported deaths as a percentage of the number of reported cases. Remember that this is not the mortality rate, but I think that it is a useful comparison between countries. This graph includes a combined line for the smaller European countries (labeled “Other W. Eur.”).
As you can see, the case fatality rate remains unusually low in Germany and the US.
The next graph is total reported cases per million.
You can see that Spain has now surpassed Italy by this metric. The US is at approximately 25% of the level of Italy and Spain in number of cases per capita (but only around 4-5% in number of deaths per capita).
The final graph is the growth rate in reported cases, showing a 4-day moving average.
As you can see, this growth rate has followed a general downward trend in all countries. The US currently has the highest growth in reported cases, but it is trending down. In several instances — the US (blue), Spain (green), and Germany (orange), you can see a “hump” in the trend line, probably corresponding to significantly increased testing, which occurred at different times in different countries.
ChiCom delenda est.
Published in Healthcare
My data source is Johns Hopkins (here). As previously noted, there appears to be a reporting error on March 12 for most of the European countries. To correct this, I interpolated the data for March 12, using the geometric mean of the corresponding figures for March 11 and 13.
Population data from Worldometer (here).
Earlier today, I heard Bill Bennett say, on Buzzfeed, that the average age of Coronavirus fatalities in Italy is 85.
Is this true?
Isn’t this information we should have for every country?
The death data from March 29 for both the US and the world were incredibly positive, and we can only hope they are not a major reporting fluke (because Sunday?) and the next couple days show that we have indeed turned the corner.
My conservative (high) estimates for total deaths are now 267k for the world and 22.3k for the US. These could be much lower if we continue to see a dramatic drop in the cases and death increases, or we could be back on exponential trend if yesterday’s data turn out to be an outlier.
To compare trends do you have a standard start point (eg date first case confirmed)?
There is one particularly hopeful interpretation of the last graph. The testing in the United States has increased significantly over the past week, but the growth rate is nevertheless declining. If the situation were not improving, one would expect increased testing to increase the growth and the rate of growth.
Earlier US numbers were not as indicative because of the spotty and low levels of testing being done due to test availability.
Zafar, good question. The charts in the OP are simply by date, throughout the month of March. I did not adjust the trend line for any of the countries.
I had done so before, in some of my prior posts. It makes some sense to do so, but it is difficult to select an objective measure. I wouldn’t want to use “first reported case,” because this could be quite idiosyncratic — for example, in the Johns Hopkins data set that I use, and among the countries evaluated, the US has the first reported case (1 case as of Jan. 22, which is when the data set starts), France is second (Jan. 24), Germany is third (Jan. 27), and Italy is tied for fourth (Jan. 31). Yet the outbreak has progressed furthest in Italy.
I experimented with several possibilities — for example, I think that I used passing 200 reported cases, and also used a deaths per million threshold later on. I did not find them very helpful, so I returned to simply reporting by calendar date.
Mike, you’re right. I was working on a graph of this when you posted. Here it is, comparing the percentage increase in reported COVID-19 deaths for: (1) March 29; (2) the prior 4 day average (Mar. 25-28); and (3) the average for the 4 days before that (Mar. 20-24).
As you note, this may be a statistical artifact, as yesterday was a Sunday. We’ll have to wait and see.
I agree with @David Carroll that the last graph is hopeful. Backing that up is a https://healthweather.us/ that was linked to a while back in a post that tracks the national reporting of flu-like occurrences, thus including reports of Covid-19 symptoms as well. The graph (below the county-by-county map) shows the expected range curve starting Feb. 8th overlaid with the observed/atypical day by day line. That line first diverged above on March 2nd, became ‘atypical’ on March 9th (5.04% observed vs 4.54% expected). It reached it’s worst case on March 16 (5.11% vs 4.2%), dropped below on March 22nd (3.65% vs 3.81%), and is now way below (1.39% vs 3.3%). So we know that social distancing is working, both with Covid-19 and the flu.
On a second point, the USAFACTS web site now includes the ability to download two county by county csv files showing day-by-day cumulative case counts and death counts. With a big thank you to @arizonapatriot for what you’ve provided, I’m hoping you can relatively easily replicate your national graphing above to hot spot locations in the U.S, with the additional datum of the population in those areas. I think that is where much of the public debate will be migrating over the next month. Here is a sample for the Bronx from both data sets, excluding dates between Jan. 22nd and March 24th.
A couple of observations on this, two states with low population density, Tennessee and Arizona, seem to be seeing younger cases. Glenn Reynolds commented on this today. The cases seem to cluster around early middle age, not elderly. I see the same in Tucson. Arizona has plenty of snowbirds and they tend to be over 65 but they (we) may be self isolating.
https://www.wate.com/health/coronavirus/coronavirus-in-tennessee-1537-covid-19-cases-7-deaths-in-state-department-of-health-reports/
This is all about “cases”; I’m asking about fatalities. All there is on the latter subject is an anecdote about a nursing home fatality. (Interesting.)
Still the strange reluctance to give statistics on age at death, when every other imaginable statistic seems to be available.
Is it possible that what is happening today is a reflection of what happened during the AIDS epidemic, when the public was misled into believing that everyone was at risk?
That time around, the reason was to maintain extremely high spending for research on a disease that actually endangered only a small minority of the population.
I speculate, this time around, the reason is to keep young people cooperating with the quarantine, when they are only minimally in danger.
But maybe there is some other reason I haven’t figured out for authorities to hold back statistics on age at death from coronavirus.
Isn’t the average age of Italians these days around 85? Ok, perhaps that’s a slight exaggeration…
Ohio gives an excellent breakdown by age of cases, hospitalizations, and deaths. Some of the results were surprising to me. On the web page under the heading called “Key Metrics” you simply select “Age.”
https://coronavirus.ohio.gov/wps/portal/gov/covid-19/home/dashboard
Age range Deaths
00-49 — 0
50-59 — 2
60-69 — 5
70-79 — 8
80+ — 24
Name of a name.
To compare trends in truthful manner, you would also need to have audits of the testing so that only the more accurate testing would be considered, and everything else should be reported with asterisks following the numbers.
The fact that both So Korea and Germany are seen by those who analyze such affairs as having properly conducted the audits of their populations, while almost all other nations have not, means that it is extremely worrisome that here in the USA, Trump has surrendered much of what is going on to the bureaucrats at the Maoist -run WHO, and the CDC, whose agency is more about helping the Big Pharmacy industry achieve record profits than the health of citizens.
One encouraging trend is that the Right Wing Hoover Institute has issued a paper critical of accepting USA’s poorly done audits of COVID. This analysis is mirrored by other papers done by more lefty outfits, and of course the recent New York Times op ed piece, which over the past week, first brought the decent audit of Germany’s population to light.
Why the need for decent audits? Well because otherwise you have our society and those across the globe shutting completely down over exaggerated and mistaken reports.
The WHO official who first set off the quarantines now so wide spread, had erroneously stated that the mortality rate of COVID 19 across all cases was some 3.4%. This statistic puts COVID 19 in the same category as the Spanish flu of 1918. However, it turns off he was off by a factor of 30.
The mortality rates of decently done surveys of a nation’s population show the actual mortality rates to be more like 0.1% to 0.7%. This still leaves COVID 19 in a serious category. Both the USA’s 1957 and 1968 influenza outbreaks were in this category. Tens of thousands of people died. I was almost one of them, in 1968. My classmate’s father did die, at the age of 38.
However in neither 1957 or in ’68 did our economy shut down.
But it could be that we will be seeing soaring rates of unemployment and further destruction of already weakened regions of our nation. Lake County Calif has lost so many small businesses due to the massive wild fires that have raged on since 2015. Although everyone whose family member or members are in the police or military are glad handing one another over this calamity, as never before have their family members gotten to be so important, the rest of us are alarmed about how our businesses are suffering.
What do you mean by an “audit of Covid testing?” And how does that help the infection or death rates?
If no one analyzes the methodology of the overall numbers that the public is continually presented by the news media, such as “Today another so many died in NY; another so many died in Italy; another so many died in Taiwan” with the mortality rate then being mentioned for each nation, then if no one is paying attention to the methods by which officials in each nation handle the various counts of the dead vs the counts of how many in each population are infected, then we know very little. (Except raw numbers which should be viewed with total caution.)
I consider that analysis to be the audit: the examination of how the methodology came about. This analysis is of extreme importance, as after all, both the Money Channel and other economic opinion makers are saying that we are now facing 20% unemployment.
If we should continue to have inflated numbers presented as the “true mortality rate,” as we have had in the USA, as few people are tested, and additionally those who are tested are quite seriously ill, which causes the mortality rate to be off by significant factors of magnitude, then not only is the public unduly alarmed, not only will our health personnel be unduly alarmed, but the “Shelter in Place” restrictions will stay in place, which will further wreck havoc with our economy.
To have a decently done mortality rate, a nation’s officials need to be testing a large segment of the population for COVID, while here in the USA many people are turned down when inquiring if they might be tested. We are told this is due to a lack of test kits, which to me seems absurd. (Inside the Fed government there are two separate agencies which had been created to handle viral pandemics: did neither consider using its annual funds to purchase test kits or masks or ventilators?)
I mentioned some of this in my recent OP on ricochet at this link: http://ricochet.com/736315/serious-indications-this-was-a-h-u-g-e-over-reaction-due-to-media-and-dems-needing-distractions/
The above is based on my thoughts on an opinion piece in The New York Times, which reflects on the extremely well done methods of data collection utilized in Germany, and the resulting excellent mortality rate of 0.7%, which happens to be a far cry from the 3.4% mortality rate from the Maoist-WHO official announced, and which triggered the lock step response from national officials across the globe.
Thanks for the explanation. The post you referenced above was excellent, too.