So the sharp increase in overall deaths on Wednesday, the exact thing that made the huge spike in total reported cases look like it was more than just a reclassification of existing cases? That never happened. And that’s not the only confusion this morning.
When the information from China’s Hubei province rolled out for Feb. 13, it broke charts and brought panic. Overnight, the total of COVID-19 cases soared from 45,000 to more than 60,000. The report from the provincial government made it clear that the total now included “clinically diagnosed cases.” That is, whereas numbers on previous days included only cases that had been confirmed by laboratory testing, these included cases where examination indicated coronavirus was likely. That made it seem impossible to determine just how the new numbers mapped to the old, and databases tracking the outbreak at news agencies, Johns Hopkins, and Daily Kos jumped to the new value with a general admission that the revised figures made mincemeat of previous data.
And the big spike in deaths—up from 110 to 247—made it hard to believe that this was simply a case of changing where the line on “case” was drawn.
But a few hours later, a more detailed report was available. Then, on Thursday afternoon, the World Health Organization released its Situation Report for Feb. 13. And while that report acknowledges the clinically diagnosed total of 14,480 new cases, it breaks out 1,508 of those cases as confirmed. So, as far as WHO is concerned, the total of COVID-19 cases on Feb. 13, 2020, was 46,997. At the same moment, Johns Hopkins data showed 60,346 cases.
So there are now two different forks in the data tracking the emergence and spread of novel coronavirus in China. Which of them is more correct? The Thursday WHO Situation Report makes it clear that it’s also not all that enamored of what it’s seeing from China right now.
This is the first time China has reported clinically diagnosed cases in addition to laboratory-confirmed cases. For consistency, we report here only the number of laboratory-confirmed cases. WHO has formally requested additional information on the clinically diagnosed cases, in particular when these have occurred in the course of the outbreak and whether suspect cases were reclassified as clinically diagnosed cases.
But, obviously, it’s hard to buy the 1,508 “confirmed” number if what it really represents is either a limit on testing material, or a limit on health care personnel. If Chinese officials are saying that the number is really 14,840, it seems like the safer thing to do is to go with that value. So, rather than split the reporting into WHO and non-WHO versions, I’m going to go with the best available total numbers from a combination of other sources. And I’ll check in with WHO again tomorrow to see if it’s been able to resolve its conflicts with how the data is being reported and come up with a value that once again puts everyone on the same page.
And then the numbers were kicked around like a football. Here’s the beginning of the Feb. 14 Hubei provincial report after a pass through Google Translate.
At 04:00 on February 13, 2020, 4,823 new cases of new coronary pneumonia (including 3,095 clinically diagnosed cases) were newly added in Hubei Province … There were 116 new deaths (including 8 clinically diagnosed cases) in the province … 690 new hospital discharges (including 214 clinically diagnosed cases) …
That was before the later update appeared from the central government declaring that the Hubei folks had screwed up to the tune of 100+ deaths. More on that in a moment.
The nice thing about the latest provincial report is that it’s clear from the beginning on the idea that, if using the old confirmed-cases-only stat, the number out of Hubei would have been 1,728. And this report also breaks down the deaths and releases between the laboratory-confirmed cases and the clinical cases—which is superhelpful and very interesting. Adding in the values from previous reports, this would make the total makeup of cases in Hubei 48,200 lab-confirmed and 16,427 clinically diagnosed. So 74% of all cases are now lab-confirmed, and 24% of all cases are clinically diagnosed. With me so far? Cool.
Here’s the good news. That 24% of clinical cases accounted for 31% of recovered cases and only 7% of deaths. Which is a pretty good indicator that, as many people were hoping, the cases that were tossed in as clinically diagnosed were, in general, less serious that those that had been lab-confirmed.
This is a good indicator that healthcare workers in the area are continuing to conduct some form of triage, in which the most serious cases are more likely to be tested. And perhaps those cases are also more likely to be getting a higher level of care. In short, I’m taking these numbers as a sign that things are not falling apart, and that—frustrated, tired, overworked, underrested, and plain old angry as local healthcare workers in Hubei must be—they are still working at this diligently.
If you, like me, tightened your panic belt up three notches last night, now is the time to let it out at least a notch. Maybe two.
The change in case reporting is even more astounding when looked at on the cases-by-day chart. But there’s something else odd about these numbers: Rather than representing a real trend, my guess is that they represent something of an accumulation of cases. The reason that there was such a big burst yesterday was not a sudden increase; it was that authorities in the region finally reached a “f#ck it” moment—one in which they just piled in most of the cases they believed to be COVID-19, but lacked time to confirm. That probably included reporting a number of deaths that occurred before Feb. 13 and that looked like coronavirus. That might explain how the big double count happened, if it was more a matter of overly tired folks throwing everything in a box.
The collection of these “looks like it” cases continued on Feb. 14, but the backlog seemed to have been largely cleared out. Now the number of reported deaths in Hubei has plunged back to numbers in line, and the update destroyed yesterday’s big spike.
So let’s pull up one more chart.
Without the extra deaths included in the Thursday report, that looks so much better. If the latest, latest, post-update numbers are correct, everything is pretty much in line with where it was on Wednesday—except that the number of recoveries is increasing.
The reported recovery time for COVID-19 has been around 10-14 days. And what do you know? The number of resolved cases is pretty much where the case levels were at the end of January. Now let’s hope this trend continues by moving the number of recovered cases into the thousands per day.
Outside of China, there were a few events that gathered a lot of press. Singapore saw a burst of cases, including nine new ones on Friday. There are a number of secondary cases there. That’s not great.
There was also a new U.S. case, in Texas. However, that case involved someone who had returned from Wuhan on a chartered flight arranged by the State Department. Those who were on the flight have been under close observation or quarantine since their return, and to have someone from that flight fall ill isn’t shocking. But it’s still concerning, especially for others who shared that long plane trip. This does not represent a secondary transmission within the U.S.
Another event getting press was a cluster of four cases in Japan. That included the second death from COVID-19 that has happened outside of China. Alarmingly, the diagnosis was not made until after the woman’s death, and for at least three of those infected in this cluster, there’s no immediate link to another known case. With the poor Diamond Princess sitting in a bay, and a series of cases from an unknown source, officials in Japan are justifiably nervous.
One other item to mention. Remember our British traveler? The one who infected five people in France and four more in the U.K.? His travels have been traced back to a small conference in Singapore on Jan. 20-22. People leaving that conference are responsible for at least 20 cases in six countries, which does suggest that the people who cancelled the world’s largest phone show earlier this week probably had the right idea.
As usual, the numbers above were the ones that seemed most trustworthy when I was putting together the charts and tables for today. No doubt by the time this post goes up, they’ll be somewhat out of date, but they should be good to a few hundred cases in China and a few dozen outside—depending on how people are doing on that darn boat.
Resources
World Health Organization 2019 Coronavirus information site.
World Health Organization 2019 Coronavirus Dashboard.
2019-nCoV Global Cases from Johns Hopkins.
BNO News 2019 Novel Coronavirus tracking site.
Worldometer / Wuhan Coronavirus Outbreak.
Some issues tackled on previous days
Why COVID-19 was not made in a bioweapon lab.
How Case Fatality Rate and other measures are calculated.
Looking at incidence in other Chinese provinces.
The death of whistleblowing Dr. Li Wenliang triggers calls for free speech.
Source: http://feeds.dailykosmedia.com/~r/dailykosofficial/~3/OHkO8ybz6hE/-There-are-either-65-000-cases-of-Covid-19-or-47-000-It-s-down-to-WHO-you-trust-and-how-you-count