(MORTALITY RATE) Mortality rate is determined by closed cases—of which we now have 112,173. The global rate has been rising—5% to 13%—for weeks, as more folks are tested. It may eventually decline again—but my source is the same as that used by THE WASHINGTON POST (Worldometers). 
PS/ I am—to put it mildly—sick and tired of being lectured by those who don't know how mortality rate is calculated that the 13% figure is "wrong." It is—of course—*not final*, but it's *not* wrong, no matter how much you wish it to be. So commenters here need to slow their roll.
PS2/ As millions and millions *more* people have been tested, the mortality rate has *risen* dramatically—so don't come on the feed and assure folks the mortality rate will drop with more testing. It certainly *may*—perhaps even "likely will"—but we don't have evidence of it yet.
PS3/ This feed does operate under the (here unproven) assumption the mortality rate will drop again because of testing (even understanding the rate is artificially low because some nations aren't doing post-mortem testing on flu deaths or—like Russia—using other causes of death).
PS4/ But it's unsettling to see so many on this feed who believe the lie told by the Trump administration in February—that mortality rate for coronavirus will be under 1%—and have therefore decided to ignore the fact that the mortality rate has *risen* alongside the testing rise.
PS5/ Mortality rate is *not* calculated using "all active cases" because that'd assume—falsely—all active patients will recover. Nor is the Worldometers data—by the organization's own admission—on "serious or critical cases" accurate. Why? Because *almost no one reports on that*.
PS6/ So if you're looking at the Worldometers "serious or critical condition" data for active cases and trying to calculate likely mortality for *active* cases from that...stop. Seriously—stop. Almost *no one* is reporting such data; Worldometers is using incomplete data on this.
PS7/ Those confused about the "serious or critical condition" data need only look at the Worldometers "countries" chart or (better) the state-by-state data on the COVID-19 Tracking Project (@COVID19Tracking). You'll see almost *no states* report "condition data" on active cases.
PS8/ The US mortality rate—again, based on closed cases—is *not* something I write about on this feed, because we have way too little data and way too few closed cases (because coronavirus hit here "late"). That's why I use global data. But here is the US data, which *will* drop:
PS9/ What many people commenting on this feed are doing is dividing U.S. (or global) deaths by U.S. (or global) cases—both closed *and* active—and then angrily demanding that the discipline of medicine treat that number as the "mortality" or "case fatality" rate. Sorry, it's not.
PS10/ So:
1⃣ Don't cite the US mortality rate—we have too few closed cases.
2⃣ When you cite the global mortality rate—13%—say it *may* decline with additional testing (thus far it's *risen* with additional testing).
3⃣ Do *not* falsely calculate mortality rate via active cases.
1⃣ Don't cite the US mortality rate—we have too few closed cases.
2⃣ When you cite the global mortality rate—13%—say it *may* decline with additional testing (thus far it's *risen* with additional testing).
3⃣ Do *not* falsely calculate mortality rate via active cases.
PS11/ Let's take a slightly deeper dive into this, to show the game that some are playing to try to falsely validate the misinformation given out by the Trump administration. Take the @COVID19Tracking for New York, which has the most data of any U.S. state for us to work from:
PS12/ Those trying to validate Trump's false claims divide deaths (114) by all positive cases (15,168) to arrive at a "mortality rate" (a misuse of the term) of 0.75%—which would mean the coronavirus is "only" 7.5 times more deadly than the seasonal flu. But's that's "fake math."
PS13/ The "mortality rate" in New York is *actually* deaths (114) divided by [deaths + recovered/discharged persons]. The problem is that *we don't have any data for recovered/discharged persons*—so we can't actually determine the mortality rate in New York. So people play games.
PS14/ The game people play—to validate Trump's misinformation—is they are assume *all current patients will recover*, meaning they turn the category of "positive cases" into (magically!) "recovered/discharged persons." And voila! You get a *totally fake* 0.75% "mortality rate."
PS15/ There *is* no accurate way to estimate how many current "positive cases" in New York will result in fatalities, versus how many will result in "recovered/discharged persons." So what this feed does, instead, is look at *global* data—where *lots* of such info *is* available.
PS16/ Using global data, we find a mortality rate (in closed cases) of 13%. But we know there are problems with that figure—namely, the same sort of problems we find in New York, inasmuch as many cases haven't "resolved" yet *and* many people haven't been tested who *need to be*.
PS17/ But global data is more illuminating than NY data—for many reasons: the US is dealing with the virus so poorly we can't assume—once hospitals are overrun—we'll have a higher-than-globe recovery rate; global data *includes* recoveries; the global rate *rose* with more tests.
PS18/ There are, of course, complications with all of this. Some of you note that the 0.1% seasonal flu mortality rate is based on estimated cases, so comparing it to the "confirmed-case" coronavirus mortality rate is problematic. Some will note the lack of post-mortem testing.
PS19/ Some will pick out one locale or another—with more complete data—and declare, "That's like the US—therefore, we'll be like them!" Others will note that the US is headed for a hospital over-run, at which point *all* our presumptions go out the window on how many we can save.
PS20/ Others will say that we don't know if you can get the coronavirus twice. Others will say that you really need to know the mortality rate by demographic. Others will say that the mortality rate can't actually be determined until long after the pandemic is over—which is fair.
PS21/ The upshot is that the rising global mortality rate—and the fact that it's *risen with more testing*—is instructive, as is the fact that it stands at 13%. "Instructive" doesn't mean "conclusive." But it tells us that *no one* should *expect* a final mortality rate under 5%.
PS22/ Moreover, there's reason to fear a mutation of the coronavirus that makes it more deadly; or hospital overruns that—with the same effect, if not cause—"make" the virus more deadly; or the fact that countries with limited healthcare are now being hit and may see a rate >13%.
PS23/ The point is *not* to be afraid, however—but to be 1) resolved, 2) clear-eyed, and 3) as cautious as possible. That means militating for a #NationalLockdown even as Trump and his cronies—who don't care about human life, that we can see—militate to *drop* social distancing.
PS24/ But for any of you still skeptical—clinging to the idea we'll "only" see a 2% mortality rate—remember that we don't yet know *what percentage of Americans will get coronavirus*. If 50% of us do, a 2% mortality rate equals... 3.3 million deaths. That's over half a Holocaust.
PS25/ So instead of attacking those who post accurate global data calculating mortality rate in the way it's actually calculated—across 112,173 "closed" cases worldwide—consider that Trump has lied to us (again) and is about to try to drop "social distancing" using that lie. /end
UPDATE/ As I was writing this thread, the global mortality rate for coronavirus—again, correctly assessing mortality rate using *only closed cases* across 117,868 data-points worldwide—rose *yet again*. It's now 14%. If that figure confuses you, read this thread for more details.
NOTE/ There are some people out there who want to hold a conversation on what the "final estimated mortality rate" will be—despite (as discussed here) there being *way* too many variables to have that conversation now, and that *not* being (at all) what *this* thread is about.
NOTE2/ In other words, if you want to have a conversation on what the "final estimated mortality rate" will be—despite knowing we've no data with which to even *begin* that conversation coherently, mid-pandemic as we are—that's fine, but I ask you hold the conversation elsewhere.
