You can't judge a trend by just a few data points, but this isn't good news... [View all]

[Bucky]

I've been following the numbers at World-O-meter. Com.

For the 6 days before yesterday there'd been a pretty steady drop in mortality rate, nationally speaking. Since April 21 there had been a pretty steady drop in covid-19 deaths. Then it suddenly jumped yesterday, the 28th.

4/21 - 2683 (mortality peak)
4/22 - 2358
4/23 - 2340
4/24 - 1957
4/25 - 2065
4/26 - 1156
4/27 - 1384
4/28 - 2470 (sudden upswing of 900 Covid 19 deaths)

Of course there's always slight variations in the overall trend. Mathematicians call that statistical clustering. But a jump of almost 900 additional deaths from the day before is what you'd expect to see with an increasing trend.

Now, this may just be a one-day outlier datapoint. But it's something to watch carefully. It seems to echo a similar upswing in new cases being reported from a few days earlier (which is, of course, a less reliable statistic because of the large numbers of CV and Covid19 cases not being tested and verified). But the new confirmed confections count looks like this:

4/17 - 32,368 (peak 2)
4/18 - 29,079
4/19 - 26,113
4/20 - 28,131
4/21 - 26,084
4/22 - 30,156
4/23 - 31,888
4/24 - 38,958 (peak 3)
4/25 - 35,419
4/26 - 26,509
4/27 - 23,196
4/28 - 25,409

So there were two national peaks followed by a general downward trend.

Now, the national "peak resource use" had been projected for April 14. This would be a few days after a big infection peak on April 9 & 10 (when there are 34,000 new cases confirmed on each day). And then came these two last peaks.

So there are two possibilities: 1, that Wednesdays are the days most sick people get tested, since the peak date are all hump days. Or 2, that we've exceeded the projected peak twice now in the 2 weeks following Easter Sunday. That was when the anti-lockdown protests began. But more importantly it's when the chatter about "opening up the economy" began.

If it was just new confirmed covid-19 cases, that might be accounted for by an increase in testing. And obviously there is a lag of a week or two from new infection to dying. However the mortality count now saw this sudden jump.

So, if yesterday's sudden spike above 2400 deaths is part of a trend rather than a statistical outlier, we might be seeing the beginning of the effects of some people taking the stay-at-home orders less seriously.