I’ve been drawing charts on COVID 19 since the onset. It’s perverse, but being a stats nerd, it’s engaging to see data coming in in real time and watch trends as they develop.
For the month of March, the public narrative about the infection and its development was pretty accurate, and matched the data closely. However, in the last two weeks…….the public narratives and the data have started to diverge. When that happens, it can mean only one thing — politicians have started to cherrypick the data for political reasons.
I hate that.
So, anyway, I figured it was time to start publishing some of my charts, and discussing the good, the bad, and the ugly. So, let’s start.
This chart very simply graphs the new cases reported each day in the US. We’ve been in a “trading range” since the 1st week of April. This looks grim, but it’s not; when we hit that first week of April peak, virtually all of the reported cases were COVID19 diagnoses as the patient presented themselves at a hospital. In other words, these were individuals who were sick enough to drag themselves to the ER.
Today, a substantial number of those new cases are derived not from ER presentation but by testing. These individuals are *not* deathly ill, and may never become deathly ill; they may be entirely asymptomatic. So, for some time, expect the number of reported cases on this chart to remain about at that level or even higher. That’s OK, as long as the rate of death decreases.
Speaking of which…….
Deaths, on the other hand….appear to be peaking. The next chart will drill down a little deeper on that to see what’s actually happening, but the interesting point on this graph is in comparison to the previous:
On the CASES chart, the peak that sort of defined the beginning of the top of the curve occurred on 4/5. On the DEATHS chart, that peak is on 4/11. So, Cases precedes Deaths by six days, approximately.
Now, let’s drill down on the last two weeks of the Deaths chart to see if it’s telling us anything:
Well, you tell me. My eyeball tells me that there’s a modest trend to the downside, but it’s modest. Let’s try it another way:
This is the same chart, but each bar is expressed as a percentage of the previous day. Looking at this, I’d have to say that our casualty rate is stuck at the present moment. So, the good news is that it’s not getting worse, but it’s not getting better, either (at least not by much.)
This chart, which I have been keeping over time, has been one of the more agonizing ones to draw, but it’s finally getting better:
So, on this chart, you can see how we (as a nation) immediately spiked up as far as Deaths/Cases were concerned, because we had the unfortunate luck (and not just from a statistical standpoint, of course) that one of our initial disease vectors found its way into a Seattle-area nursing home. That kicked our mortality WAY up initially; we then spent the next three weeks getting sick without a lot of deaths……but then the fatalities started happening. Our fatality rate increased at an uncomfortable rate until the 18th, after which the rate of growth started to taper off…….and we now have three days after peak.
Of course, as the previous charts showed, this doesn’t mean that we’re getting healthier yet….it just means that we’re starting to test, more than likely. Otherwise healthy people with a light case of COVID19 improves the data.
Bottom line is that nationally we’re midway through this, with obvious regional differences, with a very slight bias to being past peak as opposed to being pre-peak.
Hope that information helps. Tomorrow I’ll get into some international and state-by-state comparisons.
