04/13/2020
This is a graph of Covid-19 cases in Maryland. It shows total cases (thick green line), new cases per day (hick blue line) and total deaths (red line), and the average rates of change (thin blue and pink lines).
There are several things we can learn from this graph. We can make predictions about the course of the disease if we do nothing, we do a little, or we do a lot. We can see how well each action affects the course of the disease (for example, school closings, “stay at home” orders etc.). And we can see how our predictions change because of them. Here are seven things to learn.
1) first, the thin blue lines (the initial rate of doublings) show that, at the beginning of infections in Maryland, before any social distancing measures were put in place, newly detected cases were doubling about every three days, and that total cases and deaths were increasing in parallel (thin blue lines).
2). if we continued this way, we were predicted to hit 1000 new cases per day by about April 3rd, 8000 new cases per day by Easter, and about a million new cases per day by May 2nd or so. Then governor Hogan closed the schools and, two weeks later (March 28th), the effects of those closures began to show up, in that the doubling rate dropped to every seven days (thin pink lines). The prediction for hitting 1000 new cases per day was delayed to about April 10th, and we’re getting close (see #6)
3) the fact that it took about two weeks for the school closures to have an effect suggests that many of the cases being detected were people who had been infected for about two weeks before detection;
4) death rates are a lagging indicator, as they changed about one week later than newly detected and total cases (three weeks after school closures).
5). The drop in the rate of new cases happened about two days before Hogan put in the order to “stay in place (“lockdown” on the graph), and that order has not yet resulted in a further flattening of the rate of new cases. I would expect to see another change sometime around Monday the 13th. However, we’ll likely need to wait till midweek to see if Monday’s predicted drop is real because of #6.
6) there is a small drop at the beginning of each week, and the line catches up by Tuesday or Wednesday, suggesting that fewer people are tested over the weekend, or that fewer tests are finished, or that the numbers are under-reported for some other reason and catch up later. This may be why the numbers dropped below the predicted line on April 6th (a Monday), shot up over the line on Tuesday, and then dropped to the line on Wednesday, the 8th. At the moment (today the 10th), we’re scheduled to steadily go over 1000 new cases per day on Monday the 13th, but that may not show up until Tuesday or Wednesday.
7) Hopefully, the effects of the lockdown will begin to show up and we may flatten the line again so that we don’t hit 1000 new cases per day till much later.
The upshot of all this is that the results of our collective efforts at social distancing, hand washig, taying at home, losing our jobs etc, are having an effect. The effect is not immediate, but it is working, I expect that we will see a further flattening by the end of the week, and (hopefully) eventually a real drop in the cases of this disease.
In the meantime, I urge everyone to keep it up. We really are making a difference!
Disclaimer. Although I am a senior Immunologist at the National Institute of Allergy and Infectious Diseases, this is not an official NIH communication, but a personal one. : >)
