COVID-19 March 22, 2020

The march toward 100,000 cases continues with excellent adherence to the exponential model. 23,640 was the total at the close of March 21, as reported by the individual state health departments. Both the short- and long-term trends predict that we will reach the shameful achievement of 100,000 about four days from now on the 26th. We only passed 10,000 a few days ago.

California and New York are now restricting testing to health workers and those that are hospitalized. This is likely to impact the total statistics, especially since New York currently has over 40% of cases. If the change in testing protocol results in a decrease of reports, it may be hard to detangle any actual decrease (i.e., flattening of the curve) from biased testing and reporting.

At some point, the curves will flatten out. They have to. Exponential growth cannot continue forever. If you take the exponential model I've been using and extrapolate those predictions forward in time, we will surpass 1 million cases in early April, 10 million about 8 days after that, 100 million yet another 8 days later, and 1 billion by early May. That exceeds the entire population of the United States. The exponential model will break down. When it does, there are slightly more complex mathematical models to represent the behavior of the data. These models are well known and understood and generally produce an S-shaped (sigmoid) curve. Once we start to deviate from the exponential curve, I may switch to those more complex models. The question is, when will that happen?