COVID-19 Cases v. Recoveries – 3/18/2020

I’ve been thinking about how to assess COVID-19 Confirmed Contraction rates with Recovery rates… this might be key for pointing at societal success in addressing the epidemic. Not sure if the data is mature enough, but what I’m finding does indicate either some potential tampering with the data or more rapid societal recovery after the cases flatten off (hospitals become more effective at that point? Here are a few examples from the most affected countries.

These charts look like something we use in the manufacturing world called a cumulative flow diagram. The interesting thing about a cumulative flow diagram is that the vertical distance between the two lines equals the cycle time. In this case, that refers to the time to clear your patient load. The vertical distance between the two lines, however, equals the “work in progress”, which in this case, would be the number of patients who have the disease (active cases). This may reflect some percentage of the hospital beds in use (what percentage of active cases need hospitalization? Probably depends on the country).

The last chart compares Italy and Iran. This data pretty much makes the case that Iran is way understating their infections and/or overstating their recoveries. The chart tells us that right now, the cycle time for both nations is undetermined (we don’t have enough recoveries yet). But it also shows that the active cases for Italy is 2-3 times that of Iran. That’s just not credible, considering that their case loads started at the same time, Iran has 20 million more people than Italy, and Iran has about 40% of the hospital beds per 1000 people as does Italy. Italians do smoke 36% more cigarettes per year per capita than Iranians (but almost zero percent women smoke in Iran, vs 19% of women in Italy).

Finally, from the China chart (if we can believe it), we can see that the cycle time to clear the patient load is consistently about 20 days. The Singapore chart might give us some concern, though, as it appears that they closed their WIP to just a handful of cases, then it opened back up significantly. Perhaps more data will help us understand this better.