The COVID-19 Data We Have May Not Be The Data We Need

Alan ReynoldsCOVID-19 statistics that are easiest for reporters to find and explain are often the ones we keep hearing about in daily news reports. A  perennial favorite is theJohn Hopkins University graphical database which is constantly updated to add up the cumulative number of “confirmed” cases, deaths and recoveries since January 21 for separate countries and the world.To switch the focus from these familiar multi ‐​month totals to what is happening now, however, I built this simple graph of daily new deaths and daily new confirmed cases. It does suggest recent flattening, though professional optimism about reaching a peak is partly based onkey states rather than national totals.The cumulative data we see repeatedly in the daily news is not always the data we need. “Confirmed cases” in the John Hopkins data means cases that were tested. This makes it sensitive to the intensity of testing and also limits the count to those sick enough to be tested. It also exaggerates the apparent death rate by excluding the majority of COVID-19 infections, those having mil d or no symptoms.The “number recovered” would ultimately add up to 99% of infections if 1% died. Yet the number recovered seen in the John Hopkins site always remains very low. That is because states don’t keep track of what happened to every sick person, and nobody knows about the vast majority who recover at ho me.As these cumulative multi ‐​month totals go, the most relevant one is missin...
Source: Cato-at-liberty - Category: American Health Authors: Source Type: blogs