In an earlier post, I mentioned that COVID deaths have been severely underreported in Europe. Comparing recent deaths in Italy with those in the same months in previous years suggests that the true number killed is about double the reported number. This is because the conditions required to be recorded as a COVID death are stringent; if you die at home, it does not count as being caused by the coronavirus. This data bias is not specific to Italy, and it matters a lot. The official numbers are used for calibrating models that predict deaths and epidemic progression in other countries. If we calibrate models based on death rates that are too low, we risk underestimating the challenge other countries face in the coming weeks and months.
Note that this is just one of many data biases, and some might be in the direction of overestimation. But underestimation is the major threat.
More at https://public.tableau.com/profile/isaia.invernizzi...
Note that this is just one of many data biases, and some might be in the direction of overestimation. But underestimation is the major threat.
More at https://public.tableau.com/profile/isaia.invernizzi...
