Backgrounder on #COVID19 Hospitalization Data (thread)
Exante started to track #hospitalization data around the time of the #COVID19 outbreak in #Italy. Hospitalization data are not as skewed by fluctuations in amount of testing done. The challenge w/ hospitalization data is that countries & various US states report it differently.
Types of #hospitalization data: 1). *Cumulative hospitalizations*: All people who have ever been in hospital for #COVID19, including: those currently in hospital, those discharged, and those that died. From cumulative hospitalizations, one can calculate new hospitalizations.
*New hospitalizations*: useful to see whether there is acceleration of #coronavirus . New hospitalizations gives you cleaner read on that front as it is without the bias of testing.
*Current hospitalizations*: Patients currently hospitalized. Patients discharged not in this number. Used for tracking hospital capacity as hospital beds are most likely fixed. These data are used to see how hospitals are managing the #coronavirus outbreak.
Some states only report current hospitalizations, but you cannot infer cumulative or new hospitalizations from that figure. Note, you cannot take the daily change because you can’t tell those who are discharged and those who died.
So, which measure of #hospitalizations is most useful, depends on which question one is trying to answer. Note that some states report new and cumulative and some report currently hospitalized and some report all three measures, but *you can’t compare the 3 measures*.
