Case Severity Index vs. staff absence
A deeper look at data for paid subscribers.
Try this for size…
Essential background reading for this slide is COVID uncovered (part 6) and COVID uncovered (part 7). What I’ve done here is the equivalent of pulled pork. The two grey lines represent the overall mean values for each series and are placed to divide the scatterplot into equivalent quadrants.
If staff absence due to COVID is a decent proxy for disease prevalence and the ratio of MV to general bed use for COVID cases is a reasonable proxy for case severity (CSI) then the top right and bottom left quadrants make sense. Going against the grain are the top left and bottom right quadrants; the ‘grain’ being my claim that these are useful proxy measures of what I believe them to be.
There are some mighty curious features here, starting with the incredibly linear relationships for 2020/Q2, 2021/Q1 and 2022/Q1. When I see highly linear forms lurking in a cloud like this I begin to doubt the validity of the data; that is, it feels a trifle ‘massaged’… or policy driven?
The relat…