Comorbidity Analysis for In-hospital Deaths

Lessons from an undisclosed NHS Trust

In yesterday’s post I revealed the distribution of the diagnostic count made per death for 3,436 positive-testing in-hospital deaths occurring within an undisclosed UK NHS Trust over the period Mar 2020 – Sep 2021. Some excellent points were made by members that encouraged me to enhance the analysis by looking at COVID death in the pre- and post-vaccination periods as well as influenza prior to 2020. Here are those ideas bundled into one colourful bar chart.

Some people will go straight for the total numbers and declare, “crikey, just 186 influenza deaths over three years compared to 3,436 COVID deaths over 19 months – that’s a factor of 35 times more deadly!”; in doing so they will have fallen foul of a first-class fallacy. The COVID deaths over the period 2020 – 21 have not been defined by clinical symptom but have been defined by judicious application of the PCR test. We didn’t have a PCR test for influenza back in 2017 – 19 so we have no idea what such a test would have produced if every darn admission were poked up the nose back then. These are apples and pears folks; please leave them in their respective baskets.

So what do we see? Well, not a lot. I was keen on placing good money on an invariant distribution and have lost out on a decent betting opportunity! For all intents and purposes (and given the modest sample size of 186 influenza deaths) these are pretty much following the same underlying distribution. In terms of the distribution of morbidities per death new corona is old corona with added lock down.