Comorbidity Analysis for In-hospital COVID-19 Deaths

Lessons from an undisclosed NHS Trust

With a row brewing over whether CDC director Rochelle Walensky actually admitted that 75% of all COVID-19 deaths involve people with 4 co-morbidities I thought I’d brush the dust off my old electronic patient record (EPR) data dump and have a look what happened in a UK NHS Trust between Mar 2020 and Sep 2021 in terms of in-hospital death ICD10 coding.

During this period a total of 3,436 COVID-19 deaths were registered across the NHS Trust and the distribution of comorbidity in terms of total diagnoses made is attached. The data dump provided ICD10 coding for ten registry fields which is why this chart runs to a maximum of ten possible diagnoses. Mean and median age was 79.7 years and 82 years respectively.

Only 88/3,436 (2.6%) of COVID-19 deaths were categorised as COVID-19 only (i.e. no additional comorbidity) though we do need to bear in mind EPR coding isn’t what it should be! Jumping up to deaths involving COVID-19 and one other diagnosis we find 455/3,436 (13.2%), this bringing us to a cumulative total of 15.8% of COVID-19 deaths with up to one additional diagnosis. Coming at this from Rochelle’s end we find 1,241/3,436 (36.1%) of COVID-19 deaths with five or more diagnoses on the EPR, one of will be COVID-19 by definition thus leaving us with 4 comorbidities. That there 75% seems well out of line with what we observe but then we’ve got the wrinkle of vaccination status and what Rochelle Walensky was actually trying to say.

If we separate the data into vaccinated and unvaccinated cases we find 1,089/3,085 (35.3%) of unvaccinated deaths with five or more diagnoses and 152/351 (43.3%) of vaccinated deaths with five or more diagnoses. Although Fisher’s Exact Test is yielding a highly statistically significant p=0.003 for a test of association we can’t go round claiming vaccines are making things worse because we haven’t accounted for differences in base clinical risk between these two groups.

Members will ask me what I mean by unvaccinated. My definition is simple: ‘unvaccinated’ means nothing was shot into the deceased’s arm at any point; ‘vaccinated’ means something was shot into the deceased’s arm at some point before they died; this could be dose 1, dose 2 or the booster. Please don’t go confusing this simple yet truthful definition of vaccination with the cock-and-bull nonsense being dished out by the authorities!

The conclusion we should reach here is that logic would dictate we can’t really talk about COVID-19 death in terms of pure risk of SARS-COV-2 infection, nor can we easily compare vaccinated and unvaccinated clinical populations but that doesn’t stop the media, experts and politicians.