The PCR Test As A Predictor Of Acute Respiratory Conditions (part 2)
I utilise data from an unknown NHS Trust to determine the real-world value of a COVID-19 diagnosis in the EPR of in-hospital deaths
In part 1 of this series I ended by opening another can of worms. I have been using an indicator variable called COVID-19 Dx to flag in-hospital deaths that were given an ICD-10 emergency code of U07.1 COVID-19 virus identified or U07.2 COVID-19 virus unidentified assuming these would unequivocally identify genuine COVID cases. After a fair few articles it became quite clear that there’s nothing unequivocal about this diagnosis, with acute respiratory conditions being mixed with chronic respiratory disease and asymptomatic cases in the biggest coding mess you can imagine. In fact, I couldn’t make a bigger mess if I tried, and it does make you wonder how much can be ascribed to sheer ignorance and how much to deliberate obfuscation brought down to bear on the NHS by government lackeys and senior NHS executives wiping the arse of WHO instead of ensuring public safety.
After a couple of blueberry muffins, a few digestives and a gluten-free chocolate thingy washed down with long-cut Rooibos I decided to re-code the diagnostic records of all 57,557 in-hospital deaths in my data sample to identify anybody with a viral or bacterial respiratory infection of any sort – including SARS-COV-2 – to ring fence absolutely anybody and everybody with a flu-like illness with or without a secondary (bacterial) infection, as well as those suffering bacterial pneumonia and other unpleasant respiratory conditions like aspergillosis (a reaction to the aspergillus fungus). Chronic disease states such as asthma, COPD and emphysema were excluded, the objective being to nail anything respiratory and acute brought about by an airborne pathogen. That is to say, I’m seeking the sort of broad diagnostic forest in which genuine and misdiagnosed COVID-19 would be hiding.
I called the new variable DDx_VIBAC and gave it the mundane label of Respiratory Infection. So let us have a quick look at some basic stats for this newly forged indicator starting with a crosstabulation of DDx_VIBAC vs. COVID-19 Dx for all 19,457 adult in-hospital deaths (>=18 years) for the period 2020/w11 – 2021/36: