A Closer Look At Respiratory Death (part 2)

Lessons from an undisclosed NHS Trust

In my previous post in this series we observed 1,574 out of 3,435 (45.8%) adult in-hospital deaths meeting the criteria we might adopt for symptomatic COVID death; that is to say they not only tested positive prior to death but were suffering from an acute respiratory condition. It follows that we may classify the remaining 1,861 as asymptomatic COVID (testing positive with no acute respiratory condition) and bake a few crosstabulations again to see how things stack up.

The top section of the attached table provides a crosstabulation of adult in-hospital deaths over the period Feb ’20 – Sep ’21 involving non-COVID pneumonia (bacterial and viral, including influenza) and our newly established audited COVID status indicator. We find 11 asymptomatic COVID deaths associated with non-COVID pneumonia (18.0%) which suggests diagnostic tests indicated positive for both. The waters muddy further when we discover 15 symptomatic COVID deaths that are also associated with non-COVID pneumonia (24.6%).

This begs the question of whether the acute respiratory condition of these 15 patients was due to SARS-COV-2 or the other pneumonia-inducing pathogen (e.g. staphylococcus, streptococcus) that had been identified - which bug is to blame or are they both to blame? Is SARS-COV-2 infection laying the groundwork for pathogenic opportunism? Could the act of intubation and mechanical ventilation and/or subsequent medication be driving this double whammy? Clearly it is impossible to say without detailed discussion with the registrar handling the case at the time but without a doubt these cases will be chalked up to COVID death and clinical understanding will bite the dust.

The middle section of the table offers the same analysis but for bronchial conditions and diseases. This time we have 160 (14.1%) deaths associated with asymptomatic COVID and 7 deaths (0.6%) associated with symptomatic COVID. I have an uneasy feeling that intubation may be to blame for bronchial issues as much as viral infection but these are not going to be fatal.

The bottom section is tricky to fathom because we now have long term chronic respiratory illness featuring in 301 asymptomatic COVID deaths (9.4%) and 314 symptomatic COVID deaths (9.8%). On the one hand we might simplistically designate these 615 deaths as’ COVID causal’, though I would prefer the phrase ‘COVID enhanced’ since without the complication of an underlying chronic respiratory condition it is doubtful whether these people would have died. Bizarrely, the authorities don’t stop to figure such subtlety.

If you are confused this is a good sign. In a nutshell we have come to realise in three small tables that patients may be carrying more than one virus, that diagnostics tests may be in error, that medical diagnoses come with fuzzy boundaries, that clinical coders may put their own spin on a case and that death can be a multicausal matrix. A patient may come in with gunshot wounds to the chest yet die of sepsis due to a gram-negative enteric pathogen. The problem we face are authorities trying to render a Rembrandt in black and white and selling it to the public as art.