Do COVID Vaccines Work? (part 5)

In this article I utilise data from an unknown NHS Trust and further develop multivariate logistic regression models in the prediction of vaccination status prior to death

In part 4 of this series I detonated a bomb. Though statistical in nature it did have the effect of showing to us all that a grand illusion is afoot whereby the vaccines are associated with a reduction in likelihood of a COVID-19 diagnosis at death but do nothing to reduce incidence of associated respiratory illness. This may be explained through bias introduced into the system whereby unvaccinated inpatients are tested more aggressively than vaccinated counterparts – a fact supported by anecdotal evidence supplied to me in confidence. I promised to visit my pantry for more variables to develop the model proffered and in this article we are going to see these peeled, sliced, diced, mashed and boiled. Get your aprons on!

Easy Does It

I thought we ought to start out nice and simple by using the methodology established in part 4 to assess incidence of cardiac conditions and hypertension in relation to COVID and vaccination. Though COVID-19 is billed as a respiratory disease there are those who argue that it is much more than that; it is a simple matter to turn the handle and produce more binary logistic models for consideration. Herewith two more models for the pooled sample of 8,714 in-hospital deaths for those aged 18 years and older for the period 2021/w1 – 2021/w37…

COVID & Cardiac

The essential ingredient in this tabulation is the interactive term Cardiac Dx(1) by COVID Dx(1) since this represents the interaction between presence of a COVID diagnosis prior to death along with a cardiac diagnosis. If the vaccines are effective at reducing severity of COVID symptoms then we may argue that the heart isn’t going to get hammered, so we should see a benefit in terms of a reduction in the odds ratio (OR) for this interaction in the prediction of vaccination status prior to death. In plain English vaccinated folk should be less likely to die from COVID-induced cardiac pathologies by way of knock-on benefit. Except that this isn’t what we observe.

What we observe is a totally insignificant interactive term (p=0.356) that indicates vaccinated folk are faring no better than unvaccinated folk when it comes to COVID-associated cardiac death. Neither is likelihood of cardiac death any different between vaccinated and unvaccinated cohorts if we ignore test results, as witnessed by the independent variable Cardiac Dx(1) (p=0.534).

If we run this model again in split-file mode to distinguish between ‘early’ (week 1 – 18) and ‘late’ (week 19 – 37) periods we still arrive at the same conclusion, with the key interactive term fetching-up as statistically insignificant at the 95% level of confidence across both periods (p=0.072 and p=0.815 respectively):

There is the merest hint of possible vaccine benefit during the first 18 weeks (OR = 0.708, p=0.072) but further investigation would be needed. At this stage my guess is that this ambiguous result, which is statistically significant at the 90% level of confidence, is artefact arising from confounding factors that have not been trapped, but we shall have to wait and see what transpires as I develop the model further.

COVID & Hypertension

The essential ingredient in this tabulation is the interactive term COVID Dx(1) by Hypertension Dx(1) since this represents the interaction between presence of a COVID diagnosis prior to death along with a diagnosis of hypertension. What I’m doing here is using hypertension as a proxy indicator for systemic disease and/or inflammatory conditions. It also serves to highlight circulatory as well as cardiopulmonary issues. We may think of hypertension (high blood pressure) as the Swiss Army knife variable for indicating underlying health issues.

Though there is no overall difference between vaccinated and unvaccinated cohorts (Hypertension Dx(1), p=0.183) there is a statistically significant difference of note for hypertensives returning a positive test result (OR = 1.684, p=0.002). What does this mean in plain English? Well, it means that those suffering from high blood pressure and COVID prior to death were more likely to have been vaccinated than not – 68.4% more likely to be precise.

This isn’t a good result to find, and quite a serious one, for it offers an indication that vaccination is associated with increased risk of COVID infection and associated hypertension. To double-check this I ran the model again but in split-file mode:

Now this is darn interesting for the effect is not evident in the early period at the 95% level of confidence (OR = 1.392, p=0.135) but it does emerge in the late period in a rather vicious and highly statistically significant manner (OR = 3.801, p=0.001). A rough rule of thumb indicates vaccinated folk prior to death are nearly four times more likely to be suffering from COVID in conjunction with hypertension. Crikey! Are we looking at evidence of a build-up of vaccine harm over time on a systemic basis?

What makes this result utterly fascinating is that there isn’t a general effect for hypertension for either period (OR = 0.965, p=0.760 for week 1 – 18; OR = 0.855, p=0.160 for week 19 – 37); the effect only comes to light in conjunction with COVID. My guess at this stage is that the vaccines are inducing a situation that increases likelihood of COVID infection with concomitant hypertension in some people. If this is the case then regular monitoring of blood pressure for vaccinated folk becomes a necessity, and may well save lives. Ideally I want GPs to chew this heavyweight result over because I’m just a numbers guy!

Another Slice

Well, I didn’t expect this! It sure looks like I’ve got my work cut-out pursuing the matter. As it stands there’s no evidence supporting the notion that vaccination reduces the risk of COVID-associated respiratory and cardiac conditions, but there is evidence that vaccination status is linked with increased risk of COVID-associated hypertension prior to death. Inflammation is where it seems to be at so I shall be raiding my pantry once more.

One last thought that strikes me is that if the lumen of blood vessels are clogging due to renegade mRNA causing amyloidosis in some people then we’d expect to see blood pressure rise as an early warning sign of circulatory issues. There should also be a knock-on effect on lung function (e.g. lowered oxygen saturation and higher cadence), heart rate, heart rate variability and kidney function. Just how many more deaths can be prevented by acknowledging the situation and routinely monitoring the vaccinated public for signs amongst heart, lung and kidney functions is anyone’s guess. Right now the government couldn’t give a flying fig-roll.

Kettle On!