Do COVID Vaccines Work? (part 13)

I utilise data from an unknown NHS Trust in the development of a staged multivariate logistic regression model in the prediction of acute respiratory conditions

In part 12 of this series I presented a staged multivariate logistic regression model for the prediction of acute respiratory conditions using a bunch of demographic data and clinical indicators for a sample of 19,457 in-hospital deaths over the period 2020/w11 – 2021/w36 using an EPR data dump supplied by an unknown NHS Trust.

The headline result was discovery of vaccine harm, which emerged from the illusion of vaccine benefit once several confounding biases had been accounted for. I then went in pursuit of correcting bias inherent in the coding of COVID status at death as well as COVID vaccination status at death using machine learning techniques to forge what I am calling probable COVID and probable vaccination. I finished my wild foray by developing a variable called PROD (prior risk of death – a proxy for case commonality) that is used in conjunction with the total diagnoses made (a proxy for case complexity).

With the advent of these three new ingredients, along with a proxy for disease prevalence based on the pillar 1 (clinical need) testing scheme, I am ready to roll my sleeves, scrub the table and roll out the pastry for yet another model bake for the prediction of acute respiratory conditions.

A Bit Of Weeding

Before I start turning the handle again there are two loose ends I wish to tidy. I’ll start by excluding those 1,222 cases exceeding 94 years of age at death that were mentioned in my previous article, and I shall exclude the 35 DOA cases with zero diagnosis as discussed in Prior Risk Of Death (part 1). This will keep things clean and reduce my sample size for in-hospital deaths to a nice, round 18,200 for the period 2020/w11 – 2021/w36.

The ingredients

Herewith summary statistics for the 9 independent variables along with the dependent variable (Acute respiratory) for the period 2020/w11 – 2021/w36. We may note that the sample mean for acute respiratory conditions is 0.22, hence 22% of deaths during this period possessed at least one appropriate diagnosis, this representing a modest sample of 3,930 cases. For definitions and explanations as to what these variables mean and how they were derived please see earlier articles in this series.

Nine independent variables will generate 36 two-way interactions and 84 three-way interactions, these 120 terms not being overly onerous to code, so I initially opted to limit the design to three-way interactions. However, preliminary runs revealed technical problems arising from over-specification and factored sample inadequacy (i.e. too many holes) so the model was limited to the 36 two-way terms plus the 9 main effects.

NOTE: In case anybody is confused as to what is meant by ‘interaction’ an example would be the three-way interactive term Major comorbidity by Chronic respiratory by Sex, for which the factored tabulation of the death tally looks like this:

The interesting result here is the lower rate of chronic respiratory disease for females with no major comorbidity (19.0%) compared to males (24.4%). This imbalance is something that may well find its way into the logistic regression model if it proves useful in predicting the incidence of acute respiratory conditions prior to death.

A Final Pudding

Well here it is… the final model that I shall attempt with the in-hospital death data! This table of coefficients and odds ratios (OR) derives from a three stage procedure in which the demographic and clinical main effects are entered (forced) at stage 1, with their two-way interactions subject to conditional (forward) selection at stage 2. The main effect for vaccination status plus the critical interaction with probable COVID is entered (forced) at stage 3:

Only one of the main effects for clinical and demographic factors failed to reach statistical significance at stage 1, this being Age (OR = 1.01, p=0.175).

That being said Age does appear as a significant interactive term with Sex (OR = 0.99, p=0.007), Major comorbidity (OR = 1.02, p<0.001), Diagnoses (OR = 0.99, p<0.001) and Probable COVID (OR = 0.97, p<0.001) in stage 2. Thus we observe a slightly lower risk of acute respiratory conditions for older females, an elevated risk for the elderly with major non-respiratory comorbidities, a slightly elevated risk for the elderly with complex medical records and a slightly lower risk for elderly patients flagged as probable COVID. This latter finding is curious but all depends on patient management and approaches to diagnosis in the care of the elderly, which is unlikely to be equitable.

Chronic respiratory appears as a significant interactive term with Sex (OR = 0.68, p<0.001) and Probable COVID (OR = 0.01, p<0.001). Thus we observe a lower risk for onset of acute respiratory conditions for females with chronic respiratory disease, and substantially lower risk for those with chronic respiratory disease that are likely to have contracted COVID. The latter finding is curious, though prophylaxis for existing respiratory conditions may well have proven beneficial. With luck somebody may drop an illuminating comment!

Probable COVID by PROD (OR = 1.35, p<0.001) is an interesting one in that likely COVID cases are associated with an elevated risk of acute respiratory conditions as their prior risk of death score mounts. This makes sense, and reveals the impact of COVID as a common diagnosis that swings PROD.

As we may expect Probable COVID sits as a main effect with a sizeable odds ratio (OR = 128.94, p<0.001) but we need to be cautious in interpreting this for Probable COVID will be synonymous with acute respiratory conditions, since these were a dominant factor in score derivation for Probable COVID in my machine learning model. At least the odds ratio didn’t explode!

The Bottom Line

Probable Vaccination as a main effect in this model fails to reach statistical significance and yields an OR that is pretty much bang on unity (OR = 0.99, p=0.875). This nothing-to-see-here result sits in stark contrast to an earlier analysis when, back in Do COVID vaccines Work? (part 10), we saw vaccination status popping up with an odds ratio of OR = 0.08 (p<0.001), indicating a thirteen-fold reduction in the overall likelihood of a positive test result for the vaccinated cohort.

I am hoping that readers can now see what sample bias can do and just how careful we must be when handling retrospective observational data instead of RCT data. Such bias is the loophole through which is crawling all manner of so-called ‘expert’ and ‘authority’ claiming all manner of magical benefit. I cannot imagine for one minute that they’re not aware of this and must conclude that we’re facing deliberate deception.

However, the super critical vaccine interaction that links vaccines to COVID to acute respiratory conditions - Probable Vaccination by Probable COVID - certainly made the grade (OR = 4.55, p<0.001). This is telling us that there is a greatly elevated risk of acute respiratory conditions for vaccinated folk who contract COVID compared to their unvaccinated counterparts. When I say ‘greatly’ I mean a near five-fold increase in the risk factor. The conclusion I must make here is that vaccines appear to be causing precisely what they were supposed to prevent, this being borne out by a great deal of mounting evidence from around the world. Here’s another slice of sorry evidence.

So There We Go

So there we go and there we have it. My best shot at squeezing an unbiased result out of this data in honest fashion is pointing us to vaccine harm and the toxicity inherent in the spike protein itself. Though you wouldn’t know it from the MSM the evidence base for vaccine harm is substantial and growing, but a shadow has fallen over the souls who are supposed to act ethically to protect us all from harm.

As Dumbledore once nearly said, “happiness can be found in the darkest of times, if one only remembers to open the biscuit tin”.

Summary

• A staged multivariate logistic regression model has been developed for the prediction of acute respiratory conditions in a sample of 18,200 in-hospital deaths over the period 2020/w11 - 2021/w36.

• Age at death appeared as a significant interactive term with Sex (OR = 0.99, p=0.007), Major comorbidity (OR = 1.02, p<0.001), Diagnoses (OR = 0.99, p<0.001) and Probable COVID (OR = 0.97, p<0.001).

• Chronic respiratory disease appeared as a statistically significant main effect (OR = 2.15, p<0.001), and as a significant interactive term with Sex (OR = 0.68, p<0.001) and Probable COVID (OR = 0.01, p<0.001).

• Probable COVID by PROD (OR = 1.35, p<0.001) indicated that COVID cases are associated with an elevated risk of acute respiratory conditions as their prior risk of death score mounts.

• Probable COVID as a main effect was associated with a large odds ratio (OR = 128.94, p<0.001), as expected. COVID ‘proper’ is synonymous with acute respiratory conditions.

• Probable Vaccination as a main effect failed to reach statistical significance (OR = 0.99, p=0.875). Thus mRNA vaccines did not reduce the likelihood of acute respiratory conditions within this sample of patients. This result sits in stark contrast to claims of generalised vaccine benefit, which are arising from a combination of confounding factors, categorisation errors and statistical bias.

• The interactive term Probable Vaccination by Probable COVID (OR = 4.55, p<0.001) indicated a greatly elevated risk of acute respiratory conditions for vaccinated folk who contract COVID compared to their unvaccinated counterparts. Vaccines appear to be causing precisely what they were supposed to prevent.

Kettle On!