Catastrophic Health Collapse (part 1)

In this series I take a look at what ‘official’ excess death figures for England would look like if we surgically remove the hike that is the 2020 spring death spike

In a five part series that started on 6th February 2023 with an article entitled Excess Death Figures: Further Considerations (part 1) I took subscribers through an exploration of methods that may be used to derive excess death, highlighting some of the issues that arise when we try to assess whether deaths we observe each week are ‘normal’ for that time of year.

We discovered that we can get very different results depending on the exact method used such that nobody can claim they have produced the one and only authoritative set of indisputable figures. Arguably this was best illustrated by using ARIMA time series modelling to show that by moving the predictive period start date by just eight weeks we can swing from a net overall negative excess of -78,789 excess non-COVID deaths to +132,994 excess non-COVID deaths; a swing of some 211,783 deaths in England for the period 2020/w1 – 2022/w46. In terms of weekly means this equates to a swing of +1,402.6 excess non-COVID deaths… and just by shifting the modelling start date a few weeks!

The dastardly idea behind this series of articles was to reveal just how fragile any estimate of excess death may be, even if produced by a stalwart organisation such as the Office for National Statistics. Quite frankly the uncomfortable truth is that, as an applied statistician, I can sit here with my lemon tea and a bunch of official data and produce pretty much any set of figures for excess death you want to see, whilst providing arguments that support my reasoning. There never was an easier time to totally fool the public and make your findings look like the real and most superior deal.

Back To The Beginning

With that numerical exercise behind me I decided this misty morning to take the situation a stage further and reveal what happens to ‘official’ excess deaths if we surgically remove the extraordinary spring spike of 2020; a spike that I have previously labelled the CHEC death spike (catastrophic health collapse).

To avoid the pitfall inherent in determination of COVID and non-COVID death I thought we ought to base this mini study on all cause death. For this purpose I utilised a very useful date of death (DOD) dataset stretching back to 1981 that was obtained under FOIA, this being extended to 2022/w46 using the dataset obtained under FOIA by Joel Smalley. I haven’t extended past 2020/w49 for this particular analysis ‘coz vaccine rollout began in 2020/w50, this being a major confounding factor for my study.

To ensure the excess deaths I’ve generated meet with official approval I opted for using the 5-year baseline period of 2015 - 2019 in the calculation of excess death for 2020 as well as 2021, and the baseline period of 2016 - 2019 plus 2021 in the calculation of excess death for 2022. This is how the ONS go about things and what I have previously called the hopping frog method.

If we go back to the very beginning and plot out what I’m about to operate on both in terms of observed weekly counts and baseline values for all cause death for the period 2010/w1 – 2020/w49, then the data looks like this:

One thing we need to realise right up front is that these counts are not standardised data and the population of England will have grown in 12-and-a-bit years. I shall be looking at the impact of standardisation a little later after we’ve gone through the ‘official’ motions.

There’s the CHEC death spike standing out like a sore thumb, closely followed by another whopper for the 2021/22 flu season. Our eyeballs will have no doubt clocked the seasonal spikes for 2014/15, 2016/17 and 2018/19, and no doubt our brain will desire to consider spikes prior to 2010. More on this later, but for now we ought to look at the excess all cause death time series that these two curves generate:

We can now perhaps begin to see why the CHEC death spike is so darn unusual, and why it is important to understand exactly what went on back in spring of 2020 when the vulnerable and elderly were shunted out of hospital beds to end their days on an EOL care pathway. In the beginning I had thoughtlessly put this spike down to a rampaging novel and deadly virus but that argument is well and truly nobbled by the fact that SARS-COV-2 was spreading around the globe in 2019 and isn’t as deadly as initially claimed.

We may note the many troughs that will naturally occur after peaks in excess, this being a function of survivor bias. It is also worth looking at cumulative excess deaths for this adds a little more flavour:

I’ve added three green dashed lines to this plot that mark out four different stages in the health of the nation. Up to 2012/w5 things were getting better but then took a turn for the worse. Perhaps it is no coincidence that the biggest shake-up in the history of the NHS since its inception in 1948 took place back then, this being the introduction of the Health and Social Care Act (2012). It would appear government ministers and their advisors were as incompetent, selfish, ignorant and foolish back then as they have been revealed to be through recent disclosure of social media messaging. The steady and most sorry path to mounting excess death reached a plateau between 2018/w12 and 2020/w8 after which all hell broke loose.

A Spot Of Surgery

I now need to perform some surgery. After a great deal of cogitation I decided to cast away all the fancy techniques for estimation of the magnitude of an ‘event’ using ARIMA and other modelling methods and do something very simple and altogether rather more intuitively obvious – replace the CHEC death spike with baseline values for the ‘hot’ period 2020/w13 – 2020/w21. Here’s what the patient looks like as they are coming round from anaesthetic:

A rather neat incision and some decent stitching even if I say so myself! Let’s now flip straight to the cumulative excess death slide and compare the observed series with my surgically de-spiked series:

Now that’s a bit of a shocker for I wasn’t expecting a difference as big as this. Instead of a highly suspicious cliff-wall the curve for cumulative excess all cause death now rises at a steady rate that mirrors the historic rate that settled in after the bombshell that was the Health and Social Care Act (2012). To my ancient eyeballs the blue curve is far more believable. In terms of excessive excess we’re looking at up to 55,755 deaths that could be attributed to misguided policies over the period 2020/w13 – 2020/w22.

The other reason why I don’t believe the cliff-wall rise in spring of 2020 is a natural response of the public to a ‘novel’ and ‘deadly’ virus is that we now have evidence that SARS-COV-2 was going global in autumn of 2019 if not before. Colleagues, friends and family attest to suffering a rather strange flu-like illness identical to COVID-19 in the run up to Christmas 2019, myself included, yet none of us suffered anything remotely similar during the spring of 2020 nor know of anybody that got clobbered.

On a less subjective note I’ve also analysed emergency admissions and in-hospital death records for one of the UK’s largest NHS trusts that confirm a substantial rise in respiratory illness prior to the alleged outbreak, with very little to show for the ‘pandemic’ year of 2020. I may well publish a summary of this work in a future article.

In Conclusion

My conclusion at this stage is that the flu-like respiratory illness we’ve called COVID-19 was happening across the UK well before the government claim something novel and deadly dangerous was happening, which suggests (but does not prove) that the CHEC death spike is likely to be iatrogenic in nature. Not iatrogenic in the narrow sense of medical intervention directly causing death but in a wider sense of poor patient management, inappropriate use of EOL protocols, overuse of potentially lethal sedatives, failure to provide a basic standard of care, and misguided application of ill-advised government policies. In a nutshell it is the response to COVID that killed.

Kettle On!