Deaths in the United Kingdom

I have a lot of research on deaths stored up; I ws hoping to do them in one piece, but it is just too complicated, so I am going to do them one by one.  This topic became intense during the epidemic because of the excessive attribution of deaths to CV-19, how many people actually died from CV-19, how many would have died anyway from something else, how many died because of lockdowns and the fear campaign and other issues.  And of course we have the absurdities being raised by the vaccine safety nuts, which really can be completely disregarded.  It is incredibly complex analysis, because we know how many deaths there actually were, we have a rough idea of the causes of those deaths, but we really are guessing when we say “here’s what would have happened in the absence of the epidemic and the epidemic response”.  So when you talk about “excess” deaths, you have to know that second piece, and we don’t know it with any precision.

As I have suggested before, here is the best approach to understanding mortality before, during and after the epidemic.  You need a long trendline pre-epidemic, at least ten years, so you can identify pre-existing trends by age group, no more than 5 years in a sub-group, and by major cause, whether a disease or something like car accidents, homicides, drug overdoses.  You have to do your analysis of the years of the epidemic, starting with 2020, by the same age sub-groups and major causes.  You have to do all your work not just with absolute numbers but as rates.  Finally, you need to assume that CV-19 deaths are exaggerated, that many people said to die from CV-19 actually died primarily from some other cause.  Unfortunately, untangling that would require reviewing all the death certificates.  And you need to be aware of events that could impact death rates for a particular disease or other cause, like the introduction of a significant new treatment, or the rise of fentanyl use.

So now we turn to our first study, the one from the UK.  The study covered from 2005 to 2023, so a nice baseline before the epidemic, except that they only used 2012 to 2019 as the baseline from which to calculate excess deaths. The researchers looked at rates and absolute numbers by age group and cause and tried to determine excess deaths.  Their general age groups are too wide, under 30, 30 to 54, 55 to 89 and 90 and over, but they presented other data by five-year age groups.  For all age groups up to 2012 death rates were declining.  After 2012, rates diverged depending on the age group.  For under 30 year-olds, rates were stable.  For those aged 30 to 54, there was a steady increase in rates after 2012, which is concerning.  And during the epidemic these rates increased further, and remained higher in 2023.

However, looking at that 30 to 54 group, the same level of death rate might have been reached during the epidemic, based on the pre-epidemic trend.  So increases in deaths during the epidemic might have occurred anyway, and any CV-19 deaths were largely substitutive for other causes.  The rate for 55 to 89 year-olds generally declined up to 2019, although at a slower rate after 2012.  The rate rose significantly during the epidemic, as you would expect for this age group, and then declined somewhat in 2022 and 2023 but not to the levels pre-epidemic.  Death rates for those over 90 showed significant variability, and while higher during the epidemic, were not dramatically higher than those in some pre-epidemic years.

Converting rates to absolute numbers in an excess deaths analysis is tricky.  For example, if the population in an age group increases, even though the rate of death stays the same, the absolute number goes up and can look like more deaths than expected.  This should be avoided by using rates as well as absolute numbers, in fact, the rate is the key metric to focus on, not the absolute number.  In the UK, this analysis shows excess deaths for all ages above 34 during the primary epidemic years, with a decline in excess deaths in 2022 and 2023 and even going negative for some age groups over 34, likely reflecting pull-forward of death.

An example of why this kind of analysis is important is that in the UK up to 2012 there was a decline in the rate of death for people with cardiovascular disease, likely attributable to new drugs and other treatments.  That decline slowed after 2012 and has stabilized, which is concerning.  But it may be that someone who previously would have died at 75 with heart disease now dies at 80, because the deferral of death from a new treatment only lasts so long and the initial cohort affected by the new treatment eventually dies and you no longer see the decline in rates.  It is worth looking at all the charts included at the end of the text of the study, to see how variable the number and rates of death can be by age.  Those more detailed charts are found by clicking the supplementary materials link at the summary on the website.  (Medrxiv Study)