Excess Death & Heat Waves (Part 3)

A fruity response to the BBC headline that record UK excess deaths are due to the summer heat wave

I am encouraged by the consistent results from my modest approach to statistical modelling of excess all cause death for England over the period 1982 – 2022; this being an approach which currently yields an estimate of +214 additional excess deaths during weeks where at least one day is defined as a heat wave period (please see earlier articles for definitions). In turn that +214 surplus death estimate yields an estimated additional +1,710 excess all cause deaths due to hot weather during the summer of 2022, this figure being a far cry from the 14k – 16k inexplicable excess summer deaths that somebody in authority needs to take seriously.

This figure is also less than the estimate of 3,271 deaths estimated by the ONS/UKHSA for those over 65 years of age but, alas, my historic data doesn’t allow me to break deaths down by age band. I thus elected to shorten my study period to 2015 – 2022, thereby allowing me to derive baseline means for those age 65 and over, thence to derive weekly standardised excess all cause deaths for England & Wales for >=65y, which is as close to the ONS/UKHSA study as I can get.

I was anticipating an increased effect for this more vulnerable age group but I was in for a surprise. Get the coffee on, get that toast buttered and join me in a fascinating rummage…

Cake Crumbs

The first thing to consider when chopping my data period down from 1982 – 2022 to 2015 – 2022 is loss of statistical power and wider margins of error for estimates. In a twinkling I went from 91 heat wave weeks among 468 to just 28 heat wave weeks among 105. That’s a bit of a bummer for sure but I can still crank the handle! Here’s what transpired when I reached from that really simple error bar plot with two orange blobs: