Counting Deaths (Part 3)

A comparison of official counts all cause deaths by date of registration (England) with counts by date of occurrence

This article was written for crunchers who analyse weekly all cause deaths for England (or England & Wales), being a rather dry and boring exploration of the adequacy of T4253H smoothing in overcoming the administrative hiccoughs with counts by date of registration. In short, I shall explore using of smoothing of date of registration data in an attempt to produce a time series more akin to date of occurrence data. I suggest strong coffee and a packet of Bourbons to munch as I crunch!

What’s It All About, Alfie?

I am going to regurgitate my spiel on T4253H smoothing that was first discussed in this article:

For those not familiar the T4253H smoothing function the process kicks off with a running median of 4, which is centred by a running median of 2. It then re-smooths these values by applying a running median of 5, a running median of 3, and ending with Hanning running weighted averages (span 3). Residuals are computed by subtracting the smoothed series from the original series, and this whole process is then repeated on the computed residuals. Finally, the smoothed residuals are computed by subtracting the smoothed values obtained the first time through the process. A bit of a head banger I admit, but there is a useful summary here with nowt to be found on Wiki!

Getting Stuck In

Without further ado herewith a time series plot for weekly all cause deaths in England by date of registration (DOR) for the pandemic period of 2020/w12 – 2022/w40, together with the T4253H smoothed version: