One of the many aspects about the covid pandemic is that even official data sources don’t seem to agree with one another.
For example, take the deaths by vaccination status data available from the UK Office for National Statistics and the UK Health Security Agency (buried in tables in the Vaccine Surveillance Reports). You’d think that these two sources would agree with how many deaths there had been in a given time period — after all, they both get their data from the same sources…
Let’s consider deaths data for two time periods. The UKHSA deaths data is per four week period, so we’ll compare with the same four week period in the ONS data; as there’s only a limited overlap between the data we’ll compare data for weeks 32-35 (September) and 35-38 (October), given in the following two tables (unvaccinated deaths shown in pink, 2x vaccinated in green):
How come the ONS and UKHSA data don’t agree? For example, look at the final row in the second table (weeks 35-38, individuals aged 80 and over): the ONS data has 136 unvaccinated deaths, the UKHSA data 189; the ONS data has 1036 double vaccinated deaths, the UKHSA data has 1288.
I’d note as an aside not to compare the absolute numbers of deaths in the above tables— the higher deaths in the double-vaccinated reflects the much higher numbers of vaccinated in older age groups, and relatively lower numbers of vaccinated in younger age groups.
I really don’t know the answer to why there is this difference between the two data sources — as I say, they both come from the same data source so it really is rather difficult to understand.
Nevertheless, I am troubled by two specific aspects of these data:
The UKHSA deaths are consistently greater than the ONS deaths. I’m not sure why it should be this way around, but I’d note that the ONS deaths data are buried in a complex spreadsheet that’s only been released once and without any fanfare. The UKHSA data, on the other hand are released weekly in nicely presented tables and then used by the government to indicate how the disease has impacted on the country1. Now, I’m sure there’s a rational explanation for this, perhaps to do with attempts to exclude deaths that were going to occur anyway — but the explanation isn’t immediately clear from the data sources. In the table below I’ve indicated the percentage increase in deaths seen in the UKHSA data compared with the ONS data:
The UKHSA overstates unvaccinated deaths in the younger age ranges more than the older. A glance at the table above should highlight one set of ‘unusual’ data points — the relative number of additional deaths in the UKHSA data for those unvaccinated and aged 10-59 is always substantially higher than all the other data points, and the excess in the younger double-vaccinated is always slightly lower. The cynic in me might suggest that during this period there was a coordinated effort in the UK to encourage vaccination in younger age groups…
Well, not all of the data — they ignore the UKHSA tables that show that the infection rate (per 100,000) in the vaccinated is double that of the unvaccinated, but that’s a story for another day.