I've got to say, I'm not sure why this has become such a big media story. I might be wrong but my current view is that it's overblown. Flu arrived early, which is why it's breaking 'time of year' records, but it's only in the medium threshold and could peak lower than last year, let alone 2022-23.

Ah, OK thanks. I hope you'll agree the way official bodies now describing 'average 10-15,000 flu deaths pa' as accepted truth is wrong, misleading govt and public about flu deaths, and how that relates to us thinking about Covid - especially whether we shld be offering more people Covid vaccines.

Thanks for sharing these - it's clearer now what you are asking.

I would treat any attempts to attribute excess deaths to causes as a best endeavours approach (i.e. treat with a pinch of salt!) It's hard and subjective enough to measure the aggregate excess, let alone attribute to causes.

Thanks, I tend to agree to include 'involving': winter, elderly, and co-morbids all may be relevant to 'flu deaths'.
Looks like pneumonia can be parsed out as the baseline during summer, but why conflate?
Must confess I don't understand your Y axis: 3.8% and 12.7% are avged percentage, but of what?

Thank you for checking. Your comment is exactly my point. So my question as to why UKHSA, DHSC, NHS*, BBC have reverted to using 'fluEWD' is not rhetorical. When official bodies use these hugely inflated flu death stats, it misleads (and has comparatively minimised Covid deaths). Imo needs changing.

khub.net/documents/13...

It sounds like I'm comparing all-cause EWD, but I'm not! These figures specifically cite flu. Please see, one example here and another below: ukhsa-newsroom.prgloo.com/news/excess-...

These estimates quote 13,500 - 14,500 average (excess) flu deaths, and up to 20-30,000.
The huge magnitude of variance quoted for excess death modelling vs death certs is not reconcilable imo.

The excess death modelling assoc with flu appears less credible than ONS J09. Are you able to elucidate?

Thanks v much for bearing with me. I think I've not made my question clear. J09 graph helpful and what I'd expect one to use when describing annual flu death numbers, avg=c600.
CMOs, Govt, BBC. UKHSA etc seem to use excess death modelling numbers eg NOMIS FluMOMO, resulting in much higher estimates👇