8 Comments
Aug 14, 2023Liked by Paul Mainwood

Thank you for this analysis Paul.

I know CMI (and ONS, though I can't speak for them) are aware of this risk. CMI mortality monitors include a caveat "Our calculations rely on data for registered deaths, and we are conscious that during the pandemic deaths may have been registered earlier or later than in previous years. Consequently, comparisons of mortality between years during the pandemic and earlier years may not be on a like-for-like basis."

It is only recently however that analysis (including yours and work by Jean Fisch) has made it clear that this is a very likely explanation for some of the recent excess - i.e. the theoretical risk looks like it has become an issue.

I have made sure that relevant people at CMI and ONS are aware of this analysis.

Expand full comment
author

Thank you Stuart, much appreciated.

Expand full comment

I am a bit confused about the late registrations, my dad died in march and it was practically the first thing we had to do

Expand full comment
Aug 14, 2023Liked by Paul Mainwood

Sorry for your loss.

A death can't formally be registered until the cause of death is determined. For some deaths (e.g. accidents, overdoses, suicides) this can take several months. It's more of an issue where people die young.

More detail here https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/impactofregistrationdelaysonmortalitystatisticsinenglandandwales/2021

Expand full comment

Love your stuff I was in A&E at the weekend and because of suspected covid (negative) sat at the ambulance entrance to A&E for quite a while -and your posts crossed my mind - generally v efficient although one ambulance couldn't be found..

Expand full comment

thank you.

Expand full comment

yes sure, so it's the coroner is the hold up? And got worse in last view years?

Expand full comment

Yes and yes. And this new analysis is suggesting that we're now seeing a bit of recovery, with registrations consistently exceeding modelled occurrences.

Expand full comment