Thread by @chrishattoncedr, In time for the @HumanRightsCtte session, @NHSEngland have just published statistics on [...]

In time for the @HumanRightsCtte session, @NHSEngland have just published statistics on COVID-19 deaths of people with learning disabilities from the LeDeR programme. They are here
https://www.england.nhs.uk/publication/covid-19-deaths-of-patients-with-a-learning-disability-notified-to-leder/

Will have a look now.

#EveryDeathCounts

The LeDeR programme can take a notification of a death of a person with learning disabilities from any source, and includes people of any any age living (and dying) in any location in England. It does not include autistic people without learning disabilities in the figures.
2/?

Notifications to the LeDeR programme are not mandatory, so although the net is cast much wider than other sources of info on COVID-19 deaths there may be gaps (& the occasional duplicate if multiple people have notified the LeDeR programme of a death with little other info)
3/?

The LeDeR is also notified of confirmed (i.e. tested) and suspected COVID-19 deaths (and has info on all deaths notified to them during the same time period).
4/?

In total, 467 COVID-19 related deaths have been reported to the LeDeR programme so far (including 14 deaths where the date of death is not yet known).
For the same time period, 1029 people died from any cause who were reported to the LeDeR programme.
5/?

The table of weekly deaths is below. The number of COVID-19 deaths notified shows a rapid rise to a peak in the week ending 12th April, and now seems to be falling rapidly. There is a similar peak for deaths from any cause.
6/?

And this is the other table (actually a graph) of daily COVID-19 related deaths notified to LeDeR, breaking down the weekly figures into daily figures (based on date of death, not date of notification). Again, there is a clear peak with a gradual decrease over time.
7/?

I think there are a number of things we can learn from this data release.
1) Time. As with the general population (either from NHSE or ONS data), the peak of people with learning disabilities dying (from COVID or any cause) was early-mid April, with a significant drop since.
8/?

2). COVID-19 vs deaths from any cause. As with the general population, the proportion of all-cause deaths that were attributed to COVID-19 increased rapidly, to 66% of all people with learning disabilities dying at the peak (the week ending 12th April), and is now falling.
9/?

3) The number of COVID-19 related deaths from the LeDeR programme corresponds closely to the number of COVID-19 confirmed deaths in hospitals of people withj learning disabilities and autistic people in the NHSE/I data published last week. Could mean any number of things:
10/?

First, could be that LeDeR isn't picking up some deaths - until we know more about where the deaths happened from LeDeR we won't know more about potential gaps (for example, were some hospital deaths or people's deaths happening elsewhere less likely to be notified?).
11/?

Second, could be that most people with learning disabilities with COVID-19 symptoms end up in acute hospitals (there's then a separate question about people's treatment in hospital).
12/?

Third, although the NHS England data is supposed to include people with learning disabilities and autistic people, is it in fact not in daily practice including autistic people without learning disabilities?
13/?

Two massive unanswered questions. Without comparable info from previous years, we don't know how many more people with learning disabilities have died from any cause (or 'excess deaths') during the pandemic so far.
14/?

We do know that people with learning disabilities in England are likely to much die younger than others. This graph estimates the age structure for people with learning disabilities in England for 2017-18 from GP records (better for adults than kids)
https://files.digital.nhs.uk/BA/4F4C1D/health-care-learning-disabilities-1718-sum.pdf
15/?

Notice how few people with learning disabilities even make it to the age bands where the risk of dying from COVID-19 really takes off in the general population.
16/?

We also know that people with learning disabilities are 4 times more likely to die than other people of the same age and sex.
https://digital.nhs.uk/data-and-information/publications/statistical/health-and-care-of-people-with-learning-disabilities/experimental-statistics-2018-to-2019
17/?

Health and Care of People with Learning Disabilities, Experimental Statistics: 2018 to 2019 [PAS] -...

This publication includes aggregated data on key health issues for people with learning disabilities in England compared to those without learning disabilities

https://digital.nhs.uk/data-and-information/publications/statistical/health-and-care-of-people-with-learning-disabilities/experimental-statistics-2018-to-2019

So, even before COVID-19, the 'baseline' level of deaths for people with learning disabilities is already hugely 'excess'.
Without comparable published info on the number of deaths of people with learning disabilities in previous years, is there an excess over the excess?
18/?

In the shifting sands of lack of info, there is no firm ground on which to stand. One way to look at this is to look at this mortality info for people with learning disabilities from about half of GP practices in England from 2016-2019.
https://digital.nhs.uk/data-and-information/publications/statistical/health-and-care-of-people-with-learning-disabilities/experimental-statistics-2018-to-2019
19/?

Health and Care of People with Learning Disabilities, Experimental Statistics: 2018 to 2019 [PAS] -...

This publication includes aggregated data on key health issues for people with learning disabilities in England compared to those without learning disabilities

https://digital.nhs.uk/data-and-information/publications/statistical/health-and-care-of-people-with-learning-disabilities/experimental-statistics-2018-to-2019

This reports, over 3 years, 4,216 deaths of people with learning disabilities from half of GP practices in England. Shoddy back of fag packet calculations (x2, /3, /52) get you to a rate of 54 people with learning disabilities dying every week from 2016-2019.
20/?

Over the 8 weeks of LeDeR data, there were 58.4 people with learning disabilities dying a COVID-19 related death per week, and 70.3 people with learning disabilities dying per week of non-COVID attributed causes.
21/?

Even allowing for shonkiness of calculations, the overall death rate during these 8 weeks is much higher than an average from 2016-2019, which itself is a rate of deaths much higher than the 13.5 deaths per week 'expected' given people's age and sex.
22/?

A second way to think about this is to go back to the hospital COVID-19 death statistics in hospitals reported by NHSE/I last week. That reported 451 COVID-19 deaths of people with learning disabilities and/or autistic people out of 23,953 total deaths (1.9% of all deaths).
23/?

However, for 6,181 people who had died there was no record of if they were a person with learning disabilities or an autistic person or not. Taking these people out, it goes from 1.9% to 2.5% (not the 2% listed in the NHSE/I table).
24/?

NHSE/I claim that this 2% of COVID-19 hospital deaths is similar to the proportion of people with learning disabilities and autistic people in the population generally. If you take prevalence research maybe...
25/?

...but 3/4 of adults with learning disabilities aren't known as such to health or social care services, and even less is known about autistic people. What are hospital staff likely to use to make this judgement?
26/?

Probably something like GP registration as a person with learning disabilities (there is no equivalent for autistic people, where GP identification of adults is low), where 0.5% of all adults are registered as an adult with learning disabilities.
https://digital.nhs.uk/data-and-information/publications/statistical/quality-and-outcomes-framework-achievement-prevalence-and-exceptions-data/2018-19-pas
27/?

Quality and Outcomes Framework, Achievement, prevalence and exceptions data 2018-19 [PAS] - NHS...

Quality and Outcomes Framework (QOF) 2018-19 prevalence, achievements and exceptions data. Data in this publication are presented at national, regional, STP, CCG and GP practice level.

https://digital.nhs.uk/data-and-information/publications/statistical/quality-and-outcomes-framework-achievement-prevalence-and-exceptions-data/2018-19-pas

Compared to between 1.9% and 2.5% of COVID-19 deaths, this suggests (again, shonky fag packet calculations) that people with learning disabilities were 4-5 times more likely to die a COVID-19 related death in hospital than you would expect.
28/?

I've been doing this thread for a couple of hours, and it's turning out horrendously grim. If I'd have known (I hadn't looked at the LeDeR data when I started), I would have given out a warning. Sorry.
29/?

Finally, there is the second, really important, unanswered question I mentioned many tweets ago. What are we going to do? We urgently need info on how old people were when they died, where they lived as well as where they died, and how people caught COVID-19.
30/?

This isn't (only) a question of care homes: most people with learning disabilities don't live in care homes. Really minimising the risk of people getting the COVID-19 infection at all needs shedloads more support, understanding and collaboration than we've seen so far.
31/?

We need to get this in place NOW so people with learning disabilities and autistic people aren't hit in any second second (third, fourth...) spike in the way they clearly were in the first spike.
32/?

And we need to make sure that protection and safety goes hand in hand with people being supported to have a healthy and fulfilling life, with society not reverting to a divisive way of life that kills people in other ways instead of COVID-19.
33/?

Last tweet. One acid test - who will be highest priority groups of people for a vaccine?
Sorry for the length and grimness of this thread, and the shonky statistical work in the absence of better data.
#EveryDeathCounts
34/34

You can follow @chrishattoncedr.

Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: