Thread on new statistics published today by @NHSEngland on COVID-19 deaths of people with learning disabilities and autistic people in hospitals in England. It's a new table in their daily deaths statistics.
#EveryDeathCounts
https://www.england.nhs.uk/statistics/statistical-work-areas/covid-19-daily-deaths/
1/21

Took a bit of staring at it, but this is my understanding so far. How I think it works - on 24th March (six weeks ago, before the first general hospital peak of deaths but after the start of the pandemic) NHSE/I introduced a flagging system to flag, among other things...
2/21

...a 'pre-existing condition of learning disabilities and/or autism', so people are put together into this combined group and it looks like people can't differentiated.
3/21

This info covers people dying in acute hospitals, and mental health/learning disability/autism inpatient units (both NHS and independent sector). From the small number of deaths of any group recorded as happening in inpatient units (76 deaths in total), almost all the deaths
4/21

...of people with learning disabilities and/or autistic people in this dataset are likely to have happened in acute hospitals (obviously, people could have been living anywhere else, including inpatient units, before being admitted to an acute hospital).
5/21

Another thing to take into account - this dataset only records the deaths of people with learning disabilities and autistic people confirmed with a test as having COVID-19. This is consistent with the ways that NHSE/I have reported COVID-19 deaths for everyone throughout.
6/21

So, enough caveats. The figures show that from 24th March to 14th May 2020 (a 6-week period), 451 people (flagged as) with learning disabilities and/or autistic people died in hospitals in England with confirmed COVID-19. That's more than 75 people per week.
7/21

What to make of this bald statistic, standing in for 451 people's lives and deaths, and the impact on those who love them? What might these figures mean?
8/21

First, these 451 deaths represent 1.9% of all 23,953 COVID-19 deaths in hospitals in England for the same time period.
9/21

However, 6,181 people weren't flagged either way as a person with or without learning disabilities or autism - taking these people out means that 2.5% of all deaths where a person's 'status' was flagged were people with learning disabilities and/or autistic people.
10/21

I don't think there's any certainty about how many people with learning disabilities and autistic people combined are living in England, but in terms that would get people flagged in hospital services my guess is it would be a lot less than 2.5% (or 1.9%) of the population.
11/21

So it's possible that more people with learning disabilities and autistic people are dying of confirmed COVID-19 in hospitals than you would expect from the number of people in these groups living in England. Why might that be?
12/21

After all, grimly, fewer people with learning disabilities and autistic people live to the older ages where risk of death from COVID-19 death shoots up, but people are more likely to experience health conditions which might place people at greater risk.
13/21

Again, it might be that more people with learning disabilities or autistic people are getting admitted to acute hospitals, or it might be that people are less likely to received life-saving treatments once in hospital.
14/21

There is no way to know, but further analysis of this dataset for people with learning disabilities and autistic people would really help:
1) A weekly breakdown would help to see if the number of deaths has peaked and is now falling, or is following a different pattern
15/21

2) A breakdown of deaths by age, sex and BAME. Are younger people with learning disabilities and autistic people more likely to die of COVID-19 than the population generally - a really important question for how people are supported.
16/21

3) Looking at how many people with learning disabilities and autistic people who died who also were recorded as having the health conditions listed in the same table - are there any particularly high-risk health conditions for these groups?
17/21

Just to finish, as with any dataset there are limitations:
1) It's only about people who have died in hospitals, not people who died elsewhere (will @ONS be able to do similar analysis, or to what extent can LeDeR or CQC data help to fill this gap?).
18/21

2) It's only confirmed COVID-19 deaths. So it's not suspected COVID-19 deaths (likely to be more prevalent if people have died in places where less tested has occurred)c, and importantly it's not deaths from any attributed cause.
19/21

3) It doesn't tell you anything about where people were living, which is essential if we are understand how to ensure effective infection control in the different circumstances that people are living in.
20/21

But, with more analysis it could yield some useful, actionable information.
21/21
#EveryDeathCounts