21-04-2021
Q. What do we know now in March 2021 that we did not know in December? What might this mean for public policy, for services, and for future research? What is new since your last blog on this subject?
There are two pieces about deaths of people with learning disabilities in England, one from Gyles Glover at Public Health England (PHE), and one from the Office for National Statistics (ONS).
The third is a pre-publication print from the Scottish Learning Disabilities Observatory (SLDO) about deaths in Scotland. Note, at this stage the findings from the SLDO paper are under peer review and may be modified. It should not yet be relied upon to guide practice. The results it reports are highly relevant, so that it is included here while we wait for the definitive article. Only the PHE report was available when our last blog was published.
Across the UK, this work points clearly to the high risk for people with learning disabilities of coronavirus infection, and has contributed to the decisions to prioritise their access to COVID-19 immunisation. It also raises questions about the health of people with learning disabilities even before the pandemic and to the importance of the lives and deaths of people with learning disabilities.
Q. Is the risk of dying with COVID-19 greater for people with learning disabilities than it is for those with no such disabilities?
Yes. All the reports confirm that there have been proportionately more deaths of people with learning disabilities from COVID-19 than for people without such disabilities. There may be some variation in the scale of that risk given their different approaches, but all agree that there were significantly more deaths per 100,000 population overall, and a greater risk of dying for people with learning disabilities than those without. More families and friends are now having to mourn the recent death of someone they loved who had learning disabilities.
So far, not surprisingly, most available information concerns deaths in the first wave of the pandemic, mostly up to the summer, or at the latest, the autumn of 2020. Each report covers a slightly different time frame, described in more detail below.
ONS studied data from January to November 2020 on people over 30 whose details had been recorded on the 2011 census, who were alive on 24th January 2020, and for whom current Primary Care records were available.
PHE looked at data collected in the spring of 2020 from the English Learning Disabilities Mortality Review (LeDeR); NHS England’s COVID-19 patient notification system (CPNS) which records deaths in hospitals; and Care Quality Commission (CQC) statutory notifications of the deaths of people receiving social care.
The SLDO study was also based on data on people identified by the 2011 census alive on the 24th of January 2020, who were over 29 in March 2020. They were able to link this data with national death certificate data; national hospital admissions data from Public Health Scotland (PHS); and laboratory results of all Covid-19 tests also held by PHS.
The methodology of the Improvement Cymru study was described in our earlier blog. The study compared the death rate among people already known to hospital services as having learning disabilities with the general population of Wales, between 1st March and 19th November 2020.
All the studies found there was higher mortality from COVID-19 among people with learning disabilities than was seen in the general population and that the differences became more marked when age at death was taken into account. That is, the risk of death from COVID19 was elevated for all ages but the difference was even more marked for younger adults.
Q. You said more people with Learning disabilities have died than in recent years. How many extra deaths have there been, how much more likely has death been?
Excess mortality
As we pointed out in our earlier blog, even before Covid-19, people with learning disabilities, age for age, had higher death rates than the general population. Data from SLDO and Improvement Cymru show there were more deaths of people with learning disabilities in the spring of 2020 than over several previous years.
SLDO compared the deaths of adults with learning disabilities between 24th January and 15th August 2020 for each of five years starting in 2015. They found that found that, once the data had been standardised for age, sex and deprivation, the all-cause mortality ratio was 2.38 for the relevant months in the five years up to 2019 compared to 2.49 in 2020. People with learning disabilities were already more than twice as likely to die as people without disabilities of the same age and sex before COVID-19. The death rate was only slightly higher for people with learning disabilities during the first months of the pandemic. Improvement Cymru similarly found there were more deaths among people with learning disabilities in the months they studied in 2020 than in the same months in the four previous years, concluding that most, though perhaps not all the excess deaths were COVID related and that further investigation of ‘excess mortality’ may be warranted.
This bar chart from improvement Cymru shows the increased death rate in 20220 relative to the same months in earlier this year.
A key point here is that it would be a major weakness, in any post COVID-19 planning for health service delivery, to simply call for a return to the pre-COVID-19 normal. Services must aim to do better than a return to a ‘normal’ that means that people with learning disabilities continue to die younger from causes amenable to better health care. The death rate before the pandemic was already unacceptably high, so that COVID-19 widened a gap that was already there. Pandemics are not the only cause of premature death among people with learning disabilities. A study under way on the impact of the 1918-1919 Influenza pandemic on people with learning disabilities shows the weakness of only looking for a return to ‘normal’. The high rates of death seen in the years before that pandemic could be attributed to TB, overcrowding, and the consequences of living in an asylum in wartime conditions. The end of the First World War and of that pandemic did see a return to normal. A ‘normal’ where one in ten of the people in the asylum died every year, compared to almost one in four in 1918. To put that historical rate in a 21st century context, a recent study showed that for adults with learning disabilities in supported living or residential care the death rate in 2014 was about one in a hundred a year, which was still higher than the rate seen for is for the general population.
Likelihood of Death
ONS Study
ONS looked both at all disabled people,(self-defined in the 2011 census), and specifically at people with learning disabilities (identified in primary care). The general disability figures are also interesting. ONS looked at mortality among people who in the 2011 census identified themselves, or were identified by the person completing the census, as ‘disabled’. This will have included some people with learning disabilities and people with a range of physical, sensory and other disabilities or conditions. Age Standardised Mortality per 100,000, for deaths involving COVID-19, was 313.1 for women in the more disabled group, 152.8 in the less disabled group and 77.1 among those not limited by disability. For men the rates per 100,000 were 469.6 for the more disabled, 270.7 among the less disabled and 148.5 for the non-disabled. All disabled people have been at greater risk of dying from COVID19, people who are more disabled were at even greater risk.
People with learning disabilities, identified not by census returns but from GP diagnostic codes, accounted for 2,955 (5.8%) of the 50,888 deaths involving coronavirus notified to ONS between January and November 2020, although people with learning disabilities only made up 1.2% of the population studied. Those with learning disabilities had a significantly higher rate of COVID-19 mortality than those who did not, in the period up to November 2020. Age Standardised Mortality rates involving COVID-19 for women with learning disabilities were 475.8 per 100,000 compared to 118.8 for women without such disabilities; and for men, 690.6 for those with learning disabilities and 196.1 for those without. 4 times as many women and 3.5 times as many men with learning disabilities died with COVID-19, compared to the general population. Men were more likely than women to die but the gap between those with and without learning disabilities was bigger for women.
PHE Study
PHE found that LeDeR was notified of 623 deaths between early February and June 5th, 2020. They considered the probability that there was under-notification of deaths to LeDeR and suggested an estimated national total of 956 deaths. The Covid related death rate among people reported to LeDeR was 240 per 100,000, 2.3 times as many as for the general population, though the exact denominator for LeDeR is not known. Correcting for under-reporting, PHE estimate a death rate of 369 per 100,000, 3.6 times the rate in the general population.
CPNS recorded 490 deaths with COVID-19 of adults with learning disabilities up to 5th June 2020. 192 per 100,000, 2.3 times the rate in the general population in the same period. Again, the data set is incomplete, not everybody who had a learning disability will have had it recorded on their hospital notes. If those whose learning disability status is recorded as “not known” are added the rate rises to 254 per 100,000 or 4 times the general population rate.
SLDO Study
The study found a crude Covid related death rate for people with learning disabilities of 259/100,000 compared to 114/100,000 in the general population. Overall people with learning disabilities compared to those without such disabilities, were 2.3 times more likely to die with COVID-19After correcting for age, sex and the level of deprivation where the deceased person lived, 3.2 times.
Improvement Cymru compared deaths among a specific cohort of people with learning disabilities with deaths amongst all Welsh residents. The findings suggest that the age-standardised rate of deaths involving COVID-19 is around 3 times to 6 times higher in this cohort than the population as a whole. Once more, the population of people with learning disabilities was limited, in this study, to those who were already identified by hospitals as having Learning Disabilities, compared with the population of Wales, whether or not they were known to hospitals. The comparison between these two populations for age specific mortality for people under 60 was even more startling, 107 death per 100,000 people with learning disabilities but only 8.4 per 100,000 of Welsh residents generally. People with learning disabilities are dying more often with COVID-19. They are also dying younger.
So, though all the papers agree that the mortality rate from COVID-19 is higher for people with learning disabilities than for the general population, the range is from 2.3 times to nearly 13 times higher for people under 60 in the IC data.
Q. Why are the results mortality rates from COVID19 so different in these papers?
That is where it starts getting more complicated because, as we said in our earlier blog - people with learning disabilities are hard to spot in routinely collected data. The authors of each study have made a huge contribution in pulling together such information as is available, and getting it out into the public domain so that it can inform policy. Each study used a different methodology. There were differences in the sources of the data which led to differences in definition of Learning Disabilities, and in age banding. Each collected data over a slightly different time frame. What they all agree on, however, is that people with learning disabilities have been at increased risk of dying from COVID-19 compared to people without such disabilities.
Who was defined as an adult with learning disabilities was different in each study.
The Scottish study concerns people who in 2011 defined themselves, or were defined by the person who filled in the census form, as having learning disabilities. Though ONS had access to the 2011 census data, the categorisation of learning disability they used to look at learning disabilities depended on a medical diagnosis being recorded in Primary Care records. They note that 59% of those so recorded reported themselves as disabled in the census. The two populations are not the same, so we cannot tell which is the best estimate of the ‘true’ number of people with learning disabilities. PHE also looked at deaths of those on Primary Care Learning Disability Registers; at deaths reported to LeDeR who were almost all people already recognised by services as having learning disabilities; and at CPNS records of deaths in hospital where the hospital had recorded that the person had learning disabilities. They are likely therefore to have been people with more pronounced, or at least more obvious, learning disabilities. Neither LeDeR nor CPNS were designed to give an estimate of the size of the living population in which the deaths occurred. The estimates of how many people were living with learning disabilities, which differed in each study, are necessary to calculate the death rate in that population. Improvement Cymru looked at people whose hospital records at any time in their life, had specified learning disabilities or one of a number of conditions, such as Down syndrome, known to cause learning disabilities. People in good health who had never used hospital services were not included.
Time Frame
All the studies covered the earlier months of the pandemic in 2020 but they did not cover identical time- frames. PHE looked at the period between February and June 2020, Improvement Cymru 1st March to 19th November, SLDO 24th January to 15th August and ONS 24th January to 20th November, all in 2020. More deaths were seen in the studies that covered a longer period of time, which is not surprising. There were also changes in the availability of COVID testing and death certification practice over that period, as well as changes in infection control and treatment, these may have affected mortality rates.
Outcome measures
Deaths reported to LeDeR may or may not include the certificated cause of death; the way in which Covid-19 was or was not recorded on death certificates may have changed over time.
Q. I read that people with learning disabilities between 18 and 34 are over 30 times as likely to die with Covid as people without those disabilities. Is that true? How concerned should young people with learning disabilities, and their families be?
Yes, it is true that the LeDeR data showed death rates for people with learning disabilities that were 31.7 times the rate for the general population. But the numbers of people, with or without learning disabilities, dying of Covid at such a young age was extremely low. So, although the figure of 31.7 is startling, it still means that very few people with learning disabilities of that age have died from COVID-19. Because the number of younger adults dying is relatively small, both ONS and SLDO had to use fairly broad age bands, to give reliable results, and to avoid identifying individual people. Only the LeDeR and CPNS data breaks down adults under 55 into narrower age bands.
This figure is taken from the PHE report and shows both the proportionately higher death rates in the younger age groups, compared to the general population, and that the number of deaths was lower among younger people.
Figure 4.2 Age-specific rates, per 100,000 adults, to 5 June 2020 for reports of COVID-19 deaths to LeDeR and for COVID-19 deaths in the general population. Grey and black bars show rates using data as notified. Outlined white bars show estimated COVID-19 death rates for people with learning disabilities allowing under-notification.
So, we should be very concerned to insist that preventative measures, such as the availability of PPE, good hygiene, good ventilation, and timely access to testing are available even to the youngest adults with learning disabilities, and that they have priority over their age-mates in the vaccine queue. At the same time, a person between 18 and 30 is unlikely to die of COVID-19, even though the likelihood is much greater than for a person the same age without learning disabilities. Though the exact number varies, each study suggests that people with learning disabilities who are in their 50s & 60s are at considerably higher risk of dying from COVID-19 than younger people with learning disabilities.
Q. Why are people with learning disabilities more likely to die with Covid? Are they more likely to get infected, or more likely to become severely ill or die when they do?
The short answer seems to be, all three. The SLDO was able to compare positive Covid test results, hospital admission rates and death between people with learning disabilities and the general population. They found infection rates of 957 against 237 per 100,000 severe infection rates of 549 against 237 per 100,000, and for mortality 259 against 114 per 100,000. Case fatality, the death rate among people known to have COVID-19, was 30% for people with learning disabilities but 25% in the general population. So people with learning disabilities were more likely to become infected, and had a worse outcome when they were than people without such disabilities.
Q. Does the sort of place a person lives in affect their chance of dying with COVID-19?
ONS found the rate of death involving COVID-19 was 3.7 times higher than in the general population. Once they adjusted for age, and then for living in a communal establishment (such as a shared supported living setting or a care home) the discrepancy fell to 2.1 for men and 2.3 for women. They point out that living in a communal establishment was a major factor in the exposure of people with learning disabilities to Covid-19. Living with a number of other people and receiving intimate support from a number of people outside of their household does seem to be a major risk factor. ONS also looked at other factors that might affect COVID-19 mortality in people with learning disabilities and concluded that the effect of type of residence was more important than geographical, demographic or socioeconomic differences, or underlying health conditions, though each of these had some effect. None of the three papers gives details of differences in, for example, the size, designation or organisation of people’s homes that might increase or reduce the risk of contracting the infection, becoming severely ill with it, or of dying. These would be difficult to estimate given that such detailed date would not be typically recorded. They are, however, important questions.
Q. Are people with Down Syndrome more likely to die with COVID-19, compared to other people with learning disabilities?
The SLDO paper mentions people with Down syndrome in the discussion section, in a way that suggests that they may have found an increased risk for adults with Down syndrome, but the detailed results are not presented in this short pre-print. Look out for more publications from this group. International reports also suggest adults with Down syndrome are at higher risk even than other adults with learning disabilities because of biological differences in immune response, high rates of respiratory problems, obesity and dementia. When numbers are small it is difficult to find out what is happening for smaller sub-groups of people, because the analysis could make it possible to identify an individual, compromising confidentiality.
Q. Are these findings taken into account in deciding vaccination priorities?
They are now. Adults with Down syndrome should have been offered vaccination in group 4 and, on 24th February it was announced that all adults on GP Learning Disability registers across the UK should be offered the vaccine in group 6, see for example, doi: https://doi.org/10.1136/bmj.n547 , https://gov.wales/covid-19-vaccinations-individuals-learning-disability-or-severe-mental-illness-html. The advice given by the UK’s independent Joint Committee on Vaccination and Immunisation (JCVI) changed in response to this kind of data, as well as to on-line campaigning by people with learning disabilities, their families, and disability organisations.
Q. Why is this kind of numerical research useful?
If we don't count, they don't count, blogged Canadian researcher and activist Yona Lunsky. She emphasised the value of both numbers and personal stories in campaigning for access to vaccines in both Canada and the UK, Knowing and understanding the numbers is part of the ammunition activists need to make governments change their minds. It is not the only thing that is needed, and it takes hard work to effect political change. None of the public health or research groups found it easy to answer highly relevant questions about the effect of the pandemic on deaths among people with learning disabilities because of a lack of routinely collected data. Yet the results they did manage to assemble shows a group of people with high death rates before 2020, and a widening gap between them and the general population as COVID-19 spread. These results have potential implications for the way social care is organised and the way in which data is collected.
Q. Who should set the priorities for research and data collection?
Regardless of whether you think numbers or stories count for more, research urgently needs greater participation of people with learning disabilities and their families in shaping research priorities and activities. The families of people with learning disabilities have long been powerful advocates. Though the self-advocacy movement also has a long history, one of the positive things that has happened in this Covid period, is that self-advocates have also become more visible. People with learning disabilities are speaking out powerfully on television and social media. They have shown that there are indeed people with learning disabilities who are prepared to take on difficult issues like death and dying and have important things to say about them. It is ironic that it took a pandemic to achieve that, and truly terrible that so many people have died before their time, and that so many of them had learning disabilities. These days, dying is something more of us, with or without learning disabilities, seem prepared to think or talk about. In the past, the participation of people with learning disabilities in decision making at the end of life has been blunted by a reluctance to start those conversations. The increasingly powerful voice of people with learning disabilities themselves is something that we must not lose but build on. People will continue to die in ways that demand better end of life care, more open discussion and respect for personal choices,. We cannot return to ‘normal’ practice here either.
Q. Is it just in the UK that people with learning disabilities are so badly effected by COVID-19?
No, sadly, it isn't. In the United States Gleason and his colleagues report the devastating impact of COVID-19 on people with Intellectual Disabilities (who in the UK would be referred to as having learning disabilities). They found that in a population of 64,858,460 people, of whom 127,003 were known to be people with Intellectual Disabilities, having an intellectual disability was the strongest independent predictor, other than age, for COVID-19 mortality. That is, it was more strongly associated with COVID-19 mortality than kidney disease, lung disease, diabetes, high blood-pressure, lung disease, or obesity; though they point out those conditions are also more common in people with intellectual disabilities. Like the Scottish Observatory they found that people with intellectual disabilities were more likely to be diagnosed with COVID-19. An interesting feature of this study is that they found that people with intellectual disabilities were not the group most likely to be admitted to hospital, or in the group of patients who, when admitted, were most likely to die. It seems that treatment pathways were different for people with intellectual disabilities. It would seem important not just to understand what outcomes the pandemic has had on mortality in people with learning disabilities, but also the treatment pathways that were followed for those infected. The authors note that these figures give a prevalence of intellectual disabilities of about 0.2% which is much lower than the proportion of people in the US estimated to have intellectual disabilities. Health information about people with intellectual disabilities is not systematically collected. Once again only the people whose disabilities are most visible will be the people most likely to be counted.
What questions are still left unanswered?
The most obvious one is whether the second wave was as disastrous as the first one for people with learning disabilities. The early results in the ONS report are mildly encouraging. The excess mortality risk for people with learning disabilities, compared to those without, was slightly lower in the second wave than the first but this difference was not statistically significant. Both ONS and the latest Improvement Cymru report cover the period up to November 2020, when the peak had not yet been reached. There seem to have been fewer deaths in the second wave than the first, and the number of deaths each week is now falling.
Where did people with learning disabilities die during the pandemic? What treatment pathways did they follow and were they the same as for other people the same age? What was the pattern of dying? How long was the gap between the onset of symptoms, a positive test result, and death? How many of the people with learning disability who died were admitted to hospital and how many died at home? Another set of questions arises about people living in what ONS calls communal establishments. Can we identify any features of supported living/care home settings which made some settings more, or less, risky for the people whose homes they were and are?
These are all important questions with practical implications for health and social care services. The pandemic also brings other questions about the lives and deaths of people with learning disabilities into sharper focus. More people with learning disabilities have died - and have died prematurely than for many years before. What has been the impact of these deaths on families, friends, house-mates and paid care/support staff? In our last blog we raised the specific question of bereavement. What can these deaths and the response to them tell us about the value society has for the life of a person with learning disabilities?
These are also methodological issues. Data linkage studies are a great way of looking at large numbers of people, but are much harder to use to understand smaller, important subgroups, like people with Down syndrome, or younger adults. Can we develop this powerful tool to look at these smaller groups or do we need to find other ways? What denominator population should we be using? GP registers, census data, hospital registers, or something else?
How can we use what we have learned from the COVID-19 pandemic to ensure that people with learning disabilities are not exposed to greater risks from COVID-19 and of premature death more generally, and have access to the services they need throughout their lives, up to the very end? That is a rhetorical question without a clear answer, but one that calls for continued resources and commitment, more than ever before. When it comes to people with learning disabilities and death and dying, a return to the old normal is just not acceptable.
Covid-19 vaccinations for individuals with a learning disability or sever illness: Explains the criteria to use to identify individuals with a learning disability or severe mental illness in priority group 6 for COVID-19 vaccination. (2021). Cardiff Retrieved from https://gov.wales/covid-19-vaccinations-individuals-learning-disability-or-severe-mental-illness-html.
Gleason, J., Ross, W., Fossi, A., Blonsky, H., Tobias, J., & Stephens, M. (2021). The Devastating Impact of Covid-19 on Individuals with Intellectual Disabilities in the United States. NEJM Catalyst. doi:DOI: 10.1056/CAT.21.0051
Glover, G. (2020). Deaths of people identified as having learning disabilities with COVID-19 in England in the spring of 2020. Retrieved from London: my son https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/933612/COVID-19__learning_disabilities_mortality_report.pdf
Henderson, A., Fleming, M., Cooper, S. A., Pell, J., Melville, C., Mackay, D.,...Kinnear, D. (2021). COVID-19 infection and outcomes in a population-based cohort of 17,173 adults with intellectual disabilities compared with the general population. medRxiv. doi: https://doi.org/10.1101/2021.02.08.21250525
Illouz, T., Biragyn, A., Frenkel-Morgenstern, M., Weissberg, O., Gorohovski, A., Merzon, E., . . . Okun, E. (2021). Specific Susceptibility to COVID-19 in Adults with Down Syndrome. Neuromol Med. doi: https://doi.org/10.1007/s12017-021-08651-5
Lunsky, Y. (2020) If we don’t count, they don’t count – Numbers and stories about COVID, vaccines, and developmental disabilities in Canada and the UK. Retrieved from: https://www.porticonetwork.ca/web/hcardd/news/-/blogs/-if-we-don-t-count-they-don-t-count-
Mahase, E. (2021). Covid-19: All adults on learning disability register should be prioritised for vaccination, says advisory committee. BMJ, 372. doi: https://doi.org/10.1136/bmj.n547
Update estimates of coronavirus (COVID-19) related deaths by disability status, England: 24 January to 20 November 2020. (2021). Retrieved from London: <https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/coronaviruscovid19relateddeathsbydisabilitystatusenglandandwales/24januaryto20november2020>
Watkins, A. (2021). COVID-19-related deaths in Wales amongst People with Learning Disabilities from 1st March to 19th November 2020. Retrieved from Cardiff: https://phw.nhs.wales/publications/publications1/covid-19-related-deaths-in-wales-amongst-people-with-learning-disabilities-from-1st-march-to-19th-november-2020/
About the authors:
Jane Bernal: Jane Bernal is a Visiting Professor at the University of South Wales. Her background is in the Psychiatry of Learning Disabilities. Her research interest is in death and dying as it affects the lives of people with learning disabilities. Jane also has a long standing interest in ethics and communication, including how and whether people are told they may be dying. Jane also has a long standing interest in ethics and communication, including how and whether people are told they may is be dying.
Stuart Todd: Stuart is a Professor in Intellectual Disability research at the University of South Wales. His research has been based on Welsh, UK and European collaborations. His research interests span across the lives and experiences of people with intellectual disabilities and their families, including care at end of life.
28-09-2023
20-02-2022
21-10-2021
07-10-2021
03-08-2021
14-06-2021
26-05-2021
21-04-2021
24-03-2021
