It's institutional disablism - deal with it

Today saw the publication of a landmark study by Australian academics Julian Trollor, Preeyaporn Srasuebkul, Han Xu and Sophie Howlett on deaths amongst a cohort of 19,362 people with learning (intellectual) disabilities in New South Wales, Australia. Their findings are in BMJ Open here, and are free to download and read. This article in the Sydney Morning Herald is also a really good summary, and there is an easy read summary here.

Their findings are depressingly familiar. The median age of death (the age at which half of all deaths had occurred) for people with learning disabilities was 54 years (55 years for boys/men; 52 years for girls/women). This was 27 years younger than the median age of death of 81 years reported for the rest of the population. Over three-quarters of the deaths of people with learning disabilities (76%) occurred before the age of 65, compared to just 18% of the deaths of other people. Adults with learning disabilities aged 20-44 were four times more likely to die than the rest of the population of the same age.

The researchers also used death certificates to investigate what had been recorded as the causes of death for people with and without learning disabilities. The most common underlying causes of death for people with learning disabilities were respiratory system illnesses (often infections; 20% of people), problems with the circulatory disease (such as strokes, 18% of people), cancers (18% of people), nervous system problems (such as epilepsy, 13% of people), and injury/poisoning (6% of people).

The researchers concluded that over a third of deaths of people with learning disabilities (38%) were potentially avoidable, meaning that health services could have done something to prevent the health problem developing and/or stop the death from happening. This compares to 17% of the deaths of the rest of the population being classed as avoidable. One shocking detail in the paper is that when they first looked at the death certificates of the people with learning disabilities, the main cause of death on the certificate had been recorded as the person’s learning disability itself on 16% of the certificates (102 out of 637 where certificates were available) – so for 49 people their main cause of death was recorded as ‘Down syndrome’. This isn’t even taking into account the 13% of people (95 people) with learning disabilities where the researchers couldn’t find any record of cause of death at all.

Why did I write that these findings are depressingly familiar? Because similar findings have been found in other countries, wherever people have looked. In England (see Chapter 3 here for a free recent summary), the median age of death for people with learning disabilities is 13-20 years younger for men and 20-26 years younger for women compared to the general population, with the gap in life expectancy not closing over time. There are consistent gaps in age of death for people with learning disabilities compared to the general population reported in the USA, Canada, Ireland and Germany, with some variation in just how yawning the gaps are.

The most common causes of death of people with learning disabilities are also similar across countries, and people with learning disabilities are more likely to die potentially avoidable deaths. For example, in England nearly half of the deaths of people with learning disabilities (49%) were avoidable, compared to just under a quarter (24%) of the deaths of the general population. The recording of a person’s learning disability as a cause of death on death certificates has also been reported in the USA and the UK.

In England, in large part due to the #JusticeforLB campaign and other campaigners, health service ‘regulators’ have begun to pay reluctant attention to the deaths of people with learning disabilities within health services, and how health services do or don’t investigate people’s deaths. The shocking report by Mazars into how Southern Health NHS Foundation Trust (the health service where LB died) determinedly and repeatedly didn’t investigate the deaths of people with learning disabilities who died in their ‘care’ laid out just how a health organisation can operate in ways to deny reality and systematically denigrate anyone who tries to bring a dose of reality to their proceedings.

One of the broader health service institutional ‘responses’ to this report was a piece of work by one of England’s health service regulators, the Care Quality Commission, looking at how health services investigated the deaths of people with learning disabilities and/or mental health issues. This resulted in the report Learning, Candour and Accountability published in December 2016. Over time I more and more have the sensation of being shocked but not surprised at these kinds of reports. The CQC summarised the main findings like this: “We weren’t able to identify any trust that demonstrated good practice across all aspects of identifying, reviewing and investigating deaths, and ensuring that learning is implemented”. More specifically:

• “We found that families and carers often have a poor experience of reviews and investigations, and are not always treated with kindness, respect and sensitivity.”
•  “We found inconsistency in the way organisations become aware of the deaths of people in their care, with no clear systems for a provider that identifies a death to tell commissioners or other providers involved in the person's care.”
•  “Healthcare staff use the Serious Incident Framework to help them decide whether a review or investigation is needed. But this can mean investigations only happen if a serious incident has been reported, and the criteria for deciding to report an incident and the application of the framework both vary.”
•  “The quality of investigations is often poor and methods set out in the Serious Incident Framework aren't applied consistently. Specialised training and support aren't given to all staff carrying out investigations. There are problems with the timeliness of investigations and confusion about standards and timelines set out in the guidance.”
•  “There are no consistent frameworks or guidance requiring boards to keep all deaths under review, and boards only receive limited information about the deaths of people using their services. When they do receive information, they often don't challenge the data effectively. Where investigations take place, there are no consistent systems to make sure recommendations are acted on or learning is shared.”

So, despite national frameworks and guidance on investigating and reporting deaths, and a generally high national profile on candour and accountability since the Francis report into mid-Staffs, the entire health service and regulatory system operates in a way that makes the deaths of people with learning disabilities invisible and unaccountable. And these aren’t random variations in practices where there is uncertainty about what to do; they are practices heavily weighted towards the extremely shoddy end when it comes to the deaths of people with learning disabilities.

Isn’t this surely a basic, convincingly evidenced indicator of systemic institutional disablism? I’m using this phrase as analogous to the phrase ‘institutional racism’ used by Sir William McPherson in his judicial inquiry into the death of Stephen Lawrence. Institutional racism was defined in the inquiry as:

“The collective failure of an organisation to provide an appropriate and professional service to people because of their colour, culture, or ethnic origin. It can be seen or detected in processes, attitudes and behaviour which amount to discrimination through unwitting prejudice, ignorance, thoughtlessness and racist stereotyping which disadvantage minority ethnic people”.

Within the wide array of health services in England (and likely in other places too, given the international similarities in evidence concerning the health inequities experienced by people with learning disabilities), isn’t this what we’re seeing when it comes to people with learning disabilities? (I’m fairly certain similar arguments could and probably have been made with respect to other groups of people, such as people with mental health issues, other groups of disabled people, or older people with dementia). I’ve added ‘systemic’ because it’s not a single organisation, and I'm aware there are issues around the use of 'disabilities' rather than 'impairments' that need more careful thinking through:

Systemic institutional disablism is the collective failure of organisations to provide an appropriate and professional service to people because of their learning disability. It can be seen or detected in processes, attitudes and behaviour which amount to discrimination through unwitting prejudice, ignorance, thoughtlessness and stereotyping which disadvantage people with learning disabilities.

So, systemic institutional disablism isn’t only actions and behaviours that are overtly discriminatory , although as the news report of the Australian study, the Confidential Inquiry in England, and any number of the experiences of people with learning disabilities and their families show, some health professionals seemingly don’t have a problem expressing overt discrimination. Institutional discrimination can be unconscious, and can be tracked through its systematic effects on the treatment (or otherwise) of people with learning disabilities. A quick random list of evidence for systemic institutional discrimination of the health system concerning people with learning disabilities (these are for England, but there is evidence for some of these in other countries too):

• People with learning disabilities often aren’t identified as such within mainstream hospital services, making it difficult to monitor the presence or absence of discrimination in how people are treated.
• Although primary care general practitioners in England have been financially incentivised for some years to provide annual health checks for people with learning disabilities, only about half of people with learning disabilities actually get a health check.
• People with learning disabilities experience lower rates of screening for all cancers compared to people without learning disabilities.
• People with learning disabilities experience delays in the identification and treatment of health problems.
• People with learning disabilities living in community settings are far, far more likely to be prescribed antipsychotic medication in the absence of a relevant diagnosis than people without learning disabilities.
• People with learning disabilities are more likely than other people to experience emergency admissions to hospital for health problems such as constipation that, if well-managed, shouldn’t result in the person needing to go to hospital at all.
• Reasonable adjustments to hospital services for people with learning disabilities are not consistently or reliably provided.
• Despite five years of government policy, the number of people with learning disabilities in inpatient services is not reducing, and whilst in inpatient services people with learning disabilities are highly likely to be prescribed antipsychotic and tranquilising medication, and experience physical restraint, seclusion, and physical assault.
• People with learning disabilities die much younger than other people, where a much higher proportion of their deaths are potentially avoidable.
• Deaths of people with learning disabilities in health settings are extremely unlikely to be properly investigated.
• People with learning disabilities are much less likely to be involved in research about effective health interventions.

Does it matter if we call this out as systemic institutional disablism? I think it does. One of the things that disappointed me about the CQC Learning, Candour and Accountability report is that it identified the issues very clearly, but couched its recommendations in terms of technical changes to systems, on the assumption that systemic poor practice is a function of professional lack of knowledge, inattention and honest confusion. In England there have also been multiple initiatives to improve various aspects of healthcare for people with learning disabilities, all operating on similar assumptions.

But I think there is something crucial missing. Professionals within health systems have had sufficient evidence for some now – ignorance and inattention can no longer be used as reasons/excuses – indeed they can be seen as further signs of institutional disablism. As can be seen from the Home Affairs Committee report into McPherson – 10 years on, the term institutional racism was resisted by many in the police and has been the cause of deep discomfort. Progress in response to the McPherson inquiry has also been patchy and uneven. But this discomfort, self-questioning, and willingness to see the systemic whole in terms of institutional disablism is urgently needed if health services are to achieve any much-vaunted step changes in how people with learning disabilities and their families are treated. Isn’t seeing the problem the first step?

Security blanket

In the ongoing debate about the Transforming Care programme in England, a distinction is often made between inpatient services described as Assessment and Treatment Units (ATUs) and inpatient services variously described as secure and/or forensic. The argument sometimes follows from this that Transforming Care should be focused on reducing/closing the number of ATU places (largely commissioned by CCGs), because secure/forensic inpatient services (largely commissioned by NHS England specialist commissioning teams) are always going to be needed for some people with learning disabilities and/or autism. Personally, I’m not sure such a sharp distinction is possible, desirable, represents reality, or necessarily means that we need the amount of secure/forensic inpatient services that exists today.

In this blogpost I want to describe the range of inpatient services that exist, using Assuring Transformation data collated and analysed by NHS Digital. The Assuring Transformation dataset presents ongoing monthly information about inpatient services for people with learning disabilities and/or autism, collected from health service commissioners. While I think it under-represents the total number of people with learning disabilities and/or autism using inpatient services (see here for why), it does contain a lot of valuable detail about inpatient services and the people within them. As well as describing these services, I’m also interested in any clues we can find to support the distinction between ATUs and secure/forensic services.

The first graph below shows the number of people with learning disabilities and/or autism in different types of inpatient service (as labelled in the Assuring Transformation dataset), for March 2015, December 2015, May 2016 and November 2016. In November 2016, the most common type of inpatient service was ‘secure forensic’, containing (a word I use advisedly) around half of all people within the Assuring Transformation dataset (1,725 people, 50.6%). ‘Acute learning disability’ services (in effect ATUs) contained 460 people (18.3%), followed by complex care/rehabilitation services (330 people; 13.1%). ‘Forensic rehabilitation’ services weighed in with 120 people (4.8%), 155 people (6.2%) were in ‘acute generic mental illness’ inpatient services, and a further 180 people (7.1%) were in inpatient services described as ‘other specialist’ or ‘other’. The extent to which some of these esoteric distinctions between service types is meaningful, particularly to the people placed within them, is a question there to be asked, I think.

While the overall trend is pretty flat, from March 2015 to November 2016 it looks like there are some trends over time for some specific types of inpatient service. The use of acute generic mental illness, complex care/rehabilitation, and other inpatient services seems to be increasing, while the use of forensic rehabilitation and other specialist inpatient services seems to be decreasing. Trends for the most common types of inpatient service over time, secure forensic and acute learning disability, are unclear. 

Another way of looking at the nature of inpatient services is to look at the security level of the places where people with learning disabilities and/or autism are put, as in the graph below. In November 2016, nearly half of people (1,195 people; 47.3%) in inpatient services were in ‘general’ (i.e. not secure) inpatient services, and a small number of people (50 people; 2.0%) were in PICUs (Psychiatric Intensive Care Units, which are secure). Well over another quarter of people (740 people; 29.3%) were in low secure services, 475 people (18.8%) were in medium secure units, and another relatively small number of people (65 people; 2.6%) were in high secure services.

Over the relatively short period of time reported in these statistics, the number of people in low secure services seemed to be decreasing, while the number of people in general (not secure) and medium secure services seemed to be increasing.

The next graph below shows the legal status of people with learning disabilities and/or autism in inpatient services. In November 2016, 250 people (9.9%) were in inpatient services ‘informally’, i.e. not legally detained under Section according to the Mental Health Act (MHA). Around half of people (1,720 people; 50.4%) were detained under Part II of the Mental Health Act (compulsory admission to hospital, usually for assessment and/or treatment). A further 355 people (14.1%) were detained under the more serious Part III of the Mental Health Act (but without restrictions imposed by the Ministry of Justice), and almost a quarter of people (595 people; 23.6%) were detained under Part III of the Mental Health Act with restrictions imposed by the Ministry of Justice. Part III of the Mental Health Act involves ‘patients concerned in criminal proceedings or under sentence’. Finally, a small number of people (50 people; 2.0%) were detained using other parts of the Mental Health Act.

Over time, the number of people ‘informally’ in inpatient services has decreased, while patterns for people legally detained under various sections of the Mental Health Act over time are unclear.

 So far this post has shown the types of inpatient service that people are in, their levels of security, and the legal status of people within them. All of these show a very diverse picture. Are there any clues about whether services that are more secure/forensic are clearly different from services that are less secure?

The Assuring Transformation data analysis offers a couple of hints. The first is in the graph below, which shows for November 2016 the legal status of people within inpatient services at different levels of security. If more secure/forensic services are doing the job claimed for them, we would expect most people within them to be under Part III sections. We would also expect general/low secure inpatient services not to have people under Part III sections (particularly those with Ministry of Justice restrictions), as services at this level of security shouldn’t be able to manage people with these apparent levels of ‘risk’.

The graph below does show that medium and high secure services do largely contain people sectioned under Part II and Part III of the Mental Health Act. Low secure units have a similar profile to more secure units, although the numbers of people in low secure units are greater overall (there are as many people in low secure units under Part III sections with restrictions as there are in medium and high secure units combined). Even general inpatient units, while weighted more towards people there ‘informally’ or with Part II sections, still contain substantial numbers of people with Part III sections, both without restrictions (80 people) and with restrictions (120 people).

This information certainly shows that most people under Part III sections are in general/low secure services (600 people) rather than medium/high secure services (350 people). Is this because there aren’t enough medium/high secure inpatient services and they are desperately needed, or is it because less restrictive options are possible (most importantly, outside inpatient services altogether)?

One final graph below in this graphtastic post – the security level of inpatient services by the total continuous length of time people have spent in inpatient services (including being transferred between them). While this post so far has been about the claims to specialism of secure inpatient services, this final graph speaks to the ‘assessment and treatment’ claims of non-secure ‘acute’ learning disability inpatient services. If that is what these services do, why on earth are there 350 people in general (non-)secure units (29.3% of people in general non-secure units) who have been in inpatient services for 5 or more years? Why are there a further 405 people (33.9% of people) who have been in general non-secure inpatient services for a year or more? Or, to put it another way, why are nearly half of all people who have been in inpatient services for 10 years or more (175 out of 370 people; 47.3%) in general (non-)secure inpatient services, if they represent such a terrible risk?

Overall, it really doesn’t look like the different parts of the learning disability inpatient service complex are highly specialist and set up to deal effectively with different groups of people with different problems. Looks like we need assessment and treatment of the complex inpatient service system itself – why is the challenging behaviour of this system so persistent? A first start would be to ask what functions does this system serve, and for whom? Cui bono?