Winterbourne View 15 Years On - 2026: Update on Report card 1: People being admitted to inpatient units, legal status and ward security

It has been three years since I last updated a series of 'report card' blogposts about what official monthly statistics published by NHS Digital/NHS England/DHSC say about autistic people and people with learning disabilities in inpatient units. The revelation of horrendous abuse at an inpatient unit, Winterbourne View, was 15 years ago now, and a lot of ink has been spilled on various policy 'initiatives' (Transforming Care, Building The Right Support and the rest) which seem to have recently fizzled out in the absence of triumphal success. The purpose of this update is to see if they reveal any tangible evidence of change over time (or at least since 2015, when some of this information began to become available), and if any changes have been sustained as the focus of policy attention has moved elsewhere.

These blogposts will focus on information produced publicly by NHS Digital/NHS England from the Assuring Transformation dataset – for this dataset, health service commissioners provide information (sometimes retrospectively) on the number of people with learning disabilities and autistic people in specialist inpatient services and on various aspects of what is happening with or to people. I’ve collected some of the information into yearly blocks, and some of the information I report for every year rather than every month (both you and I would not survive that amount of information…). This means that most of the information goes up to March 2026. 

As with the previous series, these four blogposts will focus on:

• Who is going into inpatient services?
• How far are people from home, and how long are people spending in inpatient services?
• What planning and review is happening in inpatient services?
• What is happening about people leaving inpatient services, and where are they going?

Using a different dataset, there is also a recent blogpost on what the statistics say about people with learning disabilities and autistic people being subject to restraints in inpatient services.

So – much of the focus of the various iterations of the Transforming Care programme and its rebranded successors has been on getting people out of inpatient units, but the slower than planned reduction in the overall number of people in these units suggests that there are still substantial numbers of people coming into these units. 

Before I talk about what the statistics tell us about admissions, I'll give a brief sketch of what the statistics say about who is in hospital inpatient units. This is in the terms of what statistics are collected, so they give a far from complete or rounded picture of the people involved.

In terms of gender, in March 2026 30% of people with learning disabilities or autistic people in inpatient units were female (including trans women), 47% were male (including trans men) and 1% were non-binary - information on gender was not coded for 21% of people. For 3% of people (60 people), their gender identity is not the same as the gender assigned them at birth - how well trans people are being supported in inpatient units is unknown. The coding of gender in March 2016 was more restrictive, and reported that 24% of people with learning disabilities and autistic people were female and 76% were male.

In terms of age, in March 2026 10% of people with learning disabilities or autistic people in inpatient units (230 people) were children and young people under the age of 18, a percentage that has hovered around this level since 2017. Almost a quarter of people in inpatient units in March 2026 were young adults aged 18-24 (22%; 435 people), with 30% of people (665 people) aged 25-34 and 20% of people (445 people) aged 45-54. Fewer people (17% in March 2026; 385 people) were aged 45 or above, a decrease from 25% of people in March 2016.

Recorded ethnicity has been fairly consistent over time, although the ethnicity of around 10% of people is consistently recorded as not known. In March 2026 the ethnicity of 75% of people with learning disabilities and autistic people in inpatient units was recorded as 'White', 6% of people were recorded as 'Black or Black British', 4% of people were recorded as 'Asian or Asian British', 3% of people were recorded as 'Mixed', and 1% of people were recorded as 'Other ethnic group'.

Throughout the collection of this data, people in inpatient units have been categorised into three very broad groups: Learning disability (only); Autism (only); and Learning disability and autism. The percentage recorded across these categories has changed hugely in a decade. In March 2026, 26% of people were recorded in the Learning disability (only) category, down from 54% of people in March 2016. In March 2026, 58% of people were recorded in the Autism (only) category, up from 22% of people in March 2016. Finally, in March 2026 16% were recorded in the Learning disability and autism category, down from 23% of people in March 2016. It might be worth mentioning that 11% of people in inpatient units acquired a 'diagnosis' of autism AFTER they had been admitted to the inpatient unit.

Admissions

Every month, the Assuring Transformation statistics report how many people have come into an inpatient unit (called ‘admissions’) according to commissioners. The graph below adds these together across ten different years (each one October to September), so we can see the number of people coming into these units and whether they have changed over time.

 

The first thing the graph shows us is, pre-COVID, the overall number of admissions to inpatient services was increasing, from 1,810 admissions in 2015/16 to 2,250 admissions in 2018/19. The total number of admissions dropped to 1,745 admissions in Oct 2019 - Sept 2020, when COVID-19 hit, increased in 2020/21 to 1,830 admissions, then dropped again in 2021/22 to 1,530 admissions. From then the number of admissions has risen every year, reaching 2,240 admissions in 2024/25, which is pretty much as high as it's ever been in this dataset. In total, there have been 19,250 admissions to these inpatient units in the last ten years - we don't know how many of these are the same people being admitted more than once or different people being admitted.

The second thing to notice is that a large proportion of admissions (the red chunk) are actually transfers from other hospitals (mainly other inpatient services of various types), running consistently at almost 20% of all admissions. The third thing I want to mention is that another 20% of admissions (the lilac chunk) are re-admissions, where people had previously been in an inpatient service less than a year before. Finally, the purple chunk shows that getting on for two thirds of admissions to inpatient services (62% in 2024/25) are people who have not been in an inpatient unit for at least a year (or maybe never).

What kinds of places are people being admitted to inpatient services coming from? The graph below shows this information from the Assuring Transformation statistics, for nine one-year blocks (starting in October 2016, through to September 2025). In 2024/25, almost half of people (47%) were admitted from their ‘usual place of residence’ and almost all the rest (49%) were admitted from other inpatient and/or hospital services (more or less equally split between mental health/learning disability inpatient wards and general or A&E wards in general hospitals). 

It seems clear to me that policy wheezes designed to reduce the chances of people being admitted to inpatient services are not working within a general context of ongoing austerity and fragile community mental health and learning disability support, showing up in increased admissions, including admissions from general hospitals, and 20% of admissions being people who had been in a similar inpatient unit less than a year before. In March 2026, just 22% of people in inpatient units were on a 'dynamic support/risk register' designed to identify and support people at potential risk of going into an inpatient unit.

 

Legal status and ward security

Have any changes in admissions resulted in changes to the legal status or the ward security of people with learning disabilities and autistic people in inpatient units? A persistent argument has been that people on MHA Part III sections (particularly those with legal restrictions) and people in higher security places are more likely to need to remain in inpatient services, so the number of people in these categories should not change much even if the number of people in inpatient units reduces overall. 

The graph below shows the number of people in inpatient units at the end of March each year from 2016 to 2026, broken down by the legal status of people in inpatient units according to Assuring Transformation data.

The most common legal status for people in inpatient units is a Part II section, where the number of people dropped from March 2016 to March 2021, but has since increased again to 1,265 people in March 2026. This has almost returned to March 2016 levels, with an overall decrease of only 5% from 2016 to 2026. Overall, people on Part II sections in March 2026 now constitute well over half (57%) of people with learning disabilities and autistic people in inpatient units, compared to less than half (47%) of people just ten years earlier in 2016.

The next most common legal status is a Part III section with restrictions, where the number of people has fairly consistently dropped from March 2016 to 535 people in March 2026, a reduction of 26% over the time period. 

Less common are people with Part III sections without restrictions (which dropped by 49% from 2016 to 205 people in March 2026), people not subject to the Mental Health Act (which dropped by 38% from 2016 to 200 people in March 2026), and people in other sections (which dropped by 75% from 2016 to 10 people in March 2026). It is clear that reductions in the number of people in inpatient units have reduced for people with almost all types of legal status from 2016 to 2026.

The picture is similar when looking at ward security, as the graph below shows. The most common level of ward security is 'general', where the number of people in 'general' ward security dropped by only 3% from March 2016 to March 2026. In March 2026, over half of people with learning disabilities and autistic people in inpatient units are in units with a 'general' level of ward security (1,295 people; 58%), compared to less than half of people (47%) ten years before, in March 2026.

The next most common level of ward security is 'low secure', where the number of people dropped by 49% from 2016 to 430 people in March 2026. 

The number of people in medium secure inpatient units dropped by 36% from 2016 to 335 people in March 2026, with smaller numbers of people recorded in high secure units (75 people) and Psychiatric Intensive Care Units (95 people - an increase of 90% from 2016).

Clearly, reducing the number of people in inpatient units has not excluded people in units with higher levels of security.

Looking at these figures over a decade now, it looks like the overall slight reduction in the number of people with learning disabilities and autistic people in inpatient units stalled during the Covid-19 pandemic and has stayed pretty static ever since. Meanwhile, apart from a dip in admissions during Covid, the number of admissions to inpatient units continue to rise every year. People are being shunted around inpatient units, going through what looks like a revolving door in and out of inpatient units, and aren't getting the support they need where they live so that an inpatient unit wouldn't be seen as an option. Policy wheezes like C(E)TRs before admission and dynamic risk registers are clearly ineffectual in the face of 15 years of austerity and counting, the consequences of the Covid-19 pandemic, and a lack of attention to and investment in both social care and community mental health support that people with learning disabilities and autistic people can make use of.

In the next blogpost I'll look at a couple of consequences of this - how far people from home are people being sent, and how long people are staying in these places.