- New NHS survey out today estimates that autistic people make up 0.9% of the population – only a minor increase in same survey published over a decade ago (0.7% in 2014) [i]
- Autism charity hits out at the narrow research methods used in the Adult Psychiatric Morbidity Survey published today, which fails to properly assess the participants using official guidance for autism diagnosis
- Autism Action warn Government against using data to downplay the number of autistic people or cast doubt on existing diagnoses
The true number of autistic people is wholly underestimated in an official survey out today, says charity Autism Action, who warn against the Government using it as a ‘staging post’[ii] for a new autism strategy due next year.
The 2023/24 Adult Psychiatric Morbidity Survey (APMS) published today (27 Nov) provides statistics about trends over time for a number of conditions and includes a claim about the prevalence of autism in the general population.
This claim is undermined by weaknesses in the survey. The survey’s researchers primarily used the ADOS (Autism Diagnostic Observation Schedule) to assess for autism and collect the data for the APMS. Relying on this tool alone goes against the thorough diagnostic practices recommended by the National Institute for Health and Care Excellence (NICE)[iii]. The report itself states that there was little correspondence between “professionally diagnosed autism” and the ADOS method used by the survey’s researchers.
Better awareness and changes to diagnostic criteria have contributed to an increase in demand for autism assessment in recent years, with over 225,000 people currently on waiting lists. Government investment and service delivery have failed to keep up, meaning that thousands wait years for a diagnosis and receive little to no support even when it comes. In addition, research demonstrates that there are problems with diagnosing autism in women and girls as well as underdiagnosis in the over 50s.
Autism Action says that the results of the APMS published today don’t reflect the full picture, and that there could be fatal consequences and misdirected resources for the autism community if they are seen in isolation.
Tom Purser, CEO of Autism Action, said:
“This new survey is not quite what it makes out to be; the diagnostic methods and sample sizes are flawed and it does not represent an accurate picture of how many autistic people there are in our communities. The report itself even says that ”great caution” should be taken when interpreting their results due to the sample size, which makes this very shaky ground. The Government must avoid using this information to conveniently downplay the number of autistic people in the country. Other NHS data shows that 1.6% of people registered with GPs already have an autism diagnosis[iv], vast numbers of over 50s remain undiagnosed [v], and women and girls still face barriers to diagnosis [vi]. If the Government leans on today’s new data for their understanding, they risk underpowering policies and misdirecting resource; which could lead to increased stigma and fatal consequences for the autistic community.
“The real struggles in people’s lives are bigger than this data and the stakes are too high – we know that autistic people are five times more likely to die by suicide, and in our own research the community tells us that NHS support services are too difficult to navigate, with widespread fear and distrust of healthcare professionals.
“This week’s House of Lords Committee report ‘Time to Deliver’ has now officially highlighted the failures of the Autism Act 2009, in turn recommending a fundamental overhaul of services, including addressing the long waiting lists for assessment and inaccessible mental health and social care services.[vii]”
The consequences of doubting or rejecting somebody’s autism diagnosis can be catastrophic and it is an issue happening across the country. An inquest in Essex last month about the death of an autistic woman called Abbigail (Abbi) Smith, 26, gave a verdict of suicide – citing failure to provide basic care and neglect during discharge from a mental health inpatient unit. Evidence heard at the inquest suggested that services consistently failed to recognise and accommodate her autism, which was diagnosed in her childhood. Coroner Sonia Hayes stated a clear absence of understanding to care for, treat and communicate with Abbi as a neurodivergent, autistic person. During the delivery of the verdict, the coroner said that “Abbi’s autism was left as a footnote in her care”.
For more information, interview requests with Tom, or for background briefings, please email media@autismaction.org.uk
Notes
See 10.2 survey assessment methods, primarily using the ADOS to diagnose autism and 10.3 results section citing little correspondence with ADOS-identified autism and professionally diagnosed autism.
[ii] In his evidence to the House of Lords Autism Act Committee, Minister Kinnock said:
“We have launched an important survey, which will report back in the autumn; that will be an important first staging post towards developing the strategy in 2026. There is also a real need to get a better understanding of the drivers of this increase in demand, because that is such an important part of the strategy. By the autumn, that ongoing work will be in a much better place because we will have the results of the survey that I mentioned.”
It was later clarified that the Minister was referring to the existing APMS survey.
[iii] NHS England guidelines for autism diagnosis
“Consistent with NICE guidelines (31,32) assessment should include, at a minimum, a clinical interview, behavioural observation, integration of developmental and corroborative information, use of validated assessment tools, a broader assessment of clinical presentation (such as estimated intellectual functioning or sensory processing) and additional liaison with referrers and other involved service(s). Siblings, partners or friends may provide corroborative information.”
“The autism assessment must include a clinical interview with the person, conducted by a clinician with a medical background or a qualified mental health professional (for example, clinical psychologist or mental health nurse)(31,32). This is distinct from an
assessment with family/carers or conversations with siblings (for example, developmental history taking or asking for descriptions of current concerns). This is also distinct from semi-structured behavioural observation assessments that specifically focus on traits associated with autism (for example, the Autism Diagnostic Observation Schedule – 2; ADOS-2 (35)). The clinical interview is pivotal for putting into context scores obtained on standardised questionnaires that may be used for screening or triage, or scores on semi-structured assessment tools.”
[iv] Autism diagnostic rates – NHS England Digital
[v] Around 90% of middle-aged and older autistic adults are undiagnosed in the UK, new review finds | King’s College London
[vi] Challenges with diagnosing women and girls are highlighted in the APMS, 10.4: “Autism assessments, including ADOS, focus on observable features. This may insufficiently consider less clearly manifested autistic traits (Cook et al. 2024), which may be more likely among women than men (Hull et al. 2020). Additionally, assessments for autism may draw more on how the condition manifests in men, leading to further under identification of autism in women (Trubanova et al. 2014). For example, Rynkiewicz et al. (2016) report that autistic females present with more vivid and energetic gestures on ADOS-2 examination and thus may be at risk of underdiagnosis.”
