The future I’d like to see is one in which information about autism and about autistic people is consistent, based on agreed definitions and shared across services and sectors. This is of course an enormous challenge but the benefits for autistic people and society set out in the National Autism Project report “The Autism Dividend: Reaping the Rewards of Better Investment” cannot be achieved without progress on this and I welcome the recent moves by government, royal colleges and autism organisations to take forward work in this area after many years of lobbying.
When I first became involved in biomedical autism research in 2005 it was at a time when new technologies such as DNA analysis and brain scanning seemed likely to cast light on the mysterious and complex condition known as autism, and the possibility of biologically-based diagnosis seemed credible and new forms of intervention based on biology feasible. This was viewed with horror by many autistic people even if welcomed by families scared by stories about autism induced by environmental factors ranging from vaccines to dog shampoo and open to every quack remedy on the planet.
What emerged from the new forms of research was of course something much more complex and nuanced. Autism arises from many small biological changes, none of them unique to autism, and with an apparent association between the number of ‘risk factors’ and the autistic characteristics of each individual. What was known from twin studies, that autism is strongly genetic, has been confirmed but clearly there are no simple answers. There is much that basic research still needs to tell us, including why autism is so strongly associated with late onset epilepsy and learning disability, but the notion of autism as a simple genetic disorder, except in some rare conditions such as Rett’s Syndrome, is history.
So if we cannot yet say what autism is, then we must work to define its impact both at an individual and societal level. Autism may simply be part of the normal distribution of human diversity but there is no doubt that autistic people and their families face challenges that others do not. So to assess this impact equitably we must use common assessment tools and common terminology and apply these in consistent ways. There are international diagnostic criteria and well-researched diagnostic tools for children but access to these is patchy and a diagnosis, even where made, is not consistently recorded. Some years ago I was part of a steering group at the then Department for Children, Schools and Families overseeing analysis of data on children identified with special educational needs arising from two exemplar conditions, deafness and autism. What this showed was that attitudes and not data determined what help autistic children received. Some education authorities simply ignored autism, often justifying this as preferring “not to give children labels”, whilst others were scrupulous in linking diagnosis to assessment of need. This inconsistency helps no-one as good practice cannot be spread.
The picture for adults is worse. In theory the Autism Act gives adults a right to an assessment of need with or without a diagnosis, but in practice an assessment of need is hard to achieve in either circumstance and self-assessment returns from local authorities suggest that only around a sixth of adults assessed as needing support actually receive it. There are no common diagnostic tools for adults or any clear pathway to the recording of adult diagnosis in medical records. Hopefully this last deficit is now being addressed although it should be in the context of an autism-specific data set rather than the mental health one proposed since autism is not a mental health condition.
One of the most disappointing findings of the NAP report is the poor quality of much of the research assessed by Prof Knapp and his team. This is not just a function of the challenges of definition described above but of a whole host of weaknesses: small samples; lack of controls; lack of basic rigour and objectivity; the absence even of a plausible hypothesis; a presumption of difference as deficit particularly in research on children; and poor use of proper statistical analysis. Many researchers work only from a theoretical model of autism and not from knowledge of real people. The nadir of this is of course the infamous work of Andrew Wakefield, who used a series of just 12 carefully chosen case studies to convince a generation of parents that the measles vaccine was linked to autism, and there are many other examples of real harm done to autistic people as a result of poor research. Very little research is fully ‘co-produced’ involving autistic people at all stages from design to translation.
The absence of good data has meant that autistic people and their families have had to create their own research databases in order to address the research weaknesses described above, through for example the IAN project in the United States and similar work by the University of Newcastle in the UK which is now being supported by Autistica’s Discover research network. The latter aims to create a database of 20,000 potential research participants by 2020. It has taken the private money of billionaire Jim Simons in the US to create an autism biobank. Can you imagine if this were the case for other conditions with large societal impact? We do not even know how many autistic people there are with current estimates of 1.1% of the UK population being based on proxy data. The Department for Education believes that it knows the identity of many but by no means all of the children with an autism diagnosis.
There is no doubt in my mind that poor information lies at the heart of the disturbing picture revealed in the NAP report, in which autistic people cannot expect to receive well researched and evidence supported services. Good information systems can only be built through the will of government, institutions and researchers to do better and to work together to that end. It may serve an austerity-minded government to continue the situation in which each autistic individual has to fight for support, but it cannot be in the long-term interests of society as a whole. Work by the National Audit Office showed that increasing employment in the autistic population by only a relatively small amount produced a significant financial benefit and there will be many other similar examples across the service spectrum.
The irony is that the direct cost of creating reliable information is relatively small as systems already exist that once populated with accurate data could be used to model service design and underpin delivery. There are even large banks of data from NHS cohort studies over many years that could be interrogated using data mining techniques to look back as well as forward. A further irony of the situation in the UK is that our national health service creates a rich data environment for health research from which the autistic population is currently largely excluded. The sharing of data across service providers is of course notoriously poor across all client groups but capturing essential basic data is an important starting point. So let us hope that the Department of Health group working on this makes rapid progress and take a tough stance in demanding that public services follow through. Politely asking GP’s to record autism diagnoses in their records, as NICE has done, is not the same as mandating their use. Of course autistic people are entitled to the same data safeguards as everyone else, and seeking or revealing formal diagnosis is a matter of choice for autistic individuals, but confidence in sharing information will come from seeing the value it can provide.
It is no accident that the first research recommendation in the NAP report calls for the development of age- and gender-specific diagnostic tools and the last for the collection of data to build better information systems to support autistic people. These are the fundamentals – the alpha and omega – of a better future for autistic people and it’s one that I so much want to see.
Consultant – National Autism Project