What are the clinical and educational risk factors for poor academic outcomes in children with Autistic Spectrum Disorder?

Approximately 1% of UK school age children have an Autistic Spectrum Disorder (ASD). Children with ASD typically have longstanding difficulties using language, engaging in every day social interactions and having a general, flexible interest in the world around them. ASD is commonly diagnosed in the first years of primary school. The majority of children who receive an ASD diagnosis will go on to have difficulties throughout their school life, often leaving with no qualifications. This puts them at a significant disadvantage as adults and at greater risk for a host of mental and physical health problems. An important objective for UK child mental health specialists, teachers and researchers is to reduce these educational failings for children with ASD whilst at school, to ensure they have better skills and opportunities in later life.

In the UK, specialist Child and Adolescent Mental Health Services (CAMHS) commonly provide ASD assessments and offer treatment. CAMHS services also treat common childhood mental health conditions including anxiety, depression, conduct and hyperactivity disorders. Children without ASD who suffer these disorders often have considerable difficulties achieving their potential at school. Fortunately there are a number of effective mental health treatments including medications that can help non ASD children get back on track. These treatments increase their chances of attending school, not getting into trouble and passing exams.

Children with ASD also suffer from these disorders and at much greater rates (up to 70% in some studies) than typically developing children. ASD combined with other mental health disorders presents as a much more complex condition to manage and may significantly impact educational performance. Currently there is little evidence available to inform clinicians on how best to treat the common co-occurring conditions associated with ASD. Pharmacological treatments, developed primarly in non ASD children, may not have the same effect in improving educational performance for ASD children.

Pharmacological treatments (for example stimulants, anti-depressants and anti-psychotics) are being increasingly used by CAMHS to help manage common co-occurring disorders in children with ASD. At present, little is known which disorders or additional symptoms lead clinicians to prescribe these medications to children with ASD. Also, very few experimental trials are conducted in children with ASD that are adequately designed to measure the long term effect of treatment, therefore their influence on longer term educational outcomes like overall school attendance, exclusion rates or final exam results remains unclear.

With robust confidentiality and data protection measures in place, electronic health records(EHRs) can provide rich data to investigate the gaps in current knowledge. They can be used to determine which childhood factors lead to impaired educational performance in ASD, and how they may be modified by treatment. For this proposal, I will use EHRs via the Clinical Record Interactive Search system(CRIS) and link the records with the National Pupil Database(NPD). CRIS was developed by King College London to allow researchers to retrieve in depth information from the anonymised case notes of over 34,000 CAMHS patients. The NPD is a database managed by the Department for Education and contains all children academic records since enrolment in state school from 2002. As the fellowship applicant, under the supervision of Professor Hotopf and Dr Hayes at Kings College London, and Professor Ford at the University of Exeter, I will study whether additional mental health difficulties impact on educational performance and whether treatments that may reduce these difficulties could improve academic outcomes.

I will establish cross-sectional and longitudinal individual clinical profiles (e.g. socio-demographic, co-occurring disorders, initial symptom severity, impairment, home, neighbourhood and service use factors) for a large ASD child cohort and model clinical predictors of pharmacotherapy use (drug type, class switches, polypharmacy).

Text mining algorithms and structured fields in CRIS will provide data on a range of baseline clinical factors and pharmacological treatment. I will then explore associations between clinical profiles and academic outcomes in children with ASD, which will include academic attainment, school drop-out, attendance and exclusion rates. I will estimate the risk of these outcomes according to clinical profiles, including co-occurring disorder.

CRIS will be linked with the National Pupil Database, a longitudinal dataset containing educational and census information since 2002, for all children who have been enrolled in state education. This will provide individual level longitudinal education performance data. I will then model the clinical predictors, treatment exposures (drug type, duration of treatment), treatment response exposures ( effect sizes derived from validated psychometric scales and normalised against outcome data provided by CAMHS Outcome Research Consortium) against multiple education outcomes. I will estimate the risk of specific education outcomes according to ASD clinical profiles and co-occurring disorders and whether these risks are modified by medication use and treatment response.

Key stakeholders in research findings from this fellowship include children and their families, paediatric and mental health clinicians, educational and social providers and the general public. The proposal explores scientific questions relating to clinical characteristics, mental health treatments and educational outcomes for children with ASD. These questions have been substantially shaped by evidence on stakeholder preferences for future research in ASD. Children and families are concerned with real-life outcomes of ASD, and one of the strengths of this proposal is that the research value, design, and implementation of the project has ongoing involvement from an Oversight Committee made up of patients, CAMHS clinicians and researchers. In this fellowship, I will draw on electronic secondary mental health care records linked to educational databases. I will examine the diverse clinical profiles of children with ASD, the treatment they are given and their later educational outcomes.

The findings produced from the fellowship will be included in the IOP BRC Stakeholder Participation Theme which is an ongoing programme set up to publicize our research to staff, children and families who user services in SLAM and neighbouring Trusts. My main findings will be posted on the SLAM Biomedical Research Centre -Mental Health website which is available to and accessible by the public. I will present results through public events hosted by the Institute of Psychiatry at King's College London and other institutions including the University of Exeter ( Exeter's Café Scientifique), UCL Partners, the Royal College of Psychiatry and British Educational Research Association. In addition, the Department of Psychological Medicine, where I will be based, have an established relationship with Science Media Centre which feeds results of scientific research to the UK public via news broadcasting and papers

The proposed fellowship provides a potentially valuable case study demonstrating local and national capacity to provide 'real world' child pharmaceutical data. At present other sources of this data are limited, possibly due to the methodological challenges of pharmaceutical research in children. This information would allow clinicians and families to make more informed decisions based on a clear understanding on prescribing practices and outcomes. Potential dissemination targets in this respect will include policy makers within NIHR for their specific expertise on relevance in this area. Dissemination targets may also include pharmaceutical industry itself, facilitated as with other CRIS work by KCL Business Ltd.

The proposed research is focused on understanding the outcome and treatment processes of children with ASD receiving clinical care, which can directly inform healthcare policy and treatment guidelines. Timescales for this type of impact are likely to be relatively rapid, ideally within 5-10 years of the output. Result summaries will be fed back to relevant organizations such as NICE, and promoted within locally with the aim of directly impacting NHS policy and current patient care. This would be assisted through support of local collaborators with expertise in psychopharmacological research dissemination at KCL. This is a particularly strong domain for KCL clinician scientists, especially through the publication of its influential Maudsley prescribing guidelines. These efforts are carried out in conjunction with the IoP media support unit and the BRC-MH Stakeholder Participation

The Research division within the Department for Education (DfE) have been highly supportive of this proposal and have expressed interest in incorporating findings into their research reports and policy papers. As the fellowship progresses, I plan to become further involved with DfE policy department to discuss dissemination and potential commission of further projects.