Stratifying Genomic Causes of Intellectual Disability by Mental Health Outcomes in Childhood and Adolescence (IMAGINE-2)

The IMAGINE-2 study is a medical research study funded by the Medical Research Council (2020-2024) that aims to investigate the impact of genetic disorders that are associated with learning difficulties on children and young people’s mental health. It is a collaboration between University College London (UCL) and Cardiff University. Cardiff University will not have access to or process the NHS England data to be provided for this UCL data request. Cardiff University do not determine the purpose or the means of the data processing for the IMAGINE-2 study and are not therefore considered to be a data controller. The University of Cardiff Investigator has had no input on determining the purpose and means of workstream 1.

The UCL study team resides at the UCL Institute of Child Health department which is a joint research office between UCL and Great Ormond Street Hospital (GOSH). As such, both UCL and GOSH logos are used in the materials for this study however GOSH does not play any further role in the study and do not determine any purposes of this study in any capacity.

The IMAGINE-2 study is a follow-up project of the previous one called Intellectual Disability and Mental Health: Assessing Genomic Impact on Neurodevelopment (IMAGINE-ID) which was funded by the Medical Research Council (MRC) and Medical Research Foundation (2015-2020). A collaborator from the University of Cambridge who was involved in the IMAGINE-ID study is involved in IMAGINE-2 as a consultant only who may provide advice to the IMAGINE-2 project. The University of Cambridge will not have access to any newly collected data from the research programme and do not determine the purpose or the means of the data processing for IMAGINE-2. The MRC as funders of IMAGINE-2 do not determine the purpose or the means of the data processing and will not process any of the study data. These organisations are not therefore considered to be data controllers or data processors.

IMAGINE-2 is divided into two workstreams. Workstream 1 aims to map trajectories of developmental risk for individuals with different types of Intellectual Disability (ID). Workstream 1 is led by University College London and requires NHS England data. Workstream 2 is led by Cardiff University and involves a face-to-face follow-up study of young people seen during IMAGINE-ID who have been identified as carrying a genetic variation which is high-risk for mental health problems. Cardiff University will independently collect data from participants in IMAGINE-2 by direct contact with the identified high-risk subset of families whom they will visit at home. Cardiff University will not have access to the NHS England data.

UCL is the sole data controller who also processes NHS England data for the IMAGINE-2 study. NHS England data will be handled exclusively by UCL in the Data Safe Haven (DSH) to which only staff who are associated to UCL will have access. Access to UCL DSH will only be given to individuals who have had the appropriate UCL non-disclosure training and have signed a Non-Disclosure Agreement. No data will be exported outside the DSH.

UCL is a ‘public authority’, as defined in the Data Protection Act 2018, with a principal object of the organisation being research and its dissemination. The processing of identifiable personal data, including special category data, is necessary to carry out medical research that serves the public interest. As such, the legal bases for processing personal data are:

Article 6(1)(e) of the GDPR, ‘processing is necessary for the performance of a task carried out in the public interest’; and

Article 9(2)(j) of the GDPR ‘processing is necessary for archiving purposes in the public interest, scientific or historical research purposes’.

Section 8 of the Data Protection Act 2018 clarifies that “In Article 6(1) of the GDPR (lawfulness of processing), the reference in point (e) to processing of personal data that is necessary for the performance of a task carried out in the public interest or in the exercise of the controller’s official authority includes processing of personal data that is necessary for… (d) the exercise of a function of the Crown, a Minister of the Crown or a government department”. University College London has an established royal charter. It includes the following statement “The objects of the College shall be to provide education and courses of study in the fields of Arts, Laws, Pure Sciences, Medicine and Medical Sciences, Social Sciences and Applied Sciences and in such other fields of learning as may from time to time be decided upon by the College and to encourage research in the said branches of knowledge and learning and to organise, encourage and stimulate postgraduate study in such branches.

As a higher education establishment, the University conduct research to improve health care and service and the linkage requested is necessary for the performance of a task carried out in the public interest; i.e. improving the health outcomes of children with genetic disorders.

The IMAGINE-2 cohort consists of children and young people who were born between 1989 and 2016, and have intellectual disability (ID) or developmental delay caused in whole or in part by a known genetic variant CNV (copy number variant); or SNV (single gene variant). The study aims to delineate the course and outcomes in CNV-associated intellectual disability (ID) and single gene disorders to provide information at the point of diagnosis and onwards for families, clinicians and service providers, as well as to pave the way to greater biological understanding and the personalisation of interventions. Several recurrent ID-associated CNVs and single gene disorders have been associated with poor mental health outcomes., However, there is considerable pleiotropy (i.e. variations in a single gene may affect multiple (possibly unrelated) observable characteristics of an individual), and also incomplete penetrance for specific psychiatric diagnoses (i.e. some individuals express the associated symptom or trait while others do not, even though they carry the disease-causing gene).

The cohort includes children with genetic disorders that put them at risk of autism, attention deficit hyperactivity disorder, anxiety and psychosis among other conditions. They are also at risk for non-psychiatric disorders including sensory impairments and epilepsy. No study to date has deployed systematic sampling and assessment to determine why some, but not all, ID-related CNVs and single gene disorders are associated with poor mental health outcomes, nor have they identified risk and resilience factors modifying outcomes across this population. Assessing the relative contributions of CNV and/or SNV genetic constitution, ID severity, cognitive profile, social/ environmental risk factors, and physical comorbidities, will highlight major determinants of adjustment. Better care could be provided if those individuals at greatest risk were identified early and if preventive intervention was timely and focused on salient biological and/or social processes. Early identification has the potential to reduce the costs of long-term care, better target key services/interventions, and improve quality of life over the life-course.

Background:

Participants were eligible for the IMAGINE-ID study if they were 4 years of age or over at the point of recruitment between 2015-2019, and if they possessed a genetic variant, identified by an NHS Regional Genetic Centre, that was reported to be causing intellectual disability/ developmental delay. The vast majority of participants in IMAGINE-ID were identified as being eligible by one of 25 UK Regional Genetics Centres (RGC). The original genetic testing was ordered on the basis of unexplained learning disabilities. Eligible families were invited to participate by the paediatric team linked to the RGC. The study was advertised to patient groups through social media and at parent-supported events. Once participants had been recruited, they were invited to complete online assessments of their child’s mental health, behaviour and well-being for the Workstream 1 data collection. 3402 participants were recruited in total. 500 families, a subset of the total sample, have been seen face-to-face for more detailed evaluations by Cardiff University collaborators for Workstream 2 of the study. The new MRC grant (2020-2024) provides funds for further study and the title has been changed from “Intellectual disability and mental health: Assessing genomic impact on neurodevelopment (IMAGINE ID)” to “Stratifying Genomic Causes of Intellectual Disability by Mental Health Outcomes in Childhood and Adolescence (IMAGINE-2)” which will follow up the already-recruited cohort for 54 months, commencing 1st April 2020.

The study has established a patient, parents and carers consultation group. This group has been consulted from the inception of the study and is regularly updated. The group was established to provide feedback, comments and suggestions that have influenced the design and progress of the project.

The Workstream 1 data collection undertaken during IMAGINE-ID provided details of the children’s development, well-being, mental health and adaptive functioning. A brief account of the children’s medical history was obtained. The Adaptive Behaviour Assessment System (ABAS-3) was used to estimate the degree of developmental delay in key domains of adaptive functioning (e.g. language, self-care, motor skills). Children’s mental health was assessed by the Development and Well-Being Assessment (DAWBA), which has been employed in three national UK studies over the period January 1999-December 2017. The DAWBA is a detailed semi-structured interview and covers many areas of development, behaviour and well-being. Rates of mental health disorder and behavioural/ emotional dysfunction in the IMAGINE-ID cohort can therefore be directly compared with a national representative sample of typically developing children and young people, the dataset of the Mental Health Children and Young People (MHCYP) from the UK Data Service. Significant general health problems are described in many cohort children with mental health disorders.

Mental Health of Child and Young People (MHCYP) data request:

MHCYP data is requested for use as the control data to compare with IMAGINE-ID cohort data collected from the assessments of mental health, behaviour and wellbeing (i.e. Workstream 1 data collection). The Mental Health of Children and Young People (MHCYP) survey provides record-level, pseudonymised data on the prevalence of mental disorders in children and young people (aged 2-19 years old) living in England. This dataset contains the same measures as the IMAGINE-ID research data and covers a similar age of children and young adults. No identifiable data is requested from the MHCYP dataset.

HES data request:

The study is applying for access to Hospital Episode Statistics (HES) data to assess the broader healthcare needs of this group as a function of their genetic disorder. Pseudonymised data is requested from NHS England but will be linked to existing study data thus making the data technically identifiable. Access to these data will permit a more detailed view of the strengths and difficulties of participants, their use of services, and comorbidities. In the IMAGINE-2 study (2020-2024), a longitudinal perspective will be taken, throughout childhood and adolescence, to ascertain disease trajectory and outcomes relating to social inclusion, education and their health needs. Better characterisation of the trajectories of health risks and wider developmental impact of these diverse ultra-rare genetic disorders will inform and improve future healthcare and management.

The request is to access Hospital Episode Statistics (A&E visits, Critical Care Episodes, Admitted Patient Care Episodes and Outpatients appointments) and Emergency Care Data Set for the IMAGINE-2 cohort and a control cohort. By linking HES and ECDS data to already collected data on the IMAGINE-2 cohort mental health and family circumstances, the study aims to build a detailed picture of the more significant medical healthcare needs of this group of children and young people. The HES control cohort requested by UCL will be matched on age, sex and index of multiple deprivation.

Through analysis of the number of hospital visits, number of outpatient appointments, and length of stays in hospital, the study will examine to what extent the population of children and young people with intellectual disability or developmental delay caused by a known genetic variant relies on the NHS healthcare system more than typically developing children. The study will investigate the costs involved in caring for these children, including the costs to the children themselves (for example missing school because they have hospital appointments or are admitted to hospital). The study will examine potential links between specific ultra-rare genetic disorders and the need for specialist intervention in particular healthcare domains. For example, one third of the cohort has had seizures; in some cases, the seizures were associated with genetic anomalies that have never previously been studied in detail because of their rarity. It would not be possible in any other way to analyse cohort data at this scale, which has implications for improving future medical management in this vulnerable population. Preliminary data, from parental reports, indicates a high rate of frequent users, reflecting the complex nature and needs of many of these disorders.

Separately, UCL also require access to a standard extract of Mental Health of Children and Young People (MHCYP) 2017 and 2020 survey data. This data is not linked (nor does it have the capability to do so) to either UCL’s cohort nor the control cohort. This data will be used for comparison purposes to give UCL a snapshot of statistics into several categories (mental health, behaviour and well-being).

UCL hypothesise that developmental trajectories of children and young people in the IMAGINE-2 cohort with genetic differences will differ to those of individuals without genetic disorders, and that this will be best captured by clusters of traits indicating their mental states, mental illnesses or disorders and impairment of their cognitive development. UCL also hypothesise that the trajectories of these clusters of traits will be impacted by genetic factors and related risk factors such as socioeconomic adversity and family environment. A control cohort matched by age, sex and geographic region will indicate the differences in impact according to genetic factors and index of multiple deprivation for example.

Data linkage:

The study aims to link HES data to the existing research data from Workstream 1 in order to obtain a detailed picture of participants’ use of NHS secondary care services. The data held from Workstream 1 are genetic, medical history and mental health data of which is coded and is in non-identifiable form. The genetic data (from Regional Genetic Centre (RGC) laboratory reports) and observable individual characteristic data (through online psychiatric assessments) will be linked to medical history data in order to build a highly detailed picture of the cohort as it develops over a period of 5 years since participating families were originally recruited and interviewed. The study will compare service usage in this cohort of children to service usage by children and young people in England.

Using the diagnosis categories for each episode, the study will be able to assess the health problems that are associated with each genetic disorder and which are common across genetic disorders. The study will ascertain if there are health problems in common within and across genetic disorders which are not yet well-known or described in the literature and will contribute to the existing body of knowledge.

Using HES data about length of episodes and specialty involved, the study aims to further develop analyses of socioeconomic factors involved in these genetic disorders. This will have two outcomes; to provide information about the level of contact with health services parents might expect if their child is diagnosed with a genetic disorder, and to provide an assessment of the cost to the NHS of caring for this group of children (using NHS Reference Costs), which will be of use in decisions relating to commissioning services. Prevalence of neurodevelopmental disorders is increasing as more children survive due to better care, and with the constant development in genetic sequencing technology, more and more children in the future will have a genetic cause of their developmental delay or intellectual disability identified. Early identification and better understanding of the disease trajectory of these conditions has the potential to reduce the costs of long-term care, better target key services and interventions, and improve quality of life over the life-course.

UCL will conduct network analyses and machine learning methods to look for commonalities which could indicate the mechanism by which specific genetic disorders are associated with medical/psychiatric disorder expression and provide paths of investigation for therapies. In regards to machine learning, this is a way of statistical analysis which may be used in this study to analyse the research and HES data. Any machine learning analysis will not involve any personal nor identifiable data. Researchers will use various statistical methods in conducting these analyses, such as regression, classification and machine learning methods, time series analysis, statistical inference and natural language processing, according to the approach that is most appropriate for the data.

In order to undertake the analyses as described, the study is applying to access HES A&E, Outpatients, Critical Care and Admitted Patient Care data and ECDS for each of the consented participants, covering as much of their lifespan as is available for each dataset (a range from 1994/95 to present) in order to gain a detailed and accurate picture of their medical history and use of services throughout their lifetime. As the data relates to individuals, the geographical spread of the data requested represents the current geographic spread of participants in England.

The study is requesting the medical history of each participant. There are no alternative or less intrusive ways of obtaining these data. Whilst the study has obtained, for approximately one third of participants, a basic medical questionnaire and a brief medical history, much more detailed data are required. It would be impractical to ask families to recount every healthcare interaction they have had. Obtaining data directly from primary or secondary care providers would not be feasible due to time and complexity, and the size of the cohort.

In order to minimise the data requested, the study is requesting very limited demographic data (limited to demographics relating to health care, such as Integrated Care Board/Trust names). No variables have been requested which are not necessary for the proposed analyses.

Data:

UCL are requesting Mental Health of Children and Young People (MHCYP) 2017 and 2020 survey data via the UK Data Service (UKDS). This will enable comparison of the mental health, behaviour and well-being of the IMAGINE-2 participants with intellectual disability (ID) to the MHCYP cohort in the general population.

UCL are also requesting HES APC, HES CC, HES OP, HES A&E and ECDS data for IMAGINE-2 participants and a matched control cohort identified by NHS England who do not have recorded ID. UCL will provide NHS England with the NHS number, date of birth, postcode and gender of IMAGINE-2 participants on one occasion alongside a unique study Identification. HES data will be returned to UCL with the unique study Identification only. An equivalent pseudo-Identification will be generated for the matched control cohort.

Processing:

The MHCYP data is only available by system access via the UK Data Service (UKDS) and the data will be processed in their secure environment. The MHCYP data will not be linked to neither cohort data. In order to protect patient confidentiality in publications resulting from analysis of MHCYP data users must apply the following rules:

· zeros should be shown,

· 1-7 to be rounded to 5,

· any other numbers rounded to nearest 5,

· rounding unnecessary for averages etc.,

· percentages calculated from rounded values,

· if zeros need to be suppressed, round to 5.

The HES data will be stored and analysed securely within the UCL Data Safe Haven environment. Once the data are received, the IMAGINE-2 research team at UCL will perform exploratory data analysis and clean data as required (e.g. by removing or flagging missing data, subsetting the data for ease of analysis and any other necessary processing in order to make the data ready for analysis). The study intends to link records in the IMAGINE-ID database regarding details of the cohort’s development, well-being, mental health and adaptive functioning to HES data for this project specifically. This will be done using each cohort member’s unique study ID. There will be no requirement or attempt to re-identify individuals.

All research data held in the Data Safe Haven for analysis is kept separate from the identifying data files (also stored in the Data Safe Haven).. These data are kept in a separate secure system within Data Safe Haven. The research data and identifiable data will not be linked, with the exception of Date of birth and postcode for demographic analysis purposes if required. The raw data will not be transferred out of the Data Safe Haven. At the stage of creating publications or presentations, only aggregate and summary data with small numbers suppressed in line with the HES Analysis guide will be transferred out of the Data Safe Haven. All data will be processed by UCL.

VIRTUS LONDON 4 do not access data held under this agreement as they only supply the building. Therefore, any access to the data held under this agreement would be considered a breach of the agreement. This includes granting of access to the database(s) containing the data.

The data will be stored for use as part of the research analysis carried out by the IMAGINE-2 study only. It will not be available at record level to third parties and will not be available for any commercial use. Access to the IMAGINE ID data within the UCL Data Safe Haven is controlled by the IMAGINE ID Principal Investigator at UCL. Only authorised users have access and access is via 2-factor authentication (username, password and authentication code).

Under this Agreement, the data will only be processed by substantive employees of UCL and those with access to the data have taken information governance training and are aware of their responsibilities and obligations.

The data will be used to create analytical outputs which are subsequently intended to be used in research reports and published in peer-reviewed journals and/or presented at academic conferences appropriate for the nature of the analysis and message.

All outputs of the data analysis will be aggregated and small numbers will be suppressed in line with the HES Analysis Guide. The study will ensure that NHS England data will not be linked to any other data which would be likely to make it identifiable.

The audience for this research is primarily clinicians and researchers. However, the research teams are keen to disseminate findings to participants in the study and to the general public. The study will work with patient groups such as UNIQUE (Understanding Rare Chromosome and Gene Disorders), a registered charity which provides information and support to individuals with various genetic disorders, and raises public awareness of the conditions. The outcome of this study may include high-level insights gained through analysis of MHCYP and the IMAGINE research data to provide more information and understanding on the genetic disorders for the affected individuals and the public. UCL aim to submit for publication within 24 months of receipt of the dataset.

Planned disseminations include:

Publications: The study will use open access publication, in line with current Medical Research Council policy, to maximise the impact of peer reviewed publications. Publications will be targeted at a number of different academic and clinical audiences. UCL intend to reach out to the community of researchers working with intellectual disability through specialist publications. A broader readership will be targeted in academic psychiatry that is interested in specific findings of more general relevance through more academic publications, and higher impact journals if appropriate. Journals to be targeted include: The Lancet Psychiatry, British Journal of Psychiatry, Journal of Developmental Disorders, Journal of Intellectual Disability Research, American Journal of Psychiatry, Journal of Health Services Research and Policy, Archives of Disease in Childhood and the British Medical Journal. Both the principal investigator (PI) and the co-PIs have strong track records in these areas, including publications in a range of high impact journals.

Conferences: Researchers intend to present findings at academic conferences and other forums relating to research in neurodevelopmental disorders and behavioural phenotypes, genetics, and rare diseases in the UK, Europe and North America, including: International Society for Autism Research (INSAR), Neurodevelopmental Disorder Annual Seminar, UCL Mental Health symposium, Society for the Study of Behavioural Phenotypes Conference, Royal College of Psychiatrists Conference.

Clinical community: A main objective is to provide clinically valuable information to clinicians from a wide range of specialities, including community paediatricians, clinical geneticists, neurologists (both paediatric and adult) as well as specialists working with intellectually disabled people. The study will target these groups by presenting findings at clinically oriented conferences. The study intends to publish summaries of the findings in wide circulation journals of general interest to practitioners, including family doctors. The study will aim to ensure that its findings are distributed to clinicians and others working professionally with ID through contacts with appropriate professional and specialist Societies, including: Royal College of Paediatrics and Child Health (through publication in the Archives of Disease in Childhood); Royal College of Psychiatrists (through publication in the British Journal of Psychiatry); Members of the British Medical Association (through publication in the British Medical Journal).

Other Intellectual Disability Stakeholders: The study has worked closely with the former Chief Executive Officer (CEO) of the charity UNIQUE to advise upon recruitment of families to the study. The study is also working closely with a wide range of other parent support organizations for children with specific ID-related genetic disorders such as SWAN ('Syndromes without a Name'). This liaison ensures there is a gateway directly into the community of parents and carers of individuals with intellectual disability for whom much of our research will have direct relevance. The study has established a newsletter and a website (https://imagine-id.org/) in collaboration with patient and public involvement (PPI) groups, to make its findings available to all stakeholders with an interest in the research, and its implications for the ID community.

Wider audiences: UCL and Cardiff University are keen to promote public engagement in science at many levels. Research staff, at all sites, participate in outreach activities such as attending conferences held by support groups. Communication of the project outcomes to the general public will take place during the lifespan of the project via on-line announcements (e.g. Twitter and Facebook), as well as meetings with journalists from the popular scientific press and general press. Presentations at symposia and public lectures will broaden the impact of the research on the public. The study also intends to make available information for families in collaboration with patient-support organisations and other stakeholders.

UCL estimate outputs will start to show 6 months on receipt of the data from NHS England.

In England, there are over a million people with learning disabilities, a quarter of whom are children of school age. Most moderate to severe intellectual disability (ID) has a genetic cause. The study hopes to have beneficial impacts upon the domains of clinical practice and care services, and quality of life for affected families. Potential benefits to health and social care include better understanding of the mental health and well-being of children with ID caused in whole or in part by a known genetic variant, with the aim of enabling more efficient targeting of healthcare resources to provide the best support to them.

Clinical practice: Clinicians in the NHS increasingly request specialist genetic investigations for children with ID. Usually, the results do not translate into specific recommendations for management or prognosis relating to behavioural adjustment, although families would welcome such knowledge . The patient, parents and carers consultation group gave feedback and comments that they highly supported more research on the genetic investigations for children with ID, in order to gain more understanding and knowledge of the conditions. With the genotypic (genetic constitution) data on specified genetic disorders with standardised phenotypic (observable individual characteristic) data and longitudinal health service utilisation, this study expects to generate valuable information of the mental health and well-being of this cohort. Identification and characterisation of the trajectory of these rare disorders will be used to provide information to clinicians and health professionals who see patients with these disorders. Better evidence-based information is expected to help clinicians and families in making appropriate health and social care decisions and aid in assessment of whether to undertake interventions or prescribe medication to ameliorate symptoms. If children are able to access better care or take advantage of adjustments or interventions, this may improve outcomes (e.g. better educational attainment, improved social communication and inclusion) which in the long term can result in fewer interactions with mental health care providers to alleviate some of the pressures on the resources of the healthcare system..

Families: Unusual behaviour patterns or emotional disorders associated with ID are often ascribed to inappropriate parenting practices. Recognising common disorder-specific patterns is the first step to reassuring parents and educating clinicians/social support staff. This is expected to reduce self-blaming and stress, with resultant improved quality of life for affected families. The impact of child behaviour can be reliably and easily measured across time, and may independently predict future symptoms and psychiatric disorders, including the interactive process by which behavioural and emotional problems can undermine family/individual quality of life. Through documenting health service utilisation in conjunction with genotypic and phenotypic detail, new opportunities for intervention could arise, thus enhancing parents' economic activity (e.g. by promoting their mental health, reducing school exclusions, limiting risk of parental separation).

Health Care Professionals: Intellectual disability implies global impairments in cognitive skills, yet some developmental trajectories may be preserved (exemplified by the relatively good language skills of children with the genetic disorder Williams syndrome). Gaining knowledge about differences in ability across different genetic disorders has implications for education planning and fostering the maximisation of individual potential. Such discoveries could inform policy on the management of children with ID due to a genetic cause. Information on environmental factors influencing emergence of challenging behaviour linked to genotypic risk could point to genotype specific interventions, reducing risk of transfer to residential care and the associated costs.

These benefits will impact not just the NHS in terms of better evidence for clinical care decisions, but also every family which includes a child with intellectual disability or learning problems due to a genetic cause, both now and in the future. Through the outputs resulting from the analysis of the requested data, we aim to deliver benefits over the next 5 years (2022-2027) initially. UCL and Cardiff University will benefit from the publication of peer-reviewed journal articles which will contribute to further funding applications and research collaborations both nationally and internationally.

No yielded benefits have been attained so far to date using the NHS England data. Information on the benefits of the wider IMAGINE-ID study can be found at https://imagine-id.org/