Investigating the inter-relationship between diabetes and children's educational achievement.

Context of research
Diabetes is one of the most common chronic childhood illnesses, affecting 1 in every 700 children. There are two types of diabetes which affect how insulin regulates blood glucose levels. In this application, we focus on Type-1 diabetes which is the most common form in childhood, arising from immune-mediated beta-cell (which produce insulin) destruction. A child with Type-1 diabetes must self-inject insulin to keep blood glucose within acceptable limits, the amount needed varying with what they have eaten and physical activity. Type 1 diabetes is estimated to account for £1 billion in direct costs and £0.9 billion in indirect costs for the UK each year. A better understanding of the processes that affect self-management during childhood will help reduce the need for costly interventions.

Aims
The overall aim of the research is to better understand the interrelationship between diabetes (and diabetes management) and the educational experience. We will also explore how the relationships between diabetes and education will be affected by characteristics of the child (e.g. gender), their families (e.g. family structure) and the health services they use (e.g. characteristics of diabetes clinics). To make robust statements about these relationships a second aim must be to utilise the most advanced quantitative methods to make the most of the various datasets available.

Methodological objectives
1. We will identify and link individuals from different health and education datasets. This allows, for example, to make inference on the effect of glycaemic control (using measures from the National Paediatric Diabetes Audit dataset) on educational outcomes (using linked measures from the national educational database).
2. We will also use 'multilevel' modelling techniques which adjust for, and allow us to make inference about the 'groupings' in which the children are nested such as families, schools and diabetes clinics. This includes using multiple measures of health and education for an individual over time.
3. Where data are missing we will also use sophisticated techniques to infer what those values may be, as otherwise these individuals would simply be deleted from our analysis.

Substantive objectives
1. To understand how diabetes affects educational attainment. This includes how health measures like diabetes management affect achievement and also which individual characteristics modify these relationships.
2. To understand how diabetes changes the health and educational trajectories of individuals, in particular changes in quality of self-management through major educational and physiological transitions.
3. To quantify the effect of key groupings (these include schools, diabetes clinics and families) and their characteristics modify the relationships between health and education.

Applications & benefits
The key findings will relate to how a diagnosis of diabetes and diabetes management affects the school experience. We will disseminate these findings to clinicians through the Brecon Group network of Welsh clinics caring for children with diabetes which has developed over the last eighteen years and established the register all new cases. This will be done through a workshop held at Cardiff University for clinicians, service users, school health representatives and Diabetes Specialist Nurses. It is hoped to give practitioners a better understanding of the effects of diabetes in a context in which the young person will spend a considerable amount of their time and which will have a large impact on the quality of their self-management. We will share our findings with other researchers and policy makers through published articles in specialist journals to help develop both the methodological and substantive aspects of the understanding of the role of education and schools in the treatment of chronic childhood conditions.

Aims
A child with diabetes will learn to self-manage their blood glucose. The quality of this management will have direct health implications which may impact educational outcomes (including attendance and achievement) which in turn may affect self-management and other health outcomes. The overall aim of the research is to better understand the interface between health outcomes and educational outcomes for children with diabetes. We will consider how other factors such as the characteristics of the child, their families, and the health services they use influence these relationships. We will use advanced statistical approaches including linking datasets, missing data approaches, longitudinal analysis and variance functions to provide robust estimates of these relationships.

Methodological objectives
1. To use robust data to combine information from multiple health and education datsets.
2. To model multilevel data structures, including cross-classified clusters, longitudinal within individual trajectories of health and education measures, and modelling cluster (individual and clinic) variance functions of blood glucose management.
3. To use advanced multiple imputation to deal with missing data in complex data structures.

Substantive objectives
1. To model the effects of health on education for children with diabetes, and the characteristics of individuals.
2. To model trajectories of health and education for children with diabetes to unpick the direct effects of health on education, the reverse causality of education on health and the unobserved characteristics which simultaneously affect both health and education. To identify important time varying characteristics and effects of the timing of important transitions in health and educational trajectories.
3. To quantify the effect of cross-classified multilevel clusters such as schools, diabetes clinics, families and how characteristics of clusters modify relationships between health and education.

List of parties who will benefit from this research:
Children with diabetes
Families of children with diabetes
Clinicians working with children with diabetes or other chronic diseases
Teachers and related education practitioners e.g. SEN co-ordinators
Health and educational policy makers and policy campaigners
Educational researchers focussing on the effects of child health
Health researchers focussing on the effects of diabetes

How they will benefit from this research

The data linkage will produce a valuable data resource (stored in SAIL) for other researchers. Academic papers and conference presentations that showcase methods (longitudinal approaches, variance functions) will find application in studies looking at other childhood conditions. Dissemination of substantive findings (quantifying effects of diabetes management and covariates on educational outcomes) will support qualitative research which seeks to explain the relationships.

In 2014 new legislation was brought in obliging schools to provide sufficient care for all children with limiting health conditions so they should not fall behind their peers. Though such legislation was welcomed by policy relevant groups such as Diabetes UK, the guidance is generic and there is a lack of evidence on the specific effects of such conditions and the way in which healthcare agencies (diabetes clinics, clinicians and GPs) and educational agencies (schools, teachers, SEN co-ordinators) contribute to successful health and educational outcome for children. We aim to quantify the extent of the challenge and identify the characteristics of clinics and schools that are successful in mitigating the effects of chronic childhood illness. We hope to communicate these findings to a policy audience using articles and presentations through agencies such as Diabetes UK.

Beyond the academic and policy audience we also wish to target organisations and individuals that deal directly with these issues including diabetes clinics, schools and the families of children with diabetes.

For eighteen years The Brecon Group of all paediatric diabetes clinics across Wales have collected diagnostic data on all children aged under 15 years with newly diagnosed diabetes. They have developed strong links to enable feedback from audit and research to influence design of clinical services. We will continue to exploit these links to get feedback on our approaches and to help disseminate findings and advice resulting from our analyses.

For the education sector we firstly produce an online report for teachers. We will use agencies such as Diabetes UK and the Department for Education to help disseminate this report. We will also explore alternative approaches, such as using infographics to help make findings more amenable to a non-scientific audience.

For families we plan to produce a brochure outlining our findings and their implications which we will disseminate through clinics. In particular we will inform families of the overall relationships between diabetes and education, quantify the effects of beneficial practices such as maintaining high school attendance and advising them where they can find further relevant information.

It is hoped that through the impacts on academics, policymakers, clinicians, teachers and families, we will improve the educational achievement and overall wellbeing of children with diabetes. However it is important to also produce outputs for the young people with diabetes that have generated the data we have used. We will therefore make a short online report and video detailing our findings in an accessible format, with a comments section and fuller wiki space that will foster feedback from children with diabetes.