Introduction to ECHILD: Linked data from health, education and childrens social care

This short course is designed to give participants a practical introduction to ECHILD (Educational and Child Health Insights from Linked Data). ECHILD is a collection of linked, longitudinal administrative datasets covering health, education and children’s social care. More information about the ECHILD, and resources for researchers and the public can be found on the ECHILD website. 

The course is aimed at both analysts intending to use ECHILD and researchers who want to understand more about how the data can be used for policy relevant research. This course includes a mixture of lectures and practical sessions that will enable participants to put theory into practice.   

Day 1 will provide information on the strengths and limitations of the different component datasets of ECHILD, through case studies of the National Pupil Database, Hospital Episode Statistics, Maternity Services Data, Mental Health Services Data, and the Community Services Dataset. Interactive lectures / tutorials will teach participants how to design a research study to answer a specific research question in ECHILD, focusing on the power and complexity of working with linked datasets. We will also discuss how to extract ECHILD data from SQL tables on the ONS Secure Research Service platform and provide an overview of access arrangements. 

Day 2 will focus on a series of practical sessions (in Stata and R) allowing participants to progress through an exemplar research study using ECHILD, covering phenotyping, developing cohorts, and analysing ECHILD cohort data. 

The current course builds on our previous in person and online training courses with the inclusion of some updated information reflecting newer developments of ECHILD. ECHILD users who have previously attended this training course and have an interest in learning about emerging ECHILD research and newly available data are encouraged to attend the next ECHILD User Day, which is available for anyone with an interest in ECHILD.

The course covers: 

• Overview of the component datasets of ECHILD, with reference to case studies 
• Accessing teh data on the ONS SRS
• Strengths and limitations of ECHILD 
• Development of electronic cohorts in ECHILD
• Phenotyping in ECHILD
• Analysing ECHILD data

By the end of the course participants will: 

• Understand what data are available in ECHILD 
• Know how to access ECHILD data 
• Understand the strengths and limitations of ECHILD 
• Know how to developing electronic cohorts in ECHILD 

The course is aimed at academic or government analysts and researchers who would like to know more about ECHILD and how ECHILD could be used in their own research, or who would like to know how it is currently being used to generate policy-relevant research. 

Participants should bring their own laptop and should have access to Stata or R for the practical sessions. 

PLEASE NOTE THIS COURSE IS TAKING PLACE IN PERSON AT FRIEND'S HOUSE, LONDON - Elizabeth Fry Suite, 173-177 Euston Road, London

Course Description

This is an intermediate level course that requires some prior knowledge of epidemiological research methods and statistical analysis, and familiarity with R or Stata.

ECHILD users who have previously attended this training course and have an interest in learning about emerging ECHILD research and newly available data are encouraged to attend the next ECHILD User Day, which is open to anyone with an interest in ECHILD.

Although there is no preparatory reading is required, although participants are encouraged to familiarise themselves with the data resources for researchers available on the ECHILD website.

Participants should have a basic understanding of epidemiological research methods and statistical analysis. They should be comfortable in using R or Stata. Experience using administrative data is not required but would be an advantage.

Programme

Day 1: 

• Introduction 
• Case Studies using each component dataset 
• Developing cohorts in ECHILD 
• Getting started with ECHILD data 
• ECHILD best practice and user community

Day 2: 

• Developing e-cohorts Part 1 
• Phenotyping to define exposures, outcomes and covariates 
• Developing e-cohorts Part 2 
• Analysing cohort data 
• Developing e-cohorts Part 3 
• Leveraging longitudinal administrative data for policy evaluation and population research