LONGITUDINAL ADMINISTRATIVE DATA SPINE SCOPING PROJECT GRANT FOR THE SPF UK POPULATION LAB WAVE I

Lead Research Organisation: University of Bristol
Department Name: Social Medicine

Abstract

This project will gather evidence to inform the development of longitudinal research data resources and to respond to core recommendations from the 2017 UK Economic and Social Research Council's (ESRC) Longitudinal Studies Review. The UK has a long history of running longitudinal population studies, where members of the public are invited to take part in research based on a set of characteristics (e.g. they were all born in the same week). Researchers then follow-up on these people's health and wellbeing over time with the aim of understanding why people's health, development and life outcomes differ, and how we can improve health and social systems to further the public good. New data science opportunities and increasing digitisation of routinely collected health and social data mean there are new opportunities to improve this type of research and new opportunities to improve the public good.

This project will gather evidence to support ESRC and Health Data Research UK (HDRUK) in assessing the feasibility and potential value of a longitudinal Administrative Data Spine (ADS). An ADS can be summarised as a register containing a record of every member of the UK population, their contact details and which has been enhanced with key health and social information (such as health status, age, sex, household and neighbourhood characteristics, social and economic status). This ADS could be a research tool to:

1) select and invite members of the public into new longitudinal research studies. When doing this it is important that those invited are representative of the wider population so everyones circumstances can be studied and everyone - including the vulnerable or disadvantaged - can benefit from the improvements that longitudinal research can bring;
2) understand if existing studies represent the wider population so that policy makers can understand how to interpret study findings when designing policies; and to understand if the way in which recruitment to the study, people stopping taking part in the study, or if data quality issues have introduced error or bias into study findings and to inform statistical methods that can be used to control for bias or impaired representation;
3) link to study participants routinely generated records (e.g. their health, education, benefits, employment, criminality records) in order to collect new data for research which may be more accurate, or very technical;
4) understand the operational and scientific opportunities and to control the scientific challenges where people in UK longitudinal research take part in more than one study (2 to 3 million UK citizens are take part in longitudinal studies).
5) the ADS could also be useful to run 'population data laboratory' research designs, where longitudinal studies are used in research based on local, regional or national databases of people's records. This type of research brings the benefits of longitudinal records (detailed data and biosamples provided by people directly over time) coupled with the benefits of routine records (records which capture information on health events or the use of government services over time);

An ADS would need to be constructed from diverse sources, including those held by government departments, the Office for National Statistics and the NHS. It is anticipated that some of this information will be held at a UK level; some at a devolved administration level; and others at a local/city/regional level. This projects, and other projects running at the same time, aim to gather the views and experience of academic researchers, health professionals, government officials and importantly the public on what such a research resource could do, how it should be constructed and what measures would need to be put in place to allow it to operate with the support of data owners and the public.

Planned Impact

NA

Publications

10 25 50