Data Linkage
What is routinely collected data?
Routinely collected data can be thought of as data that has been collected for purposes other than research. Examples of routinely collected health data include GP records, A&E attendance, hospital appointments and disease registries. These will contain a wide variety of information connected to each contact a patient has with the health service, such as the reason for any visit, whether they had a test such as an x-ray and what medicine they were prescribed. Patient data is vital for managing an individual’s care, but it can also be used in other ways such as planning health services, improving diagnosis and treatment, and evaluating the effectiveness of policy, provided enough safeguards are in place.
Sharing routinely collected health data for health professionals and researchers
It has not been routine for health data to be shared between different health care providers, such as from your GP to a hospital consultant. That is why patients have often had to repeat their medical history and list of medications each time they see a new nurse or doctor. It is only recently that the sharing of health data between these medical teams has become more commonplace
There has been a history over the last decade of researchers using routinely collected health data to investigate treatments and their impact on health outcomes. Much research is completed using a single dataset such as hospital admissions to ensure the treatments people receive are the most appropriate. This is because there are several challenges to use routinely collected health data.
Several security measures are in place to allow researchers to use even a single dataset. Identifiable information such as names and addresses are removed. The researcher must be trained to handle data safely to achieve a ‘trusted’ status. In many case, data are not moved to the researcher’s computer but instead are stored securely within a central databank. Researchers are only allowed to access the data once their project has been approved by a committee comprised of ethical experts, and public advisors.
Adding value from other routinely collected data
Routinely collected data is not, however restricted to the field of healthcare. Data is also generated routinely by other organisations, for example, councils, charitable organisations and the census. There are also other types of data that can be used rather than the traditional format of a table of rows and columns, for instance maps. So how can they be used to help us with healthcare research?
You will have heard of the saying “you are what you eat”. Well, what you eat is determined by many other factors such as where you live, where you go to school, and your family circumstances, such as being unemployed or living with a disability. Our health is determined by the wider built, natural, economic, and other “structural determinants” that influence us on a daily basis. So, your health is determined by where you live. We could also say: “you are where you live”.
It is only by linking together a wide variety of multiple datasets for the same person that we can look at the structural determinants of health to investigate what impact these are having on people’s health and well-being. Previously unnoticed relationships can be identified giving greater depth of clarity and understanding that wouldn’t otherwise be possible from analysis using a single source of data alone.
King’s College have a video on YouTube which explains the evolution of healthcare records and the benefits of data linkage for research which can be viewed here.