Health Data Can Save Lives — So Why Is It So Hard to Use?
Health data collected through the NHS has enormous potential to improve care and answer important questions. Yet accessing and linking this data can be a long and complex process. When access is slow, opportunities to generate important insights that could improve healthcare may be delayed, and researchers may be discouraged from undertaking new studies.
In a new paper published in the International Journal of Population Data Science (IJPDS), researchers describe their experience of conducting two large studies - LAUNCHES QI and CHAMPION* - that linked several national datasets to investigate the lifetime outcomes of people born with congenital heart disease. Drawing on their experience, the authors outline the challenges and delays they encountered, offer practical advice for other researchers planning similar studies, and suggest system improvements that could help protect sensitive data while enabling research to progress more efficiently.
To access routine health data, the two studies required multiple applications and approvals from 12 different regulators. These included research ethics committees, the Confidentiality Advisory Group, and each organisation responsible for the requested datasets. While these processes are essential for ensuring data is accessed safely and appropriately, the authors identified substantial duplication and friction that could be improved without compromising ethical oversight.
The scale of the administrative burden was considerable. Setting up the first study, LAUNCHES, required 39 lengthy forms and took two and a half years before the research team received the data. Data access costs alone totalled £55,650. As the LAUNCHES data was later reused for the CHAMPION study, the researchers expected a simpler process the second time. However, CHAMPION still required 44 forms, nearly two years to receive initial approvals, and a further 18 months to obtain updated or new datasets. These updates cost a further £24,456.
Administrative work also continued after the data was received. Several regulators required annual renewals, and any study amendments to the studies needed prior approval. For LAUNCHES, these renewals and amendments required an additional 75 forms, while CHAMPION required 92 more.
The researchers also encountered several delays and challenges along the way. These included repeated requests for similar information across multiple applications, with duplicated questions, limited regulator resources to process requests, uncertainty around application procedures, organisational changes affecting data custodianship, and errors identified in the data once received.
Based on their experience, the authors suggest several relatively simple improvements that could help streamline the system. These include using the same application forms or questions across regulators, providing sample datasets to reduce data errors, simplifying processes for study amendments, and more streamlined feedback during the application process to reduce repeated back-and-forth between researchers and regulators.
While national initiatives - including recent government and independent reviews, and proposals for a national health data service – aim to improve the UK’s health data research infrastructure, such reforms may take years to implement. In the meantime, practical improvements within the current system could help reduce delays and support research that has the potential to deliver meaningful benefits for public health.
*LAUNCHES QI: Linking AUdit and National datasets in Congenital HEart Services for Quality Improvement
CHAMPION: Congenital Heart Audit: Measuring Progress In Outcomes Nationally
Click here to read the full article
Julie Taylor, Senior Research Coordinator, Clinical Operational Research Unit, Department of Mathematics, University College London, UK