Big Data has become a sort of buzzword, but has already become part of our everyday lives, determining everything from how politicians campaign to the ads you see on the web. It is also at the forefront of lowering costs and expanding access in a variety of domains such as education, where insights are helping shape how young people around the world learn without geographic and cost barriers. The ability to screen for and identify disease on a vast scale promises to significantly reduce costs, while simultaneously allowing for the universal access that the public seeks. And so, as debates about healthcare access intensify, we are curious why Big Data has not yet found its way into the realm of health.
Our review revealed that central to the lack of widespread adoption within healthcare was a lack of communication. Physicians, researchers, and entrepreneurs are walled off from each other, and existing systems do not promote the sort of collaboration required for implementing Big Data in practice. Electronic Health Records are central to this problem, as they work to improve billing, and so are not interoperable with each other and often indecipherable to researchers as well. And so, we propose a new framework to think about patient records management: as fundamentally decentralized and linked through unique Patient IDs. This will allow for a diversity of data types to come online and be seamlessly integrated into clinical decision making and active research without the hurdles of database access and security that we currently face.
Our primary concern behind this new revolution in healthcare is equity. Our past work has focussed on how African-American patients, amongst other ethnic groups, are disadvantaged by our current healthcare system. And so, we approached the question of how to ensure equitable outcomes. Our survey found that many healthcare Big Data projects across the world systematically exclude ethnic minorities in their sample population. This helps reproduce many of the existing disparities in research that primarily benefit white patients - and can lead to spurious results in ethnic minorities. This focussed our recommendations on the broadening of datasets for research, along the lines of the All of Us project, in order to ensure researchers and clinicians are truly able to serve all patients equally.
Our review builds on our collective experience within the realms of computational biology and clinical informatics, and current literature, to identify the many problems with the way digital healthcare exists in its modern form around the world. From fragmentation to discrimination, we demonstrate how current initiatives in different nations fail to meet the challenges of serving an increasingly globalized world. We propose several solutions, drawing from the world in past proposals for universal patient IDs and new technologies such as blockchain, that would enable a Big Data future that can both empower physicians to make better decisions for their patients and for researchers to gain greater clinical insight into disease. We hope that our work is useful for researchers, clinicians, and regulators alike as they think about developing the framework on which to build our healthcare future – and how to best implement our shared principles of security, equity, and efficacy.
Blog post written by Raag Agrawal & Sudhakaran Prabakaran.
Please sign in or register for FREE
If you are a registered user on Research Data at Springer Nature, please sign in