As the healthcare industry strives to converge all the data sources required to manage population health, the mass of data needed to do it well, and to both clinically and analytically inform, will require something of a science project. Let’s call it gravity.
Dave McCrory first proposed the idea of “Data Gravity” back in 2010. The analogy is rather simple. As the mass of the object (think population health data stores) grows, it becomes exponentially difficult to move. The more initiatives, data points and products centered around these masses, the quicker they grow and take on gravity of their own. The gravity is in the form of dependencies that manifest with expanding support and reliance on the data sources.
A quick Google search of “population health products” brings up 67,300,000 results. The reality is there aren’t quite that many immovable objects in the technical universe of population health. However, the initial placement of clinical, claims and ancillary data sources has the ability to produce a mass unlike most big data initiatives the healthcare industry has experienced to date.
You may recall back in 2013, the impact of the statement, “A full 90% of the world’s data has been generated over the last two years.” Any credible population health product will not only have the mass of detailed data required to support the effort, but also may need live or near live data feeds. A single component supporting the effort may impact its ability to be successful. Think about your organization. Can that aging data warehouse, or EMR system in constant need of care, support population health? How many workflows and data streams touch those data sources now?
These dependencies and challenges make the first design choices the most important for population health products. Population health technologies and the infrastructure supporting the products will most likely vary across vendors, so it will take a true vendor partner to pull back the curtain and share their design and even a technology transition plan.
As healthcare technologists, we all agree that population health and the longitudinal record will improve health care. However, the data feeds and warehouses should be carefully chosen, and all or most of the technical debt involved in supporting them should be solved to ensure their viability in the short and long-term. A successful vendor partnership will not only support these efforts, but should also include sharing, and better yet partnering, on how it’s to be done.
The technology of population health shouldn’t be a mystery.
AJ Dandrea is operations manager of the cloud hosting organization at 3M Health Information Systems
Do you want to achieve better care and outcomes at a lower cost? Try putting population health management into action.
2. SINTEF. (2013, May 22). Big Data, for better or worse: 90% of world’s data generated over last two years. ScienceDaily. Retrieved November 20, 2015 from http://www.sciencedaily.com/releases/2013/05/130522085217.htm