A recent quote attributed to Intermountain CIO Marc Probst describes the current phenomenon in the healthcare industry of organizations fixating on analytics as a “bright shiny object” and “magic” solution. This hits on a critical issue. Analytics tools and surrounding processes do have amazing potential to transform healthcare – and will do things in population health in the next 10 years that we can’t even imagine today. However, healthcare organizations can’t get so focused on analytics that they cut corners on the foundational work that makes meaningful analytics possible, or fail to look beyond the immediate analytics applications they are interested in today. CIOs, CMIOs, and the teams working on population health projects need to first make sure that they are assembling their data properly, and secondly that they are building population health architectures that will be scalable and flexible enough to meet future needs that may not even have been contemplated yet.
Getting your data house in order means speed, accuracy and completeness - making sure you’re tapping into all of the relevant data sources that are available to you efficiently and without degrading their quality. At Orion Health, we have done population health projects all over the U.S. and across the world, and we break them down into six steps:
• Acquisition of data
• Aggregation of data (in a patient-centric, normalized fashion)
• Analytics of data to stratify the population
• Access to the macro-level analytics and micro-level longitudinal patient record by various stakeholders
• Action on the stratified population (e.g. enrollment onto a condition-specific, community-wide care pathways)
• Adoption of the systems and processes to support all of the above. Without adoption by all members of the circle of care (providers, clinicians, and patients themselves), success will be limited.
These “Six A’s” build on each other, yet the critical foundation of any population health initiative is in those first two A’s – acquisition and aggregation. Analytics – shiny as it is – works best when it comes later, atop a solid and comprehensive foundation.
Healthcare organizations have so many data sources to tap into and more coming on-line all the time. Many use legacy or proprietary interfaces that make accessing complete data very difficult, and some may not even offer real-time access to data. Traditional sources like EHRs, registration systems, lab systems and pharmacy systems have always been difficult to aggregate together into “apples-to-apples” data sets associated with the right patients, and as we add more new data sources from multiple organizations across the community or from genomic and biometric data, the complexity and the scale of that challenge escalates dramatically. So, for most population health initiatives, acquiring and aggregating all of the necessary data will be the hardest and most important work for the foreseeable future and will probably represent the majority of effort on the project.
In addition to the acquisition and aggregation of data, the other challenge that many organizations gloss over at their peril as they architect their population health systems is keeping up with innovation. Population health is in its infancy relative to the critical role it will play in the decades to come and we’ve only scratched the surface of the types of data we will gather and the applications of it that we will employ. While we can’t possibly know today everything what we will someday want to do with this data, we can be sure that that the massive onslaught of data will continue to intensify, and that we will need to be prepared to be as flexible as possible to accommodate the applications, integrations and analytics to come.
Future-proofing IT infrastructure isn’t easy, but we at Orion Health are focusing on a few things to make sure we are as well positioned as possible: super-scalability using technologies like the NoSQL Cassandra distributed DBMS, support for the latest high-performance interfaces like FHIR, and basing it all on a standards-based platform with open APIs so we can adapt to and accommodate the exciting innovations to come.
One organization we have worked with that really understood this bifurcated challenge of future-proofing while quickly aggregating the underlying data for today’s needs is Cal INDEX. Cal INDEX is one of the biggest health information exchange (HIE)/population health initiatives in the world, serving one of the largest and fastest growing populations in the U.S. They knew they needed super-clean, super accurate data for analytics, and they wanted to provide universal access to comprehensive, up-to-date patient information to providers across the state of California. Consequently, those first two A’s (acquisition and aggregation) were a top priority. Our Orion Health Open Platform was a perfect choice for them, because it’s built on one of the most mature and robust data integration engines on the market, we have worked with virtually every data source there is in healthcare, and it is highly scalable and flexible to accommodate new data models as they emerge.
Since Cal INDEX also has a very long-term view of the value their initiative will provide to generations of Californians, this scalability and flexibility were essential requirements for their project. They needed open APIs to connect to all of the current and future front-end applications they might employ, and they needed high-capacity, high-performance database technology to handle the very large volumes of data they expect to be moving through their systems. Cal INDEX is off to a great start and has already aggregated an unprecedented amount of data as it promises to become a model for the future of HIE and population health initiatives.