Navigating the Cs of Interoperability

“The story of any one of us is, in some measure, the story of us all” –Frederick Buechner

Every patient’s unique healthcare journey does not start nor end with a single encounter.  It is the culmination of a lifetime of genetic inheritance, personal choice, and care received.  Each of these unique journeys contribute to the population health story. As the steward of our population’s healthcare data, we assume responsibility for each of those stories being as accurate and comprehensive as possible.

Disparate systems, evolving technology landscapes, and data collection barriers start us on an unstable foundation.  Without connectivity, consistency, and continuity in our data, we are only able to see partial truths.  The stories we tell are incomplete.  

Connectivity is the cornerstone of interoperability.  All organizations need a dependable infrastructure of data pipelines to connect their disparate silos of information.  A multitude of solutions are currently available to make this construction seamless.  Most modern applications, including EHRs, provide APIs from which data can be easily extracted.  Whether they are housed on premises or in the cloud, API access reduces the amount of maintenance and scripting required.  Additionally, a multitude of vendors are now available with pre-configured connections to almost all traditional EHRs.

Consistency becomes the next challenge.  Many different factors can influence how the same piece of information is represented.  These may include differences in data entry processes, data asset types, application workflow, transcription, and variations in data transformation scripting.  Evolving healthcare data standards, such as HL7 / FHIR, provide an excellent baseline from which to consolidate multiple data assets into a single version of the truth.  Most connectivity vendors also provide a single, cross platform, schema solution for data delivery.

Continuity remains the final bridge to bring our stories together.  Even consistent, connected data will remain isolated without a way to identify patients across platforms.  Customized fields and business processes can be used to incorporate a single patient identifier within a practice, or even across a hospital system.  However, these solutions are subject to human error and may not be supported by all applications.  Patient matching algorithms based on discrete data elements are more flexible and reliable.  Ongoing advances in machine learning and AI techniques continue to drive more sophisticated and precise matching solutions.

With connectivity, consistency, and continuity in place, our clinical data warehouse is now a treasure trove of accurate, comprehensive, unique journeys.  We have a robust vocabulary to tell clear and concise stories about the health of our population and those who care for them.