Clinical Decision Support (CDS) can improve care by delivering context-sensitive clinical guidance at the right time to the right care provider. Nevertheless, authoring and executing CDS artifacts within or in conjunction with an electronic health record (EHR) has historically been an onerous and highly labor intensive task. The cost and organizational commitment involved in creating, maintaining and governing clinical guidance are formidable. The current paucity of executable CDS content reflects such impediments.
Further complicating matters, CDS content has historically been near impossible to share across organizational boundaries, both in the academic setting and in the commercial market. Standards-based terminology and process models required to support knowledge interoperability across heterogeneous platforms have been lacking until recently. The lack of content portability hinders greater adoption by EHR systems, inflates costs, and creates an unsustainable environment. Fortunately, the introduction of standards-based and computable knowledge artifact specifications is beginning to address these limitations.
Developed in response to an Office of the National Coordinator (ONC) Standards & Interoperability initiative called Health eDecisions, the HL7 Clinical Decision Support Knowledge Artifact Specification defines an implementation-agnostic XML standard for the representation for order sets, simple event-condition-action rules, and structured documentation templates (e.g., forms). The specification ensures that content semantics, behavior, and expressiveness are fully defined and can be retained when exchanged between systems. The specification, while by no means complete, addresses a core set of CDS artifacts – the low hanging fruit in a potentially vast opportunity.
Recently, Cognitive Medical Systems and Motive Medical Intelligence demonstrated to the Dept. of Veterans Affairs (VA) and the ONC for Health Information Technology the most comprehensive implementation of the standard of which we are aware. The demonstration illustrated how a healthcare system might purchase third party, implementation-agnostic “content,” transform it into runtime, executable code that can be seamlessly integrated into a standards-based health information architecture.
“We showed how domain knowledge and content can be imported into a system automatically,” said Claude Nanjo, Cognitive Medical Systems Chief Scientist. “We then translate and present that knowledge at the point of care.”
The demo validated that the Clinical Decision Support Knowledge Artifact Specification is expressive enough to facilitate structured knowledge capture and content integration. It proved that knowledge interoperability between parties can be supported technically and that the traditionally steep barriers to CDS adoption can be lowered.
In the future, if EHRs were built to support it, the Knowledge Artifact Specification could pave the way for automatically integrating CDS content into a wide range of clinical workflows and systems. Content vendors could write CDS content once and deploy across EHRs and health care organizations. For healthcare providers, implications include lower cost to institute and maintain CDS, faster delivery of the latest clinical guidance, and more effective support of evidence based clinical care.
Cognitive’s CEO Emory Fry, MD, hopes the Centers for Medicare and Medicaid Services will make the knowledge artifact specification part of Meaningful Use, encouraging EHR systems to provide standard interfaces for consuming the CDS knowledge artifacts their customers clearly desire.