This is the home to an emerging Clinical Decision Support (CDS) platform that seeks to provide the knowledge management, business intelligence and predictive analytic technologies required for advanced cognitive support and workflow optimization. An Event Driven Architecture (EDA), combined with a Service Oriented Architecture (SOA), is used to deploy and manage these capabilities. The EDA initiates appropriate analytic processing in response to real-time events, a feature critical in effective Clinical Decision Support. Triggers can be messages, for example, HL7 transaction sets are used to communicate laboratory results or patient monitor waveforms that require Complex Event Processing to be handled effectively. The initiated workflows are then managed using SOA components; each service ensuring that core business logic is well abstracted, reusable, and encapsulated behind standards-based interfaces. A workflow engine provides the advanced process orchestration and state management critical for executing complex clinical guidelines and treatment plans. A Production Rule engine (Drools) is utilized to a) capture and encode clinical domain expertise, b) ensure process validity with respect to declarative constraints, and c) provide flexible control over application and middle tier behavior.
The CDS Collaboratory Architecture (CCA) is designed to facilitate behavior change and improve health care outcomes for patients. This is a data-driven, goal-oriented objective, and consequently, the infrastructure is designed to request and persist clinical data from distributed government and "civilian" networks using the Federal Health Architecture (FHA) Nationwide Health Information Network (NwHIN) architecture. It aggregates this distributed clinical information with local data, such as site-specific research databases or legacy Electronic Medical Records (EMR), to create a virtual warehouse ideally suited for developing individual and population-based action plans. A design principle unique to the Socrates approach is that rule and workflow processing is done in a patient-specific context, each session can be dynamically instanced and provisioned with select knowledge bases and individualized preferences. This design, while resource intensive, ensures personalized, high-performance rule evaluations and workflow orchestration.
CCA manages the data structure and semantics required to aggregate heterogenous, distributed data into a collection of normalized, canonical data. Furthermore, Socrates recognizes that not all analyses of this operational data store are best approached with predicate logic; some require alternative inference techniques. To expand the analytic capabilities available, we extended the Drools rule engine with a Predictive Model Markup Language (PMML) interface, a standard used to represent predictive models such as support vector machines, neural networks, or cluster algorithms. By enabling the Drools engine to import and emulate the functionality of alternative inference models, Socrates expands the reasoning techniques available to solve the problem at hand.
CCA includes several prototypical, standards-based applications designed to facilitate knowledge management and promote usability in the clinical environment. In the coming months, the project will contribute the following applications and demonstrate how they can be layered upon existing health information systems.
...
...
CCA is an ambitious attempt to deliver high quality, advanced clinical decision support using commodity, open source infrastructure. It is a platform for researching how information technology can help change patient / provider behavior to better achieve and sustain health.