Park Street’s CompatX clinical informatics tools provide a powerful query and application development environment. CompatX uses international standard knowledge bases and clinical vocabularies to allow the creation of queries, analyses, and measures, and protocols that are independent of database structure. At the same time, CompatX makes it possible for analysts and query writers to express clinical logic in terms of concepts in the knowledge base – in other words, in terms of clinically meaningful concepts rather than database identifiers. CompatX makes it possible to efficiently build sophisticated queries and analytics that meet the unique demands of clinical informatics.
Park Street is working to create a community of institutions that can share standards-based clinical measures, protocols, analyses, and applications developed upon the CompatX platform. This approach is expected to dramatically reduce deployment costs and risks for a broad range of clinical tools and content.
Portable, interoperable clinical measures, protocols, and applications are the principal design objective of the CompatX toolset.
Portability requires that we become independent of database structures that are particular to specific event types and data representations that are driven by data sources. It requires normalizing on-standard local coding systems for event types and observation values, correcting incoherent local representations of time-stamped data, and overcoming local challenges involving the management of patient identity. It requires a focus on the use of industry-standard vocabularies and ontologies. It calls for a single, coherent representation for these ontologies that integrates local knowledge systems and supports mappings between local and standard concepts. Finally, portability demands that we take advantage of commonly-accepted rules for organizing and grouping data, and that we offer a simple, standard mechanisms for deidentifying patient data and representing patient states and intervals of time.
CompatX is based upon a sotware development language we call OCCL — Open Clinical Concept Language. OCCL is an open specification for an domain-specific query language for clinical informatics, intended to support both general-purpose application building and ad-hoc query usage. OCCL implicitly assumes that data is organized using standard clinical vocabularies. OCCL takes advantage of latest work in domain-specific language design (see Martin Fowler, for example).
OCCL has the following features:
