# i2b2 — Informatics for Integrating Biology and the Bedside ## Overview i2b2 is an open-source clinical data warehousing and cohort discovery framework originally developed in 2004 as an NIH National Center for Biomedical Computing at Harvard and Mass General Brigham, now deployed at more than 400 academic medical centres worldwide. Its core architecture organises clinical data into a cell-based model where a query processor communicates with domain-specific data cells (demographics, diagnoses, medications, laboratory results, genomics) via a standardised XML messaging layer, allowing institutions to expose only chosen data domains. It is primarily used for cohort discovery: clinicians and researchers can build patient cohorts interactively by combining clinical criteria without writing SQL, and the system supports temporal queries such as "diagnosis A followed within 30 days by procedure B". ## Relationship to OMOP CDM and HL7 FHIR i2b2 and [[OMOP CDM]] coexist at several major sites, including [[AP-HP]], where the i2b2 CDW was deployed before the [[OMOP CDM]]-based EDS warehouse. The two systems serve complementary roles: i2b2 for local cohort discovery and feasibility, [[OMOP CDM]] for standardised federated research. Established ETL pathways exist to transform i2b2 data into [[OMOP CDM]]. Integration with [[HL7 FHIR]] is supported through the i2b2-on-FHIR project, allowing FHIR-based data ingestion and query translation. ## Federated Use The SHRINE (Shared Health Research Information Network) layer sits on top of i2b2 to enable federated cohort queries across multiple sites without moving patient-level data. In the United States the ACT (Accrual to Clinical Trials) network connects over 80 academic medical centres via SHRINE/i2b2. Equivalent federated deployments exist in Europe, including at several ECRIN-linked sites. ## Connections - Compatible with: [[OMOP CDM]] (ETL pathway), [[HL7 FHIR]] (i2b2-on-FHIR) ## Resources - https://www.i2b2.org - https://github.com/i2b2