The i2b2 Cohort Discovery Tool provides an easy-to-use, self-service method for MCW researchers to query the CTSI’s Clinical Research Data Warehouse (CRDW) to answer the question “Does the CRDW contain a cohort of patients with certain characteristics?”. Search criteria can include patient demographics, ICD-9 coded diagnoses, CPT and HCPCS coded procedures, laboratory test results, medication orders by class or ingredient, clinical encounter metadata and MCW Tissue Bank biospecimen variables.
i2b2 queries return an approximate number of distinct patients matching the search criteria. No patient identifiers are revealed. All i2b2 users create their own library of queries which can be re-run as the data warehouse expands or re-used as components within new queries.
The Froedtert Health System (FH) and Children’s Hospital of Wisconsin (CHW) have separate data warehouses and toolsets. Generally, the FH system contains ~10% children and CHW contains ~10% adults.
All completed i2b2 queries transmit to the CTSI’s Honest Broker data extraction tools which are also separate for FH and CHW. These toolsets enables research teams to generate pre-defined data tables in de-identifed form to facilitate cohort feasibility or retrospective data analysis. Research teams with an approved MCW or CHW IRB protocol are allowed to request identified data tables from their respective IRB for a variety of projects.
Hypotheses and Feasibility vs. Data Content
Until very recently, EMR systems like Epic have been primarily designed to improve patient safety, clinical workflow management and billing compliance. Using this data for research purposes can be challenging. We strongly encourage researchers to meet with us in the early stages of a project to gauge the feasibility of clinical queries and to help identify any nuances in the clinical data that could affect their interpretation of the data.
Examples of Clinical Data Types Available
* Lifestyle (tobacco, alcohol, drugs, and contraceptives)