At UCSF, IT has partnered with the Clinical & Translational Science Institue (CTSI) and the Institute for Computational Health Sciences (ICHS) to build a de-identified version of its Caboodle - Clinical Data Warehouse (CDW) to support UCSF research investigators to have greater self-service access to EHR data.
Current UCSF policy interpretation of HIPAA Rules requires that all research uses of identifiable patient data be accessed through an "honest broker" that will verify IRB authorization for data access and that will limit the data to the minimum necessary data to meet the research needs.
The new de-identified CDW (DI-CDW) would therefore not be subject to HIPAA restrictions on research use, and would not require IRB approval or an honest broker intermediary. The DI-CDW will accelerate the research process through permitting UCSF investigators to find the data they need as soon as they need it. This will accelerate research progress and encourage an exploratory approach to hypothesis generation.
In this session, we will cover all aspects of the project including technical architecture, de-identification methodologies, validation techniques, privacy, policy and external audit requirements. This session would be interesting to anyone that is interested in creating data sets that need to have PHI and/or PII removed, or is simply interested in the general issue of protecting personal information.
None.
None