Toward a fully de-identified biomedical information warehouse.

Area of Science:

• Biomedical Informatics
• Health Data Management
• Clinical Research Data

Background:

• The Ohio State University Medical Center's Information Warehouse stores diverse clinical and research data.
• Data use is regulated by federal privacy laws and Institutional Review Board (IRB) oversight.
• The Information Warehouse acts as an "Honest Broker" for de-identified data.

Purpose of the Study:

• To evaluate de-identification schemes for a new, directly accessible de-identified data warehouse.
• To ensure regulatory compliance while facilitating data querying and updating.
• To address the impact of date-shifting on other data elements.

Main Methods:

• Performance evaluation of various de-identification techniques.
• Analysis of the effects of date-shifting on diagnosis and procedure codes.
• Development of solutions for data integrity issues caused by de-identification.

Main Results:

• Findings on the performance of different de-identification schemes are reported.
• The impact of date-shifting on data elements like diagnosis and procedure codes was analyzed.
• Potential solutions for date-shifting challenges were considered.

Conclusions:

• De-identification schemes require careful evaluation for research data usability.
• Date-shifting can affect data integrity, necessitating specific mitigation strategies.
• The developed de-identified data warehouse aims to support research and education effectively.