Transformation and Evaluation of the MIMIC Database in the OMOP Common Data Model: Development and Usability Study

Area of Science:

• Biomedical Informatics
• Data Science in Healthcare
• Critical Care Medicine

Background:

• The increasing volume of big data in intensive care units (ICUs) necessitates advanced analytical tools for real-time patient monitoring and electronic health record analysis.
• The Medical Information Mart for Intensive Care (MIMIC) database is a key open-access resource for ICU research.
• Common Data Models (CDMs) like the Observational Medical Outcomes Partnership (OMOP) CDM facilitate data standardization and sharing, improving database searchability and analytical efficiency.

Purpose of the Study:

• To convert the MIMIC database into the OMOP common data model format.
• To assess the advantages of the MIMIC-OMOP transformation for data analysts and researchers.
• To create a standardized, accessible dataset for intensive care research.

Main Methods:

• The MIMIC database (v1.4.21) was mapped to the OMOP CDM (v5.3.3.1) through structural and semantic alignment.
• The mapping process involved three phases: conception, implementation, and evaluation, focusing on aligning local terminologies to OMOP standards.
• A documented, tested, and open repository was developed to support the transformation and facilitate community contributions.

Main Results:

• 64% of MIMIC data items and 78% of source concepts were standardized into the OMOP CDM.
• The resulting MIMIC-OMOP dataset demonstrated robust support for community contributions during a datathon with 160 participants.
• Over 15,000 requests were executed within a 1-minute maximum duration, highlighting efficient data accessibility.

Conclusions:

• The MIMIC-OMOP dataset is the first freely available, de-identified dataset for reproducible intensive care research.
• This data transformation methodology is generalizable to other medical domains.
• The OMOP CDM enhances the utility of large healthcare datasets for research and analysis.