Development and usage of an anesthesia data warehouse: lessons learnt from a 10-year project

Area of Science:

• Health Informatics
• Clinical Data Management
• Anesthesiology Research

Background:

• Developing robust clinical data warehouses is crucial for advancing medical research and improving patient care.
• Anesthesia information management systems generate vast amounts of data that require structured storage and analysis.
• Hospital discharge reports provide valuable administrative and clinical context for patient outcomes.

Purpose of the Study:

• To describe the development and implementation of a ten-year anesthesia data warehouse at Lille University Hospital.
• To share lessons learned and provide guidance for similar data warehouse projects.
• To enhance data accessibility and facilitate faster, easier querying for research and clinical insights.

Main Methods:

• Integration of data from an anesthesia information management system and hospital discharge reports.
• Implementation of a data warehouse with high accuracy for administrative (daily) and monitoring (secondly) data.
• Utilization of datamarts for faster query execution and secondary data computation.

Main Results:

• Successfully integrated 636,784 anesthesia records for 353,152 patients between 2010 and 2021.
• Identified and reported key concerns and barriers encountered during the project development.
• Provided 8 practical tips for overcoming implementation challenges.
• Implemented the data warehouse within the OMOP (Observational Medical Outcomes Partnership) common data model.

Conclusions:

• The developed anesthesia data warehouse provides a valuable resource for research and clinical analysis.
• The OMOP common data model integration facilitates broader data standardization and interoperability.
• Future work includes disseminating OMOP model use in anesthesia and critical care and exploring federated learning for multicenter studies.