The Analytic Information Warehouse (AIW): a platform for analytics using electronic health record data.

Area of Science:

• Health Informatics
• Clinical Data Analytics
• Electronic Health Records (EHR)

Background:

• Quality improvement initiatives require robust methods for analyzing complex clinical data.
• Extracting meaningful clinical phenotypes from Electronic Health Records (EHRs) presents significant challenges.
• Existing data infrastructure often lacks standardization for derived variable computation and reuse.

Purpose of the Study:

• To develop an analytics platform for specifying and detecting clinical phenotypes and derived variables within EHR data.
• To facilitate quality improvement investigations by enabling efficient data analysis.
• To create a reusable and scalable architecture for clinical data warehousing.

Main Methods:

• Developed an Analytic Information Warehouse (AIW) architecture.
• Implemented data transformation into a common data model for variable reuse.
• Ensured correctness and consistency in computing derived variables and long-term data curation.
• Created secure, high-performance software for processing large EHR datasets.

Main Results:

• The AIW architecture has been implemented and deployed in a production environment.
• The developed software is available as open source.
• The platform was utilized in a project to reduce 30-day hospital readmissions, analyzing over 100 phenotypes against 5 years of data from multiple institutions.

Conclusions:

• A widely accessible platform for managing and detecting phenotypes in EHR data can accelerate research.
• This approach can significantly enhance the use of EHR data for quality improvement and comparative effectiveness studies.
• The open-source nature of the platform promotes broader adoption and collaboration in clinical informatics.