An Interactive Information Visualization System for Temporal Queries in a Large-scale COVID-19 EHR Dataset...

Area of Science:

• Clinical Informatics
• Health Data Science
• Computational Epidemiology

Background:

• Secondary use of Electronic Health Records (EHR) is crucial for COVID-19 research.
• Challenges exist in accessing and analyzing large-scale EHR data for clinical research.
• Existing informatics tools often require specialized database expertise, creating barriers for researchers.

Purpose of the Study:

• To introduce COVID-SPHERE, a self-service web application for secondary use of EHR data in COVID-19 research.
• To facilitate complex temporal query building and data visualization for clinical researchers.
• To reduce informatics barriers and enhance the efficiency of clinical and translational research.

Main Methods:

• Developed a flexible EHR concept framework with hierarchical concepts and ontologies.
• Implemented dynamic, real-time MongoDB query generation from a single-click interface.
• Integrated interactive, faceted visualizations for longitudinal patient activity and health record analysis.
• Deployed on a server managing over 5 TB of data, including 30 billion health records from 8.8 million patients over 15 years.

Main Results:

• COVID-SPHERE successfully supported multiple published research studies on COVID-19 epidemiology, treatment outcomes, and long-term sequelae since November 2020.
• Demonstrated generalizability across various COVID-related research topics.
• Facilitated efficient cohort discovery and data-driven insights for surveillance and intervention.

Conclusions:

• COVID-SPHERE significantly reduces informatics barriers, empowering researchers to utilize EHR data more effectively.
• The application enhances the efficiency of clinical research and promotes data-driven insights for public health.
• The architecture serves as a scalable model for future public health informatics systems and large-scale clinical research data warehouses.