Characterizing Basic and Complex Usage of i2b2 at an Academic Medical Center

Area of Science:

• Biomedical Informatics
• Clinical Research Informatics

Background:

• The i2b2 platform evolved to support complex queries for patient cohort discovery.
• Little is known about whether advanced query features meet researcher needs.

Purpose of the Study:

• To evaluate the sophistication of queries used by researchers on the i2b2 platform.
• To determine if complex i2b2 features align with researcher requirements for cohort discovery.

Main Methods:

• Categorized 6,662 queries from 609 researchers over three years as 'basic' or 'complex'.
• Analyzed query complexity in relation to requests for patient cohort re-identification.
• Assessed the usage of implemented data domains.

Main Results:

• The majority of queries, including those for cohort re-identification, were basic.
• Complex i2b2 features were underutilized by researchers.
• Common data domains (e.g., diagnoses) were used more frequently than novel, extensively implemented domains.

Conclusions:

• Focusing on optimizing basic query performance for common data domains better serves researcher needs.
• Investment in novel data domains and complex query features may not align with current research practices.
• Enhancing the performance of fundamental cohort discovery tools is crucial for supporting the research community.