Electronic Medical Record Search Engine (EMERSE): An Information Retrieval Tool for Supporting Cancer Research.

Area of Science:

• Medical Informatics
• Clinical Research Informatics

Background:

• The Electronic Medical Record Search Engine (EMERSE) is a software tool designed to facilitate research.
• It supports critical areas such as cohort discovery, population health analysis, and data abstraction for clinical trials.
• EMERSE is currently operational at three major academic medical centers, with ongoing implementation efforts at additional sites.

Purpose of the Study:

• To describe the utilization of the EMERSE tool in supporting cancer research.
• To evaluate the effectiveness and reach of EMERSE across various academic medical centers.

Main Methods:

• Peer-reviewed publications utilizing EMERSE were identified through online searches and direct user communication.
• Usage data from audit logs at three academic medical centers were analyzed to summarize system use.
• Natural Language Processing (NLP) tool MetaMap was employed to characterize search terms and map them to semantic types.

Main Results:

• A total of 326 peer-reviewed publications involving EMERSE were identified by August 2019.
• Oncology research accounted for a significant portion, with 105 (32.2%) publications.
• EMERSE usage logs indicated extensive use, with nearly 3,500 users logging in over 100,000 times collectively. The most frequent semantic type mapped from search queries was 'disease or syndrome'.

Conclusions:

• EMERSE has proven to be a valuable asset for advancing cancer research.
• Successful deployment across multiple sites highlights its utility, though site-specific implementation challenges were noted.