Development and Validation of Natural Language Processing Algorithms in the ENACT National Electronic Health Record...

Area of Science:

• Biomedical Informatics
• Clinical Research Informatics
• Health Services Research

Background:

• Electronic health record (EHR) data offer rich clinical insights but often contain unstructured text.
• Accessing information within clinical narratives requires natural language processing (NLP) techniques.
• The Evolve to Next-Gen Accrual to Clinical Trials (ENACT) network aims to broaden access to EHR data for translational research.

Purpose of the Study:

• To describe the implementation and deployment of NLP infrastructure within the ENACT network.
• To make NLP-derived clinical information accessible and queryable across the network.
• To support translational research by unlocking the potential of clinical text.

Main Methods:

• Formation of the ENACT NLP Working Group to address the need for NLP capabilities.
• Description of implementation practices, logistics, and utilized NLP tools and technologies.
• Extension of the ENACT ontology for standardization and querying of NLP-derived data.
• Multisite evaluations of NLP algorithms to ensure performance and reliability.

Main Results:

• Successful implementation and deployment of NLP infrastructure across the ENACT network.
• Standardized and queryable access to NLP-derived clinical information is now available.
• Multisite evaluations demonstrated the efficacy of the NLP algorithms.

Conclusions:

• The ENACT network has successfully integrated NLP to leverage clinical narrative data.
• This advancement democratizes access to a broader range of EHR data for research.
• Lessons learned can guide other national data networks in deploying NLP for clinical text analysis.