Standardization of Clinical Progress Notes in Japan with FHIR.

Area of Science:

• Medical Informatics
• Natural Language Processing
• Healthcare Standards

Background:

• Electronic Medical Records (EMRs) are crucial for clinical documentation.
• Standardizing progress notes is essential for interoperability and AI training.
• Existing EMR structures may not align with international healthcare data standards.

Purpose of the Study:

• To establish a standardized format for progress notes as a training dataset for LLMs.
• To analyze the compatibility of Japanese EMRs with the Consolidated Clinical Document Architecture (C-CDA) on Fast Healthcare Interoperability Resources (FHIR).
• To identify challenges in achieving progress note standardization in Japan.

Main Methods:

• Reviewed existing clinical documentation guidelines.
• Analyzed the structure of the EMR system at the Medical School of Ehime University.
• Mapped Subjective, Objective, Assessment, and Plan (SOAP) sections to C-CDA sections.
• Introduced a DocumentReference extension for handling attachments like DICOM images.

Main Results:

• Successfully mapped EMR SOAP sections to C-CDA sections.
• A DocumentReference extension was implemented for attachments.
• Substantial discrepancies were identified between Japanese EMRs and C-CDA on FHIR.
• The study underscored the need for multi-vendor analysis.

Conclusions:

• A foundational step was taken towards creating LLM training data from EMRs.
• Significant gaps exist between current Japanese EMR practices and the C-CDA on FHIR standard.
• Further research involving multiple vendors is critical for advancing progress note standardization in Japan.